Estimation system for rack force of steer-by-wire system

Abstract

The present disclosure provides an estimation system for a rack force of a steer-by-wire system. The estimation system comprises a signal preprocessing module, a rack force estimation module, and an estimation result post-processing module. In any traveling situation, a generalized rack force estimated by said estimation system is kept highly consistent with a tie rod force collected by a sensor, and the estimation is rapid, thereby making the experience in road sense feedback close to the EPS characteristics.

Description

A rack force estimation system for steer-by-wire systems

[0001] This disclosure claims the benefit of Chinese patent application No. 202410064942X, filed on January 17, 2024. The entire text of the aforementioned Chinese patent application is incorporated herein by reference. Technical Field

[0002] The present disclosure relates to the technical field of steering systems, and in particular to a rack force estimation system for a wire-controlled steering system. Background Art

[0003] Steer-by-wire (SBW) systems eliminate the mechanical connection between the steering wheel and the steering actuator, thereby decoupling the upper and lower steering mechanisms. This helps improve the chassis responsiveness of autonomous driving systems, saves space, and facilitates features such as variable gear ratios that are unavailable with traditional Electric Power Steering (EPS). However, this mechanical decoupling prevents direct road feedback from being transmitted directly to the steering wheel, hindering the driver's perception of road conditions. This is detrimental to driving, and thus, ensuring real-time road feel feedback becomes a challenge.

[0004] Road feel is the most important means for drivers to perceive the vehicle's motion and environment, second only to vision. It is an important supplement to vision. There are mainly the following ways to provide road feel feedback:

[0005] 1. Sensor measurement method: A force sensor is installed on the steering rod to obtain the rod force, and then obtain the road feel feedback torque. However, this method adds additional costs and places new requirements on mechanical structure and reliability.

[0006] 2. Data Fitting: This method uses data fitting of vehicle state and parameter information to generate a nonlinear function that describes road feel. The fitting coefficients are then modified based on the road feel requirements of different drivers to simulate the driver's preferred road feel. This method allows for flexible design of road feel, but requires extensive data fitting, and the resulting road feel feedback often doesn't match the EPS system's road feel.

[0007] 3. Dynamic Model Estimation: This method starts with a physical model, constructing the steering system and performing estimations using mathematical models. Two common approaches are: One involves building a tire model to estimate the tire's self-aligning torque, which incorporates road surface information and can be used as road feel feedback; the other involves building a rack model to estimate the rack force, which also incorporates road surface information. The tire models in both approaches are overly complex, and tires are highly nonlinear, making them unsuitable for practical engineering applications.

[0008] To obtain realistic road feel and facilitate implementation, estimating rack force is a superior option. However, existing rack force estimation methods perform poorly on rough roads (such as cobblestones and speed bumps). Therefore, it is necessary to develop a highly robust, generalizable, and fast estimation system.

[0009] Summary of the Invention

[0010] In order to overcome the shortcomings of the existing technology, the present disclosure provides an estimation system for the rack force of a wire-controlled steering system. Based on precise modeling, the Kalman filter algorithm is used to set correct parameters, so that the algorithm converges quickly and achieves accurate and timely estimation under any driving conditions.

[0011] To achieve the above objectives, a rack force estimation system for a steer-by-wire system is designed, which includes a signal preprocessing module, a rack force estimation module, and an estimation result post-processing module.

[0012] The signal preprocessing module includes a rack displacement signal processing unit, a rack speed signal processing unit and a motor torque signal processing unit;

[0013] The rack displacement signal processing unit is used to convert the steering motor rotation angle into rack displacement;

[0014] The rack speed signal processing unit is used to convert the rotation speed of the steering motor into the rack speed;

[0015] The motor torque signal processing unit is used to convert the motor torque into rack driving force;

[0016] The rack force estimation module includes a steering system dynamics model unit and a Kalman filter estimation unit;

[0017] The steering system dynamics model unit is used to establish an equivalent system dynamics model;

[0018] The Kalman filter estimation unit derives a state space equation based on the equivalent system dynamics model, and expands the rack force to be estimated into a corresponding state quantity based on the state space equation;

[0019] The estimation result post-processing module includes an estimation result filtering processing unit and an estimation result phase compensation processing unit;

[0020] The estimation result filtering processing unit is used to select different filtering coefficients for filtering processing based on the first logic strategy;

[0021] The estimation result phase compensation processing unit is used to select different phase compensation parameters based on the second logic strategy to perform phase compensation processing.

[0022] Preferably, the equivalent system dynamics model is to regard the steering motor, reduction mechanism, rack mechanism and steering column of the steer-by-wire system as an equivalent rack system, and the equivalent dynamic differential equation of the equivalent rack system is:

[0023] Where M is the equivalent system mass, B is the equivalent system damping coefficient, T m is the motor torque, F f is the equivalent system friction, F r is the force exerted by the pull rod on the rack, g m is the transmission ratio of the motor reduction mechanism, r pg is the transmission ratio from the rack to the motor reduction mechanism, x r is the rack displacement, is the rack speed, Rack acceleration.

[0024] Preferably, the steering system dynamics model unit is further used to convert the equivalent system friction force F in the equivalent dynamics differential equation into f and the force F exerted by the pull rod on the rack r The sum of is a generalized rack force F R , where: F R =F r +F f ;

[0025] Based on the generalized rack force F R The equivalent system dynamics differential equation becomes:

[0026] Rewrite the equivalent dynamic differential equation into the corresponding state space equation:

[0027] Among them, x r is the first state variable of the equivalent rack system and is the second state variable of the equivalent rack system.

[0028] Preferably, the Kalman filter estimation unit is also used for reconstructing the state space equation, including:

[0029] The generalized rack force F R The third state variable F of the equivalent rack system is defined as R ;

[0030] Assumptions k is a zero or non-zero constant, and the state space equation is reconstructed as:

[0031] The reconstructed state-space equation is discretized to obtain the discrete state-space equation:

[0032] Among them, Δ t is the sampling time of the rack force estimation module, I is the identity matrix;

[0033] Based on the Kalman filter design, the discrete state space equation is simplified into the following form:

[0034] in,

[0035] The following estimation iterative formula is designed based on the Kalman filter principle:

[0036] Among them, Q and R are diagonal matrices, the initial value of P is a diagonal matrix, H is the measurement matrix, and Z k is the actual measured value.

[0037] Preferably, the true measurement value Z k The rack displacement output by the signal preprocessing module The measurement matrix H is (1 0 0).

[0038] Preferably, when the equation When k is a non-zero constant, the value range of k is (0, 2000), and the estimation result of the generalized rack force in the Kalman filter estimation unit is multiplied by the non-zero constant k1 to obtain the true generalized rack force, and the value range of the non-zero constant k1 is (-10000, 10000).

[0039] Preferably, the rack displacement signal processing unit is further configured to subject the rack displacement signal to amplitude limiting processing first, and then to first-order low-pass filtering processing, to obtain a target rack displacement signal;

[0040] The rack speed signal processing unit is further configured to process the rack speed signal by firstly performing amplitude limiting processing and then performing first-order low-pass filtering processing to obtain a target rack speed signal;

[0041] The motor torque signal processing unit is further configured to process the motor torque signal by firstly subjecting it to amplitude limiting processing and then subjecting it to first-order low-pass filtering processing to obtain a target motor torque signal.

[0042] Preferably, the estimation result filtering processing unit is further configured to perform first-order low-pass filtering on the estimated generalized rack force, and select the corresponding filtering coefficient according to the rack velocity, rack acceleration, steering wheel angle and steering wheel angular velocity.

[0043] Preferably, the estimation result phase compensation processing unit is further configured to perform phase advance processing on the estimated generalized rack force, and select the phase compensation parameter according to the rack velocity and the rack acceleration, including:

[0044] A two-dimensional lookup table is obtained according to the rack speed and the rack acceleration, wherein the rack speed or the rack acceleration in the two-dimensional lookup table is positively correlated with the phase compensation parameter.

[0045] Preferably, the Kalman filter estimation unit is further used to estimate the rack displacement and the rack speed.

[0046] Compared with the prior art, the present disclosure provides a rack force estimation method for a wire-controlled steering system. Based on precise modeling and utilizing a Kalman filter algorithm, correct parameters are set to enable the algorithm to converge quickly and achieve accurate and timely estimation under any driving conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] FIG1 is a block diagram of a rack force estimation system of a steer-by-wire system according to the present disclosure.

[0048] FIG2 is a schematic structural diagram of a steering actuator module disclosed herein.

[0049] FIG3 is a schematic diagram of an equivalent system of a steering actuator module according to the present disclosure.

[0050] FIG4 is a comparison diagram of rack force estimation results disclosed herein.

[0051] FIG5 is a comparison diagram of rack force estimation results disclosed herein. DETAILED DESCRIPTION

[0052] Example 1

[0053] The present disclosure is further described below with reference to the accompanying drawings.

[0054] As shown in Figure 1, a rack force estimation system for a steer-by-wire system includes a signal preprocessing module, a rack force estimation module, and an estimation result post-processing module;

[0055] The signal preprocessing module includes a rack displacement signal processing unit, a rack speed signal processing unit and a motor torque signal processing unit;

[0056] The rack displacement signal processing unit is used to convert the steering motor rotation angle into rack displacement;

[0057] The rack speed signal processing unit is used to convert the rotation speed of the steering motor into the rack speed;

[0058] The motor torque signal processing unit is used to convert the motor torque into rack driving force;

[0059] The rack force estimation module includes a steering system dynamics model unit and a Kalman filter estimation unit;

[0060] The steering system dynamics model unit is used to establish an equivalent system dynamics model;

[0061] The Kalman filter estimation unit derives a state space equation based on the equivalent system dynamics model, and expands the rack force to be estimated into a corresponding state quantity based on the state space equation;

[0062] The estimation result post-processing module includes an estimation result filtering processing unit and an estimation result phase compensation processing unit;

[0063] The estimation result filtering processing unit is used to select different filtering coefficients for filtering processing based on the first logic strategy;

[0064] The estimation result phase compensation processing unit is used to select different phase compensation parameters based on the second logic strategy to perform phase compensation processing.

[0065] As shown in Figure 2, it is a structural diagram of the steering actuator module, which includes a steering motor, a reduction mechanism, a rack mechanism, and a steering column. Since these four parts are mechanically connected, the equivalent system dynamics model regards the steering motor, reduction mechanism, rack mechanism, and steering column under steer-by-wire as an equivalent rack system. The equivalent system of the steering actuator module is shown in Figure 3. r is the force exerted by the pull rod on the equivalent rack, a is the equivalent rack acceleration, v is the equivalent rack velocity, F f is the equivalent rack friction force, F b is the equivalent rack damping force, F t is the motor driving force, and the equivalent dynamic differential equation of the equivalent rack system is:

[0066] Where M is the equivalent system mass, B is the equivalent system damping coefficient, T m is the motor torque, F f is the equivalent system friction, F r is the force exerted by the pull rod on the rack, g m is the transmission ratio of the motor reduction mechanism, r pg is the transmission ratio from the rack to the motor reduction mechanism, x r is the rack displacement, is the rack speed, Rack acceleration.

[0067] As an implementable manner, the steering system dynamics model unit is further used to convert the equivalent system friction force F in the equivalent dynamics differential equation intof and the force F exerted by the pull rod on the rack r The sum of is a generalized rack force F R ,in:

[0068] F R , F R =F r +F f ;

[0069] Based on the generalized rack force F R The equivalent system dynamics differential equation becomes:

[0070] Rewrite the equivalent dynamic differential equation into the corresponding state space equation:

[0071] Among them, x r is the first state variable of the equivalent rack system and is the second state variable of the equivalent rack system.

[0072] As an implementable manner, the Kalman filter estimation unit is also used for reconstructing the state space equation, including:

[0073] The generalized rack force F R The third state variable F of the equivalent rack system is defined as R ;

[0074] Assumptions k is a zero or non-zero constant, and the state space equation is reconstructed as:

[0075] The reconstructed state-space equation is discretized to obtain the discrete state-space equation:

[0076] Among them, Δ t is the sampling time of the rack force estimation module, I is the identity matrix;

[0077] Based on the Kalman filter design, the discrete state space equation is simplified into the following form:

[0078] in,

[0079] The Kalman filter algorithm principle includes two parts: time update and measurement update. Based on the Kalman filter principle, the following estimation iterative formula is designed:

[0080] Among them, Q and R are diagonal matrices, the initial value of P is a diagonal matrix, H is the measurement matrix, and Z k is the actual measured value.

[0081] As a practical approach, the true measurement value Z k The rack displacement signal output by the signal preprocessing module The measurement matrix H is (1 0 0).

[0082] As a feasible approach, when the equation When k is a non-zero constant, the value range of k is (0, 2000), and the estimated result of the generalized rack force in the Kalman filter estimation method is multiplied by the non-zero constant k1 to obtain the true generalized rack force. The value range of the non-zero constant k1 is (-10000, 10000), and the specific values ​​of k and k1 are specific values ​​after debugging through simulation tests.

[0083] As an implementable manner, the rack displacement signal processing unit is further configured to first subject the rack displacement signal to amplitude limiting processing, and then subject the rack displacement signal to first-order low-pass filtering processing to obtain a target rack displacement signal;

[0084] The rack speed signal processing unit is further configured to process the rack speed signal by firstly performing amplitude limiting processing and then performing first-order low-pass filtering processing to obtain a target rack speed signal;

[0085] The motor torque signal processing unit is further configured to first subject the motor torque signal to amplitude limiting processing and then to first-order low-pass filtering to obtain a target motor torque signal. As an achievable approach, the estimation result filtering unit is further configured to apply first-order low-pass filtering to the estimated generalized rack force, selecting the corresponding filter coefficient based on the rack velocity, rack acceleration, steering wheel angle, and steering wheel angular velocity.

[0086] As an implementable approach, the estimation result phase compensation processing unit is further configured to perform phase advance processing on the estimated generalized rack force, and select the phase compensation parameter according to the rack velocity and the rack acceleration, including:

[0087] A two-dimensional lookup table is obtained according to the rack speed and the rack acceleration, wherein the rack speed or the rack acceleration in the two-dimensional lookup table is positively correlated with the phase compensation parameter.

[0088] As an implementable manner, the Kalman filter estimation unit is further configured to estimate the rack displacement and the rack velocity.

[0089] The present invention utilizes the Kalman filter algorithm to perform real-time estimation of the rack force. Due to the use of the Kalman filter algorithm, real-time optimization of the gain is achieved, thereby achieving the goals of high robustness and rapid estimation. The estimation performance on bad roads is excellent, and real road feel feedback can be achieved under any driving conditions.

[0090] The present disclosure can not only estimate the rack force, but also estimate the rack displacement and rack speed. At the same time, the estimated rack force can be used for control optimization of the angle control of the wire-controlled steering system.

[0091] The present disclosure can also be used for rack force estimation in an electric power steering (EPS) system, which is of great significance for torque closed loop and feel optimization of the EPS system.

[0092] As shown in Figure 4, the horizontal axis is time, and the unit of the horizontal axis is s (seconds), and the vertical axis is the rack force, and the unit of the vertical axis is N (Newton). It is a comparison chart of the generalized rack force estimated by the present invention and the tie rod force collected by the sensor when driving on an asphalt road. The dotted line is the generalized rack force estimated by the present invention, and the solid line is the tie rod force collected by the sensor. It can be seen from the figure that the generalized rack force estimated by the present invention and the tie rod force are highly consistent and have the same trend.

[0093] As shown in Figure 5, the horizontal axis is time, the unit of the horizontal axis is s, the vertical axis is rack force, the unit of the vertical axis is N, Figure 5 is a comparison diagram of the estimated generalized rack force and the pull rod force collected by the sensor when the present invention is driving on a cobblestone road, wherein the dotted line is the generalized rack force estimated by the present invention, and the solid line is the pull rod force collected by the sensor; it can be seen that on a bad road, the generalized rack force estimated by the present invention is still highly consistent with the pull rod force, which indicates that the present invention has strong robustness and generalization.

[0094] In actual vehicle tests, the generalized rack force estimated by the present invention maintains a high degree of consistency with the pull rod force collected by the sensor under any driving conditions, and the estimation is fast, and the experience in road feel feedback is close to EPS characteristics.

[0095] Although the specific embodiments of the present disclosure have been described above, those skilled in the art will appreciate that these are merely illustrative and that various changes or modifications may be made to these embodiments without departing from the principles and essence of the present disclosure. Therefore, the scope of protection of the present disclosure is defined by the appended claims.

Claims

1. A system for estimating the rack force of a steer-by-wire system, characterized in that: The estimation system includes a signal preprocessing module, a rack force estimation module, and an estimation result postprocessing module; The signal preprocessing module includes a rack displacement signal processing unit, a rack speed signal processing unit, and a motor torque signal processing unit; The rack displacement signal processing unit is used to convert the rotation angle of the steering motor into a rack displacement; The rack speed signal processing unit is used to convert the rotation speed of the steering motor into a rack speed; The rack force estimation module includes a steering system dynamics model unit and a Kalman filter estimation unit; The steering system dynamics model unit is used to establish an equivalent system dynamics model; The Kalman filter estimation unit derives a state space equation based on the equivalent system dynamics model, and expands the rack force to be estimated into corresponding state variables based on the state space equation; The estimation result postprocessing module includes an estimation result filtering processing unit and an estimation result phase compensation processing unit; The estimation result filtering processing unit is used to select different filtering coefficients based on a first logic strategy for filtering processing; The estimation result phase compensation processing unit is used to select different phase compensation parameters based on a second logic strategy for phase compensation processing.

2. The rack force estimation system for a steer-by-wire system according to claim 1, wherein: The equivalent system dynamic model takes the steering motor, reduction mechanism, rack mechanism, and steering column under steer-by-wire as an equivalent rack system, and the equivalent dynamic differential equation of the equivalent rack system is as follows: where M is the equivalent system mass, B is the equivalent system damping coefficient, T m is the motor torque, F f is the equivalent system friction force, F r is the force exerted by the tie rod on the rack, g m is the transmission ratio of the motor reduction mechanism, r pg is the transmission ratio from the rack to the motor reduction mechanism, x r is the rack displacement, is the rack speed, is the tooth rack acceleration.

3. The rack force estimation system of a steer-by-wire system according to claim 2, wherein: The steering system dynamics model unit is also used to take the sum of the equivalent system friction force F f and the force F r exerted by the tie rod on the rack as a generalized rack force F R , where: F R = F r + F f ; Based on the generalized rack force F R The dynamic differential equation of the equivalent system is changed to: Rewrite the equivalent dynamic differential equation into a corresponding state - space equation: where x r is the first state variable of the equivalent rack system and is the second state variable of the equivalent rack system, and y1 and y2 are the output variables of the state space equation, corresponding to x respectively r , 4. The estimation system for rack force of a steer-by-wire system according to claim 2, wherein: The Kalman filter estimation unit is further used to reconstruct the state space equation, including: Define the generalized rack force F R as the third state variable of the equivalent rack system Hypothesis k is zero or a non-zero constant, and the state space equation is reconstructed as: Among them, y3 is the newly added output of the state space equation, corresponding to F R ; Discretize the reconstructed state-space equation to obtain the discrete form of the state-space equation as follows: where Δ t is the sampling time of the rack force estimation module, and I is the identity matrix; Based on the Kalman filter design, the discrete-form state space equation is simplified into the following form: where x k+1 is the state variable at the next moment, x k is the state variable at the previous moment, u k is the system input, y k+1 is the output for the next moment, G is the state matrix, F is the input matrix, C is the output matrix, Based on the Kalman filter principle, the following estimation iteration formula is designed: Among them, Q and R are diagonal matrices, P pre is the predicted covariance matrix, P is the corrected error covariance matrix, the initial value of P is a diagonal matrix, H is the measurement matrix, Z k is the true sensor measurement value, H′ is the transpose matrix of the H matrix, K is the Kalman gain, G′ is the transpose matrix of the G matrix, x′ k+1 is the estimated value of the state variable, x k+1 is the corrected value of the state variable.

5. An estimation system for the rack force of a steer-by-wire system, according to claim 4, characterized in that: the true measurement value Z k is the rack displacement output by the signal preprocessing module The measurement matrix H is (1 0 0).

6. The rack force estimation system of a steer-by-wire system according to claim 4, wherein: When the equation When k is a non-zero constant, the value range of k is (0, 2000), and the estimation result of the generalized rack force in the Kalman filter estimation unit is multiplied by a non-zero constant k1 to obtain the true generalized rack force, and the value range of the non-zero constant k1 is (-10000, 10000).

7. An estimation system for the rack force of a steer-by-wire system according to claim 1, characterized in that: The rack displacement signal processing unit is further used to first perform amplitude limit processing on the rack displacement signal, and then perform first-order low-pass filtering processing to obtain a target rack displacement signal; The rack speed signal processing unit is further used to first perform amplitude limit processing on the rack speed signal, and then perform first-order low-pass filtering processing to obtain a target rack speed signal; The motor torque signal processing unit is further used to first perform amplitude limit processing on the motor torque signal, and then perform first-order low-pass filtering processing to obtain a target motor torque signal.

8. The estimation system for the rack force of a steer-by-wire system according to claim 3, characterized in that: The estimation result filtering processing unit is further used to perform first-order low-pass filtering processing on the estimated generalized rack force, and select the corresponding filtering coefficient according to the rack speed, rack acceleration, steering wheel angle, and steering wheel angular velocity.

9. The rack force estimation system of a steer-by-wire system according to claim 3, wherein: The estimation result phase compensation processing unit is further used to perform phase lead processing on the estimated generalized rack force, and select the phase compensation parameter according to the rack speed and rack acceleration, including: A two-dimensional look-up table is obtained according to the rack speed and the rack acceleration, and the rack speed or the rack acceleration in the two-dimensional look-up table is positively correlated with the phase compensation parameter.

10. The rack force estimation system of a steer-by-wire system according to claim 4, wherein: The Kalman filter estimation unit is further configured to estimate the rack displacement and the rack speed.

Images