JP2019035649A - Rotation transmission device having torque measuring device - Google Patents

Abstract

To achieve structure for improving reliability regarding torque measurement precision.SOLUTION: A computing unit 26 calculates a torque signal from output signals of a pair of sensors 19, 22 in which the detection sections are allowed to oppose on surfaces to be detected of a pair of encoders 9, 10, and removes an error component from the toque signal by performing filtering processing by an adaptive filter using an LMS algorithm. Further, the computing unit 26 stops learning the adaptive filter when n-order components of a frequency of a processing signal and a rotation frequency of a torque transmission shaft agree each other.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a rotation transmission device with a torque measuring device that is incorporated in, for example, an automatic transmission for an automobile and transmits torque and measures the magnitude of torque to be transmitted.

  The rotational speed of the shaft that constitutes the automatic transmission for automobiles and the magnitude of torque transmitted by this shaft are measured, and the measurement results are used as information for performing shift control of the transmission or engine output control. It has been conventionally used. Japanese Patent Laid-Open No. 57-169641 discloses an apparatus for converting the amount of elastic torsional deformation of a shaft into a phase difference between output signals of a pair of sensors and measuring the magnitude of torque based on the phase difference. Is described. Such a conventional structure will be described with reference to FIG.

  In the conventional structure, a pair of encoders 2 are fitted and fixed at two positions in the axial direction of the torque transmission shaft 1 that transmits torque during operation. The magnetic characteristics of the surface to be detected, which is the outer peripheral surface of the pair of encoders 2, change alternately and at equal pitches in the circumferential direction. In addition, the pitch at which the magnetic characteristics of the detected surfaces change in the circumferential direction is equal between the pair of detected surfaces. Further, both sensors 3 are supported by a housing (not shown) in a state where the detection portions of the pair of sensors 3 are opposed to the respective detection surfaces. Both of these sensors 3 change their output signals in response to changes in the magnetic characteristics of the part where their own detection units are opposed.

  The output signals of both sensors 3 as described above periodically change as the pair of encoders 2 rotate together with the torque transmission shaft 1. The frequency (and period) of this change takes a value commensurate with the rotational speed of the torque transmission shaft 1. For this reason, the rotational speed of the torque transmission shaft 1 can be obtained based on this frequency (or period). When the torque transmission shaft 1 is elastically twisted and deformed along with the transmission of torque by the torque transmission shaft 1, the pair of encoders 2 are relatively displaced in the rotational direction. As a result, the phase difference between the output signals of the pair of sensors 3 changes. The phase difference takes a value commensurate with the torque (the elastic torsional deformation amount of the torque transmission shaft 1). For this reason, torque can be calculated | required from a phase difference using this relationship. In the structure shown in the figure, the output signals of the pair of sensors 3 are respectively transmitted to a computing unit (not shown) through the harness 4, and the computing unit calculates the rotational speed of the torque transmission shaft 1 and the transmitted torque.

JP-A-57-169641 JP 2008-39537 A

  However, in the conventional structure as described above, when each encoder 2 has a shape error such as a magnetization pitch error or a geometric error such as eccentricity, an error component is included in the output signals of the pair of sensors 3. It will be. For this reason, an error component is also included in the torque obtained from the output signals of the pair of sensors 3. Therefore, if some countermeasure is not taken, there is a possibility that the accuracy of torque measurement will be reduced.

  Therefore, it is conceivable to remove an error component by using an adaptive filter using an LMS algorithm as described in JP 2008-39537 A. The adaptive filter using the LMS algorithm has a characteristic of learning a cancel signal having an n-order component of the rotation frequency of the rotator, which has a correlation with the error component, and removing the error component from the target signal. Here, since the frequency based on the error of the magnetized pitch of the encoder matches the primary component of the rotation frequency of the encoder (torque transmission shaft), the adaptive pitch using the LMS algorithm can be used to An error component based on the error can be removed.

  However, for example, a torque transmission shaft constituting an automatic transmission for an automobile vibrates in the rotational direction under the influence of torque fluctuations of a drive source, so that a target signal to be filtered is a vibration signal. Therefore, in a situation where the frequency of the target signal matches the n-th order component of the rotational frequency of the torque transmission shaft, not only the error component but also the signal representing the actual torque is canceled, and the torque is obtained. (The torque value after correction becomes zero). As a result, the reliability regarding the measurement accuracy of torque becomes low.

  The present invention has been made in view of the above-described circumstances, and an object thereof is to realize a structure capable of improving the reliability related to the measurement accuracy of torque.

The rotation transmission device with a torque measuring device of the present invention includes a torque transmission shaft, a pair of encoders, a pair of sensors, and a calculator.
The torque transmission shaft transmits torque during use.
The pair of encoders are made of permanent magnets in which S poles and N poles are alternately arranged in the circumferential direction on each detected surface, and are synchronized with the torque transmission shaft directly or in use. Supported by a rotating member.
The pair of sensors is a portion that does not rotate even when it is used, with each detection unit facing the detection surface of the pair of encoders and changing the output signal according to the magnetic flux density passing through the detection unit. It is supported by.
The arithmetic unit performs a signal calculation function for obtaining a processing signal from the output signals of the pair of sensors and a filtering process by an adaptive filter using an LMS algorithm in order to remove an error component included in the processing signal. And a filter function.
In particular, in the rotation transmission device with a torque measuring device according to the present invention, the computing unit determines whether the frequency of the processing signal and the n-order component (n is a positive integer) of the rotation frequency of the torque transmission shaft match each other. If not, learning of the adaptive filter is stopped.

  In the present invention, when the frequency of the processing signal and the n-order component of the rotational frequency of the torque transmission shaft match, the frequency of the processing signal and the n-order component of the rotational frequency of the torque transmission shaft match. It is also possible to execute the filtering process based on the learned value learned before.

  In the present invention, a torque signal calculated based on a phase difference between output signals of the pair of sensors can be employed as the processing signal. Alternatively, a signal representing a phase difference between the output signals of the pair of sensors, a signal representing a phase difference ratio (phase difference / period), or the like can be employed as the processing signal.

  According to the rotation transmission device with a torque measuring device of the present invention as described above, it is possible to improve the reliability related to the measurement accuracy of torque.

FIG. 1 is a cross-sectional view of a rotation transmission device with a torque measuring device according to a first example of an embodiment. FIG. 2 is a block diagram of an adaptive filter that performs a filtering process on the torque signal. FIG. 3 is a schematic diagram for explaining the relationship between whether or not the frequency of the torque signal matches the rotational frequency of the torque transmission shaft and whether or not learning is performed by the adaptive filter. FIG. 4A shows a case where the frequency of the torque signal and the rotational frequency of the torque transmission shaft do not match with respect to the present embodiment, and FIG. FIG. 4C shows the case where the frequency of the torque signal and the rotation frequency of the torque transmission shaft match for the comparative example. 4 (A) to 4 (C), the upper stage shows the signal before the filtering process, the middle stage shows the signal to be removed by the filtering process, and the lower stage shows the signal obtained after the filtering process. . FIG. 5: is a schematic diagram which shows the power transmission mechanism (powertrain) provided with the rotation transmission apparatus with a torque measuring device regarding the 2nd example of embodiment. FIG. 6 is a diagram corresponding to FIG. 5 and showing a third example of the embodiment. FIG. 7 is a diagram corresponding to FIG. 5 and showing a fourth example of the embodiment. FIG. 8 is a schematic side view showing an example of a rotation transmission device with a torque measuring device having a conventional structure.

[First example of embodiment]
A first example of the embodiment will be described with reference to FIGS.
The rotation transmission device with a torque measuring device of this example is used by being incorporated in an automatic transmission for an automobile, and includes a housing (mission case) 5, a torque transmission shaft 1 a that functions as a counter shaft, and a counter gear. A functioning input gear 6 and output gear 7, a pair of rolling bearings 8a and 8b, a first encoder 9 and a second encoder 10, a pair of sensor units 11a and 11b, and a calculator 26 are provided.
In the following description of this example, “one side” in the axial direction refers to the right side of FIG. 1, and “the other side” in the axial direction refers to the left side of FIG. 1.

  The torque transmission shaft 1a is made of an alloy steel such as carbon steel in a hollow cylindrical shape, and is subjected to heat treatment such as quenching and tempering so that the surface hardness of the torque transmission shaft 1a is HV400 or more, The surface carbon concentration is 0.2% or more. In this example, the input gear 6 for inputting torque to the torque transmission shaft 1a is separated from the torque transmission shaft 1a at a portion closer to the other end in the axial direction of the torque transmission shaft 1a (a portion closer to the left end in FIG. 1). An output gear 7 for outputting torque is provided at a portion closer to one end of the torque transmission shaft 1a in the axial direction (portion closer to the right end in FIG. 1) and separately from the torque transmission shaft 1a. In addition, both sides of the torque transmission shaft 1a sandwiching the portion where the input gear 6 and the output gear 7 are installed (the other end portion in the axial direction and the one end portion in the axial direction) are paired by a pair of rolling bearings 8a and 8b. The housing 5 is rotatably supported.

  The input gear 6 and the output gear 7 are helical gears or spur gears made of alloy steel such as carbon steel, and are provided separately from the torque transmission shaft 1a. For this reason, with respect to the fitting portion of the input gear 6 and the output gear 7, a cylindrical surface fitting portion for ensuring concentricity (coaxiality) and an involute spline engaging portion for preventing relative rotation, A configuration is used that is adjacent in the axial direction.

  The pair of rolling bearings 8a and 8b are, for example, deep groove type and angular type ball bearings, tapered roller bearings, cylindrical roller bearings, radial needle bearings, self-aligning roller bearings, and the like (in the example shown, ball bearings). Each is composed of an annular outer ring 12a, 12b and inner rings 13a, 13b and a plurality of rolling elements 14. The outer rings 12 a and 12 b are stationary wheels that do not rotate during use, and are fitted and fixed to the housing 5. The inner rings 13a and 13b are rotating wheels that rotate during use, and are fitted and fixed to the torque transmission shaft 1a. The rolling element 14 is held by a cage between the outer ring raceway formed on the inner peripheral surface of the outer ring 12a, 12b in the axial intermediate portion and the inner ring raceway formed on the outer peripheral surface of the inner ring 13a, 13b in the axial intermediate portion. In this state, it is provided so as to be able to roll freely. Further, in this example, the rolling bearings 8a and 8b are mutually opposite in contact angle.

  The first encoder 9 is supported and fixed at the other axial end of the torque transmission shaft 1a. For this reason, the first encoder 9 can rotate (synchronously) with the other axial end of the torque transmission shaft 1a. On the other hand, the second encoder 10 is externally fitted and fixed to one end of the torque transmission shaft 1a in the axial direction. For this reason, the 2nd encoder 10 can rotate with the axial direction one end part of the torque transmission shaft 1a (synchronously).

  The first encoder 9 includes an annular screw ring 15 such as a nut that is screwed and fixed to the other axial end of the torque transmission shaft 1a, and rubber, synthetic resin, and the like fixed to the outer peripheral surface of the screw ring 15. It is composed of an encoder body 16 made of permanent magnets such as rubber magnets and plastic magnets, in which magnetic powder is dispersed in a polymer material to form a cylindrical shape as a whole. On the other hand, the second encoder 10 includes a screw ring 17 such as a nut that is screwed and fixed to one axial end of the torque transmission shaft 1a, and an encoder body 18 made of a permanent magnet fixed to the outer peripheral surface of the screw ring 17. It consists of and.

  Examples of magnetic powders contained in the encoder bodies 16 and 18 include ferrite-based magnetic powders such as strontium ferrite and barium ferrite, and rare-earth element magnetic powders such as samarium-iron, samarium-cobalt, and neodymium-iron-boron. Can be adopted. The outer peripheral surfaces of the encoder bodies 16 and 18, each of which is a detected surface, have the same diameter and are arranged coaxially with each other. In addition, on the outer peripheral surfaces of the encoder bodies 16 and 18, S poles and N poles are alternately arranged at equal pitches in the circumferential direction, and the magnetic characteristics are alternately arranged at equal pitches in the circumferential direction. It is changing. The total number of magnetic poles (S poles and N poles) on the outer peripheral surfaces of the encoder bodies 16 and 18 coincide with each other.

  The sensor unit 11a is supported and fixed to an outer ring 12a constituting the rolling bearing 8a. The first sensor 19, a synthetic resin sensor block 20 supporting the first sensor 19, and the sensor block 20 on the inner side. And a held metal sensor cap 21. On the other hand, the sensor unit 11b is supported and fixed to the outer ring 12b constituting the rolling bearing 8b, the second sensor 22, the sensor block 23 made of a synthetic resin that supports the second sensor 22, and the sensor block. And a metal sensor cap 24 holding the inside 23.

  Magnetic detection elements such as Hall elements, Hall ICs, and MR elements (including GMR elements, TMR elements, and AMR elements) are incorporated in the detection units of the first sensor 19 and the second sensor 22. In the state where the sensor units 11a and 11b are supported as described above, the detection unit of the first sensor 19 is brought close to and opposed to the detected surface of the first encoder 9 (the outer peripheral surface of the encoder body 16), and the second The detection unit of the sensor 22 is placed in close proximity to the detected surface of the second encoder 10 (the outer peripheral surface of the encoder body 18). For this reason, the first sensor 19 changes the output signal according to the magnetic flux density (magnetic flux passing through the detection surface / area of the detection surface) passing through its detection unit, and the second sensor 22 The output signal is changed according to the magnetic flux density passing through the detection unit. In this example, the output signals of the first sensor 19 and the second sensor 22 can be transmitted to the computing unit 26 through the harnesses 25a and 25b, respectively.

  The computing unit 26 of this example has a function of accurately obtaining the torque transmitted by the torque transmission shaft 1a using the output signals of the first sensor 19 and the second sensor 22. Specifically, the computing unit 26 has a function of removing an error component included in a torque signal representing torque calculated based on the phase difference between the output signals of the first and second sensors 19 and 22. . Hereinafter, functions provided in the calculator 26 will be described in detail.

  When torque is transmitted by the torque transmission shaft 1a, the portion between the input gear 6 and the output gear 7 of the torque transmission shaft 1a is elastically torsionally deformed, so that both ends of the torque transmission shaft 1a in the axial direction. The parts (first and second encoders 9, 10) are relatively displaced in the rotational direction (circumferential direction). As a result of the relative displacement of the first and second encoders 9 and 10 in the rotational direction in this way, the output signals of the first and second sensors 19 and 22 corresponding to the torsion angle of the torque transmission shaft 1a. The phase difference between the two changes. The phase difference takes a value commensurate with the torque (the elastic torsional deformation amount of the torque transmission shaft 1a). Therefore, the computing unit 26 uses the data representing the relationship (table, map, etc.) between the torque and the phase difference obtained in advance based on the torsional rigidity of the torque transmission shaft 1a, and the signal calculation function provided by itself. Thus, a torque signal d having a waveform as shown by a solid line in the upper stage of FIG. 4 is calculated from the phase difference between the output signals of the first and second sensors 19 and 22.

  In this example, as described above, the magnitude and direction of the torque transmitted by the torque transmission shaft 1a are obtained based on the phase difference between the output signals of the first and second sensors 19 and 22, In order to accurately determine the magnitude and direction of the torque, the accuracy of the detected surfaces of the first and second encoders 9 and 10 needs to be good. On the other hand, the accuracy with respect to the position of the boundary portion where the characteristics of the detected surfaces of the first and second encoders 9 and 10 change may not be sufficient due to an error in the magnetization pitch. Therefore, in this example, the torque signal d is processed by the adaptive filter 27 as shown in FIG. 2 so that an error based on the magnetization pitch (an error related to the position of the boundary portion between the N pole and the S pole) is removed. I have to. The adaptive filter 27 used in this example uses an LMS algorithm. Hereinafter, the filter function of the adaptive filter 27 will be described focusing on the case of removing the magnetization pitch error.

The boundary position between the N pole and the S pole existing on the detection surface of each of the first and second encoders 9 and 10 at the part where the detection units of the first and second sensors 19 and 22 are opposed to each other is actually The amount of change due to twisting of the torque transmission shaft 1a is superimposed on the error component due to the magnetization pitch error or the like. For this reason, as shown in the upper part of FIG. 4, the torque signal d (solid line), which is a calculated value obtained from the output signals of the first and second sensors 19 and 22, is the torque transmitted by the torque transmission shaft 1a. A signal obtained by adding a signal d _d (dashed line) representing an actual torque corresponding to the magnitude and an error component d _n (broken line) due to an error in the magnetization pitch of the first and second encoders 9 and 10. (D = d _d + d _n ). Thus, the adaptive filter 27, by subtracting the error component (variation) d _n from the torque signal d, will be asked to signal d _d representing the actual torque.

Meanwhile, in order to operate the adaptive filter 27, a reference signal x having a correlation with the error component d _n is required. If you have a reference signal x, the adaptive filter 27 by the self-learning, with the same characteristics as the transmission characteristics of the actual signal flow "d _n → d" to form a FIR (finite impulse response) filter. In other words, the adaptive filter 27, by learning the error component d _n by using the reference signal x, obtain {y (k) to be described later} cancellation signal y. Therefore, by subtracting the cancellation signal y from the torque signal d, it becomes equivalent to removing the error component d _n from the torque signal d (d-d _n). If this way, eliminate the error component d _n, the adaptive filter 27, instead of filtering of the torque signal d sent the main root of the signal (upper half portion in FIG. 2), sub-root (FIG. 2 The cancel signal y is calculated based on the reference signal x sent to the lower half). Then, the cancel signal y is subtracted from the torque signal d.

Here, the influence of the error in the magnetization pitch of the first and second encoders 9 and 10 coincides with the primary component of the rotational frequency of the torque transmission shaft 1a. For this reason, the reference signal x is, for example, a sine wave, a triangular wave, a rectangular wave, or a M wave with one cycle if the first and second encoders 9 and 10 have M pulses per rotation, It can be self-generated as a pulse wave or the like. Then, if called for cancellation signal y based on the obtained reference signal x, the cancellation signal y, by subtracting from the torque signal d representing the torque, correction signal representative of the signal d _d representing the original torque e { E (k)} to be described later is obtained. The processing for obtaining the correction signal e by obtaining the cancel signal y in the adaptive filter 27 and further subtracting the cancel signal y from the torque signal d representing the torque is performed by the following equations (1) to (3). Based on.

In the equations (1), (2), and (3), k is a data number in time series, and N is the number of taps of the FIR filter used as the adaptive filter 27. W represents a filter coefficient of the FIR filter, w _k represents a filter coefficient used when k-th data processing is performed, and w _{k + 1} represents a filter used when processing the next data series (k + 1-th). Each coefficient is represented. In other words, the FIR filter is an adaptive filter in which the filter coefficients are sequentially updated appropriately according to the equation (3).

The reference signal x input to the adaptive filter 27 may be a signal correlated with the n-th order (n is a positive integer) component of the rotational frequency of the torque transmission shaft 1a. One impulse signal per 10 revolutions may be used. Therefore, it is assumed that the reference signal x is a 1-impulse signal and the number N of taps of the adaptive filter 27 is equal to the number of pulses per rotation of the first and second encoders 9 and 10. In this case, the reference signal x used for calculation at the instant of time series k is expressed by the following equation (4).

In the above equation (4), the position j where the reference signal x becomes an impulse having the value 1 is shifted one by one to the right as the time series k advances, and shifts to the N-1th rightmost position. In the next time series, a new impulse value appears at the leftmost zeroth. That is, the reference signal x is a data string in which the position of the impulse having the value 1 is circulated from the 0th to the (N-1) th. Substituting the above expression (4) into the above expressions (1) and (3), the following expressions (5) and (6) are obtained.

  When the adaptive filter is operated by a normal non-synchronous LMS algorithm, it is necessary to repeat the calculations shown in the equations (1), (2), and (3), whereas the adaptive filter is operated by the synchronous LMS algorithm. In the case of operation, it is only necessary to perform the calculations shown in the equations (5), (6) and (2).

Here, the torque transmission shaft 1a vibrates in the rotational direction under the influence of torque fluctuations of a drive source such as an engine, so that the torque signal d to be filtered is a vibration signal. Therefore, when not subjected whatsoever measures, the frequency of the torque signal d, in a situation where the n-order component of the rotational frequency of the torque transmission shaft 1a coincides, not only the error component d _n, the signal representing the original torque Even up to d _d is canceled and the torque cannot be obtained (the corrected torque value becomes zero).

Specifically, with reference to FIG. 4C showing the conventional example, as shown in the upper part, the frequency of the torque signal d, which is the calculated value, is the n-order component of the rotational frequency of the torque transmission shaft ( In the example shown in the figure, when it coincides with the primary component), d (= d _d + d _n ) is learned as the error component learning value (cancellation signal y) as shown in the middle stage. For this reason, as shown in the lower part, the correction value (correction signal e) obtained after the filtering process becomes zero.

Therefore, in this example, the calculator 26 determines whether or not the frequency of the torque signal d matches the n-order component of the rotational frequency of the torque transmission shaft 1a. As shown in FIG. 3, when the frequency of the torque signal d and the n-order component of the rotational frequency of the torque transmission shaft 1a do not match each other, learning of the adaptive filter 27 is executed and they match. If so, learning of the adaptive filter 27 is stopped.
The n-order component of the rotational frequency of the torque transmission shaft 1a is obtained using the output signal of the first sensor 19 or the output signal of the second sensor 22. That is, since the rotational speed of the torque transmission shaft 1a is obtained from the output signal of the first sensor 19 or the output signal of the second sensor 22, the n-order component of the rotational frequency of the torque transmission shaft 1a is obtained using this rotational speed. Can be sought.

If the frequency of the torque signal d and the n-th order component of the rotational frequency of the torque transmission shaft 1a do not coincide with each other, the cancel signal y learned in this state is subtracted from the torque signal d to obtain the original torque. A modified signal e representing the signal d _d representing is obtained. Specifically, as shown in the upper part of FIG. 4A, the frequency of the torque signal d (solid line), which is a calculated value, and the n-order component of the rotational frequency of the torque transmission shaft 1a (primary component in the illustrated example). ), The learning value _dn (dashed line) is obtained as shown in the middle of FIG. Therefore, as shown in the lower part of the figure, d _d (= d−d _n ) is obtained as the correction value (correction signal e).

On the other hand, when the frequency of the torque signal d and the n-th order component of the rotational frequency of the torque transmission shaft 1a coincide with each other, learning of the adaptive filter 27 is stopped. Then, the calculator 26 uses the cancellation signal y, which is a learned value learned before (eg, immediately before) when the frequency of the torque signal d and the n-th order component of the rotational frequency of the torque transmission shaft 1a coincide with each other. Find e. Specifically, as shown in the upper part of FIG. 4B, the frequency component of the torque signal d (solid line) that is the calculated value and the n-order component of the rotational frequency of the torque transmission shaft 1a (primary component in the illustrated example). If the error component is learned as it is, the learning value becomes d as in the conventional example {see FIG. 4C}. For this reason, in this example, without performing such learning, instead of the n-order component of the frequency of the torque signal d and the rotational frequency of the torque transmission shaft 1a as shown in the middle part of FIG. As shown in the lower part, d _d (= d−d _n ) is obtained as a correction value (correction signal e) using the learning value d _n (broken line) learned before the two match.

As for the adaptive filter 27, the filter coefficient w _k used at the start of calculation may be self-adapted if it starts to move even if zero is substituted. It is also possible to obtain and substitute that value. Alternatively, the filter coefficient last used in the previous process may be stored in a storage unit such as an EEPROM and used at the time of restart. Furthermore, the data represented by the first input signal can be used as the initial value of the filter coefficient.

なお、（７）式中のαも、フィルタ係数を自己適正化させていくための更新量を決定するパラメータとなるが、０＜α＜１の範囲であれば良く、前記μよりも設定が容易である。また、参照信号ｘを自己生成する場合には、前記（７）式中の分母の値は既知であり、μの最適値を事前に算出しておくこともできる。計算量削減の観点からは、予め（７）式でこのμを算出しておき、このμを定数として前記（３）式でフィルタ係数を自己適正化させるのが望ましい。 In addition, μ in the equation (3) is a value called a step size parameter that determines an update amount when the filter coefficient is self-optimized, and is usually about 0.01 to 0.001. Actually, however, the validity of the adaptive operation can be examined and set in advance, or it can be sequentially updated using the following equation (7).

Note that α in the equation (7) is also a parameter for determining an update amount for self-optimizing the filter coefficient, but it may be in a range of 0 <α <1, and is set more than the μ. Easy. When the reference signal x is self-generated, the denominator value in the equation (7) is known, and the optimum value of μ can be calculated in advance. From the viewpoint of reducing the amount of calculation, it is desirable to calculate this μ in advance using equation (7) and to self-optimize the filter coefficient using equation (3), where μ is a constant.

  In this example having the above-described configuration, the first and second encoders 9, 10 are included in the torque signal d calculated from the phase difference between the output signal of the first sensor 19 and the output signal of the second sensor 22. Even when an error component based on the pitch error is included, the error component can be removed by the filter function of the calculator 26. For this reason, the reliability regarding the measurement accuracy of torque can be improved. In addition, in this example, the torque cannot be calculated even under a situation where the vibration frequency of the torque signal d caused by the torque fluctuation of the engine and the n-order component of the rotational frequency of the torque transmission shaft 1a coincide with each other (correction value). Can be zero). Therefore, it is possible to sufficiently ensure the reliability related to the measurement accuracy of the torque.

[Second Example of Embodiment]
In the second example of the embodiment, a case where the rotation transmission device with a torque measuring device of the present invention is applied to a specific structure of an automatic transmission for a vehicle will be described with reference to FIG. In the illustrated example, the power output from the engine 28 that is a power source is a torque converter 30, a forward / reverse switching mechanism 31, a belt-type transmission mechanism 32, a counter gear mechanism 33, and the like that constitute a continuously variable transmission 29. Is transmitted to a drive wheel (not shown).

  The engine 28 has various structures depending on the number of cylinders and their arrangement, and the frequency of vibration in the rotational direction of the engine output shaft (crankshaft) 34 varies depending on the structure. For example, in a four-cylinder engine employed in many automobiles, vibration in the rotational direction of the engine output shaft 34 has a frequency twice the rotational speed of the engine output shaft 34 (secondary component of the rotational frequency).

  The torque converter 30 has a function of mechanically transmitting torque (lock-up mechanism) in addition to a function of transmitting torque input from the engine output shaft 34 of the engine 28 via a fluid. The torque converter 30 also has a function of amplifying the transmitted torque.

  The forward / reverse switching mechanism 31 includes a plurality of gears (not shown), and switches between forward and reverse of the automobile.

  The belt-type transmission mechanism 32 is configured by spanning an endless belt 39 between a primary pulley 36 fixed to the input shaft 35 and a secondary pulley 38 fixed to the output shaft 37. The belt-type transmission mechanism 32 adjusts the transmission ratio between the input shaft 35 and the output shaft 37 in a stepless manner by relatively changing the groove width of the primary pulley 36 and the groove width of the secondary pulley 38. Can do.

  The counter gear mechanism 33 includes a counter shaft 40 that is disposed in parallel with the output shaft 37 that constitutes the belt-type transmission mechanism 32. An input gear 41 and an output gear 42 each functioning as a counter gear are fixed to the counter shaft 40 in a state of being separated from each other in the axial direction. The counter shaft 40 rotates in the direction opposite to the output shaft 37 by engaging the input gear 41 of the counter shaft 40 with the output gear 43 fixed to the output shaft 37 of the belt-type transmission mechanism 32. In this example, the present invention is applied using such a counter shaft 40 as a torque transmission shaft.

  The vibration in the rotational direction of the engine output shaft 34 based on the torque fluctuation of the engine 28 reaches the counter shaft 40 constituting the counter gear mechanism 33 via the torque converter 30, the forward / reverse switching mechanism 31, the belt-type transmission mechanism 32, and the like. Communicated. Therefore, as in the first example of the embodiment, the counter shaft 40 is twisted in the rotational direction by the first sensor 19 and the second sensor 22 (see FIG. 1) in order to obtain the torque transmitted by the counter shaft 40. When detecting and obtaining a torque signal representing the torque transmitted by the counter shaft 40 by the computing unit 26 (see FIG. 1) based on the phase difference between the output signal of the first sensor 19 and the output signal of the second sensor 22 The torque signal has a secondary component of the rotational frequency of the engine output shaft 34 caused by the vibration of the engine output shaft 34.

  Here, considering the case where the torque converter 30 is mechanically transmitting torque by a lock-up mechanism, the ratio of the rotation frequency of the engine output shaft 34 and the rotation frequency of the counter shaft 40 is the same as that of the engine output shaft 34. It is determined by the transmission ratio of the mechanism existing between the counter shaft 40 and the counter shaft 40. Therefore, when the gear ratio between the engine output shaft 34 and the counter shaft 40 is near an integer multiple of 0.5 (0.5, 1.0, 1.5...), The rotational frequency of the counter shaft 40 And the frequency of the torque signal of the torque transmitted by the counter shaft 40 (= second-order component of the rotational frequency of the engine output shaft 34) matches.

  Therefore, in this example, when the gear ratio of the mechanism existing between the engine output shaft 34 and the counter shaft 40 is an integral multiple of 0.5, the computing unit 26 uses the adaptive filter 27 (see FIG. 2). Stop learning. On the other hand, when the gear ratio of the mechanism existing between the engine output shaft 34 and the counter shaft 40 does not become an integer multiple of 0.5, the computing unit 26 performs learning of the adaptive filter 27.

  In the above description, the case where the engine 28 is a four-cylinder engine is taken as an example. Therefore, the vibration of the engine output shaft 34 has a frequency twice the rotational speed of the engine output shaft 34 (secondary component of the rotational frequency). Had. On the other hand, considering the case where the engine 28 is an N cylinder, the vibration of the engine output shaft 34 has a frequency N / 2 times the rotational speed of the engine output shaft 34. When the transmission ratio of the mechanism existing between the counter shaft 40 and the counter shaft 40 is an integral multiple of 2 / N, the rotation frequency of the counter shaft 40 and the frequency of the torque signal of the torque transmitted by the counter shaft 40 coincide with each other. Will do. Therefore, when the gear ratio of the mechanism existing between the engine output shaft 34 and the counter shaft 40 is an integral multiple of 2 / N, the calculator 26 stops learning of the adaptive filter 27. On the other hand, when the gear ratio of the mechanism existing between the engine output shaft 34 and the counter shaft 40 does not become an integer multiple of 2 / N, the computing unit 26 performs learning of the adaptive filter 27. About another structure and an effect, it is the same as the 1st example of embodiment.

[Third example of embodiment]
A third example of the embodiment will be described with reference to FIG. In this example, a case where a parallel shaft gear type multi-stage transmission 44 as shown in FIG. 6 is used as an automatic transmission for an automobile will be described.

  The illustrated multi-stage transmission 44 has an input shaft 35a and an output shaft 37a arranged in parallel to each other. On the input shaft 35a, first- to fourth-speed drive gears 45a to 45d and a reverse gear 46a are supported so as to be idle. On the other hand, the driven gears 47a to 47d and the reverse gear 46b of 1st to 4th gears are fixed to the output shaft 37a, respectively. A reverse gear 46c is fixed to an intermediate shaft (not shown) disposed between the input shaft 35a and the output shaft 37a.

  The gear 45a and the gear 47a form a first-speed gear pair, the gear 45b and the gear 47b form a second-speed gear pair, the gear 45c and the gear 47c form a third-speed gear pair, The gear 45d and the gear 47d form a 4-speed gear pair. The three gears 46a to 46c form a reverse gear pair. The illustrated multi-stage transmission 44 has four shift stages, and the gear stage is switched by selecting a gear fastened to the input shaft 35a.

  A clutch 48 is provided between the input shaft 35 a constituting the multi-stage transmission 44 as described above and the engine output shaft 34 of the engine 28. When the clutch 48 is engaged, the power of the engine 28 is transmitted to the input shaft 35a. In this example, the present invention is applied using the output shaft 37a constituting the multi-stage transmission 44 as a torque transmission shaft.

  In the multi-stage transmission 44 as described above, under the situation where the clutch 48 is engaged, the ratio between the rotational frequency of the engine output shaft 34 and the rotational frequency of the output shaft 37a is the speed ratio of the multi-stage transmission 44 (whichever Depending on whether the gear pair transmits torque). For this reason, when the engine 28 is an N cylinder, when the speed ratio of the multi-stage transmission 44 is an integral multiple of 2 / N, the computing unit 26 (see FIG. 1) includes the adaptive filter 27 (see FIG. 2). ) Stop learning. On the other hand, when the gear ratio of the multi-stage transmission 44 is not an integral multiple of 2 / N, the adaptive filter 27 is learned. About another structure and an effect, it is the same as the 1st example and 2nd example of embodiment.

[Fourth Example of Embodiment]
In the fourth example of the embodiment, a model that is more typical than the second and third examples of the embodiment will be described with reference to FIG.
As shown in the figure, a structure is considered in which a speed change mechanism 50 is provided between a torque transmission shaft 1b to be measured for torque and a drive source 49 that generates torque transmitted by the torque transmission shaft 1b and generates torque fluctuation. . In the case of this example as well, as in the first example of the embodiment, the torque signal of the torque transmitted by the torque transmission shaft 1b based on the torque fluctuation of the drive source 49 becomes a vibration signal. The frequency of the signal is calculated, and it is determined whether or not this frequency matches the n-th order component of the rotational frequency of the torque transmission shaft 1b. When the frequency of the torque signal and the n-th order component of the rotational frequency of the torque transmission shaft 1b do not match each other, the computing unit 26 (see FIG. 1) performs learning of the adaptive filter 27 (see FIG. 2). On the other hand, if they match, learning of the adaptive filter 27 is stopped. About another structure and an effect, it is the same as the 1st example of embodiment.

  The torque transmission shaft that constitutes the rotation transmission device with the torque measuring device of the present invention is not limited to the rotation shaft that constitutes the power train of the automobile, but, for example, the rotation shaft of the windmill (main shaft, rotation shaft of the speed increasing device), rolling mill Roll neck, rolling shaft of rolling stock (axle, rotating shaft of speed reducer), rotating shaft of machine tool (main shaft, rotating shaft of feed system), construction machinery / agricultural machinery / home appliance / motor rotating shaft, etc. The rotating shafts of various mechanical devices can be targeted.

Further, when configuring a power train of an automobile, for example, an input shaft (turbine shaft) to which torque is input from a torque converter or a countershaft can be targeted.
In addition, the form of the transmission in the case of configuring the transmission by incorporating the rotation transmission device with the torque measuring device of the present invention is not particularly limited, and various continuously variable transmissions such as an automatic transmission (AT), a belt type and a toroidal type. It is possible to employ a transmission that performs a shift by vehicle-side control, such as a machine (CVT), an automated manual transmission (AMT), a dual clutch transmission (DCT), or a transfer.
The relationship between the installation position of the transmission and the drive wheels is not particularly limited, and there are front engine front wheel drive vehicles (FF vehicles), front engine rear wheel drive vehicles (FR vehicles), four wheel drive vehicles, and the like. It becomes a target.
Further, the power source placed on the upstream side of the transmission does not necessarily have to be an internal combustion engine such as a gasoline engine or a diesel engine, and may be an electric motor used in a hybrid vehicle or an electric vehicle, for example.

  In the above-described embodiment, the detection surface of the encoder has a cylindrical shape, and the structure in which the detection portion of the sensor is opposed to the detection surface in the radial direction has been described as an example. Can adopt a structure in which the detection surface of the encoder has a ring shape, and the detection portion of the sensor faces the detection surface in the axial direction. The pair of encoders may be supported directly on the torque transmission shaft, or may be supported via other members such as a sleeve, a gear, and a bearing inner ring. Further, the pair of encoders and the pair of sensors are not limited to a structure in which they are arranged apart from each other on both sides in the axial direction of the torque transmission shaft, but an inner shaft disposed on the inner diameter side of the torque transmission shaft is used. A pair of encoders and a pair of sensors can also be placed close together.

  In addition, the rolling bearing for rotatably supporting the torque transmission shaft is not limited to a ball bearing, and it is possible to use conventionally known rolling bearings such as a tapered roller bearing, a cylindrical roller bearing, and a needle bearing. it can.

DESCRIPTION OF SYMBOLS 1, 1a, 1b Torque transmission shaft 2 Encoder 3 Sensor 4 Harness 5 Housing 6 Input gear 7 Output gear 8a, 8b Rolling bearing 9 First encoder 10 Second encoder 11a, 11b Sensor unit 12a, 12b Outer ring 13a, 13b Inner ring 14 Roll Moving body 15 Thread ring 16 Encoder body 17 Thread ring 18 Encoder body 19 First sensor 20 Sensor block 21 Sensor cap 22 Second sensor 23 Sensor block 24 Sensor cap 25a, 25b Harness 26 Calculator 27 Adaptive filter 28 Engine 29 Continuously variable transmission 30 Torque converter 31 Forward / reverse switching mechanism 32 Belt-type transmission mechanism 33 Counter gear mechanism 34 Engine output shaft 35, 35a Input shaft 36 Primary pulley 37, 37a Output system Shaft 38 Secondary pulley 39 Endless belt 40 Counter shaft 41 Input gear 42 Output gear 43 Output gear 44 Multi-stage transmission 45a-45d Gear 46a-46c Gear 47a-47c Gear 48 Clutch 49 Drive source 50 Transmission mechanism

Claims (3)

A torque transmission shaft that transmits torque during use;
Each of the detected surfaces is made of a permanent magnet in which S poles and N poles are alternately arranged in the circumferential direction, and is supported by a member that rotates directly or synchronously with the torque transmission shaft when used. A pair of encoders;
A pair of sensors supported by a portion that does not rotate even when used, with each detection unit facing the detection surface of the pair of encoders and changing an output signal according to the magnetic flux density passing through the detection unit; ,
A signal calculation function for obtaining a processed signal from the output signals of the pair of sensors, and a filter function for performing a filtering process by an adaptive filter using an LMS algorithm in order to remove an error component included in the processed signal. And an arithmetic unit having
The computing unit determines whether or not the frequency of the processing signal and the n-order component (n is a positive integer) of the rotational frequency of the torque transmission shaft match each other. Stop learning,
Rotation transmission device with torque measuring device.   When the frequency of the processing signal and the n-order component of the rotational frequency of the torque transmission shaft match, learning is performed before the frequency of the processing signal and the n-order component of the rotational frequency of the torque transmission shaft match. The rotation transmission device with a torque measuring device according to claim 1, wherein filtering processing is executed based on the learning value.   The rotation transmission device with a torque measuring device according to any one of claims 1 and 2, wherein the processing signal is a torque signal calculated based on a phase difference between output signals of the pair of sensors.

Images