JP2010038896A - Speed detector, positioning detector, and positioning device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a positioning device that stably detects the position of a slider with simple structure at low cost, while achieving high design flexibility. <P>SOLUTION: The positioning device includes a slider 1 position-controlled in the X-axis direction, and a platen 10 in which a magnetic polar gear is formed on the surface facing the slider 1 to include the slider 1 and a planar motor, while being equipped with a location detector 300 comprising a resolver 30 which outputs signal depending on the X-axis directional position of the slider 1 and an acceleration sensor 31 for detecting X-axis directional acceleration of the slider 1. The location detector 300 determines the position of the slider 1 based on the output of both the resolver 30 and the acceleration sensor 31. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a speed detection device, a position detection device, and a positioning device using them, and more specifically, determines the speed and position of a moving body based on the outputs of an acceleration sensor and a position sensor.

  FIG. 27 is a block diagram showing an example of a conventional positioning device. Quadrilateral sliders A and B are mounted on the quadrilateral platen 10, and the positions of the sliders A and B are individually controlled on the platen 10 in the X-axis direction and the Y-axis direction, respectively. On the opposing surfaces of the platen 10 and the sliders A and B, magnetic pole teeth (hereinafter simply referred to as teeth) having a predetermined pitch are formed to constitute a planar motor. By driving the planar motor, the sliders A and B move to designated positions, respectively.

  Bar mirrors are provided on the three sides of the sliders A and B. A plurality of laser interferometers are fixedly disposed on each side 10 a to 10 d of the platen 10. The laser interferometers arranged on one opposing side 10a, 10b detect the position of the sliders A, B in the X-axis direction, and the laser interferometers arranged on the other opposing side 10c, 10d are sliders A, B The position in the Y-axis direction is detected. These laser interferometers detect the positions of the sliders A and B based on the interference between the emitted laser light and the reflected light that is reflected by the bar mirrors of the sliders A and B and returned. Patent Document 1 listed below describes a positioning device that detects the position of a slider using a laser interferometer.

JP2007-163418A

  However, when the above-described system using the laser interferometer is used for detecting the position of the slider, there is a problem that a large amount of cost is required because the laser interferometer and the bar mirror are expensive. Depending on the accuracy of position detection and the scale of the positioning device, it may reach 30 million yen per slider. Furthermore, the laser light source used in the laser interferometer has a short life and needs to be replaced after about 10,000 to tens of thousands of hours.

  In addition, in a system using a laser interferometer, the position cannot be detected if there is something that blocks the laser beam, so there is a problem that the degree of freedom of design of the entire positioning device is low. For example, in the example of FIG. 27, the position of the slider B in the X-axis direction is detected by a laser interferometer arranged on the side 10b of the platen 10, but the slider B is behind the slider A (ie, between the slider A and the side 10a). ), The laser beam from the laser interferometer on the side 10b is blocked by the slider A, and the position cannot be detected. Therefore, the movable range of the sliders A and B is limited, and the positioning device must be designed in consideration of this point.

  Furthermore, since a system using a laser interferometer uses laser light, there is a problem that the influence of the surrounding environment such as temperature and air fluctuation is large.

  The present invention eliminates the problems of the prior art, can stably detect the position of the slider with an inexpensive and simple configuration, and has a high degree of design freedom, and a speed detection device and a position required for the positioning device. An object is to provide a detection device.

In order to achieve the above object, the invention of claim 1
In a speed detection device that detects the speed of a moving object,
An acceleration sensor for detecting the acceleration of the moving object;
A first integrator for time integrating the output of the acceleration sensor;
A first high-pass filter that passes only high frequency components of the output of the first integrator;
A position sensor that outputs a signal corresponding to the position of the moving body;
A differentiator for differentiating the output of this position sensor with respect to time,
A first low-pass filter that passes only low frequency components of the output of the differentiator;
A first error controller that feeds back an error generated in the output of the first integrator to the input of the first integrator based on the difference between the output of the differentiator and the output of the first integrator. When,
A first computing unit for determining a speed of the moving body based on outputs of the first high-pass filter and the first low-pass filter;
A speed detecting device characterized by comprising:

The invention of claim 2
The speed detection device according to claim 1,
A delay time adjustment unit for adjusting a delay time generated in the output of the acceleration sensor is provided.

The invention of claim 3
The speed detection device according to claim 2,
The delay time adjustment unit adjusts the delay time generated in the output of the acceleration sensor so as to coincide with the delay time generated in the output of the position sensor.

The invention of claim 4
The speed detection device according to claim 2 or 3,
The delay time adjustment unit is
A value obtained by multiplying the output of the acceleration sensor by a predetermined gain is added to the output of the first integrator.

The invention of claim 5
The speed detection device according to any one of claims 1 to 4,
A second integrator for time integrating the output of the speed detector;
A second high-pass filter that passes only high frequency components of the output of the second integrator;
A second low-pass filter that passes only low frequency components of the output of the position sensor;
A second error controller that feeds back an error generated in the output of the second integrator to the input of the second integrator based on the difference between the output of the position sensor and the output of the second integrator. When,
A second computing unit for determining a position of the moving body based on outputs of the second high-pass filter and the second low-pass filter;
It is the position detection apparatus characterized by comprising.

The invention of claim 6
A slider whose position is controlled in at least one of the X-axis and Y-axis orthogonal to each other, a magnetic pole tooth formed on a surface facing the slider, and a platen constituting the slider and a planar motor;
A position detection device according to claim 5,
The position sensor is a resolver that outputs a signal corresponding to the position of the slider in the one-axis direction;
The acceleration sensor detects acceleration in the one-axis direction of the slider;
The positioning device is characterized in that the position detecting device obtains the position of the slider based on outputs of the resolver and the acceleration sensor.

The invention of claim 7
7. The positioning device according to claim 6, wherein at least one of the first and second high-pass filters and the first and second low-pass filters includes switching means for switching a cut-off frequency.

The invention of claim 8
8. The positioning device according to claim 7, wherein the switching unit switches the cut-off frequency according to a movement state of the slider.

The invention of claim 9
9. The positioning device according to claim 7, wherein the switching unit switches the cutoff frequency to a relatively low frequency when the speed of the slider is equal to or higher than a predetermined value, and the speed of the slider is equal to or lower than the predetermined value. In some cases, the cutoff frequency is switched to a relatively high frequency.

The invention of claim 10
10. The positioning device according to claim 6, wherein the acceleration sensor is arranged at a distance equal to a direction perpendicular to the one axis from a center line parallel to the one axis of the slider. 10. And

The invention of claim 11
10. The positioning device according to claim 6, wherein the acceleration sensor is arranged at a diagonal corner of the slider.

According to the invention of claim 1,
Refer to the velocity signal obtained based on the output of the acceleration sensor for the component in the high frequency region and the velocity signal obtained based on the output of the position sensor for the component in the low frequency region. By obtaining the speed signal, a speed signal with less error can be obtained in both the high frequency area and the low frequency area, and the speed can be detected with high accuracy. Further, since the first error controller is provided, it is possible to reduce the influence of the integration error due to the first integrator on the detection speed, and to detect the speed with higher accuracy.

According to the invention of claim 2,
Since the delay time adjustment unit that adjusts the delay time generated in the output of the acceleration sensor is provided, the accuracy of the signal obtained based on the output of the acceleration sensor can be improved by adjusting the delay time.

According to the invention of claim 3,
The delay time adjustment unit adjusts the delay time that occurs in the output of the acceleration sensor to match the delay time that occurs in the output of the position sensor, so it can eliminate the difference in delay time that occurs in the output of the acceleration sensor and the position sensor, Speed can be detected with higher accuracy.

According to the invention of claim 4,
Since the delay time adjustment unit adds a value obtained by multiplying the output of the acceleration sensor by a predetermined gain to the output of the first integrator, the delay time adjustment unit can be realized with a simple configuration.

According to the invention of claim 5,
The position of the moving body is referred to the position signal obtained based on the output of the speed detection device according to any one of claims 1 to 4 for the high frequency region component, and the position sensor is used for the low frequency region component. By obtaining with reference to the position signal obtained based on the output, a position signal with little error can be obtained in both the high frequency region and the low frequency region, and the position can be detected with high accuracy. Further, since the second error controller is provided, it is possible to reduce the influence of the integration error due to the first integrator on the detection position, and to detect the position with higher accuracy.

According to the invention of claim 6,
By using a resolver instead of a laser interferometer as a sensor for detecting the position of the slider, it is possible to stably detect the position of the slider with an inexpensive and simple configuration, and a positioning device with a high degree of freedom in design. Can be realized.
Further, by detecting the position of the slider with the position detection device according to claim 5, the accuracy of the feedback signal to the planar motor is improved, and the slider can be positioned with high accuracy.

According to the invention of claim 7,
Since the switching means for switching the cutoff frequencies of the first and second high-pass filters and the first and second low-pass filters is provided, the cutoff frequencies of the high-pass filter and the low-pass filter can be made variable.

According to the invention of claim 8,
Since the cut-off frequency is switched according to the moving state of the slider, the optimum cut-off frequency can be selected according to the state of acceleration, deceleration, etc. of the slider.

According to the invention of claim 9,
The cutoff frequency is set to a relatively low frequency when the slider speed is equal to or higher than the predetermined value, and is switched to a relatively high frequency when the slider speed is equal to or lower than the predetermined value. Therefore, the slider speed ripple and acceleration ripple are reduced. It is possible to reduce the waiting time during the positioning operation.

According to the invention of claim 10,
Since the acceleration sensor is arranged at a distance equal to the direction perpendicular to the one axis from the center line parallel to the one axis direction of the slider, the acceleration in the one axis direction can be accurately obtained by taking the average of the output of the acceleration sensor. It can be detected. Further, the acceleration of the slider in the uniaxial direction can be obtained regardless of the acceleration in the rotation direction of the slider, and the calculation for obtaining the acceleration can be simplified.

According to the invention of claim 11,
Since the acceleration sensors are arranged at the diagonal corners of the slider, the distance between the acceleration sensors can be increased, and the acceleration in the rotation direction of the slider can be detected with higher accuracy. If an acceleration sensor that can detect accelerations of two or more axes is used as the acceleration sensor, acceleration in both the X-axis and Y-axis directions can be detected with a small number of acceleration sensors.

  In the present invention, a resolver is used as a sensor for detecting the position of the slider. Since the resolver is composed only of a core and a coil, the structure is simple and inexpensive compared to other sensors such as a laser interferometer, and it has high environmental resistance. Furthermore, since the resolver has no short-lived parts like the laser light source of the laser interferometer, maintenance is easy. First, the operation principle of the resolver will be described.

  FIG. 1 is an explanatory diagram of an operation principle in which a resolver outputs a signal corresponding to a position. The resolver sensor unit 200 includes a core 201 and a coil 202 wound around the core 201. The sensor unit 200 is disposed on the platen 10 via a predetermined gap. Teeth are formed on the surface of the core 201 facing the platen 10 at the same pitch L as the teeth of the platen 10.

  Between the platen 10 and the core 201, an impedance Z corresponding to the relative position of the teeth of the platen 10 and the teeth of the core 201 exists. When a rectangular voltage having a constant amplitude is input to the coil 201 as an excitation voltage, the core 201 is excited and a magnetic circuit B is formed between the coil 201 and the platen 10. Since the magnetic circuit B is affected by the impedance Z, as a result, the voltage between the terminals of the coil 202 changes according to the relative position of the platen 10 and the core 201. Therefore, the voltage between the terminals of the coil 202 is taken out and its amplitude is used as a detection signal.

FIG. 1A shows a relative position where the facing area between the teeth of the core 201 and the teeth of the platen 10 is maximum, that is, the phase difference is 0 °. At this time, the impedance Z is minimized and the detection signal is maximized.
On the other hand, FIG. 1B shows a relative position where the facing area between the teeth of the core 201 and the teeth of the platen 10 is minimum, that is, the phase difference is 180 °. At this time, the impedance Z is maximized and the detection signal is minimized.

  When the sensor unit 200 moves on the platen 10, the impedance Z changes in a sine wave shape. Therefore, the detection signal changes in a sine wave shape with a pitch L as shown in FIG.

  The resolver is configured by preparing two sets of such sensor units 200 and shifting the phase with respect to the platen 10 by 90 ° as shown in FIG. The phase difference of the resolver with respect to the platen 10, that is, the relative position of the platen 10 with respect to the teeth is obtained by taking one of the sensor units 200 as the sin phase and the other as the cos phase and taking the arc tangent of the detection signals obtained from these sensor units. It is done. The relative position of the platen 10 with respect to which tooth is obtained by counting the number of sine wave peaks that the detection signal repeats from the origin position. As described above, a signal corresponding to the position from the origin position can be extracted by the resolver.

FIG. 3 is a top view showing a configuration in which the present invention is applied to a linear motor as a positioning device as the first embodiment.
The slider 1 is mounted on the platen 10 using an air bearing. Teeth 10a are formed on the opposing surfaces of the platen 10 and the slider 1 at a constant pitch L in the X-axis direction to constitute a planar motor. The position of the slider 1 is controlled in the X-axis direction on the platen 10 by this planar motor. In this figure, some of the teeth of the platen 10 are omitted.

  The slider 1 has a rectangular shape, and each side of the slider 1 is disposed on the platen 10 along either the X axis or the Y axis perpendicular to the X axis.

  A resolver 30 that outputs a detection signal corresponding to the position in the X-axis direction is fixed to one side surface along the X-axis of the slider 1. The resolver 30 is disposed at the center of the side surface of the slider 1 in the X-axis direction. The output of the resolver 30 can be regarded as a signal corresponding to the X position at the center of the slider 1. On the surface of the resolver 30 facing the platen 10, teeth are formed at a constant pitch L in the X-axis direction, and these teeth are arranged to face the teeth of the platen 10.

  The slider 1 moves to the target position on the platen 10 by servo control. A position command to be a target position at the center of the slider 1 is given to the planar motor, and a speed command is generated by feeding back the current position of the center of the slider 1 detected based on the output of the resolver 30. Further, the current speed calculated from the current position of the slider 1 detected based on the output of the resolver 30 is fed back to the speed command, thereby performing servo control of the slider 1. In order to move the slider 1 accurately and smoothly, the accuracy of the feedback signal to the planar motor is important.

  By the way, at the slider position calculated based on the output of the resolver 30, a fixed pattern expressed by the sum of n-order components n / L (n: 1 to ∞) with the tooth pitch L of the platen 10 as a basic period. Error occurs. Therefore, when the slider 1 is moved at the speed v, a position detection error occurs at the frequency f = nv / L at the position detected based on the output of the resolver 30. Therefore, even if the output of the resolver 30 is used as it is, a highly accurate feedback signal cannot be generated, and the slider 1 cannot be moved smoothly with high accuracy.

As an example, the case where the slider 1 is moved at a constant speed v will be described. Assume that the true current position, true current speed, and true current acceleration of the slider 1 are x, v, and a (a = 0 for constant speed motion). The position detection values of the slider 1 obtained based on the output of the resolver 30 and x _R, the speed detection value and the detected value of acceleration calculated based on the detected position value x _R v _R, and a _R.

The position detection value x _R is expressed by the following equation. Δx _n and φ _n are the error and phase of the n-th order component, respectively.

The speed detection value v _R of the slider 1 is expressed by time differentiation of the position detection value x _R.

Because

The acceleration detection value a _R of the slider 1 is expressed by a second-order time derivative of x _R.

Since the slider 1 moves at a constant speed,
And the second term on the right side of Equation (4) is zero. Therefore, the detected acceleration value a _R,
It becomes.

As represented in Equation (5), when calculating the detected acceleration value a _R based on the output of the resolver 30, the slider 1 is despite the uniform motion, the acceleration detected value a _R is not zero .

That is, if the slider 1 is servo-controlled with reference to the position detection value x _R obtained based on the output of the resolver 30, an extra acceleration a _R is given to the slider 1. Therefore, even if the servo control is performed so that the slider 1 moves at a constant speed, the slider 1 is actually moved while being shaken at the acceleration a _R.

A workpiece to be positioned is mounted on the slider 1. When the acceleration a _R is applied to the slider 1, a force of F = M · a _R (M: workpiece mass) is applied to the workpiece mounted on the slider 1. The force F may cause the workpiece to be displaced on the slider 1 or cause fatigue failure.

According to the equation (5), the amplitude of the acceleration a _R is proportional to the square of the frequency f = nv / L. That is, as the frequency f of the error component of the position detection value x _R obtained based on the output of the resolver 30 is high, the acceleration a _R increases, resulting in greater force F workpiece is subjected.

Therefore, in order to reduce the force F received by the workpiece by moving the slider 1 accurately and smoothly, components in the frequency region lower than the predetermined frequency f ₀ are used based on the output of the resolver 30 in the servo control of the slider 1. A feedback signal is generated, and a feedback signal is generated by another mechanism for components in a frequency region higher than the frequency f ₀ .

Returning to FIG. 3, the slider 1 is provided with an acceleration sensor 31 that detects the acceleration of the slider 1. In this embodiment, to generate a feedback signal for the components of the frequency range higher than the frequency f ₀ by using the acceleration sensor 31. Unlike the resolver 30, the acceleration detected by the acceleration sensor 31 has no fixed pattern error due to the tooth pitch L of the platen 10. Therefore, even if the detected acceleration is time-integrated to obtain the speed signal and position signal of the slider 1, no error occurs in a high frequency region.

  The acceleration sensor 31 is embedded in the center of the upper surface of the slider 1, and the upper surface of the slider 1 is a smooth surface. Outputs from the resolver 30 and the acceleration sensor 31 are input to the calculation unit 32. The resolver 30, the acceleration sensor 31, and the calculation unit 32 constitute a position detection device 300 that detects the current position of the slider 1.

  FIG. 4 is a block diagram of a servo control unit for positioning the slider. A position command Pcmd of the slider 1 is given from a host device (not shown). The position command Pcmd is a signal indicating an X coordinate to which the slider 1 is to be moved, that is, a target position. The position command Pcmd is input to the position control unit 100. The position control unit 100 determines the moving speed until the slider 1 reaches the target position, and outputs a speed command Vcmd to the speed control unit 101. The speed control unit 101 generates a current command Icmd corresponding to the speed command Vcmd and outputs it to the amplifier 102. The amplifier 102 supplies a current Io corresponding to the current command Icmd to the motor 103. The motor 103 moves the slider 1 using the current Io as a drive current. As the slider 1 moves, the workpiece 104 mounted on the slider 1 is positioned.

The position detection device 300 detects the detection signal X and the current acceleration a corresponding to the current position of the slider 1 by the resolver 30 and the acceleration sensor 31. The position detection device 300 generates a position feedback signal Pfb and a velocity feedback signal Vfb based on the detection signal X and the acceleration a. The position detection device 300 feeds back the generated position feedback signal Pfb to the position control unit 100 and feeds back the speed feedback signal Vfb to the speed control unit 101. The position control unit 100 and the speed control unit 101 perform servo control so that the deviation between the feedback signals Pfb and Vfb, the position command Pcmd, and the speed command Vcmd becomes zero.
In this figure, the position control unit 101 and the speed control unit 102 are shown only by functional blocks, but in actuality, they are constituted by a CPU.

FIG. 5 is a block diagram of the position detection apparatus 300. Detection signal X outputted from the resolver 30 is a position information, which corresponds to x _R in the formula (1). The detection signal X is time-differentiated by a differentiator 32a and converted into a speed signal v _R. The velocity signal v _R includes various frequency components as shown in the equation (3). Low pass filter 32b is a filter which passes only frequency components below the cut-off frequency f _0. The speed signal v _R is input to the low-pass filter 32b, and frequency components higher than the frequency f ₀ are removed. The output v _R ′ of the low-pass filter 32b is output to the calculation unit 32c.

The acceleration a output from the acceleration sensor 31 is time-integrated by an integrator 32d and converted into a speed signal va. The high-pass filter 32e is a filter that passes only frequency components higher than the same cutoff frequency f ₀ as the low-pass filter 32b. Speed signal va is input to a high-pass filter 32e, a low frequency component is removed than the frequency f _0. The output va ′ of the high pass filter 32e is output to the calculation unit 32c.

The computing unit 32c adds the speed signal v _R ′ input from the low-pass filter 32b and the speed signal va ′ input from the high-pass filter 32e to generate a speed signal Vfb. The speed signal Vfb is fed back to the speed control unit 101 as a speed feedback signal indicating the current speed of the slider 1.

Adding the 'speed signal va and' velocity signal v _R is the frequency components lower than the frequency f ₀ utilizes the speed signal v _R 'which is detected by the resolver 30, the frequency component higher than the frequency f ₀ Corresponds to complementing with the velocity signal va ′ detected by the acceleration sensor 31. As a result, the speed signal Vfb with less error can be generated even in a frequency region higher than the frequency f ₀ .
Moreover, by utilizing the speed signal Vfb as a speed feedback signal, the component of the frequency range higher than the frequency f ₀ with reference to the speed signal va 'detected by the acceleration sensor 31, a frequency range lower than the frequency f ₀ With respect to these components, speed feedback control of the slider 1 is performed with reference to the speed signal v _R ′ detected by the resolver 30.

  By the way, the acceleration a output from the acceleration sensor 31 has an offset error and a gain error. Therefore, if the acceleration a is simply integrated by the integrator 32d, an integration error resulting from these errors is superimposed on the speed signal va. Therefore, a mechanism for reducing this integration error is required.

An error controller 32f is provided as a mechanism for reducing the integration error of the integrator 32d.
An integration error resulting from an offset error or a gain error is a component in a low frequency region. For this reason, in the low frequency region, the speed signal obtained based on the resolver 30 has a smaller error and is closer to the true value than the speed signal obtained based on the acceleration sensor 31.
Therefore, the error controller 32f calculates an offset error and a gain error included in the acceleration a based on the deviation of the speed signal va from the speed signal v _R and feeds back to the input of the integrator 32d.

Figure 6 is a diagram showing details and structure around the error controller 32f, a configuration example in which the control band of the error control by the frequency f _C = ω ₀ / 2π. In this figure, the differentiation process is expressed by a transfer function s, and the integration process is expressed by a transfer function 1 / s.
The speed signal v _R and the speed signal va are input to the error controller 32f. The error controller 32f calculates a difference between the input speed signal v _R and the speed signal va, and multiplies the difference by a feedback gain ω ₀ . Further, the processing of (s + ω ₀ ) / s is performed and added to the input of the integrator 32d. Thereby, the integration error superimposed on the speed signal va is reduced, and the accuracy of the speed feedback signal Vfb can be increased.
The frequency f _C is set sufficiently lower than the cutoff frequency f ₀ described above. Therefore, although superimposed on the speed signal v _R , the frequency domain error component removed by the low-pass filter 32b does not affect the operation of the error controller 32f.

  The high pass filter 32e and the low pass filter 32b will be described in more detail. FIG. 7 is a block diagram showing the configuration of the high-pass filter 32e and the low-pass filter 32b. The high-pass filter 32e and the low-pass filter 32b are expressed by digital function blocks.

The high-pass filter 32e and the low-pass filter 32b digitally convert the input speed signals va and v _R with a period T, and perform processing represented by the following equation (6). The process of Expression (6) is a general IIR filter process. The IIR filter is a digital filter that can obtain filter characteristics with a small order, and has an advantage that the calculation load of the CPU can be reduced. Equation (6) is a first-order IIR filter.
p _i : current input value q _i : current output value p _i-1 : input value before one cycle q _i-1 : output value before one cycle a, b, c: coefficients

In FIG. 7, the speed signals va and v _R correspond to the input values p _i in the equation (6), and the speed signals va ′ and v _R ′ correspond to the output values q _i . z ⁻¹ is an inverse function of the z function and corresponds to a time delay of one period T. The coefficients a _H , b _H , and c _H correspond to the coefficients a, b, and c in Expression (6), respectively, and are determined so that the entire 32e becomes a high-pass filter with a cutoff frequency f ₀ . Similarly, the coefficients a _L , b _L , and c _L correspond to the coefficients a, b, and c in Equation (6), respectively, and are determined so that the entire 32b becomes a low-pass filter with a cutoff frequency f _0. .

In the high pass filter 32e, the multiplier 40 multiplies the speed signal va by the coefficient b _H and outputs the result to the adder 41. Further, a speed signal one cycle before the speed signal va is obtained by the z ⁻¹ operator 42, and this speed signal is multiplied by a coefficient c _H by the multiplier 43 and output to the adder 44. Further, a speed signal one cycle before the speed signal va ′ is obtained by the z ⁻¹ operator 45, the speed signal is multiplied by a coefficient a _H by a multiplier 46, and output to the adder 44. The adder 44 adds the outputs from the multipliers 43 and 46 and outputs the result to the adder 41. The adder 41 adds the outputs from the adders 40 and 44 and outputs the result as a speed signal va ′. As described above, the processing of Expression (6) is executed.

Similarly, in the low-pass filter 32 b, the multiplier 50 multiplies the speed signal v _R by the coefficient b _L and outputs the result to the adder 51. Further, a speed signal one cycle before the speed signal v _R is obtained by the z ⁻¹ operator 52, the speed signal is multiplied by a coefficient c _L by the multiplier 53, and output to the adder 54. Further, a speed signal one cycle before the speed signal v _R ′ is obtained by the z ⁻¹ operator 55, the speed signal is multiplied by a coefficient a _L by a multiplier 56, and output to the adder 54. The adder 54 adds the outputs from the multipliers 53 and 56 and outputs the result to the adder 51. The adder 51 adds the outputs from the adders 50 and 54 to obtain a speed signal v _R ′. As described above, the processing of Expression (6) is executed.

The determination of the cutoff frequency f ₀ will be described.
FIG. 8 is a diagram for explaining the process of determining the cutoff frequency f ₀ , where (a) is an error component of the speed signal v _R and (b) is an error of the acceleration signal obtained by differentiating the speed signal v _R. component, exemplifies the 'error component, (d) the speed signal v _R' of the error component of the acceleration signal obtained by differentiating the (c) is the velocity signal v _R. All the horizontal axes are the speeds of the slider 1.

The speed signal v _R and the acceleration signal obtained by differentiating the speed signal v _R with respect to time include an n-th order (n = 1 to ∞) error component, as shown in (a) and (b). Yes. The magnitude of the error component of each order changes according to the speed of the slider 1. In FIG. 8, only n = 1, 2, 3, and 6th order components are shown as representatives. Moreover, (a) of FIG. 8 is equivalent to what represented Formula (3) and (b) illustrated Formula (5).

First, a speed range for moving the slider 1 is determined. Thereafter, the cutoff frequency of the low-pass filter 32b is provisionally determined. Then, the velocity signal v _R ′ that has passed through the low-pass filter 32b is obtained, and further, an error component included in the acceleration signal obtained by time differentiation of the velocity signal v _R ′ is obtained. Among the obtained error components, the maximum absolute value of the error components generated within the speed range of the slider 1 is extracted. The absolute value of the extracted error component is compared with an allowable acceleration value that can be tolerated by the work mounted on the slider 1. If the absolute value of the error component exceeds the allowable acceleration value, the cutoff frequency of the low-pass filter 32b is provisionally determined to be a lower frequency and compared with the allowable acceleration value again. A series of processing is repeated until the maximum value of the error component is equal to or less than the allowable acceleration value, and the maximum frequency fmax at which the maximum value of the error component is equal to or less than the allowable acceleration value is obtained. The cutoff frequency f ₀ is set to a value lower than the maximum frequency fmax.

By appropriately setting the cut-off frequency f ₀ in this way, as shown in FIGS. 8C and 8D, an acceleration signal obtained by time differentiation of the speed signal v _R ′ and the speed signal v _R ′. Can be largely removed.

Returning to FIG. The output of the calculator 32c is time-integrated by the integrator 33d and converted into a position signal xa. High-pass filter 33e is of the same construction as the high-pass filter 32e, a filter which passes frequency components higher than the cut-off frequency f _0. Position signal xa is input to the high-pass filter 33e, a low frequency component is removed than the frequency f _0. The output xa ′ of the high pass filter 33e is output to the calculator 33c.

The low-pass filter 33b has the same configuration as the low-pass filter 32b, and is a filter that allows a frequency component lower than the cutoff frequency f ₀ to pass through. The detection signal X of the resolver 30 is input to the low-pass filter 33b, and the frequency component higher than the frequency f ₀ is removed. The output x _R ′ of the low-pass filter 33b is output to the calculator 33c.

The calculator 33c adds the position signal x _R ′ input from the low-pass filter 33b and the position signal xa ′ input from the high-pass filter 33e to obtain a position signal Pfb. The position signal Pfb is fed back to the position controller 100 as a position feedback signal indicating the current position of the slider 1.

Adding a 'position signal xa and' position signal x _R utilizes a position signal x _R 'which is detected by the resolver 30 for frequency components lower than the frequency f _0, the frequency component higher than the frequency f ₀ Corresponds to complementing with the position signal xa ′ detected by the acceleration sensor 31. As a result, the position signal Pfb with less error can be generated even in a frequency region higher than the frequency f ₀ .
Moreover, by utilizing the position signal Pfb as a position feedback signal, the component of the frequency range higher than the frequency f ₀ with reference to the position signal xa 'detected by the acceleration sensor 31, a frequency range lower than the frequency f ₀ The position feedback control of the slider 1 is performed with reference to the position signal x _R ′ detected by the resolver 30 for this component.

  Incidentally, an error controller 32f is provided to reduce the integration error of the integrator 32d, but the integration error cannot be completely removed from the speed signal va. Therefore, a slight integration error remains in the speed signal Vfb generated from the speed signal va. Therefore, if the speed signal Vfb is simply integrated by the integrator 33d, the integration error resulting from the integration error remaining in the speed signal Vfb is superimposed on the position signal xa. Therefore, an error controller 33f is provided to prevent such accumulation of integration errors.

  The error controller 33f has the same configuration as the error controller 32f. The position signal xa and the resolver detection signal X output from the integrator 33d are input to the error controller 33f. The error controller 32f calculates an integration error included in the speed signal Vfb based on the deviation of the position signal xa from the detection signal X, and feeds it back to the input of the integrator 33d. Thereby, the integration error of the integrator 33d is reduced and accumulation of integration errors is prevented.

The differentiator 32a, integrators 32d and 33d, low-pass filters 32b and 33b, high-pass filters 32e and 33e, calculators 32c and 33c, and error controllers 32f and 33f constitute the calculator 32.
The resolver 30, the acceleration sensor 31, the differentiator 32a, the low-pass filter 32b, the calculator 32c, the integrator 32d, the high-pass filter 32e, and the error controller 32f constitute a speed detection device 33a.
Further, the position detection device 300 is configured by the speed detection device 33a, the low-pass filter 33b, the calculator 33c, the integrator 33d, the high-pass filter 33e, and the error controller 33f.
In FIG. 5, the calculation unit 32 is shown only in the functional block diagram, but actually the calculation unit 32 is configured by a CPU or the like in which a program is downloaded.

FIG. 9 is a diagram showing transfer characteristics from the actual speed of the slider 1 to the speed signals v _R ′, va ′, Vfb as operation waveforms of the speed detection device 33 a, and (a) is based on the output of the resolver 30. velocity signal obtained Te v 'transfer characteristic, (b) the speed signal va obtained based on the output of the acceleration sensor 31' _R is the transfer characteristic of the transfer characteristic, (c) the speed signal Vfb of. In this figure, the cutoff frequency f ₀ is set to about 100 Hz.

As shown in FIG. 9A, the velocity signal v _R ′ has a frequency component higher than about 100 Hz attenuated by the action of the low-pass filter 32b.
On the other hand, as shown in FIG. 9B, the speed signal va ′ has a frequency component lower than about 100 Hz attenuated by the action of the high-pass filter 32e.

When (a) and (b) of FIG. 9 are added, the speed signal Vfb shown in (c) of FIG. 9 is obtained. In the speed signal Vfb, the high frequency component attenuated by the low-pass filter 32b is complemented by the high frequency component that has passed through the high-pass filter 32e, and the amplitude is maintained over a wide frequency range from a low frequency component to a high frequency component.
In this way, the speed signal constituting the speed signal Vfb is switched from the speed signal v _R ′ to the speed signal va ′ with the cutoff frequency f ₀ as a boundary.

  It is also possible to obtain the speed signal and the position signal based on the output of the acceleration sensor 31 in all frequency regions. However, the offset error of the acceleration a is likely to accompany changes with time, and the amount of change is difficult to predict. For this reason, with respect to the components in the low frequency region, the speed signal and the position signal obtained based on the output of the resolver 30 have less error with time. Therefore, the speed signal and position signal are not obtained by the acceleration sensor 31 for all frequency regions, but the resolver 30 is used for components in the low frequency region, and the speed signal and position obtained by the acceleration sensor 31 are used for components in the high frequency region. The feedback control of the slider 1 is performed with reference to the signal.

The linear motor of the present embodiment is configured as described above, and in particular, the speed detection device 33a is
The speed of the slider 1 is referred to the speed signal va ′ obtained based on the output of the acceleration sensor 31 for the high frequency region component, and the speed signal v obtained based on the output of the resolver 30 for the low frequency region component. By obtaining with reference to _R ′, a speed signal with less error can be obtained in both the high frequency region and the low frequency region, and the speed of the slider 1 can be detected with high accuracy. Further, since the error controller 32f is provided, it is possible to reduce the influence of the integration error by the integrator 32d on the speed signal Vfb, and to detect the speed of the slider 1 with higher accuracy.

In addition, the position detection device 300 is
The position of the slider 1 is referred to the position signal xa ′ obtained based on the output of the speed detector 33a according to claim 1 for the high frequency region component, and based on the output of the resolver for the low frequency region component. Since the position signal x _R ′ obtained in this way is obtained, a position signal with little error can be obtained in both the high frequency region and the low frequency region, and the position of the slider 1 can be detected with high accuracy. Can do. Further, since the error controller 33f is provided, it is possible to reduce the influence of the integration error by the integrator 33d on the position signal Pfb, and to detect the position of the slider 1 with higher accuracy.

Further, since the resolver 30 is used as a sensor for detecting the position of the slider 1 instead of the laser interferometer, the position of the slider 1 can be stably detected with an inexpensive and simple configuration, and the degree of freedom in design. It is possible to realize a positioning device having a high height.
Further, by detecting the position of the slider 1 with the position detection device 300, the accuracy of the feedback signals Pfb and Vfb to the planar motor is improved, and the slider 1 can be positioned with high accuracy.

  In this embodiment, the acceleration sensor 31 is embedded in the upper surface of the slider 1, but the acceleration sensor 31 may be built in the slider 1.

  In this embodiment, the low-pass filters 32b and 33b and the high-pass filters 32e and 33e are composed of the first-order IIR filters, but may be second-order or higher-order filters. If a high-order IIR filter is used, the cutoff characteristic of the filter can be improved.

In this embodiment, the cut-off frequency f ₀ of the low-pass filters 32b and 33b and the high-pass filters 32e and 33e is determined based on the error component included in the acceleration value obtained by time differentiation of the speed signal v _R ′. However, in order to calculate the time derivative of the speed signal v _R ′, Δx _n and φ _n (n = 1 to ∞) must all be known. In practice, it is difficult to keep track of all [Delta] x _n and phi _n of _{(n = 1~∞), Δx n} , typical values of phi _n (e.g., n = 1, 2, 3, 6 ) May be used for calculation. Alternatively, the frequency f ₀ may be obtained empirically by experiments or the like.

The distribution of frequency components included in the position feedback signal Pfb changes according to the movement state of the slider 1.
For example, when the slider 1 is stopped, the slider 1 has neither a position change nor an acceleration change, so that the position feedback signal Pfb has many components in a low frequency region. For this reason, the position signal x _R ″ obtained by the resolver 30 is dominant as the position signal constituting the position feedback signal Pfb.

  On the other hand, immediately after the start of movement of the slider 1 or immediately after it stops, the change in the acceleration of the slider 1 becomes steep, and the position feedback signal Pfb has a high frequency region component. Therefore, the position signal xa ″ obtained by the acceleration sensor 31 is dominant as the position signal constituting the position feedback signal Pfb.

That is, the proportion of the resolver 30 and the acceleration sensor 31 with respect to the position feedback signal Pfb varies depending on the movement state of the slider 1. This change is due to the position feedback signal Pfb, it is generated by combining the low-pass filter 33b and the high-pass filter 33e having a common cut-off frequency f _0.

  The change in the ratio of the resolver 30 and the acceleration sensor 31 in the position feedback signal Pfb means that the ratio of the resolver 30 and the acceleration sensor 31 in the detection error included in the position feedback signal Pfb also changes.

  In the above example, the error due to the detection system of the acceleration sensor 31 becomes dominant immediately after the slider 1 stops, and the error due to the detection system of the resolver 30 becomes dominant after a while after the stop. The error caused by the detection system of the resolver 30 and the error caused by the detection system of the acceleration sensor 31 are errors inherent to each detection system. Therefore, the detection error of the position feedback signal Pfb changes with time according to the movement state of the slider 1.

  Therefore, even when the slider 1 is located at the same point on the platen 10, the position feedback signal Pfb obtained immediately after the slider 1 stops at that point and the position obtained after a while after stopping. The feedback signal Pfb is different. That is, the position feedback signal Pfb changes with time. For this reason, if the position feedback signal Pfb is fed back and the slider 1 is servo-controlled, the slider 1 moves according to the time change of the error of the position feedback signal Pfb.

Eventually, the slider 1 stops at a position where the position detection error of the resolver 30 is on. When performing some work while the slider 1 is stopped, a waiting time is required until the position of the slider 1 falls within an allowable error range. This waiting time becomes shorter as the cutoff frequency f ₀ is higher. However, if the cut-off frequency f ₀ is set high, position feedback control is performed with reference to the position signal from the resolver 30 up to a high frequency range, and the speed ripple and acceleration ripple during the movement of the slider 1 increase.

FIG. 10 is a diagram showing operation waveforms when the cutoff frequency f ₀ is high and low.
FIG. 10A shows the time differential value of the position command Pcmd input to the position control unit 100. The time differential value of the position command Pcmd is controlled so that the time differential value of the position command until reaching the target position of the slider 1 becomes a trapezoidal wave. The slider 1 accelerates from time t1 to t2, moves at a constant speed from time t2 to t3, and decelerates from time t3 to t4. At time t4, the slider 1 arrives at the target position, and the update of the position command is completed.

(B1) to (b3) in FIG. 10 are actual operation waveform diagrams of the slider 1 when the cutoff frequency f ₀ is set to a high frequency, where (b1) is the speed of the slider 1 and (b2) is This is the acceleration of the slider 1. (B3) is the distance between the slider 1 and the target position, the solid line is the actual position of the slider 1, and the broken line is the position command value Pcmd. Note that the slider 1 is finally stabilized at a position deviated from the target position because of a position detection error of the resolver 30.
By setting the cutoff frequency f ₀ to a high frequency, the ratio of the resolver 30 to the position feedback signal Pfb is increased, and the slider 1 quickly approaches the target position. Therefore, the waiting time is short. However, large ripples occur in speed and acceleration.

On the other hand, (c1) to (c3) in FIG. 10 are operation waveform examples of the actual slider 1 when the cutoff frequency f ₀ is set to a low frequency, where (c1) is the speed of the slider 1 and (c2 ) Is the acceleration of the slider 1, and (c3) is a comparison with the position command value Pcmd of the slider 1.
By setting the cut-off frequency f ₀ to a low frequency, speed and acceleration ripples are greatly reduced. However, since the ratio of the resolver 30 to the position feedback signal Pfb becomes low, it takes time until the slider 1 reaches the target position, and the waiting time becomes long.

  When performing the positioning operation of the slider 1, that is, when the slider 1 is moved to the target position by changing the position command to the position control unit 100 over time, the slider 1 is quickly moved after the update of the position command is completed. It is important to shorten the waiting time by reaching the target position.

In the second embodiment, as compared with the first embodiment, the cutoff frequency f ₀ of the low-pass filters 32b and 33b and the high-pass filters 32e and 33e is made variable, thereby suppressing and positioning the velocity ripple and acceleration ripple during slider movement. The waiting time during operation can be shortened.

  FIG. 11 is a block diagram of the servo control unit of the second embodiment. Hereinafter, the same components as those in the first embodiment are denoted by the same reference numerals, and the description thereof is omitted.

  The position detection device 300 ′ detects the detection signal X and the current acceleration a corresponding to the current position of the slider 1 by the resolver 30 and the acceleration sensor 31. The position detection device 300 ′ generates a position feedback signal Pfb ′ and a speed feedback signal Vfb ′ based on the detection signal X and the current acceleration a, and feeds back to the position control unit 100 and the speed control unit 101. The position control unit 100 and the speed control unit 101 perform servo control so that deviations between the feedback signals Pfb 'and Vfb' and the position command Pcmd and the speed command Vcmd are zero.

  The switching signal generation unit 105 generates the switching signal S based on the position command Pcmd. The switching signal generator 105 generates a speed signal Vj by differentiating the position command Pcmd with respect to time, and compares the absolute value of the speed signal Vj with a predetermined threshold value Vth. The switching signal generator 105 outputs an H signal as the switching signal S when the speed signal Vj is less than or equal to the threshold Vth, and outputs an L signal when the speed signal Vj exceeds the threshold Vth.

The switching signal S is input to the position detection device 300 ′. The position detection device 300 ′ includes a low-pass filter and a high-pass filter that can switch the cutoff frequency to the frequency f _L or f _H (f _L <f _H ). The position detection device 300 ′ switches the cutoff frequency of these filters by the switching signal S.

FIG. 12 is a block diagram of the position detection apparatus 300 ′. The low-pass filters 32b ′ and 33b ′ can switch the cutoff frequency to the frequency f _L or f _H. Further, the high-pass filters 32e ′ and 33e ′ can switch the cutoff frequency to the frequency f _L or f _H. Each filter switches the cutoff frequency to the frequency f _L when the switching signal S is the L signal, and switches the cutoff frequency to the frequency f _H when the switching signal S is the H signal. That is, if the speed signal Vj is equal to or smaller than the threshold Vth is the frequency f _H is selected as the cutoff frequency, if the speed signal Vj exceeds the threshold value Vth is the frequency f _L is selected as the cutoff frequency.

Low-pass filter 32 b ', the switching signal S is removed frequency components higher than the frequency f _L from the speed signal v _R which is input when the L signal, the speed signal when the speed signal S is H signal v _R removing frequency components higher than the frequency f _H from.

The high-pass filter 32e ′ removes a frequency component lower than the frequency f _L from the input speed signal va when the switching signal S is an L signal, and from the speed signal va when the speed signal S is an H signal. Remove frequency components lower than f _H.

The output v _R ″ of the low-pass filter 32b ′ and the output va ″ of the high-pass filter 32e ′ are input to the calculator 32c. Calculator 32c adds the "speed signal va and the" speed signal v _R, generates a speed signal Vfb '. The speed signal Vfb ′ is used as a speed feedback signal indicating the current speed of the slider 1.

Similarly, the low-pass filter 33b ′ removes a frequency component higher than the frequency f _L from the input position signal X when the switching signal S is an L signal, and the position signal when the speed signal S is an H signal. A frequency component higher than the frequency f _H is removed from X.

The high-pass filter 33e ′ removes a frequency component lower than the frequency f _L from the input speed signal Vfb ′ when the switching signal S is an L signal, and the speed signal Vfb ′ when the speed signal S is an H signal. removing frequency components lower than the frequency f _H from.

The output x _R ″ of the low-pass filter 33b ′ and the output xa ″ of the high-pass filter 33e ′ are input to the calculator 33c. The calculator 33c adds the position signal x _R ″ and the position signal xa ″ to generate a position signal Pfb ′. The position signal Pfb ′ is used as a position feedback signal indicating the current position of the slider 1.

The high pass filter 32e ′ and the low pass filter 32b ′ will be described in more detail.
FIG. 13 is a block diagram showing the configuration of the high-pass filter 32e ′ and the low-pass filter 32b ′. In this figure, switching means 47 to 49 and 57 to 59 are added to the high pass filter 32e and the low pass filter 32b of the first embodiment.

The coefficients a _HH , b _HH , and c _HH are coefficients that make the entire high-pass filter 32e ′ a high-pass filter with a cutoff frequency f _H , and the coefficients a _HL , b _HL , and c _HL have the cutoff frequency f _{L as} a whole. This is a coefficient for a high-pass filter.
The coefficients a _LH , b _LH , and c _LH are coefficients that make the entire low-pass filter 32b ′ a low-pass filter having a cutoff frequency f _H , and the coefficients a _LL , b _LL , and c _LL are the coefficients that make the entire low-pass filter 32b ′ have a cutoff frequency f _L. It is a coefficient used as a low-pass filter.

When the switching signal S is an L signal, the switching means 47 to 49 and 57 to 59 select coefficients a _HL , b _HL , c _HL , a _LL , b _LL , and c _LL , respectively. When the switching signal S is H signal, respectively switching means 47～49,57～59 coefficients _{a HH, b HH, c HH} , a LH, b LH, selects c _LH. Multipliers 46, 40, 43, 56, 50 and 53 multiply the coefficients selected by the switching means 47 to 49 and 57 to 59, respectively. Thereby, when the switching signal S is the L signal, the cutoff frequency of the high-pass filter 32e ′ and the low-pass filter 32b ′ is set to the frequency f _L , and when the switching signal is the H signal, it is set to the frequency f _H.

Frequency f _L is set to a sufficiently lower frequency than the maximum frequency fmax. The frequency f _H is 'seek, the maximum frequency fmax' maximum frequency fmax allowed when limiting the maximum speed of the slider 1 to the speed Vth to a value below.

In the positioning device configured as described above, when the speed of the slider 1 exceeds the speed Vth, the switching signal S becomes the L signal, and the cutoff frequencies of the low-pass filters 32b ′ and 33b ′ and the high-pass filters 32e ′ and 33e ′ are the frequency f _L. It becomes. By setting the cutoff frequency to a low frequency, the ratio of the position signal generated based on the acceleration sensor 31 to the position feedback signal Pfb ′ increases, and the speed ripple and the acceleration ripple are suppressed.

When the slider 1 decelerates to the speed Vth or less, the switching signal S is switched to the H signal, and the cutoff frequencies of the low-pass filters 32b ′ and 33b ′ and the high-pass filters 32e ′ and 33e ′ become the frequency f _H. By setting the cutoff frequency to a high frequency, the ratio of the position signal generated based on the resolver 30 to the position feedback signal Pfb ′ increases, and the slider 1 quickly stops at the target position.

  14A and 14B are diagrams showing operation waveforms of the slider 1 according to the present embodiment, in which FIG. 14A shows the speed of the slider 1 and FIG. 14B shows the acceleration of the slider 1. (C) is the distance between the slider 1 and the target position, the solid line is the actual position of the slider 1, and the broken line is the position command value.

In FIG. 14A, the speed ripple is significantly reduced as compared with FIG. 10B1. Further, in FIG. 14B, the acceleration ripple is greatly reduced as compared with FIG. 10B2. Thereby, the slider 1 moves smoothly to the target position.
Further, in FIG. 14C, the slider 1 reaches the target position more quickly than in FIG. 10C3. Therefore, the waiting time during the positioning operation is shortened.

  According to the present embodiment, there are provided switching means 47 to 49 and 57 to 59 for switching the cut-off frequencies of the low-pass filters 32b ′ and 33b ′ and the high-pass filters 32e ′ and 33e ′ as well as having the same effect as the first embodiment. Therefore, the cutoff frequency of these filters can be made variable.

  Further, since the cut-off frequency is switched according to the movement state of the slider 1, the optimum cut-off frequency can be selected according to the state of acceleration, deceleration, etc. of the slider 1.

The cutoff frequency is set to a relatively low frequency f _L when the speed of the slider 1 is equal to or higher than the predetermined speed Vth, and is switched to a relatively high frequency f _H when the speed of the slider 1 is equal to or lower than the speed Vth. The speed ripple and acceleration ripple of the slider 1 can be reduced, and the waiting time during the positioning operation can be shortened.

In the present embodiment, in determining the cutoff frequency f _H, the velocity Vth of switching the cutoff frequency previously determined previously, the cutoff frequency from the maximum frequency fmax 'the slider 1 is allowed when moving at this speed Vth to determine the f _H. However, the cutoff frequency f _H leave the previously determined, may determine the velocity Vth from the moving speed of the slider 1 is allowed by the cut-off frequency.
Further, the slider 1 is moved while changing the cutoff frequency f _H as a parameter may be determined while checking whether acceleration a which is output from the acceleration sensor 31 satisfies the allowable acceleration value.

  In this embodiment, the switching signal S is switched based on one threshold value (speed Vth). However, different threshold values are used for switching from the L signal to the H signal and for switching from the H signal to the L signal. You may prepare.

  Further, in this embodiment, the switching signal S is switched according to the time differential value of the position command to the slider 1, that is, the speed, but it may be switched according to other than the speed. For example, the switching signal S may be switched to the L signal before the position command of the slider 1 is updated, and the switching signal S may be switched to the H signal a predetermined time before the updating of the position command is completed. Since there is a time difference from the end of the update of the position command until the slider 1 actually reaches the target position, it may be switched to the H signal at the moment when the update of the position command is completed.

In this embodiment, the speed signal Vj generated from the position command Pcmd is used to generate the switching signal S. However, the speed command Vcmd or the actual speed of the slider 1 may be used. However, since the speed command Vcmd and the actual speed of the slider 1 include various disturbances and correction amounts for the disturbances, when the switching signal S is generated based on these signals, the speed serving as a threshold value. When hysteresis is provided for Vth, the switching process is stabilized.
Alternatively, the switching signal S may be generated by a host device that outputs the position command Pcmd and directly output to the position detection device 300 ′.

  In this embodiment, the number of stages switched by the switching means 47 to 49 and 57 to 59 is two. However, a plurality of stages may be switched while holding more coefficients. If the number of stages can be switched, the number of cutoff frequency stages that can be switched by the low-pass filters 32b 'and 33b' and the high-pass filters 32e 'and 33e' can be increased. If the cutoff frequency is switched to a higher frequency immediately before the slider 1 stops, the slider 1 can be stopped more smoothly.

FIG. 15 is a diagram showing a configuration in which the present invention is applied to a two-dimensional planar motor as a third embodiment.
A slider 1 ′ is mounted on the platen 10 using an air bearing. The position of the slider 1 ′ is controlled on the platen 10 in the X-axis direction, the Y-axis direction, and the θ-axis direction. Teeth 10b are formed in a lattice shape with a constant pitch L in the X-axis direction and the Y-axis direction on the surface of the platen 10 facing the slider 1 '. In the figure, only a part of the teeth 10b of the platen 10 is shown.

  The slider 1 ′ has a rectangular shape, and each side of the slider 1 ′ is disposed on the platen 10 so as to be along either the X axis or the Y axis perpendicular to the X axis.

  FIG. 16 is a layout diagram of the core of the slider 1 ′. A plurality of cores 1 a to 1 d are provided on the surface of the slider 1 ′ facing the platen 10, and constitutes a flat motor with the platen 10. The cores 1a and 1b are formed with teeth having a constant pitch L in the X-axis direction, and move the slider 1 'in the X-axis direction. On the cores 1a and 1b, teeth with a constant pitch L are formed in the Y-axis direction, and the slider 1 'is moved in the Y-axis direction. By driving the planar motor, the slider 1 ′ moves to the designated X position and Y position on the platen 10. Further, by generating thrust in the cores 1a and 1b in the opposite direction, thrust is generated in the θ-axis direction of the slider 1 '. Hereinafter, the core 1a is referred to as an X1 core, the core 1b is referred to as an X2 core, and the cores 1c and 1d are referred to as a Y core.

  Resolvers 30a to 30d are fixed to the center of each side surface of the slider 1 '. Resolvers 30a and 30b are arranged on opposite sides along the X axis of the slider 1 ', and resolvers 30c and 30d are arranged on opposite sides along the Y axis. The resolvers 30a and 30b output detection signals corresponding to positions in the X-axis direction, and the resolvers 30c and 30d detect detection signals corresponding to positions in the Y-axis direction.

The average of the outputs of the resolvers 30a and 30b is taken and set as a detection signal X _R corresponding to the current position in the X-axis direction at the center of the slider 1 ′. The average of the outputs of the resolvers 30c and 30d is taken as a detection signal Y _R corresponding to the current position in the Y-axis direction at the center of the slider 1 ′. Further, the rotation angle θ _R of the slider 1 ′ is obtained from the difference between the outputs of the resolver 30a and the resolver 30b.

Acceleration sensors 31a to 31d are embedded in the upper surface portion of the slider 1 ′. The acceleration sensors 31a and 31b detect acceleration in the X-axis direction, and the acceleration sensors 31c and 31d detect acceleration in the Y-axis direction. Based on the outputs of the acceleration sensors 31a to 31d, the acceleration A _{X in} the X-axis direction, the acceleration A _{Y in the Y-} axis direction, and the acceleration A _θ in the θ-axis direction of the slider 1 ′ are obtained.

FIG. 17 is a diagram showing the arrangement of the acceleration sensors 31a to 31d on the slider 1 ′. The acceleration sensors 31a to 31d are attached at the following positions when the center point C of the slider 1 ′ is the origin. Expressions (7-1) to (7-4) represent the position of the acceleration sensor in polar coordinates.
Acceleration sensor 31a: (r _X1 , θ _X1 ) (7-1)
Acceleration sensor 31b: (r _X2 , θ _X2 ) (7-2)
Acceleration sensor 31c: (r _Y1 , θ _Y1 ) (7-3)
Acceleration sensor 31d: (r _Y2 , θ _Y2 ) (7-4)

The outputs of the acceleration sensors 31a to 31d are expressed as follows. B _{X1 to} B _Y2 are offset errors included in the output of each acceleration sensor.

When the equations (8-1) to (8-4) are arranged, the accelerations A _X , A _Y , A _θ are expressed as follows.
Accelerations A _X , A _Y , _Aθ are obtained by calculating Expressions (9) to (11) based on the outputs of the acceleration sensors 31 a to 31 d.

Returning to FIG. 15, the outputs of the resolvers 30 a to 30 d and the acceleration sensors 31 a to 31 d are input to a calculation unit 320 (not shown). Resolver 30 a to 30 d, the acceleration sensor 31 a to 31 d, in the calculating portion 320, constituting the position detecting device 301 detects the current position and the rotation angle theta _R in the X-axis direction and the Y-axis direction of the slider 1 '(not shown) . Based on the output of the position detector 301, the slider 1 ′ is servo-controlled in the X-axis direction, the Y-axis direction, and the θ-axis direction.

FIG. 18 is a block diagram of a servo control unit for positioning the slider 1 ′. Position command Pxcmd the slider 1 'from a host device (not shown), Pycmd, given a P _theta cmd. The position commands Pxcmd and Pycmd are signals indicating the X coordinate and Y coordinate to which the slider 1 ′ is to be moved, that is, the target position. The position command P _θ cmd is a signal indicating the target rotation angle of the slider 1 ′. Position command _Pxcmd, Pycmd, P θ cmd is respectively inputted into the position control section. Each position control unit, a position command Pxcmd, Pycmd, speed command Vxcmd corresponding to P _theta cmd, generates _Vycmd, V θ cmd, and outputs to the respective speed control unit. Each speed controller, the inputted speed command Vxcmd, current command Irx corresponding to _Vycmd, V θ cmd, Iry, generates Ir _theta, and outputs to the command distributor 106.

The X1 amplifier, the X2 amplifier, and the Y amplifier are amplifiers corresponding to the X1 core, the X2 core, and the Y core, respectively. Command distributor 106, the input current command Irx, Iry, based on Ir _theta, effective current command Ix1cmd given to each amplifier, Ix2cmd, generates a Iycmd. The X1 amplifier supplies a current Iox1 according to the effective current command Ix1cmd to the X1 amplifier, the X2 amplifier supplies a current Iox2 according to the effective current command Ix2cmd to the X2 amplifier, and the Y amplifier supplies a current Ioy according to the effective current command Iycmd. Is supplied to the Y amplifier. Each core moves the slider 1 'using the currents Iox1, Iox2, and Ioy as drive currents. As the slider 1 'moves, the workpiece mounted on the slider 1' is positioned.

The position detection device 301 detects the detection signals X _R and Y _R and the rotation angle θ _R according to the current positions of the slider 1 ′ in the X-axis direction and the Y-axis direction by the resolvers 30a to 30d. Further, the position detection device 301 detects accelerations A _X , A _Y , and A _θ of the slider 1 ′ in the X axis direction, the Y axis direction, and the θ axis direction by the acceleration sensors 31a to 31d.

Position detecting device 301, the detection signal X _R, Y _R, the rotation angle theta _R and acceleration A _X, based on A _Y, A _theta, position feedback signal Pxfb, Pyfb, P _θ fb and the speed feedback signal Vxfb, Vyfb, V _θ fb is generated. Position detector 301 generates the position feedback signal Pxfb, Pyfb, the P _theta fb, respectively fed back to a corresponding position control unit. The position detecting device 301, the generated speed feedback signal Vxfb, the _Vyfb, V θ fb, respectively fed back to a corresponding speed control unit.

Each position control unit performs feedback signal Pxfb, Pyfb, _P θ fb and position command Pxcmd, Pycmd, the servo control so that the deviation becomes zero each of the P _theta cmd. Further, each speed controller, the speed feedback signal _Vxfb, Vyfb, V θ fb and the speed command Vxcmd, the servo control so that the deviation becomes zero each of the _Vycmd, V θ cmd performed.
In FIG. 18, the position control unit, the speed control unit, and the command distributor 106 are shown only by functional blocks, but in practice they are configured by a CPU.

FIG. 19 is a diagram illustrating a configuration of the command distributor 106. The command distributor 106 subtracts Ir _θ from the current command Irx to generate an effective current command Ix1cmd. Further, the command distributor 106 adds the current commands Irx and Ir _θ to generate an effective current command Ix2cmd. Thus, an effective current command reflecting the current command Ir _θ of the slider 1 ′ is given to the X1 core and the X2 core, and the rotation angle of the slider 1 ′ is controlled. The current command Iry is output as the effective current command Iycmd.

  FIG. 20 is a block diagram showing the position detection device 301. The acceleration calculation unit 321 and the calculation units 322 to 324 constitute the calculation unit 320. The calculation units 322 to 324 have the same configuration as the calculation unit 32 in the first embodiment.

The acceleration calculation unit 321 receives the outputs of the acceleration sensors 31a to 31d. Acceleration calculation unit 321 performs the calculation of equation (9) to (11), the acceleration A _X of the slider 1 ', calculates a A _Y, A _theta. The calculated accelerations A _X , A _Y , A _θ are output to the calculation units 322 to 324, respectively.

The calculation unit 322 receives the acceleration A _X and the detection signal X _R and outputs a position feedback signal Pxfb and a velocity feedback signal Vxfb in the X-axis direction.
Similarly, the calculation unit 323 receives the acceleration A _Y and the detection signal Y _R and outputs a position feedback signal Pyfb and a velocity feedback signal Vyfb in the Y-axis direction.
Further, the calculation unit 324 receives the acceleration A _θ and the rotation angle θ _R and outputs a position feedback signal P _θ fb and a rotation speed feedback signal V _θ fb in the θ-axis direction of the slider 1 ′.

  Each feedback signal output from the arithmetic units 322 to 324 is used as a feedback signal for servo control of the slider 1 '.

  Note that the arithmetic unit 320 is configured by a CPU or the like in which a program is downloaded. Other configurations are the same as those of the first embodiment.

  According to the present embodiment, the accuracy of the position feedback signal and speed feedback signal of the slider 1 'in the X-axis direction and the Y-axis direction can be improved. Furthermore, since a plurality of resolvers and acceleration sensors for detecting the one-axis direction of the slider 1 ′ are provided, the acceleration in the θ-axis direction of the slider 1 ′ can be obtained. By using this, the accuracy of the position feedback signal and the velocity feedback signal around the θ axis of the slider 1 ′ can be increased, and the slider 1 ′ can be controlled with high accuracy.

The acceleration sensors 31a and 31b are
The second term on the right side of Equation (9) = 0. By this arrangement, it is possible to obtain the acceleration A _X regardless acceleration A _theta. To simplify the calculation of equation (9), by eliminating the influence of the acceleration A _theta, it is possible to improve the accuracy of the acceleration A _X.
Similarly, the acceleration sensors 31c and 31d are
The second term on the right side of Equation (10) = 0. By this arrangement, it is possible to obtain the acceleration A _Y regardless acceleration A _theta. To simplify the calculation of equation (10), by eliminating the influence of the acceleration A _theta, it is possible to improve the accuracy of the acceleration A _Y.

  FIG. 21 is a diagram illustrating an example of the arrangement of the acceleration sensors 31a to 31d that satisfy Expression (12) and Expression (13). Cx and Cy are center lines of the slider 1 'in the X-axis direction and the Y-axis direction, respectively. The acceleration sensors 31a and 31c and the acceleration sensors 31d and 31b are arranged at a distance dx in the X-axis direction with the center line Cx interposed therebetween. Further, the acceleration sensors 31a and 31d and the acceleration sensors 31c and 31b are arranged at a distance dy in the Y-axis direction with the center line Cy interposed therebetween.

  In the present embodiment, the calculation units 322 to 324 have the same configuration as the calculation unit 32 of the first embodiment, but the same configuration as the calculation unit 32 ′ in the second embodiment may be used.

In this embodiment, the rotation angle θ _R of the slider 1 ′ is determined based on the difference between the outputs of the resolver 30a and the resolver 30b. However, the rotation angle of the slider 1 ′ is determined based on the difference between the outputs of the resolver 30c and the resolver 30d. θ _R may be obtained. Furthermore, an average of the rotation angle obtained from the resolver 30a and the resolver 30b and the rotation angle obtained from the resolver 30c and the resolver 30d may be used as the rotation angle θ _R of the slider 1 ′.

  FIG. 22 is a top view showing Embodiment 4 of the present invention. In this embodiment, the type and arrangement of the acceleration sensor are changed with respect to the third embodiment.

  In Example 3, the acceleration sensors 31a to 31d each detected acceleration in one axis direction. However, in recent years, acceleration sensors that can detect accelerations of two or more axes with one device have been put into practical use.

  In FIG. 22, reference numerals 31e and 31f denote acceleration sensors that detect accelerations in the biaxial directions of the X-axis direction and the Y-axis direction. The acceleration sensor 31e is attached to the corner of the side where the resolvers 30a and 30c of the slider 1 'are disposed. The acceleration sensor 31f is attached to the corner of the side where the resolvers 30b and 30d of the slider 1 'are disposed.

This embodiment includes a position detection device 301 ′ (not shown in FIG. 22). The position detection device 301 ′ detects the detection signals X _R and Y _R and the rotation angle θ _R according to the current positions of the slider 1 ′ in the X-axis direction and the Y-axis direction by the resolvers 30a to 30d. Further, the position detection device 301 ′ detects accelerations A _X , A _Y , A _θ of the slider 1 ′ in the X-axis direction, the Y-axis direction, and the θ-axis direction by the acceleration sensors 31e, 31f. The configuration of the servo control unit in the X-axis direction, the Y-axis direction, and the θ-axis direction is the same as that in the third embodiment.

FIG. 23 is a diagram showing a position detection device 301 ′ in the present embodiment.
The position detection device 301 ′ includes an acceleration calculation unit 321 ′ and calculation units 322 to 324.
The acceleration calculation unit 321 ′ receives accelerations in the biaxial directions of the X axis direction and the Y axis direction from the acceleration sensors 32e and 32f, respectively. The acceleration calculation unit 321 ′ regards the outputs of the acceleration sensors 32e and 32f as the outputs of two acceleration sensors (for the X-axis direction and for the Y-axis direction) arranged at the same position, and the expressions (9) to ( 11) The calculation is performed. Other configurations are the same as those of the third embodiment.

According to the present embodiment, the same effect as in the third embodiment is obtained, and furthermore, since the acceleration sensors 32e and 32f are arranged at diagonal corners of the slider 1 ′, the distance between the acceleration sensors can be increased. can be can be the acceleration a _theta of theta-axis direction of the slider 1 'is detected more accurately.
Further, by using an acceleration sensor capable of detecting acceleration of two or more axes as the acceleration sensors 32e and 32f, acceleration in both the X-axis and Y-axis directions can be detected with a small number of acceleration sensors.

  FIG. 24 is a diagram illustrating a position detection apparatus 300 ″ according to the fifth embodiment. In this embodiment, a delay time adjustment unit 32g is added to the first embodiment.

There are inherent delay times tdr and tda before the detection signal X output from the resolver 30 and the acceleration a output from the acceleration sensor 31 reach the arithmetic unit 32 ″. The delay time tda of the acceleration a. If the difference between the delay time tdr of the detection signal X and the detection signal X occurs, simply using the acceleration a and the detection signal X reaching the calculation unit 32 ″ as they are will cause a difference in delay time (tgap = tda−tdr). A phase shift occurs between the speed signal va ′ and the speed signal v _R ′, and an error occurs in the current speed Vfb. An error also occurs in the current position Xfb calculated based on the current speed Vfb.
Therefore, by reducing the difference (tgap) between the delay time of the acceleration a and the detection signal X, the current speed Vfb and the current position Xfb can be detected with higher accuracy.

As a mechanism for reducing the difference (tgap) in the delay time between the acceleration a and the detection signal X, the speed detection device 33a ″ is provided with a delay time adjustment unit 32g. The delay time adjustment unit 32g detects the delay time tda of the acceleration a. A time advance or time delay is caused approximately to the speed signal va so as to coincide with the delay time tdr of X. As a result, the phase shift between the speed signal va and the speed signal v _R is eliminated, and as a result The phase shift between the signal va ′ and the speed signal v _R ′ is eliminated.

  The delay time adjustment unit 32g receives the acceleration a corrected by the error controller 32f. The delay time adjustment unit 32g generates an adjustment signal vad by multiplying the acceleration a by a predetermined gain Kff, and adds the adjustment signal vad to the output of the integrator 32d. The speed signal obtained by adding the adjustment signal vad is input to the high-pass filter 32e and the error controller 32f as the speed signal va.

  The gain Kff is a value corresponding to the difference (tgap) between the acceleration a and the delay time of the detection signal X, and is obtained experimentally or empirically. The gain Kff is held in advance in the delay time adjustment unit 32g. Setting a positive value for the gain Kff causes a time advance in the speed signal va, and setting a negative value for the gain Kff causes a time delay in the speed signal va.

FIG. 25 is a diagram showing the adjustment of the delay time by the delay time adjustment unit 32g, and shows the speed signal va. The horizontal axis represents time [s], and the vertical axis represents speed [m / s].
Line L shows the case where the delay time is not adjusted for speed signal va, line Lr shows the case where time advance is caused in speed signal va, and line Ls shows the case where time delay is caused in speed signal va. is there.
When the gain Kff = 0 is set, it is the same as not adjusting the speed signal va, and is equivalent to the configuration of the first embodiment.

When the delay time tdr of the detection signal X is larger than the delay time tda of the acceleration a, in order to increase the delay time tda and make it coincide with the delay time tdr, a negative value is set for the gain Kff and the speed signal va is set. Causes a time delay.
On the other hand, when the delay time tdr of the detection signal X is smaller than the delay time tda of the acceleration a, in order to reduce the delay time tda and match the delay time tdr, a positive value is set for the gain Kff, and the speed signal Cause va to advance in time.

Note that setting a positive value for the gain Kff corresponds to estimating the future speed signal va for the time corresponding to the gain Kff based on the current acceleration a. Setting a negative value for the gain Kff corresponds to estimating the past speed signal va for a time corresponding to the gain Kff based on the current acceleration a.
By providing such a delay time adjustment unit 32g, the delay time tda of the acceleration a can be approximately adjusted. Other configurations are the same as those of the first embodiment.

FIG. 26 is a diagram illustrating an example of the effect of the delay time adjustment unit 32g when there is a difference between the delay time tdr of the resolver 31 and the delay time tda of the acceleration sensor a. FIG. 26A shows the position command value Pcmd. 26 (b1) and (b2) show the case where the delay time adjustment unit 32g is not provided, and (c1) and (c2) show the case where the delay time adjustment unit 32g is provided. 26, (b1) and (c1) are deviations between the position command Pcmd and the detected current position Pfb, and (b2) and (c2) are deviations between the current position Pfb and the true value of the current position of the slider 1. It is.
Note that (b1) and (b2) in FIG. 26 are deviations between the position command Pcmd and the detected current position Pfb. Similarly, since (c1) and (c2) in FIG. 26 are deviations between the current position Pfb and the true value of the current position of the slider 1, it is desirable that the values be near zero.

  As shown in FIG. 26A, according to the position command value Pcmd, the slider 1 reaches the target position at about t = 0.3 [s], and the motor 103 should stop.

  On the other hand, in (b1) of FIG. 26, it takes about t = 0.6 [s] until the deviation between the position command Pcmd and the current position Pfb is stabilized to zero. In (b2), it takes a long time of about t = 1.0 [s] until the deviation between the current position Pfb and the true value is stabilized to zero. This means that the slider 1 continues to move after t = 0.3 [s] and the position is not stable. If it takes time for the position of the slider 1 to become stable, there is a problem that the time at which the work on the work mounted on the slider 1 can be started is delayed, leading to a decrease in throughput.

  On the other hand, in (c1) and (c2) of FIG. 26, the time until the deviation is stabilized to 0 is shortened to t = 0.5 [s], which is compared with (b1) and (b2) of FIG. Has been greatly improved.

  According to the present embodiment, the acceleration sensor 31 has the same effect as that of the first embodiment and includes the delay time adjustment unit 32g that adjusts the delay time tda generated in the acceleration a. Therefore, the acceleration sensor 31 such as the speed signal va and the speed signal va '. The accuracy of the signal obtained based on the output can be increased.

  The delay time adjustment unit 32g adjusts the delay time tda generated in the acceleration a so as to coincide with the delay time tdr generated in the detection signal X output from the resolver 30. The difference in delay time that occurs can be eliminated. Thereby, the current speed Vfb and the current position Pfb can be detected with higher accuracy.

Further, since the delay time adjusting unit 32g adds the adjustment signal vad obtained by multiplying the acceleration a by a predetermined gain Kff to the output of the integrator 32d, the delay time adjusting unit can be realized with a simple configuration.
Since the delay time adjustment unit 32g can be configured by software, it can be realized at low cost without changing hardware. If the delay time adjustment unit 32g is configured by software, the gain Kff can be easily set or changed.

  In addition, although the present Example was demonstrated as a modification of the said Example 1, it can apply similarly to the said Examples 2-4. That is, in the second to fourth embodiments, by adding the delay time adjustment unit 32g, the current speed Vfb and the current position Pfb can be detected with higher accuracy when there is a difference between the delay time of the acceleration a and the detection signal X. It becomes like this.

It is explanatory drawing of the operation principle of the position detection by a resolver. It is a figure which shows the structure of a resolver. It is a top view which shows Example 1 of this invention. It is a block diagram of the servo control part which positions a slider. 2 is a block diagram of a position detection device 300. FIG. It is a figure which shows the detail of the error controller 32f, and the structure of the periphery. It is a block diagram of the high pass filter 32e and the low pass filter 32b. It is a diagram for explaining the cutoff frequency f _0. Speed signal v _R ', va', illustrates a transfer characteristic of Vfb. It is a figure which shows the operation | movement waveform when the cutoff frequency f is high and low. It is a block diagram which shows the servo control of the positioning device of Example 2 of this invention. It is a block diagram of position detection apparatus 300 '. It is a block diagram of the high pass filter 32e 'and the low pass filter 32b'. It is a figure which shows the operation | movement waveform of the slider 1 of Example 2. FIG. It is a top view which shows Example 3 of this invention. FIG. 6 is a layout diagram of a core of a slider 1 '. It is a figure which shows arrangement | positioning of the acceleration sensors 31a-31d. It is a block diagram of the servo control part which positions slider 1 '. 3 is a diagram showing a configuration of a command distributor 106. FIG. It is a block diagram which shows the position detection apparatus 301 of Example 3. FIG. It is a figure which shows the position of the acceleration sensors 31a-31d. It is a top view which shows Example 4 of this invention. It is a block diagram which shows the position detection apparatus 301 'of Example 4. FIG. It is a figure which shows position detection apparatus 300 "of Example 5. FIG. It is a figure which shows adjustment of the delay time by the delay time adjustment part 32g. It is a figure which shows an example of the effect of the delay time adjustment part 32g. It is a block diagram which shows an example of the positioning device of a prior art example.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Slider 10 Platen 30 Resolver 31 Accelerometer 32 Arithmetic unit 32a Differentiator 32b Low pass filter 32c Arithmetic unit 32d Integrator 32e High pass filter 32f Error controller 32g Delay time adjustment unit 33a Speed detection device 33b Low pass filter 33c Calculator 33d Integrator 33e High-pass filter 33f Error controller 100 Position controller 101 Speed controller 102 Amplifier 103 Motor 104 Work 300 Position detection device

Claims (11)

In a speed detection device that detects the speed of a moving object,
An acceleration sensor for detecting the acceleration of the moving object;
A first integrator for time integrating the output of the acceleration sensor;
A first high-pass filter that passes only high frequency components of the output of the first integrator;
A position sensor that outputs a signal corresponding to the position of the moving body;
A differentiator for differentiating the output of this position sensor with respect to time,
A first low-pass filter that passes only low frequency components of the output of the differentiator;
A first error controller that feeds back an error generated in the output of the first integrator to the input of the first integrator based on the difference between the output of the differentiator and the output of the first integrator. When,
A first computing unit for determining a speed of the moving body based on outputs of the first high-pass filter and the first low-pass filter;
A speed detection device comprising:   The speed detection apparatus according to claim 1, further comprising a delay time adjustment unit that adjusts a delay time generated in the output of the acceleration sensor.   The speed detection apparatus according to claim 2, wherein the delay time adjustment unit adjusts a delay time generated in the output of the acceleration sensor to coincide with a delay time generated in the output of the position sensor. The delay time adjustment unit is
4. The speed detection apparatus according to claim 2, wherein a value obtained by multiplying the output of the acceleration sensor by a predetermined gain is added to the output of the first integrator. The speed detection device according to any one of claims 1 to 4,
A second integrator for time integrating the output of the speed detector;
A second high-pass filter that passes only high frequency components of the output of the second integrator;
A second low-pass filter that passes only low frequency components of the output of the position sensor;
A second error controller that feeds back an error generated in the output of the second integrator to the input of the second integrator based on the difference between the output of the position sensor and the output of the second integrator. When,
A second computing unit for determining a position of the moving body based on outputs of the second high-pass filter and the second low-pass filter;
A position detection device comprising: A slider whose position is controlled in at least one of the X-axis and Y-axis orthogonal to each other, a magnetic pole tooth formed on a surface facing the slider, and a platen constituting the slider and a planar motor;
A position detection device according to claim 5,
The position sensor is a resolver that outputs a signal corresponding to the position of the slider in the one-axis direction;
The acceleration sensor detects acceleration in the one-axis direction of the slider;
The positioning device, wherein the position detection device obtains the position of the slider based on outputs of the resolver and the acceleration sensor.   The positioning device according to claim 6, wherein at least one of the first and second high-pass filters and the first and second low-pass filters includes a switching unit that switches a cut-off frequency.   The positioning device according to claim 7, wherein the switching unit switches the cutoff frequency in accordance with a movement state of the slider.   The switching means switches the cutoff frequency to a relatively low frequency when the slider speed is equal to or higher than a predetermined value, and sets the cutoff frequency to a relatively high frequency when the slider speed is equal to or lower than the predetermined value. The positioning device according to claim 7, wherein the positioning device is switched.   The positioning device according to claim 6, wherein the acceleration sensor is arranged at a distance equal to a direction perpendicular to the one axis from a center line parallel to the one axis of the slider. .   The positioning device according to claim 6, wherein the acceleration sensor is arranged at a corner of the slider.