JP2014219340A - Offset correction method and offset correction device - Google Patents

Abstract

An offset correction method capable of appropriately correcting an offset of a geomagnetic sensor is provided. An offset correction method includes: a first detection result at a first time of a geomagnetic sensor installed on a moving body; and a first detection time of the gyro sensor installed on the moving body from the first time. The detection result of the geomagnetic sensor at the second time using the detection results up to a second time later than the time of the first time, the predicted detection result, and the second time And comparing with the second detection result of the geomagnetic sensor in determining whether the result of the comparison is within a predetermined range, and when determining that the result of the comparison is not within the predetermined range, Correcting the offset of the output of the geomagnetic sensor based on using the predicted detection result. [Selection] Figure 5

Description

  The present invention relates to an offset correction method and an offset correction apparatus.

  In a navigation system or a positioning system, an amount of relative movement from a reference position is estimated using an azimuth angle calculated from the output of a geomagnetic sensor or a gyro sensor. In order to calculate the correct azimuth, it is necessary to correct the offset from the output of the geomagnetic sensor. Therefore, it is common to calibrate the geomagnetic sensor in advance and correct the offset.

  However, there is a problem that the offset changes due to a change in magnetization caused by a change in temperature of components in the terminal device on which the geomagnetic sensor is mounted or a change in environment where the terminal device is placed. If calibration is performed, the offset can be corrected. However, depending on the use situation of the user, there is a possibility that the use of the terminal device is interrupted and there is no room for performing calibration again on the spot.

  Therefore, in Patent Document 1, it is proposed to correct the offset error of the geomagnetic sensor by using the difference between the change amount of the detection value of the gyro sensor in the minute time and the change amount of the detection value of the geomagnetic sensor in the minute time. .

JP 2011-112500 A

  However, in Patent Document 1, one of the geomagnetic sensor and the gyro sensor performs straight-ahead determination, and a sensor that is not used for determination in the case of straight-ahead travel is corrected for a sensor that is used for determination in the case of non-straight-ahead travel. . Since the geomagnetic sensor is easily affected by external magnetic field changes such as magnetization, the straight traveling determination itself becomes inappropriate, and there is a possibility that an appropriate correction result cannot be obtained. If the correction cannot be made properly, there is a possibility that the direction cannot be obtained appropriately.

  The present invention has been made in view of the above problems, and an object of the present invention is to propose a new technique capable of appropriately correcting the offset of the geomagnetic sensor.

  Application Example 1 The offset correction method according to this application example includes the first detection result of the first time of the geomagnetic sensor installed in the moving body and the first detection result of the gyro sensor installed in the moving body. The detection result of the geomagnetic sensor at the second time using the detection result from the time until the second time after the first time, and the predicted detection result; , Comparing the second detection result of the geomagnetic sensor at the second time to determine whether the comparison result is within a predetermined range, and the comparison result is not within the predetermined range The offset of the output of the geomagnetic sensor is corrected using the predicted detection result.

  According to this application example, the first detection result at the first time of the geomagnetic sensor and the detection result of the gyro sensor from the first time to the second time after the first time are used. Thus, the detection result of the geomagnetic sensor at the second time is predicted. The gyro sensor is not affected by magnetization. Therefore, the result at the second time of the geomagnetic sensor predicted using the detection result from the first time to the second time of the gyro sensor is the case where the offset of the geomagnetic sensor has not changed due to the influence of magnetization. It can be said that it corresponds to the detection result of. If the result of comparing the predicted detection result with the second detection result is not within the predetermined range, it is considered that the offset has changed. Therefore, the offset can be corrected more appropriately by using the predicted detection result. In this specification, “offset (of the output of the geomagnetic sensor)” means an error with respect to a true value (a true geomagnetic vector generated by the earth's magnetic field) included in the output of the geomagnetic sensor.

  Application Example 2 In the offset correction method described in the application example, when it is determined that the comparison result is within the predetermined range, the bias of the gyro sensor is corrected using the second detection result. May be included.

  According to this application example, when it is determined that the comparison result is within the predetermined range, the bias of the gyro sensor is corrected using the second detection result. Since the second detection result when the comparison result is within the predetermined range is appropriate, the bias of the gyro sensor can be corrected more accurately by using the second detection result. Thereby, when the corrected detection result of the gyro sensor is used for the subsequent prediction, the prediction can be performed more accurately. In this specification, “bias (of a gyro sensor or acceleration sensor)” means a zero point bias of the gyro sensor (or acceleration sensor). That is, it means an output generated when the gyro sensor (or acceleration sensor) is stationary.

  Application Example 3 In the offset correction method described in the application example described above, when it is determined that the comparison result is not within the predetermined range, the offset detection method and the second time are used to determine the first detection result. The detection result of the geomagnetic sensor at the third time may be predicted using the detection result of the gyro sensor up to a third time after the second time.

  According to this application example, when it is determined that the comparison result is not within the predetermined range, the detection result at the third time is predicted using the predicted detection result. If the comparison result is not within the predetermined range, the predicted detection result is more appropriate than the second detection result. Therefore, by using the predicted detection result for prediction at the third time, it is more appropriate. Predictions can be made.

  Application Example 4 In the offset correction method according to the application example, the geomagnetic sensor at the second time using the corrected offset when it is determined that the comparison result is not within the predetermined range. To obtain a new second detection result, the new second detection result, and the third time after the second time to the third time after the second time. The detection result of the geomagnetic sensor at the third time may be predicted using the detection result of the gyro sensor.

  According to this application example, when it is determined that the comparison result is not within the predetermined range, the output of the geomagnetic sensor at the second time is corrected using the corrected offset. Then, the detection result at the third time is predicted using the obtained new second detection result. Since the corrected offset is more appropriate than the offset before correction, an appropriate prediction can be performed by using the new second detection result for prediction at the third time.

  Application Example 5 In the offset correction method described in the application example, when it is determined that the result of the comparison is within the predetermined range, the offset detection method includes the second detection result and the second time. The detection result of the geomagnetic sensor at the third time may be predicted using the detection result of the gyro sensor up to the third time after the second time.

  According to this application example, when it is determined that the comparison result is within the predetermined range, the detection result of the geomagnetic sensor at the third time is predicted using the second detection result. By using the detection result of the geomagnetic sensor for the prediction at the third time, it is possible to suppress the prediction accuracy from being lowered due to accumulation of the bias of the gyro sensor in the predicted detection result.

  Application Example 6 The offset correction apparatus according to this application example includes the first detection result at the first time of the geomagnetic sensor installed on the moving body, and the first gyro sensor installed on the moving body. And a geomagnetism prediction unit that predicts a detection result of the geomagnetic sensor at the second time using a detection result from the time to a second time after the first time, and the geomagnetism prediction unit. A geomagnetic offset determination unit that compares the predicted detection result with a second detection result of the geomagnetic sensor at the second time and determines whether the comparison result is within a predetermined range; When the geomagnetic offset determination unit determines that the result of the comparison is not within a predetermined range, the offset of the output of the geomagnetic sensor is calculated using the predicted detection result. Comprising an offset correction unit positive for the.

  According to this application example, the geomagnetism prediction unit includes the first detection result of the geomagnetic sensor at the first time and the gyro sensor from the first time to the second time after the first time. The detection result of the geomagnetic sensor at the second time is predicted using the detection result. The gyro sensor is less stable and less stable than the geomagnetic sensor. Also, it is not affected by magnetization. Therefore, the result at the second time of the geomagnetic sensor predicted using the detection result from the first time to the second time of the gyro sensor is the detection result when the offset of the geomagnetic sensor is appropriately corrected. It can be said that it is equivalent. If the result of comparison between the predicted detection result and the second detection result by the geomagnetic offset determination unit is not within a predetermined range, it is considered that the offset has changed. Therefore, the offset correction unit corrects the offset based on the predicted detection result, so that a more appropriate offset can be obtained.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a diagram illustrating a state in which a user wears a navigation device according to a first embodiment and a schematic configuration of the navigation device. The block diagram which shows the function structure of a navigation apparatus. The conceptual diagram which represented the process part as a functional block. The flowchart which shows the flow of a movement locus | trajectory calculation process. The flowchart which shows the flow of an azimuth | direction calculation process. (A) is a figure which shows the movement locus | trajectory of a comparative example, (B) is a figure which shows the azimuth | direction angle calculated in the comparative example. The figure which shows the movement locus | trajectory in 1st Example. (A) is a figure which shows the azimuth calculated from the detection result (actual value) of the geomagnetic sensor in the first embodiment, and (B) is a comparison with the azimuth calculated from the predicted value in the first embodiment. FIG. The figure which shows the comparison with the detection result (actual value) of the geomagnetic sensor in 1st Example, and a predicted value. (A) corresponds to the X axis, (B) corresponds to the Y axis, and (C) corresponds to the Z axis. The figure which shows the movement locus | trajectory in 2nd Example. The figure which shows the comparison with the correction value and prediction value of a geomagnetic sensor in a 2nd Example. (A) corresponds to the X axis, (B) corresponds to the Y axis, and (C) corresponds to the Z axis.

  Hereinafter, with reference to the drawings, an embodiment when the present invention is applied to a portable navigation device which is a kind of electronic apparatus will be described. However, embodiments to which the present invention can be applied are not limited to the embodiments described below. Note that the drawings to be used are appropriately enlarged or reduced so that the part to be described can be recognized.

(First embodiment)
1. Schematic Configuration FIG. 1 is an explanatory diagram of a state in which a portable navigation device 1 according to this embodiment is worn by a user who is a pedestrian and a schematic configuration of the navigation device 1. The navigation device 1 is an electronic device having a function of estimating a stride of a user who is a pedestrian, estimating a moving direction, calculating a moving trajectory, and the like according to an instruction operation from the operation unit 20, as shown in FIG. For example, it is used by being worn on the user's waist (right waist or left waist).

  The navigation device 1 includes a sensor unit 40 having an acceleration sensor 41, a gyro sensor 43, and a geomagnetic sensor 42. Then, the user's walking is detected based on the component of the acceleration vector detected by the acceleration sensor 41 (hereinafter simply referred to as “acceleration”), and the stride is estimated in accordance with a model formula representing the correlation between the acceleration and the stride. To do. Further, the moving direction of the user is estimated based on the geomagnetic vector detected by the geomagnetic sensor 42 and the angular velocity detected by the gyro sensor 43. Then, the user's movement trajectory is calculated by sequentially calculating the user's position using the estimated stride and movement direction, and is displayed on the display unit 30.

  The acceleration sensor 41 is a sensor that detects a user's acceleration vector, and detects an acceleration vector of three orthogonal axes. The acceleration sensor 41 may be either a strain gauge type or a piezoelectric type, and may be a MEMS (Micro Electro Mechanical Systems) sensor. The gyro sensor 43 is a sensor that detects the angular velocity of the user, and detects the angular velocity of three orthogonal axes. The gyro sensor 43 is arranged and set so as to have the same axial direction as the acceleration sensor 41.

  The geomagnetic sensor 42 is a sensor that detects a geomagnetic vector based on the earth's magnetic field, and detects the magnetic flux density of three orthogonal axes. The geomagnetic sensor 42 may be either one using an MR (Magneto resistive) element or one using an MI (Magneto impedance) element, or may be one using a Hall element. The geomagnetic sensor 42 is arranged and set so as to have the same axial direction as the acceleration sensor 41. In FIG. 1, the acceleration sensor 41, the gyro sensor 43, and the geomagnetic sensor 42 are illustrated as independent sensors. However, the acceleration sensor 41, the gyro sensor 43, and the geomagnetic sensor 42 may be integrated sensors. Good.

  In this embodiment, when viewed from the user, the forward direction in the front-rear direction is the “positive direction of the X axis”, the right direction in the horizontal direction is the “positive direction of the Y axis”, and the downward direction (vertical direction) is the “Z axis. This will be described as an orthogonal three-axis coordinate system with a “positive direction”. In order to make the explanation easy to understand, the axial directions of the detection axes of the acceleration sensor 41, the gyro sensor 43, and the geomagnetic sensor 42 and the positive directions of the respective axes are defined as the X axis, the Y axis, the Z axis, and the positive direction. It is explained as the same. Of course, it may be different in practice. Since the detection axes of the acceleration sensor 41, the gyro sensor 43, and the geomagnetic sensor 42 and the axial directions of the X axis, the Y axis, and the Z axis are fixed to the user in a relative relationship, the acceleration sensor 41, From the detection results of the gyro sensor 43 and the geomagnetic sensor 42, the respective values of X, Y, and Z can be calculated. Hereinafter, the coordinate system of these sensors is referred to as a “sensor coordinate system”. The coordinate system of the sensor is a coordinate system generally referred to as b-frame (body frame).

2. Functional Configuration FIG. 2 is a block diagram showing a functional configuration of the navigation device 1. The navigation device 1 is a computer system that includes a processing unit 10, an operation unit 20, a display unit 30, a sensor unit 40, and a storage unit 60, and each unit is connected by a bus 70.

  The processing unit 10 is a functional unit that comprehensively controls each unit of the navigation device 1 according to various programs such as a system program stored in the storage unit 60, and is realized by a processor such as a CPU (Central Processing Unit). . In the present embodiment, the processing unit 10 calculates a user's movement locus according to the movement locus calculation program 601 stored in the storage unit 60 and causes the display unit 30 to display the movement locus.

  The operation unit 20 is an input device configured by, for example, a touch panel or a button switch, and outputs a pressed key or button signal to the processing unit 10. By operating the operation unit 20, various instruction inputs such as a movement locus calculation start instruction operation, a reset instruction operation, and a power-off instruction operation are performed.

  The display unit 30 is configured by an LCD (Liquid Crystal Display) or the like, and is a display device that performs various displays based on display signals input from the processing unit 10. The display unit 30 displays a movement locus, time information, and the like.

  The sensor unit 40 is a functional unit having a sensor for detecting a user's walking state, and includes an acceleration sensor 41, a gyro sensor 43, and a geomagnetic sensor 42, for example. Values detected by the acceleration sensor 41, the gyro sensor 43, and the geomagnetic sensor 42 are output to the processing unit 10 as needed (hereinafter referred to as “output”).

  The storage unit 60 is a storage device including a memory such as a ROM (Read Only Memory), a flash ROM, or a RAM (Random Access Memory), for example, and a system program for the processing unit 10 to control the navigation device 1 or a movement Various programs and data for realizing the locus calculation function are stored. Further, a work area for temporarily storing a system program executed by the processing unit 10, various processing programs, data being processed in various processing, processing results, and the like is formed.

3. The data configuration storage unit 60 stores a movement locus calculation program 601 that is read by the processing unit 10 and executed as a movement locus calculation process as a program. Further, the movement locus calculation program 601 includes an offset correction program 610 executed as an offset correction process (see FIG. 5) as a subroutine. Furthermore, a walking speed range determination table 602, a threshold frequency determination table 603, and correlation model expression data 604 are stored as predetermined data. Further, sensor data 605, estimated stride data 606, walking direction data 607, movement trajectory data 609, and offset data 611 are stored as data updated as needed.

  In the movement trajectory calculation process, the processing unit 10 performs walking detection and stride estimation based on the detection result of the acceleration sensor 41, calculates an azimuth by the azimuth calculation process, and uses the estimated stride and azimuth. This is a process of calculating a movement trajectory. The azimuth calculation process is a process in which the processing unit 10 uses the angular velocity of the gyro sensor 43 to predict a change in the detection result of the geomagnetic sensor 42 and calculates the azimuth based on the predicted detection result. In the movement trajectory calculation process, the azimuth calculation process is performed using the offset corrected by the offset correction process. The offset correction process, the movement trajectory calculation process, and the azimuth calculation process will be described later in detail using a flowchart.

  FIG. 3 is a conceptual diagram showing the processing unit 10 as a functional block. The processing unit 10 includes a walking speed determination unit 11, a filter processing unit 12, a walking detection unit 13, a stride estimation unit 14, a walking direction estimation unit (orientation calculation device) 15, and a movement trajectory calculation unit 16. Configured. Each functional unit can be configured as an independent circuit, or can be configured as software processing by performing digital arithmetic processing using a processor.

  The walking speed determination unit 11 is a functional unit that determines the walking speed range of the user based on the amplitude of the Z-axis acceleration detected by the acceleration sensor 41 and the time required for one step detected by the walking detection unit 13. is there.

  The filter processing unit 12 is a functional unit that extracts low-frequency components by performing filter processing on the acceleration vector, the geomagnetic vector, and the angular velocity detected by the acceleration sensor 41, the geomagnetic sensor 42, and the gyro sensor 43. The filter processing unit 12 may perform the filtering process by variably setting the threshold frequency according to the walking speed range output from the walking speed determination unit 11.

  The walking detection unit 13 is a functional unit that detects the user's walking by detecting the peak of the Z-axis acceleration output from the filter processing unit 12. The stride estimation unit 14 estimates the stride according to the correlation model expression stored in the storage unit 60 using the Z-axis acceleration output from the filter processing unit 12 when one step is detected by the walk detection unit 13. It is a functional part.

  The walking direction estimation unit 15 is a functional unit that estimates the user's walking direction (azimuth angle) using the geomagnetic vector and the angular velocity output from the filter processing unit 12. The movement trajectory calculation unit 16 is a functional unit that calculates a user's movement trajectory using the stride output from the stride estimation unit 14 and the walking direction output from the walking direction estimation unit 15.

  The walking direction estimation unit 15 according to the present embodiment includes a geomagnetic prediction unit 17, a geomagnetic offset determination unit 18, a geomagnetic offset correction unit 19, and an orientation calculation unit 23.

  The geomagnetism prediction unit 17 detects the detection result (first detection result) of the geomagnetic sensor 42 at the first time and the detection of the gyro sensor 43 from the first time to the second time after the first time. The detection result (second detection result) of the geomagnetic sensor 42 at the second time is predicted using the result (angular velocity).

  The geomagnetic offset determination unit 18 compares the second detection result (predicted detection result) predicted by the geomagnetism prediction unit 17 with the detection result (second detection result) of the geomagnetic sensor 42 at the second time. Then, it is determined whether or not the comparison result is within a predetermined range.

  The geomagnetic offset correction unit 19 corrects the offset included in the output of the geomagnetic sensor 42 using the predicted detection result when the geomagnetic offset determination unit 18 determines that the comparison result is not within the predetermined range.

  The azimuth calculation unit 23 calculates the azimuth (azimuth angle) of the user at the second time based on the correction result of the geomagnetic offset correction unit.

4). Process Flow FIG. 4 is a flowchart showing the flow of the movement trajectory calculation process. FIG. 5 is a flowchart showing the flow of the azimuth calculation process. The movement trajectory calculation process and the offset correction process according to the present embodiment are performed by reading and executing the movement trajectory calculation program 601 and the offset correction program 610 stored in the storage unit 60, respectively. It is executed in the device 1. The movement trajectory calculation process is a process of starting execution when the processing unit 10 detects that a power-on operation has been performed by the user via the operation unit 20. In the following azimuth calculation processing, it is assumed that the output value of the sensor unit 40 is updated and stored as needed in the sensor data 605 of the storage unit 60.

  In the movement trajectory calculation process (FIG. 4), first, the filter processing unit 12 starts the filter process for the output values of the axes of the acceleration sensor 41, the geomagnetic sensor 42, and the gyro sensor 43 at the set threshold frequency. (Step A1).

  Next, in step A2, the sensor unit 40 is calibrated. Specifically, the processing unit 10 estimates the bias included in the acceleration sensor 41 and the gyro sensor 43 and the offset included in the output of the geomagnetic sensor 42 and stores the offset in the offset data 611 of the storage unit 60. These biases and offsets can be estimated by a known method. For example, the offset of the geomagnetic sensor 42 includes the output of the geomagnetic sensor 42 obtained by changing the attitude of the geomagnetic sensor 42 (the attitude of the navigation device 1 on which the geomagnetic sensor 42 is mounted) in a three-dimensional space, and a known magnetic force ( (Determined by the position on the earth). In the present specification, the output of the sensor unit 40 subjected to the correction is referred to as “detection result”.

  When the offset of the geomagnetic sensor 42 and the bias of the gyro sensor 43 are stored in the offset data 611, an offset correction process (step A3), an azimuth calculation process (step A4), and a walking detection process (step A5) are started.

First, the offset correction process (step A3 in FIG. 4) will be described with reference to FIG. The offset correction process is a process performed by the walking direction estimation unit 15. The geomagnetism prediction unit 17 performs geomagnetic vector prediction (step S20). Specifically, the detection result of the geomagnetic sensor 42 at the time t _k (first time) and the detection result of the gyro sensor 43 at the time t _{k + 1} according to the expressions (1-1) and (1-2). (A change in angle from time t _k to time t _{k + 1} ) is used to predict a detection result of the geomagnetic sensor 42 at time t _{k + 1} (the predicted detection result is also referred to as a prediction result).

In the mathematical expressions in this specification, the component of the geomagnetic vector is represented by a small letter “m”, and each component of the angular velocity is represented by a small letter “ω”. In addition, the corresponding component of each axis is indicated by a lowercase subscript. For example, _mx is a component of the geomagnetic vector corresponding to the X-axis direction. Further, the unit matrix is represented by an uppercase “I” and the transposed matrix is represented by a superscript “T”. Δt is a time interval (time interval from time t _k to time t _{k + 1} ) at which the gyro sensor 43 outputs an angular velocity, and exp () is a natural logarithm base e (= 2.71828...) ) Indicates the power of the value in parentheses.

  Expression (1-1) is an approximate expression assuming that the geomagnetic force is constant regardless of the position on the earth. Actually, the geomagnetic force differs depending on the position on the earth, but since the change in the geomagnetic force is minute within the moving range of the moving body (pedestrian or car), it can be assumed to be constant.

Next, the geomagnetism prediction unit 17 determines whether or not a predetermined time has been reached (step S30). If it is determined that the predetermined time (second time) has not been reached, the process returns to step S20 (step S30: No) and prediction is performed again. When the prediction is made after returning from step S30, the prediction is performed using the prediction result of the previous time (time t _{k + 1} ) and the detection result of the gyro sensor 43 of the current time (time t _{k + 2} ).

Although the subsequent processing may be performed using the result of the one-time prediction, that is, the time t _{k + 1} as the second time, in general, the angular velocity obtained by one detection is very small. Then, since the geomagnetic vector predicted from one detection result (angular velocity) becomes almost constant each time (changes are close to zero), the actual angular change cannot be reflected and there is a possibility that correct correction cannot be performed. is there. By repeating the prediction and accumulating the detection result of the gyro sensor 43 for a predetermined time, the detection result of geomagnetism can be predicted more accurately. The predetermined time only needs to be such that a change in the azimuth angle of the moving body can be detected. For example, in the case of a pedestrian, the predetermined time may be 0.5 seconds or 1 second. Moreover, it is good also as the timing which the walk detection part 13 detected 1 step instead of time, and you may determine the frequency | count (for example, 50 times) which performed prediction.

In the above prediction, a relative angle change from the first time to the second time is added to the geomagnetic vector (information indicating the absolute azimuth angle) at the first time (time t _k ). It can be said. Since the angle change from the first time to the second time is used, the detection result of the geomagnetic sensor 42 can be obtained regardless of whether the vehicle travels straight (no azimuth change) or non-straight (change azimuth). Can be predicted appropriately. In addition, prediction can be performed even when the moving body is moving or stationary. Therefore, there is no need to remove the navigation device 1 attached to the user and perform calibration again.

  If it is determined that the predetermined time has been reached (step S30: Yes), in step S40, the geomagnetic offset determination unit 18 obtains the prediction result at the second time and the detection result of the geomagnetic sensor 42 at the second time. Compare. Specifically, the difference between the prediction result and the detection result is calculated for each component of the geomagnetic vector at the second time. The difference calculated here may be a difference between the components or an absolute value of the difference. Moreover, the sum of squares of the difference of each component may be sufficient.

  Next, in step S50, the geomagnetic offset determination unit 18 determines whether or not the comparison result in step S40 is within a predetermined range. Specifically, for example, the absolute value of the difference between the components of the geomagnetic vector obtained in step S40 is compared with the threshold value, and the magnitude of the threshold value is determined. That is, the X-axis component, the Y-axis component, and the Z-axis component are each compared with a threshold value, and if all are equal to or less than the threshold value, it is determined that they are within a predetermined range. The predetermined range can be set according to the accuracy of the azimuth angle that is finally required, and is, for example, 0 μT (Tesla) or more and 5 μT or less in absolute value. When it is set as the threshold, it may be 5 μT. When the determination is performed using the square sum of the differences, the values in the predetermined range may be compared with the square sum (for example, if the predetermined range is an absolute value of 0 μT to 5 μT, the sum of squares). Is 75 μT).

  The output of the gyro sensor 43 is not affected by the magnetization. That is, the detection result of the gyro sensor 43 is not affected by the magnetization. Therefore, by using the detection result of the gyro sensor 43, the detection result of the geomagnetic sensor 42 that should be obtained when the offset correction is appropriate can be accurately predicted. When there is no change in the offset of the geomagnetic sensor 42 (which is negligible), the predicted detection result and the detection result obtained from the actual output of the geomagnetic sensor 42 substantially coincide. If the offset of the geomagnetic sensor 42 changes due to a change in magnetization or the like, the difference between the two increases. That is, if the difference is larger than the threshold (if not within the predetermined range), it can be determined that the offset of the geomagnetic sensor 42 has changed.

  If it is determined that it is not within the predetermined range (step S50: No), the process proceeds to step S60. When it determines with it being within the predetermined range (step S50: Yes), it progresses to step S70.

  In step S60, the geomagnetic offset correction unit 19 corrects (updates) the offset of the geomagnetic sensor 42 using the comparison result in step S50. Specifically, the comparison result (difference) is added to the offset of the geomagnetic sensor 42 stored in the offset data 611, and the process proceeds to step S80.

  When it determines with it being within the predetermined range in step S50 (step S50: Yes), it progresses to step S70. In step S70, the bias of the gyro sensor 43 is estimated. Specifically, the bias of the gyro sensor 43 is estimated using the detection result of the geomagnetic sensor 42 at the second time, and the bias of the gyro sensor 43 stored in the offset data 611 is updated with the estimated bias.

  A known method can be used to estimate the bias. For example, the bias can be estimated by performing a Kalman filter calculation using the azimuth angle at the second time obtained according to Equation (2) described later as an observed value. The detection result (geomagnetic vector) of the geomagnetic sensor 42 at the second time may be an observed value. When there is no change in the offset of the geomagnetic sensor 42, that is, when the detection result of the geomagnetic sensor 42 is appropriate, a more accurate bias can be estimated by estimating the bias of the gyro sensor 43 using the detection result. In addition, when the bias changes, it is possible to prevent an appropriate detection result from being obtained because correction is not performed. When the bias is updated, the process proceeds to step S80.

  Next, in step S80, the walking direction estimation unit 15 determines whether to end the offset correction process. If it is determined that the offset correction process is to be terminated (step S80: Yes), the offset correction process is terminated. If it is determined not to end the offset correction process (step S80: No), the process returns to step S20.

  If it returns to step S20, the next prediction will be started. If it is determined in step S50 that it is within the predetermined range, the next prediction is started using the detection result of the geomagnetic sensor 42. If it is determined in step S50 that it is not within the predetermined range, the next prediction is started using the prediction result at the second time. If the prediction is started from the second time, the prediction is performed until the third time (step S30), and the prediction result at the third time is compared with the detection result (step S40).

  In this embodiment, when it is determined in step S50 that it is not within the predetermined range, it can be said that there is a change in the offset. When there is a change in the offset, the detection result at the second time includes the offset. Therefore, more accurate prediction can be performed by performing the next prediction using the detection result predicted based on the detection result of the gyro sensor 43.

  On the other hand, if it is determined in step S50 that it is within the predetermined range, it can be said that the offset is not changed or the change is acceptable. Therefore, the detection result of the geomagnetic sensor 42 is an appropriate value. Usually, a slight bias remains in the detection result of the gyro sensor 43. Since this bias is accumulated for a predetermined period in step S30, if the prediction is further repeated based on the previous prediction result, the bias is gradually accumulated and there is a possibility that the magnitude cannot be ignored. By using the detection result of the geomagnetic sensor 42, that is, by not using the prediction result for prediction for the next predetermined period, accumulation of the bias of the gyro sensor 43 in the prediction result can be suppressed.

  Returning to the movement trajectory calculation process of FIG. 4, the azimuth calculation process (step A4) and the walking detection process (step A5) will be described. In step A4, the bearing calculation unit 23 calculates the bearing angle. Specifically, the azimuth calculation unit 23 calculates an azimuth angle according to the following equation (2) based on the determination result of step S 50, and stores it in the walking direction data 607 of the storage unit 60. In the present embodiment, the azimuth at the time when walking is detected in the walking detection process (step A5) is calculated. The bearing calculation unit 23 corrects the output of the geomagnetic sensor 42 stored in the sensor data 605 with the offset stored in the offset data 611 and uses the obtained detection result (new second detection result). To calculate the azimuth. That is, if it is determined in step S50 that it is not within the predetermined range (step S50: No), the azimuth angle is calculated using the offset updated in step S60.

Here, “Ψ” is an azimuth angle, “θ” is a roll angle obtained from the detection result of the gyro sensor 43, and “φ” is a pitch angle obtained from the detection result of the gyro sensor 43.

  In addition, the azimuth angle calculated | required from Formula (2) is a rotation angle of the sensor coordinate system with respect to the N direction in the NED (North East Down) coordinate system which made magnetic north N (North). By correcting the magnetic north declination from true north (the direction of the north pole), it is possible to calculate the azimuth angle based on true north.

  The roll angle and pitch angle in Equation (2) are rotation angles from the NED coordinate system. When obtaining the azimuth angle at the second time, a value obtained by integrating the angular velocities from the first time to the second time is used.

  Returning to FIG. 4, in step A5, the gait detector 13 determines whether or not a peak is detected from the Z-axis acceleration after the filtering process and the bias correction are performed. And the number of steps is updated. Then, the walking detector 13 calculates the time from the previous peak to the current peak as the time required for one step.

  Next, the walking speed determination unit 11 determines the walking speed range of the user with reference to the walking speed range determination table 602 stored in the storage unit 60 (step A6). The walking speed range determination table 602 includes the walking speed range, which is a determination condition, of the amplitude of the Z-axis acceleration or the time required for one step and three steps of walking speed ranges (“low speed”, “medium speed”, “high speed”). It is an associated table. The walking speed determination unit 11 determines a walking speed range corresponding to the amplitude of the Z-axis acceleration and the time required for one step calculated in step A5.

  Next, the stride length estimation unit 14 refers to the correlation model equation data 604 stored in the storage unit 60, and estimates the stride of the user (Step A7). The correlation model formula data 604 stores a walking speed range and stride estimation model formula data in association with each other. Since the relationship between the walking speed range and the stride estimation model formula and the stride estimation model formula are known, the description thereof will be omitted. The stride estimation unit 14 selects stride estimation model formula data associated with the walking speed range determined in step A6. Then, the user's stride is estimated according to the selected stride estimation model expression data, and the estimated stride data 606 in the storage unit 60 is updated.

  Next, the movement trajectory calculation unit 16 uses the latest position stored in the movement trajectory data 609, the stride stored in the estimated stride data 606, and the walking direction stored in the walking direction data 607. Then, the current position and speed of the user are calculated (step A8). The movement locus data stores the output time of the calculation result, the calculated position, speed, and walking direction. The position (absolute position) at the start of movement can be obtained by a known method. For example, the absolute position may be acquired by communication with the outside of the navigation device 1. Then, the movement trajectory calculation unit 16 updates the movement trajectory data 609 in the storage unit 60 with the calculated position, speed, and walking direction (step A9).

  Next, in step A10, the movement trajectory calculation unit 16 determines whether or not to end the movement trajectory calculation process. When it is determined that the movement trajectory calculation process is to be ended (step A10: Yes), the movement trajectory calculation process is ended. If it is determined not to end the movement trajectory calculation process (step A10: No), the process returns to step A4 (and step A5).

5. Example An experiment in which a subject wearing the navigation device 1 of the present embodiment walks the following walking test and calculates a movement trajectory was performed. The walking test route walks 60m straight from the starting point to the north (approximately 11.3 degrees from true north), turns 180 degrees, and travels 60m straight to the south (approximately -168.7 degrees from true north). It is a route that walks and arrives at the end point.

  First, as a comparative example, the movement locus was calculated from the integration result of the azimuth angle and the stride calculated using the detection result of the geomagnetic sensor 42.

  FIG. 6A is a diagram showing a movement trajectory of the comparative example, and FIG. 6B is a diagram showing an azimuth angle calculated from the detection result of the geomagnetic sensor in the comparative example. The symbol “N” shown in FIG. 6A represents the direction of true north. The same applies to the following drawings. Before the start of walking, the sensor unit 40 was calibrated, and the offset of the geomagnetic sensor 42 was calculated in advance. However, as shown in FIG. 6A, the azimuth angle starts to shift from the middle of the movement locus. Further, as shown in FIG. 6B, the variation in the azimuth angle in the second half (after the direction change) is large and unstable. From these results, it is considered that the geomagnetic sensor 42 is affected by some influence during walking, and the offset changes, and the detection result deviates from the true value.

  On the other hand, as a first embodiment, the detection result of the geomagnetic sensor 42 is predicted using the angular velocity measured from the gyro sensor 43, and the movement locus is calculated using the predicted detection result itself.

  FIG. 7 is a diagram showing a movement locus of the first embodiment. As shown in FIG. 7, the movement trajectory almost coincided with the actual walking test route, and the deviation of the azimuth angle was almost unknown. It can be estimated that the azimuth angle calculated from the predicted three-axis components of the geomagnetic sensor 42 substantially coincides with the true value.

  FIG. 8 is a diagram showing the calculation result of the azimuth angle in the first embodiment. FIG. 8A is a diagram showing the azimuth angle calculated from the detection result of the geomagnetic sensor. FIG. 8B shows the azimuth angle calculated from the predicted detection result of the geomagnetic sensor 42. Comparing FIG. 8 (A) and FIG. 8 (B), in the first half of the walk (before the direction change), FIG. 8 (A) tends to rise slightly to the right, and the second half (after the direction change) FIG. 8B is more stable and has less variation. From this, it can be seen that by using the predicted detection result, it is possible to avoid the deterioration of the azimuth calculation accuracy due to the change in the offset and to calculate the azimuth. However, if the prediction is repeated, the error in the angular velocity of the gyro sensor 43 accumulates, and the prediction accuracy of the detection result of the geomagnetic sensor 42 may decrease.

  FIG. 9 is a diagram showing each component of the geomagnetic vector in the first embodiment. In FIG. 9, M is a component of the detection result of the geomagnetic sensor 42 (actual value M), and P is a component of the predicted geomagnetic vector (predicted value P). 9A shows a comparison between the actual measurement value M of the component in the X-axis direction and the predicted value P, and FIG. 9B shows a comparison between the actual measurement value M of the component in the Y-axis direction and the predicted value P. FIG. 9C is a comparison between the actual measurement value M and the predicted value P of the component in the Z-axis direction. Although the degree is different, the difference between the actual measurement value M and the predicted value P increases in the latter half of any axis. Since the positive direction of the X-axis is the forward direction of the subject, the change in the component in the X-axis direction represents the change in the traveling direction. Therefore, in this embodiment where the direction is changed by 180 degrees, the graph is 0 μT if the detection result is correct. It is almost rotationally symmetric with respect to this point. As shown in FIG. 9 (A), the predicted value P is about −40 μT in the first half and about +40 μT in the second half, which is a result reflecting the movement of the subject. On the other hand, the measured value M was about -40 μT in the first half and about +15 μT in the second half, and the first half and the second half were asymmetric. From this result, it was found that the predicted value P was more correct than the actually measured value M. It is considered that the measured value M of the three-axis component of the geomagnetic sensor 42 is affected by some influence, and the offset of the output of the geomagnetic sensor 42 has shifted.

  Furthermore, as a second example, the movement trajectory was calculated by the method described in the embodiment. That is, the predicted detection result of the geomagnetic sensor 42 is compared with the actual detection result of the geomagnetic sensor, and if it is determined that the comparison result is not within the predetermined range, the offset of the geomagnetic sensor 42 is corrected and the movement trajectory is determined. Calculated.

  FIG. 10 is a diagram illustrating a movement locus of the second embodiment. The movement locus 80 is the movement locus of the second embodiment. The movement locus 82 is the movement locus of the first embodiment. From FIG. 10, it can be seen that the movement trajectories of the two examples are almost the same. Therefore, it can be seen that the correct azimuth angle can be obtained by correcting the offset of the geomagnetic sensor 42 by the method of the present embodiment.

  FIG. 11 is a diagram showing each component of the geomagnetic vector in the second embodiment. In FIG. 11, C is a value (correction value C) obtained by correcting the actual measurement value M by the method of this embodiment. Reference numeral P is the same as in FIG. 11A shows a comparison between the correction value C in the X-axis direction and the predicted value P, and FIG. 11B shows a comparison between the correction value C in the Y-axis direction and the predicted value P. FIG. (C) is a comparison between the correction value C in the Z-axis direction and the predicted value P. As shown in FIGS. 11A, 11 </ b> B, and 11 </ b> C, the correction value C approaches the predicted value P as compared with FIG. 9 by correcting the offset using the prediction result. I understand.

6). The embodiment to which the present invention is applicable is not limited to the above-described embodiment, and can be appropriately changed without departing from the spirit of the present invention. Hereinafter, modified examples will be described. In addition, about the structure same as the said embodiment, the same code | symbol is attached | subjected and repeated description is abbreviate | omitted.

(Modification 1)
In the above embodiment, when the offset is updated, the next prediction is started using the prediction result at the second time, but the output of the geomagnetic sensor 42 at the second time stored in the sensor data 605 is updated. The azimuth angle may be calculated using the detection result obtained by correcting with the offset. Thereby, the result similar to the case where the next prediction is performed using the prediction result at the second time can be obtained.

(Modification 2)
In the above embodiment, the geomagnetic sensor 42 that detects orthogonal three-axis geomagnetic vectors is used. However, when the azimuth is considered on a plane (two dimensions) without considering the tilt, the geomagnetic sensor 42 It is only necessary to detect the geomagnetic component in the Y-axis direction. When the geomagnetic sensor detects components in the X-axis direction and the Y-axis direction, a relational expression between the geomagnetic component and the posture angle is expressed by Expression (3). The detection result of the geomagnetic sensor 42 can be predicted using the equation (4) obtained by substituting the equation (3) into the equation (1-1).

Here, δΨ = ω _Z * Δt.

(Modification 3)
In the above embodiment, the detection result of the geomagnetic sensor 42 is predicted every time the detection result from the gyro sensor 43 is obtained. However, the prediction may be performed only at the timing of correcting the offset of the geomagnetic sensor 42. That is, the frequency of performing the prediction may be less than the output frequency of the gyro sensor 43. Below, the formula used for prediction of this modification is shown. The timing (interval) for performing the prediction is the same as the predetermined time in the embodiment. From the first time (here, time t = 1) to the second time (here, time t _{k + 1} ), the angle (posture) change is calculated by equation (5), and the equation at the second time is calculated. Prediction is performed using (6).

Here, [Δθ] _k is the same as in equation (1-2).

  When a mobile object such as an automobile, which has a small change in angular velocity, is used as a target, instead of calculating Equation (5) for each detection result of the gyro sensor 43, 10 detection results are calculated for each axis component. You may use what was averaged as angular velocity (omega) of each axis | shaft. The amount of calculation is reduced by averaging.

(Modification 4)
In the above embodiment, the azimuth angle is calculated using the roll angle and the pitch angle obtained from the detection result of the gyro sensor 43 (Equation (2)), but the azimuth angle is calculated using the detection result of the acceleration sensor 41. Also good. For example, the rotation angle (roll and pitch) from the NED (North East Down) coordinate system to the sensor coordinate system is calculated from the detection result of each axis of the acceleration sensor 41 and the gravitational acceleration vector when the moving body is stationary. can do.

(Modification 5)
In the said embodiment, although the person (namely, pedestrian) is illustrated as a moving body, you may mount an azimuth | direction calculation apparatus in a motor vehicle. It may be mounted on a bicycle. Moreover, you may mount | wear with an autonomous walking robot or an animal. If the azimuth calculating device is carried by a person, the vehicle is regarded as a moving body in any case of riding in a car, riding a bicycle, or acting on foot.

(Modification 6)
In the above-described embodiment, walking detection is performed, and a walking trajectory is calculated by estimating walking speed and stride. However, processing other than the azimuth calculation processing in the moving trajectory calculation processing can be performed by a known inertial navigation calculation. For example, the movement trajectory may be calculated by calculating the velocity vector and the position by prediction calculation, integration, or the like using the detection result of the acceleration sensor 41 and the direction obtained by the direction calculation process.

(Modification 7)
In the above embodiment, the absolute position is acquired by communication with the outside, but GPS (Global Positioning System) may be used. In this case, the navigation device 1 further includes a GPS position calculation unit (not shown). The GPS position calculation unit uses a GPS satellite signal transmitted from a GPS satellite, which is a kind of positioning satellite, at a predetermined time interval (for example, once per second), the user's position, moving speed, and moving direction. This is a position calculation circuit for calculating. The GPS position calculation unit calculates the absolute position of the user by performing a known position calculation calculation using a pseudo distance between the navigation device 1 and the GPS satellite.

  In addition, using the GPS position calculation unit, the user's moving speed and moving direction are calculated by performing a known moving speed / moving direction calculation calculation using the time change of the Doppler frequency accompanying the movement of the navigation device 1 and the GPS satellite. May be. Using the calculated moving speed, moving direction, and absolute position, the walking speed and position may be corrected in the movement trajectory calculation process, and a bias in inertial navigation calculation may be estimated.

(Modification 8)
In the above embodiment, the navigation device 1 has the display unit 30 that displays the movement trajectory, but the display unit 30 may not be provided. When the display unit 30 is not provided, information stored in the movement locus data 609 may be transmitted by short-range wireless communication, and the movement locus may be displayed on a mobile device such as a smartphone or a wristwatch type display device. The user may be notified of locus and position information by voice.

(Modification 9)
In the above embodiment, the navigation device 1 has been described as including the azimuth calculation device (walking direction estimation unit 15) and the movement trajectory calculation unit 16, but the azimuth calculation device may be an independent device. For example, the azimuth calculation device can be configured as a device including the geomagnetic sensor 42, the gyro sensor 43, and the processing unit 10 that executes the azimuth calculation program. It is also possible to record the output using the geomagnetic sensor 42 and the gyro sensor 43 and execute the azimuth calculation process later using the output using a personal computer or the like.

  DESCRIPTION OF SYMBOLS 1 ... Navigation apparatus 10 ... Processing part 11 ... Walking speed determination part 12 ... Filter processing part 13 ... Walking detection part 14 ... Stride estimation part 15 ... Walking direction estimation part (orientation calculation apparatus) 16 ... Movement locus calculation part 17 ... Geomagnetic prediction Unit 18: Geomagnetic offset determination unit 19 ... Geomagnetic offset correction unit 20 ... Operation unit 23 ... Direction calculation unit 30 ... Display unit 40 ... Sensor unit 41 ... Acceleration sensor 42 ... Geomagnetic sensor 43 ... Gyro sensor 60 ... Storage unit 70 ... Bus 80 , 82 ... Movement locus 601 ... Movement locus calculation program 602 ... Walking speed range determination table 603 ... Threshold frequency determination table 604 ... Correlation model formula data 605 ... Sensor data 606 ... Estimated step length data 607 ... Walking direction data 609 ... Movement locus Data 610 ... Offset correction Program 611: Offset data.

Claims (6)

The first detection result at the first time of the geomagnetic sensor installed on the moving body and the second of the gyro sensor installed on the moving body after the first time from the first time Predicting the detection result of the geomagnetic sensor at the second time using the detection result up to the time;
Comparing the predicted detection result with a second detection result of the geomagnetic sensor at the second time to determine whether the result of the comparison is within a predetermined range;
When it is determined that the result of the comparison is not within a predetermined range, the offset of the output of the geomagnetic sensor is corrected using the predicted detection result;
including,
Offset correction method. The offset correction method according to claim 1,
Correcting the gyro sensor bias using the second detection result when it is determined that the result of the comparison is within the predetermined range;
Offset correction method. The offset correction method according to claim 1, further comprising:
When it is determined that the comparison result is not within the predetermined range, the predicted detection result and detection of the gyro sensor from the second time to a third time after the second time Using the result to predict the detection result of the geomagnetic sensor at the third time,
Offset correction method. The offset correction method according to claim 1, further comprising:
When it is determined that the result of the comparison is not within the predetermined range, the output of the geomagnetic sensor at the second time is corrected using the corrected offset to obtain a new second detection result;
The geomagnetism at the third time using the new second detection result and the detection result of the gyro sensor from the second time to a third time after the second time. Including predicting sensor detection results,
Offset correction method. In the offset correction method according to any one of claims 1 to 4,
When it is determined that the result of the comparison is within the predetermined range, the second detection result and the gyro sensor from the second time to a third time after the second time. Using the detection result to predict the detection result of the geomagnetic sensor at the third time,
Offset correction method. The first detection result at the first time of the geomagnetic sensor installed on the moving body and the second of the gyro sensor installed on the moving body after the first time from the first time A geomagnetism prediction unit that predicts the detection result of the geomagnetic sensor at the second time using the detection results up to the time;
The detection result predicted by the geomagnetism prediction unit is compared with the second detection result of the geomagnetic sensor at the second time to determine whether the comparison result is within a predetermined range. A geomagnetic offset determination unit;
An offset correction unit that corrects an offset of the output of the geomagnetic sensor using the predicted detection result when the geomagnetic offset determination unit determines that the comparison result is not within a predetermined range;
including,
Offset correction device.