DE102011102400A1 - Method for smoothing transmission input shaft speed signal of dual clutch gear box in drive train of motor car, involves determining prediction weight based on slippage rotation speed after fading of oscillation of speed signal - Google Patents

Abstract

The method involves supplying a transmission input shaft speed signal to a filter, where the speed signal comprises a prediction weight as a function of slippage rotation speed of a clutch. The prediction weight is adjusted to a constant value independent of slippage rotation speed for recognition of oscillation of the speed signal, where the constant value is less than 70 percent. The prediction weight is determined again based on the slippage rotation speed after fading of the oscillation of the speed signal, where the prediction weight represents weight factor of the filtered speed signal.

Description

The invention relates to a method for smoothing a transmission input shaft speed signal of a transmission, in particular a dual-clutch transmission, wherein the transmission input speed signal is fed to a filter having a prediction weighting function of a slip speed of a clutch and a method for detecting a vibration of an electrical signal.

In modern motor vehicles increasingly automated clutches are used. The use of such clutches not only has the advantage of improved ride comfort, but experience has shown that it is used more often in gears with a long gear ratio, which reduces fuel consumption and environmental impact.

From the DE 10 2008 030 481 A1 For example, there is known a method of controlling slip of a vehicle clutch in which the difference between a speed signal of a drive-side input shaft of the clutch and a filtered speed signal of a transmission-side output shaft of the clutch is maintained at a target value. The following steps are taken into account: calculating a moving average from a raw signal corresponding to an instantaneous speed of the transmission-side output shaft of the clutch, wherein the moving average is an averaging width of n interrupts at a pickup frequency for a respective gear. Furthermore, a derivative of the moving average is calculated from the moving average calculated in the current interrupt and from a moving average calculated in the preceding interrupt. From the calculated derivation of the moving average value and from a smoothed derivative of the difference signal calculated in the previous interrupt, a smooth derivative of the speed signal is calculated with a PT1 filter. From a filtered speed signal calculated from the previous interrupt and from the smoothed derivative, a forecast speed signal for the current interrupt is calculated, wherein a weighted average is calculated from the raw signal and the speed signal filtered in the previous interrupt, and finally from the weighted average and the forecast speed signal filtered speed signal is determined. The filtered speed signal is used for the slip control of the clutch.

The forecast speed represents a prediction weight of the filter and is a function of the slip speed, the prediction weighting for small slip speeds becoming zero or at least very small and reaching a saturation value above a certain slip limit. This dependency makes sense, since at low slip speeds no strong vibrations of the transmission input speed occur and a low prediction weighting of the signal delay caused by the filter is kept small. If, however, an oscillation of the transmission input shaft, in which the slip develops a sufficient amplitude, the prediction weight is also excited to a vibration. Due to this vibration of the prediction weighting, the filter can not eliminate the vibration of the transmission input speed.

The invention is therefore based on the object of specifying a method in which the fluctuation of the filter parameters is reliably prevented.

According to the invention, the object is achieved in that, upon detection of a vibration of the transmission input speed signal, the prediction weight is set independently of the slip speed, wherein preferably after a decay of the vibration, the prediction weight is again determined as a function of the slip speed. This has the advantage that it is reliably prevented that, in the case of juddering vibrations of the transmission input speed signal with low slip, a filter parameter also begins to oscillate. This improves the effectiveness of the filter.

Advantageously, the prediction weight is set to a constant value. This reliably prevents the filtered transmission input speed signal from more or less following the oscillation of the unfiltered transmission input speed signal.

In one embodiment, the prediction weighting is a factor in weighting the filtered transmission input speed signal.

In one variant, the prediction weight is set to a constant value> 50%, preferably> 70%. Setting to such a constant high value ensures that oscillation of a filter parameter is prevented.

A development of the invention relates to a method for detecting a vibration of an electrical signal, wherein a sign change of the deviation of the electrical signal is detected by a reference signal, the extreme values of the positive and negative deviation of the electrical signal from the reference signal are determined, wherein after detection of a positive and / or negative extreme value, an associated partial amplitude is set to the detected extreme value and from the sum of the partial amplitude held at the positive extreme value and the amount of the partial amplitude held at the negative partial value, a vibration width of the electrical signal is determined. This has the advantage that vibrations of any electrical signal can be detected. If there is no natural reference signal, for example a filtered or unfiltered signal, given and this can not be constructed, it may be possible to consider its time derivative instead of the signal of interest.

Advantageously, a vibration of the electrical signal is detected when the vibration width of the electrical signal exceeds a predetermined threshold value. In particular, the determined extreme values (amplitudes) are taken into account in order to reliably determine the vibration range.

In one embodiment, a time between two consecutive sign changes of the electrical signal is determined, which is considered as half the period of the oscillation of the electrical signal, and from which the frequency of the electrical signal is determined. In this case, the frequency of the oscillation can be reliably calculated from the second observed change of sign of the frequency of the oscillation. The frequency can still be slightly filtered for smoothing, for example with a PT1 filter.

In a further development, the reference signal is surrounded by a demarcation area, wherein only deviations of the electrical signal from the reference signal are detected, which lie outside the demarcation area. This demarcation area is used for noise suppression and is placed around the reference signal, so that small deviations of the electrical signal do not lead immediately to a vibration detection. Only vibrations that extend beyond the bounds are considered significant. This reduces the number of signals to be processed to a manageable level.

In one variant, the measurement of the deviation and / or the time between two successive changes of sign begins when the electrical signal leaves the cutoff region around the reference signal. In this case, only waveforms of the signal deviation are detected as oscillations, in which case the signal deviation passes completely through the demarcation range twice within a predetermined period of time.

In a further development, the time between two successive changes of sign of the electrical signal is determined by means of a counter, wherein, in particular, upon reaching a maximum value of the counter, without a sign change of the electrical signal has occurred, the measured time is discarded. Thus, it can be determined with certainty when the oscillation of the electrical signal has subsided.

The invention allows numerous embodiments. One of them will be explained in more detail with reference to the figures shown in the drawing.

It shows:

1 : Schematic representations of a drive train

2 : Representation of a vibration of a transmission input shaft speed as a function of the slip speed according to the prior art

3 : Representation of the vibration of the transmission input shaft speed using the method according to the invention

4 : Principle of the determination of the vibration amplitude

5 : Principle of determining the period of oscillation Identical features are identified by the same reference numerals.

1 shows the structure of a known vehicle drive train with a dual-clutch transmission. The drive motor 1 leads to the dual-clutch transmission 2 , The dual-clutch transmission 2 has two sub-transmission line, wherein the first sub-transmission train of a first clutch 3 which consists of a first clutch actuator 14 is pressed. Via a transmission input shaft 4 is the first clutch 3 on a partial transmission 5 which contains the odd gears of the transmission. A second sub-transmission train has a second clutch 6 on that with a second clutch actuator 15 connected is. The coupling 6 is via a second transmission input shaft 7 with the second partial transmission 8th connected, which contains the even gears of the transmission. The two partial transmissions 5 and 8th are in a differential 9 merged, which each have a propeller shaft 10 . 11 with a drive wheel 12 . 13 of the motor vehicle is connected.

Opposite to the first transmission input shaft 4 is a speed sensor 17 arranged while opposite to the second transmission input shaft 7 a second speed sensor 18 is installed. The speed sensors 17 and 18 are the same like the partial transmission 5 and 8th or the clutch actuators 14 and 15 with a control unit 16 connected. This controller 16 is formed in a known manner as a microprocessor with associated memory devices. The memory devices contain programs with which the clutch actuator 14 or the clutch actuator 15 as well as the partial transmission 5 and 8th be controlled. In addition, in the control unit 16 a filter 20 contain.

The speed sensors 17 . 18 each guide a speed signal 100 the transmission input shafts 4 . 7 to the control unit 16 further, of which only one signal 100 in 2 is shown. The slip speed, which is the difference between the speed of the drive output shaft 22 and the speed of the transmission input shaft 4 . 7 represents is in the controller 16 from the signals of the speed sensor 19 , which is the drive output shaft 22 is opposite, and the signals of the speed sensor 17 the transmission input shaft 4 or the speed sensor 18 the transmission input shaft 7 determined.

In 2 is the initial problem evident, which is to be improved according to the prior art. Next to the signal 100 the speed of a transmission input shaft 4 . 7 is the signal 101 represented, which corresponds to the slip speed. Furthermore, the predication weighting is a signal 102 shown and for completeness, the speed of the drive motor 1 103 characterized, while the filtered speed signal by the reference numeral 104 is marked. How out 2 can be seen, between the time t = 24 s and t = 25 s a pronounced oscillation of the signal 100 the transmission input shaft speed available. The slip speed 101 varies between 0 rpm and up to 100 rpm during this period. The prediction weighting attempts in vain to the oscillation of the transmission input speed signal 100 to follow and remains essentially at levels below 20%.

In 3 It is assumed that it has been detected that the transmission input shaft speed signal 100 performs a vibration. This is particularly recognized by the fact that at time t = 24.15 s the amplitude 105 the transmission input shaft speed signal 100 increases. The more detailed method of detecting whether it is a vibration will be explained later. Upon detection of the vibration of the transmission input speed signal 100 becomes the signal of the prediction weight 102 driven over a fast ramp to a value of 75% and fixed there. After the vibration of the transmission input speed signal 100 is sufficiently decayed at the time t = 24.65 s, the prediction weight becomes 102 again depending on the slip speed 101 certainly. Is through the control unit 16 found that, in turn, a vibration through the transmission input shaft speed signal 100 is executed, this routine is restarted and the prediction weighting 102 set to a constant value.

In the following, will now be explained in more detail how the presence of a vibration in the speed signal 100 the transmission input shaft 4 . 7 is recognized. For this purpose, the method is considered on the simple example of a damped sine wave, as in 4 is shown. For simplicity, in 4 not signal and reference signal shown, but their difference, so the signal deviation 200 , It should be noted that in 4 only the unfiltered size is shown. The two dashed lines in 4 set the lower or upper limit of the demarcation area 201 These two limits do not necessarily have to be symmetrical about zero. The following are the in the 4 considered with 1 to 10 times. At time 1, all amplitudes are zero. The signal deviation 200 leaves the demarcation area for the first time 201 and the short-term amplitude 202 the signal deviation 200 increases until the signal deviation 200 their first positive extreme 203 (Maximum) has reached. Subsequently, the short-term amplitude 202 to this positive extreme 203 recorded.

At time 2 has the signal deviation 200 the demarcation area 201 completely traverses and reaches the area below the lower limit of the demarcation area 201 , This takes over the long-term amplitude 204 the value of the short-term amplitude 202 , which is then reset to zero. Furthermore, the second negative short-term amplitude begins 205 the negative signal deviation 200 to follow until this is their first negative extreme 206 (Minimum) has reached.

At time 3 has the signal deviation 200 the demarcation area 201 crossed again and has become positive again. The short-term amplitude 202 follows the signal deviation. At the same time, the negative long-term amplitude takes over 207 the new negative extreme value 206 the negative short-term amplitude 205 , which goes back to zero. Now, the two long-term amplitudes, the positive long-term amplitude 204 and the negative long-term amplitude 207 nonzero values. The sum of the amounts of these values is determined by means of the PT1 filter 20 of the control unit 16 the swing width 208 the vibration of the speed signal is calculated.

Analogous to the sequence of situations described at the times 1, 2 and 3, the Determining the vibration width 208 at times 4 to 9.

How out 4 apparent, has the signal deviation 200 since time 9, the demarcation area 201 no longer completely crossed. Therefore, since time 9 all amplitudes have stopped at their values. The time between times 9 and 10 is the longest period that can be measured. Because the vibration deviation 200 the demarcation area 201 has not left again, the vibration of the transmission input shaft speed signal 100 is considered decayed. After this time has elapsed, all amplitudes are reset to zero and the routine can begin again.

In 5 is shown how the period of the oscillation of the transmission input speed signal is determined. For this purpose, the same vibration is considered as it was assumed to determine the vibration amplitude. Once the signal deviation 200 at time 1 the demarcation area 201 leaves in a positive direction, begins a counter 21 of the control unit 16 from 1 to run in the positive direction. At time 2, half the period T / 2 is initialized with its adjustable maximum value. Now the signal deviation has 200 the demarcation area 201 crosses and is located below the lower limit of the demarcation area 201 , The half period T / 2 is then to the current count of the counter 21 is set and the counter starts counting again from -1 in the negative direction. The signal deviation 200 has the demarcation area at time 3 201 go through again. Since the counter has counted in the negative direction, T / 2 is set to the absolute value of the current counter reading. This procedure is repeated until time 5. Shortly before times 6 and 8, the counter exceeds the current value of T / 2 and causes an immediate takeover of the period. Between times 9 and 10 leaves the signal deviation 200 the demarcation area 201 no more. At time 10 have the counter 21 and half the period T / 2 reaches the maximum possible value. The counter 21 is reset to its resting value zero, where it remains until the signal deviation 200 the demarcation area 201 left again. Half the period T / 2 remains at its maximum value. The period of the oscillation of the transmission input shaft speed signal 100 is calculated by a PT1 filtering of T / 2 and then multiplied by 2.

The presented method allows the detection of vibrations of any signal around a reference signal. If an oscillation is detected, it supplies the amplitude and period of this oscillation.

LIST OF REFERENCE NUMBERS

1

drive motor

2

Double clutch

3

first clutch of the dual-clutch transmission

4

first transmission input shaft

5

first partial transmission

6

second clutch of the dual-clutch transmission

7

second transmission input shaft

8th

second partial transmission

9

differential

10

propeller shaft

11

propeller shaft

12

drive wheel

13

drive wheel

14

first clutch actuator

15

second clutch actuator

16

control unit

17

Speed sensor of the first transmission input shaft

18

Speed sensor of the second transmission input shaft

19

Speed sensor of the drive motor

20

filter

21

counter

22

Drive output shaft

100

Signal of transmission input shaft speed

101

Slip speed signal

102

Prädikationsgewichtung

103

Speed signal of the drive motor

104

filtered speed signal of the transmission input shaft

105

Amplitude of the transmission input shaft speed signal

200

signal deviation

201

Abgrenzbereich

202

positive short-term amplitude

203

first positive extreme value of the signal deviation

204

positive long-term amplitude

205

negative short-term amplitude

206

first negative extreme of the signal deviation

207

negative long-term amplitude

208

working stroke

QUOTES INCLUDE IN THE DESCRIPTION

This list of the documents listed by the applicant has been generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited patent literature

• DE 102008030481 A1 [0003]

Claims (10)

Method for smoothing a transmission input speed signal of a transmission, in particular of a dual clutch transmission, in which the transmission input speed signal is fed to a filter having a prediction weight as a function of a slip speed of a clutch, characterized in that upon detection of a vibration of the transmission input speed signal ( 100 ) the prediction weighting ( 102 ) regardless of the slip speed ( 101 ), wherein preferably after a decay of the oscillation the prediction weighting ( 102 ) again depending on the slip speed ( 101 ) is determined. Method according to Claim 1, characterized in that the prediction weighting ( 102 ) is set to a constant value. Method according to Claim 2, characterized in that the prediction weighting ( 102 ) a factor for weighting the filtered transmission input speed signal ( 100 ). Method according to Claim 2 or 3, characterized in that the prediction weighting ( 102 ) is set to a constant value> 50%, preferably> 70%. Method for detecting a vibration of an electrical signal, in particular according to claim 1, characterized in that - a reference signal is detected, - the extreme values ( 203 . 206 ) of positive and negative deviation ( 200 ) of the electrical signal ( 100 ) are determined by the reference signal, - after detection of a positive and / or negative extreme value ( 203 . 206 ) an associated partial amplitude ( 204 . 207 ) to the detected extreme value ( 203 . 206 ), and - from the sum of the positive extremes ( 203 ) held partial amplitude ( 204 ) and the amount of negative value ( 206 ) held partial amplitude ( 207 ) a vibration width ( 208 ) of the electrical signal ( 100 ) is determined. Method according to Claim 5, characterized in that an oscillation of the electrical signal ( 100 ) is detected when the vibration width ( 208 ) of the electrical signal ( 100 ) exceeds a predetermined threshold. A method according to claim 5 or 6, characterized in that a time (t) between two successive sign changes of the electrical signal ( 100 ) determined as half the period duration (T / 2) of the oscillation of the electrical signal ( 100 ), from which the frequency of the electrical signal ( 100 ) is determined. Method according to at least one of the preceding claims, characterized in that the reference signal is from a demarcation area ( 201 ), with only deviations ( 200 ) of the electrical signal ( 100 ) are detected by the reference signal, which outside the demarcation range ( 201 ) lie. Method according to at least one of the preceding claims, characterized in that the measurement of the deviations ( 200 ) and / or the time (t) between two consecutive sign changes begins when the electrical signal ( 100 ) the demarcation area ( 201 ) leaves the reference signal. Method according to claim 9, characterized in that the time (t) between two successive changes of sign of the electrical signal ( 100 ) by means of a counter ( 21 ), in particular when a maximum value of the counter ( 21 ), without a change of sign of the electrical signal ( 100 ), the measured time is discarded.

Images