Game crossing detection using a shaft torque sensor

Abstract

Methods for detecting play, including: (a) Measuring torque transmitted between components across a torsional discontinuity, (b) Measuring drivetrain twist across said discontinuity, (c) Using a ratio of drivetrain twist divided by torque and drivetrain stiffness to determine whether the drivetrain is entering, exiting, or in a play zone.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention

[0001] This invention relates generally to detecting lash crossing in the driveline of a motor vehicle, particularly using a device that generates signals indicative of the magnitude of the transmitted torque and the direction of rotation of the transmitted torque. 2. Description of the state of the art

[0002] Automotive drive trains contain meshing gear teeth that exhibit a dead zone when the direction of the torque applied to them changes, as shown graphically in Fig. 1 and Fig. 2 shown.

[0003] Such a deadband causes a discontinuity in the transmitted torque and excitation of the driveline due to the shafts' degree of compliance. For this reason, the deadband is commonly referred to as "backlash" or, since it occurs during the torque direction change, as "backlash crossing." If there were no backlash associated with the meshing gear teeth, the shaft twist-torque relationship could be expressed using the torsion spring equation (Hooke's law): t = Ks q, where Ks is the shaft stiffness.

[0004] Backlash is one of the significant factors contributing to drivability problems. There have been ongoing efforts in the automotive control engineering community to address the problem of backlash by predicting, detecting, avoiding, and mitigating the effects of backlash.

[0005] Document DE 601 04 247 T2 relates to a method for taking up play in a drive system when a load change occurs, the drive system comprising: a drive motor, an electric motor and a control system, characterized in that the method comprises the following steps: determining whether a load change is imminent, determining whether a drive or braking torque is required from the electric motor, sending a single pulse to the electric motor when a load change occurs, wherein a torque pulse from the electric motor takes up the play in the drive train before a torque stage from the drive motor begins.

[0006] Document DE 10 2006 039 678 A1 relates to a method for operating a prime mover in a drive train subject to play and elasticity to prevent a load shock. In this method, a speed increase of the prime mover is initially permitted during an acceleration phase and then reduced during a deceleration phase. The switching time from the acceleration to the deceleration phase is determined at the start of the method.

[0007] Document DE 10 2007 013 335 A1 relates to a method for backlash estimation and handling clunk in a driveline using multivariable active driveline damping. When the estimated driveline axle torque is substantially non-zero, the nominal parameters for the transmission mode or gear in the state estimator are used. When the estimated driveline axle torque is substantially zero, the parameters are switched to neutral parameters, the backlash estimator indicates a neutral backlash state, and the backlash angle is tracked until it accumulates an expected amount of backlash in the driveline. During a backlash transfer time, the active damping controls the speeds of the driveline components such that the effect of backlash cancellation is minimized. After backlash cancellation occurs, the desired axle torque is used by the system.

[0008] Document DE 10 2009 051 474 A1 relates to an apparatus and a method for determining an estimated value for driveline lash, the method comprising determining whether the driveline is going through driveline lash; when the driveline is going through driveline lash, determining when the driveline lash is balanced; determining when an estimated lash state changes from a neutral state to a positive or negative state, determining the time difference between when the driveline lash is balanced and when the estimated lash state changes from a neutral state to a positive state or a negative state, and updating the adaptive lash estimate when the aforementioned time difference is not equal to zero.

[0009] With the measures available in current production vehicles, reliably detecting deadband crossing is a significant challenge. Several publications examine Hooke's Law to derive driveline torque based on shaft position sensors, commonly used for angular velocity and known as "speed sensors." Others use the speed differential across the unlocked torque converter as an indication of the direction of rotation at the transmission input.

[0010] The numerical and resolution-related problems associated with estimation methods based solely on speed sensor position (tooth count) measurements remain limited to higher speed operation and are susceptible to various noise factors: synchronization, low-speed resolution, engine and road disturbances, metrological and numerical errors, to name a few.

[0011] Based on this situation, the object of the present invention is to develop a backlash detection system using shaft torque that overcomes the disadvantages of the prior art. In particular, a device is to be provided that generates signals indicating the magnitude of the transmitted torque and the direction of rotation of the transmitted torque. SUMMARY OF THE INVENTION

[0012] According to the invention, this object is achieved by independent claim 1. Advantageous embodiments are specified in the dependent claims.

[0013] A method for detecting lash and adjusting driveline variables includes measuring torque transmitted between components across a torque discontinuity, measuring driveline twist across the discontinuity, using a ratio of driveline twist divided by torque and an inverse function of driveline stiffness to determine whether the driveline is entering, exiting, or in a lash zone, and using a measured driveline twist and a measured driveline torque in the lash zone to set reference values ​​for driveline twist and driveline torque.

[0014] The method avoids the numerical and resolution-related problems associated with estimation methods based solely on speed sensor position measurements, high-speed operation, and susceptibility to various noise factors, including synchronization, low-speed resolution, engine and road disturbances, and measurement errors.

[0015] The scope of applicability of the preferred embodiment will become apparent from the following detailed description, claims, and drawings. It should be understood that the description and specific examples are for illustrative purposes only, while indicating preferred embodiments of the invention. Various changes and modifications to the described embodiments and examples will become apparent to those skilled in the art. DESCRIPTION OF THE DRAWINGS

[0016] The invention will be better understood with reference to the following description and the accompanying drawings, in which: Fig. Figure 1 is a schematic diagram illustrating two shafts of a drive train with deadband; Fig. Figure 2 is an end view of meshing gear teeth showing backlash between the gears; Fig. Figure 3 is a diagram showing the change in driveline torque and twist from negative to positive twist, including discontinuities where entry into and exit from a clearance occur; Fig. Figure 4 is a schematic diagram of an electronic control unit for controlling the torque converter clutch; Fig. 5 is an algorithm for clearance crossing and size detection using shaft torque and twist measurements; and Fig. 6 Time curves of a backlash crossing detection using the relationship between a rough estimate of the twist and the measured shaft torque. DESCRIPTION OF THE PREFERRED EMBODIMENT

[0017] Fig. Figure 1 shows a drive train 10 with two shafts 12, 14, which are connected by a torque discontinuity 16, in which play can occur. The discontinuity 16 is to be formed by the meshing gears 18, 20 of Fig. 2, with each gear attached to one of the shafts 12, 14. A production-grade magneto-elastic shaft torque sensor 24, capable of measuring torque directly at a transmission input or output shaft, enables numerous vehicle control functions. The dynamic response of the shaft torque sensor 24 occurs in a fraction of a millisecond and far exceeds the accuracy of currently available on-board torque estimates. Consequently, this torque measurement can enable reliable backlash detection as well as useful feedback for automotive drivability control.

[0018] Fig. 3 is a diagram showing the change of the driveline torque ┬ and the twist θ from a negative torsion angle to a positive torsion angle, including discontinuities θn, θp upon entering or exiting a clearance zone 26, i.e. where clearance crossing occurs.

[0019] Theoretically, a zero-crossing torque measurement on a shaft close to the main source of deadband, which typically occurs at the differential, should directly indicate lash crossing. In practice, however, several sources of uncertainty exist that can be eliminated by combining information from the torque sensor and speed sensor signals. An offset in the speed sensor reading, either due to measurement noise or a measurement location on the driveline, can cause lash crossing to occur at a non-zero torque signal. Using the tooth count information from the speed sensors in the speed range around the zero value can help accurately define lash crossing and, in turn, zero the torque measurement offset.

[0020] The discontinuity of the flank Ks, as in Fig. 1, can be used to determine when the driveline is in the contact-free zone. More specifically, the torque-twist relationship is as follows: Ks(θ−θp),θ>θp>0 τ=0,θn<θ<θp,|θn|=|θp| Ks(θ−θn),θ<θn<0 where: θ is the measured angle of one side of the torque discontinuity with respect to the other side, θ p the reference play angle is in the positive direction, θ n the reference play angle is in the negative direction, and τ is the torque on the shaft element.

[0021] When the drive train is in dead zone 26, the flank Ks is zero.

[0022] Fig. Figure 4 illustrates a control unit 50 having an electronic microprocessor 54 having access to an electronic memory 56 containing control algorithms, a data communications bus 58 interconnecting components of the control unit, inputs 60 for communication of the control unit with various powertrain sensors, and outputs 62 transmitting signals from the control unit in response to the results generated by the execution of the algorithms. The RAM memory component 56 contains reference data associated with the torque sensor 24, including ε, and reference data related to the driveline 10, including Ks, θp, θn, a signal 55 representing a torque ┬ measured at the transmission input or output, and the transmission angular displacement or twist θ1, θ2. The input signals to the control unit 50 include a wheel angular displacement or twist θ3, typically transmitted from the ABS module.

[0023] There are many ways this can be used to determine the play zone by measuring shaft torque and shaft twist based on the measured position count of the differential output shaft speed and wheel speed sensors.

[0024] The most direct way to determine that the driveline is in the zero-contact (backlash) range is to observe that the torque measurement (τ) is constant while the twist measurement (or estimate) value (θ) changes. Note that using only the torque signal to detect when the backlash is at zero crossing may not be sufficient because the signal may contain offset or measurement noise.

[0025] One way would be to calculate the ratio (τ / θ) and, if it is close to 0, conclude that the driveline is in the clearance region.

[0026] Another method, less susceptible to numerical errors and other noise factors, is the use of the inverse ratio (θ / τ). This ratio becomes significant as the driveline enters the non-contact region because the measured torque is close to zero. An algorithm 30 of this method is described with reference to Fig. 5. In step 32, a test is performed to determine whether the measured torque |τ| < ε, where ε is the measurement error range specified for the torque sensor 24 in the environment where the torque sensor indicates that zero torque is being transmitted in the driveline 10. ε is the maximum allowable torque measurement error of the torque sensor 24 for a zero-torque measurement where backlash crossover occurs. For example, a torque sensor specification may have a tolerance ε of + / -5 Nm for torque values ​​up to 100 Nm and 5% error for larger torques.

[0027] If the result of test 32 is false, control proceeds to step 44, which is explained below.

[0028] If the result of test 32 is true, a test is performed at step 36 to determine whether |θ / τ| > δ, where δ is the edge detection threshold, which must be calibrated accordingly with Ks. Since τ = Ks θ, θ / τ = 1 / Ks. Therefore, the angular displacement threshold δ is related to 1 / Ks. To ensure lower susceptibility, |θ / τ| should be calibrated as a scalar larger than the estimated 1 / Ks. Crucially, in the region where backlash traversal occurs, the torque τ is approximately zero; dividing the angular displacement θ by the torque ┬ therefore results in a very large number for θ / τ.

[0029] Therefore, if the result of test 36 is true, drive train 10 operates in clearance zone 26. Step 36 can be made less vulnerable by, for example, also checking | θ / (τ-ε) | > δ, | θ / (τ+ε) | > δ for more robust detection in noise.

[0030] In step 38, a test is performed to determine whether the backlash flag is zero. If the backlash flag is zero, no change in the backlash state has occurred since the last execution of algorithm 30. In step 40, the backlash flag is set to 1 to indicate entry into backlash zone 26, and in step 46, the backlash flag is reset to 0 to indicate exiting the backlash zone. Since driveline 10 is in the backlash zone according to test 36, and since the driveline exited the backlash zone during the last execution of the algorithm according to the result of test 38, step 40 refers to the driveline just entering backlash zone 26.

[0031] It is helpful to determine whether driveline 10 is entering or exiting game traversal game zone 26. If step 36 indicates that driveline 10 is currently operating in game zone 26, different logic paths are subsequently taken depending on whether a transition occurs or the previous game state continues.

[0032] If the results of steps 36 and 38 are both true, indicating that the drive train is between θ p and θ n enters the play zone, in step 40, the offset (if any) of the driveline angular displacement can be eliminated either (i) by immediately resetting it to 0, (ii) by filtering it to 0, or (iii) by gradual adjustment over time. For example, if θ is expected p and θ n occur at 3 degrees, but the following statements of algorithm 30 indicate that θ p or θ nat 5, 6, 5, 4 degrees, the currently measured shaft twist θ can be compared with the stored model values ​​θp and θ n be compared, and θp and θn can be set in the memory 56 in accordance with the measured twist θ.

[0033] At step 40, the game flag is set to 1 to indicate that the game state has changed since the last execution of algorithm 30. After executing step 40, control proceeds to step 34, where execution of algorithm 30 is terminated.

[0034] If the result of step 36 is true and the result of step 38 is false, indicating that the driveline 10 is operating in the slack zone 26, the torque offset is adjusted at step 42, i.e., the memory 56 is updated to account for the offset. After step 36 indicates that the driveline is operating in the slack zone 26, the torque measurement provided by the torque sensor 24 should be "0." If this measurement is not zero, the measured torque is the torque offset or torque error, which can be used incrementally or immediately corrected in the torque error memory 56.

[0035] After executing step 42, control proceeds to step 34, where the execution of the algorithm is terminated.

[0036] If the result of test 36 is false, indicating that torque and twist are low and the driveline 10 is not operating in the backlash region 26, a test is performed at step 44 to determine whether the backlash flag is equal to 1. If the result of test 44 is false, control proceeds to step 34, where execution of the algorithm terminates.

[0037] If the result of test 44 is true, the backlash flag of zero indicates that the backlash state has not changed since the flag was set after entering backlash zone 26 during the last execution of algorithm 30. Consequently, the driveline leaves backlash zone 26. At step 46, the offset (if any) of the driveline twist θ can be removed, i.e., eliminated either (i) by immediately resetting the twist to 0, (ii) by filtering it to 0, or (iii) by gradual adjustment over time. For example, if θ p and θ nshould be expected to occur at 3 degrees, but the following statements of algorithm 30 indicate that θ p or θ n If the shaft twists occur at 5, 6, 5, and 4 degrees, the currently measured shaft twist θ can be compared with the stored model values ​​θp and θn, and θp and θn can be adjusted in memory 56 in accordance with the measured twist θ. At step 46, the backlash flag is reset to 0.

[0038] At step 48, the driveline twist gap |θp - θn| can be determined for the current gear ratio in which the transmission is operating. More specifically, θp and θn can be determined, and the deviation from the driveline twist estimate or measurement is eliminated after recognizing that θp and θn are symmetric about 0. After leaving the play zone 26, the driveline traverses θp or θn, as shown in Fig. 3. If the measured angle at the transitions θp or θn does not match the expected values ​​of θp and θn stored in memory 56, these values ​​can be adjusted. For example, if θp is 4.0 degrees and θn is 2.0 degrees, |θp - θn| is set to 3.0 degrees, i.e., they converge to the corrected center value of 3.0 degrees.

[0039] The time curves in Fig. 6 shows the shaft torque measurement 70, the backlash detection flag (top) 72, a rough estimate of the driveline twist 74, the signal 76 resulting from dividing the estimated driveline twist 74 by the torque measurement 70, and the limit value δ 78.

[0040] Since vehicle data indicative of the torsion angle signal generated by the driveline shaft speed sensors is not available, the estimated shaft twist 74 is obtained by coarse integration of a wheel speed sensor and the transmission output shaft speed sensor signals.

[0041] The preferred embodiment has been described in accordance with the provisions of the patent statutes. However, it should be understood that alternative embodiments may be practiced otherwise than as specifically illustrated and described.

Claims

[1] A method for detecting backlash, comprising: (a) Measuring torque transmitted between components across a torsional discontinuity, (b) measuring a driveline twist across said discontinuity, (c) Using a ratio of driveline twist divided by torque and a driveline stiffness to determine whether the driveline is entering, exiting, or in a play zone. [2] The method of claim 1, further comprising: Using measured driveline twist and driveline torque in the clearance zone to set reference driveline twist and driveline torque values. [3] The method of claim 1, wherein step (a) further comprises: Using a torque sensor to measure torque transmitted between components across the torsional discontinuity. [4] The method of claim 1, wherein step (a) further comprises: Using a torque sensor to measure torque transmitted between components across the torsional discontinuity, and Determine that an absolute magnitude of torque is less than a measurement error tolerance range specified for the torque sensor when the torque sensor indicates that the torque is zero. [5] The method of claim 1, wherein step (b) further comprises: Using angular displacement sensors placed on opposite sides of the torsional discontinuity to measure driveline twist across said discontinuity. [6] The method of claim 1, wherein step (c) further comprises: Using an absolute value of a ratio of driveline twist divided by torque and an inverse function of driveline stiffness to determine whether the driveline is entering, exiting, or in a play zone. [7] The method of claim 1, wherein steps (a) and (c) further comprise: Using a torque sensor to measure a torque transmitted between components across the torsional discontinuity, and Determine that an absolute value of a ratio of driveline twist divided by a difference between the torque and a torque sensor specified Measurement error tolerance range is smaller than an inverse function of the driveline stiffness when the torque sensor indicates that zero torque is being transmitted. [8] The method of claim 1, further comprising: Using the driveline twist measured when the driveline enters the play zone to set a reference driveline twist corresponding to the driveline entering the play zone. [9] The method of claim 1, further comprising using the driveline torque measured during operation of the driveline in the play zone to set a reference driveline torque corresponding to operation of the driveline in the play zone. [10] The method of claim 1, further comprising using the driveline twist measured upon exit of the driveline from the play zone to set a reference driveline twist corresponding to exit of the driveline from the play zone. [11] The method of claim 1, further comprising using an absolute magnitude of a measured driveline twist difference upon exit of the driveline from the play zone to set a reference driveline twist difference corresponding to an exit of the driveline from the play zone.

Images