Electric power steering fault diagnosis and self-adaption system based on torque offset

Abstract

The invention discloses an electric power steering fault diagnosis and self-adaption system based on torque offset, which relates to the technical field of fault diagnosis and comprises a data acquisition unit, a model establishment unit, a torque offset evaluation unit, a fault judgment unit and a power output adjustment unit. According to the method, the return-to-power torque model is constructed based on the vehicle dynamics law and experience adjustment, the model parameters are optimized to obtain the return-to-power torque model, meanwhile, the real-time output torque of the EPS system is calculated according to the steering wheel torque signal, the deviation coefficient of the EPS system is comprehensively calculated, and the accuracy of the EPS system is improved. The deviation coefficient is evaluated according to a preset fault grading interval, a multi-grade fault judgment mechanism is established according to the change rule of the deviation coefficient, dynamic health evaluation of the EPS system motor performance is achieved, the evaluation serves as the basis of control strategy adjustment, the output assistance value can be dynamically adjusted, return control can be enhanced or a safety mode can be switched, and the reliability of the EPS system motor performance is improved. Therefore, the redundancy control capability and the driving safety are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of fault diagnosis, and in particular to an electric power steering fault diagnosis and adaptive system based on torque offset. Background Art

[0002] Existing electric power steering systems (EPS) generally adopt a control architecture with torque sensors, vehicle speed sensors and motor drivers as the core. The control strategy is mainly based on a preset power assist mapping table or model, combined with driver input and vehicle speed information, to control the motor to provide power assist torque of corresponding size and direction. At the same time, the system usually has a basic self-diagnosis module embedded in it to monitor the status of key components such as the motor driver, power supply, current sensor, etc., such as typical faults such as overvoltage, overcurrent, and short circuit. In addition, some EPS systems have a return torque compensation function to improve the vehicle's return performance and steering stability. This function applies a certain return torque compensation to the motor based on input parameters such as steering wheel angle and vehicle speed. However, this compensation model is usually a static, independent module that only serves dynamic handling performance and is not used in fault monitoring or power assist adjustment strategies;

[0003] The existing EPS system has the following main problems:

[0004] 1. Self-diagnosis is limited to electrical parameters: Current self-diagnosis mainly identifies typical hardware faults based on motor voltage, current, resistance and other parameters, and cannot effectively identify performance degradation problems such as partial winding failure, response delay caused by temperature rise, and increased transmission gap.

[0005] 2. Separation of diagnosis and power-assistance control strategies: The system usually triggers a fixed degradation mode after fault diagnosis and cannot make fine-grained adjustments based on different degrees of performance degradation, which is not conducive to driving experience and safety assurance.

[0006] 3. Failure to use the return torque model for health assessment: The return torque model is only used for control purposes and is not used to analyze the deviation between actual output and predicted values, thus failing to form an effective dynamic health assessment mechanism.

[0007] 4. Lack of a diagnostic mechanism that integrates model-driven and data-driven approaches: Traditional methods fail to integrate predictive models with actual sensor data, failing to capture soft fault symptoms such as torque offset trends. Consequently, diagnostic granularity and foresight are limited.

[0008] In view of the above technical defects, a solution is now proposed. Summary of the Invention

[0009] The purpose of the present invention is to achieve a dynamic health assessment of the EPS system motor performance as a basis for adjusting the control strategy, dynamically adjust the output power value, strengthen the return control or switch to the safety mode, thereby improving the redundant control capability and driving safety.

[0010] To achieve the above objectives, the present invention adopts the following technical solution: an electric power steering fault diagnosis and adaptation system based on torque offset, comprising a data acquisition unit, a model building unit, a torque offset evaluation unit, a fault judgment unit, and a power output adjustment unit, wherein:

[0011] The data acquisition unit is used to obtain real-time sensor data during the vehicle's driving process through the sensor group, eliminate invalid data in the real-time sensor data, classify it, and send it to the torque offset evaluation unit and the power output adjustment unit according to the classification results;

[0012] The model building unit is used to obtain real-time sensor data and build a self-centering assist torque model based on vehicle dynamics and empirical adjustments. The real-time sensor data is used as training samples to train the self-centering assist torque model. The ideal self-centering assist torque that should be output under ideal conditions is obtained by table lookup or function fitting. The model parameters are then optimized to obtain the self-centering assist torque model.

[0013] The torque deviation evaluation unit is used to obtain real-time sensor data, specifically the steering wheel torque signal, calculate the real-time output torque of the EPS system based on the steering wheel torque signal, calculate the difference between the real-time output torque and the ideal return-to-center assist torque, and simultaneously obtain the aging coefficient of the EPS system, comprehensively calculate the deviation coefficient of the EPS system, and send it to the fault judgment unit;

[0014] The fault judgment unit is used to obtain the deviation coefficient of the EPS system and obtain a preset fault classification interval to evaluate the deviation coefficient, so as to generate a fault judgment result and send it to the power output adjustment unit;

[0015] The power output adjustment unit is used to obtain the fault judgment result and trigger a preset hierarchical response mechanism. The hierarchical response mechanism analyzes the fault judgment result and adapts the corresponding power output strategy according to the fault level. The main controller command is generated according to the power output strategy and sent to the EPS motor and driver. The EPS motor and driver output power torque according to the main controller command;

[0016] The actual power assist back-calculation unit is used to obtain the feedback current of the EPS motor and driver, calculate the actual execution torque based on the feedback current, calculate the difference between the actual execution torque and the desired return power assist torque, and perform secondary optimization of the return power assist torque model based on the difference to obtain an accurate return power assist torque model.

[0017] Furthermore, the sensor group includes a steering wheel torque sensor, a steering wheel angle sensor and a vehicle speed sensor, wherein:

[0018] The steering wheel torque sensor is used to collect the steering wheel torque signal applied by the driver in real time;

[0019] The steering wheel angle sensor is used to obtain the steering wheel rotation angle;

[0020] The vehicle speed sensor is used to obtain the current speed of the vehicle;

[0021] The steering wheel torque signal is sent to the torque offset evaluation unit, and the steering wheel rotation angle and the current speed of the vehicle are sent to the power output adjustment unit.

[0022] Furthermore, the specific process of obtaining the return-to-center assist torque model is as follows:

[0023] S101. Modeling is performed based on the tire self-aligning torque, where the tire self-aligning torque is the core input of the self-aligning assist. The specific expression is as follows:

[0024] Talign = -D*sin(C*arctan(B*α-E*α-arctan(B*α))), where Talign is the tire aligning torque and α is the tire slip angle, calculated based on the steering wheel rotation angle and the current vehicle speed. B, C, D, and E are preset formula parameters whose specific values ​​are calibrated through tire testing.

[0025] S102. Construct a steering wheel return dynamics equation based on the steering system dynamics. The specific expression is as follows: Where θ is the steering wheel rotation angle, J is the moment of inertia, B is the damping coefficient, K is the torsional stiffness, Tdriver is the driver input torque, which is 0 when returning to the center position, Tassist is the motor assist torque, and Tfric is the friction torque;

[0026] S103. Set up rule-based assistance strategy:

[0027] When the vehicle speed is less than or equal to the preset speed threshold, that is, under low-speed conditions, when the steering wheel rotation angle θ is greater than the preset angle threshold, and the rotation angular velocity When the direction is consistent with the return direction, power assistance is provided, and the motor power assistance torque Tassist is expressed as follows: Where kp and kd are PD control gains respectively, Tcomp is the torque to compensate for friction and damping, and its specific value is a constant set according to the vehicle speed;

[0028] When the vehicle speed is greater than the preset speed threshold, that is, under high-speed conditions, damping control is used to suppress excessive return to the positive position. The motor assist torque Tassist is expressed as follows:

[0029] S104: Acquire key parameters, including the speed threshold Vthreshold, the angle threshold θthreshold, and the PD control gain. Adjust the key parameters through actual vehicle testing. The specific adjustment method is as follows:

[0030] After fixing the vehicle speed, gradually increase kp until the return time meets the standard;

[0031] Adjust kd to suppress oscillations;

[0032] Repeatedly test different vehicle current speeds and steering wheel rotation angles to optimize the parameter mapping table;

[0033] S105. Calibrate the Coulomb friction Tcoulomb and the viscous friction Tviscous through bench tests, and establish a friction torque model expressed as follows:

[0034] S106 , obtaining multiple sets of real-time sensor data and integrating them into training samples, and obtaining the ideal self-centering assist torque that should be output under the ideal state through table lookup or function fitting, and optimizing the model parameters to obtain a self-centering assist torque model.

[0035] Furthermore, the specific process of generating the fault judgment result is as follows:

[0036] S201 , obtaining the real-time output torque Tactual and the ideal return-to-center assist torque Tassist of the EPS system, and calculating the difference between the real-time output torque and the ideal return-to-center assist torque: ΔT = |Tactual-Tassist|;

[0037] S202: Obtain an aging coefficient η of the EPS system. The aging coefficient η is calculated based on the historical driving distance of the vehicle and the equipment life using a collective exponential decay model. The deviation coefficient of the EPS system is calculated according to the following formula: Where e and d are preset weight factors, and the deviation coefficient is used to evaluate the difference between the actual braking state of the EPS system and the ideal braking state;

[0038] S203: Obtain a preset fault classification interval (R1, R2, R3, R4). If the deviation coefficient of the EPS system is less than or equal to R1, it indicates that the EPS system is in a normal state.

[0039] If the deviation coefficient of the EPS system is greater than R1 and less than R2, it means that the EPS system is slightly deviated at this time;

[0040] If the deviation coefficient of the EPS system is greater than R2 and less than R3, it means that the EPS system is in a moderate fault at this time;

[0041] If the deviation coefficient of the EPS system is greater than R3 and less than R4, it means that the EPS system is in a serious fault;

[0042] If the deviation coefficient of the EPS system is greater than or equal to R4, it means that the EPS system is in an emergency failure.

[0043] Furthermore, the hierarchical response mechanism includes a hierarchical sensing module, a policy storage module, and a policy matching module, wherein:

[0044] The hierarchical sensing module is used to obtain the fault judgment results, analyze the fault judgment results, and generate the fault level feature Ai, where i = 1, 2, 3, 4;

[0045] The strategy storage module is used to preset the corresponding power output strategy according to the degree of torque deviation during vehicle operation, as follows:

[0046] If the fault level characteristic quantity is A1, corresponding to a mild offset, the standard power assist mapping is used;

[0047] If the fault level characteristic is A2, corresponding to a moderate fault, the power assist amplitude is reduced by 5-10%, the proportion of the return torque is increased, and the directional sense consistency is ensured;

[0048] If the fault level characteristic is A3, it corresponds to a serious fault, the maximum power output is limited, and an abnormal warning signal is sent through the instrument;

[0049] If the fault level characteristic is A4, corresponding to an emergency fault, the main power assist motor is turned off, the mechanical backup is enabled, and the redundant power assist strategy is engaged;

[0050] The strategy matching module is used to find the corresponding power-assistance output strategy according to the fault level characteristic quantity, and generate the main controller instruction according to the power-assistance output strategy.

[0051] Furthermore, the specific process of calculating the difference between the actual execution torque and the assist torque is as follows:

[0052] S301. Obtain the feedback current I of the EPS motor and driver, and calculate the actual execution torque according to the following formula: Tmodel = kl*I, where kl is a preset proportional coefficient;

[0053] S302 : Obtain the ideal self-aligning assist torque Tassist and calculate the difference: ΔT′=|Tmodel-Tassist|.

[0054] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0055] 1. This torque-offset-based electric power steering fault diagnosis and adaptation system enables multi-level fault self-diagnosis. By real-time monitoring and graded judgment of the deviation between the motor's actual output torque and the theoretical torque calculated based on the return-torque model, it can accurately identify varying degrees of EPS system performance anomalies and faults, thereby improving the accuracy and timeliness of fault diagnosis.

[0056] 2. This torque-offset-based electric power steering fault diagnosis and adaptive system improves system safety and reliability. By dynamically adjusting the power steering output strategy based on the fault level, it effectively prevents sudden changes, interruptions, or excessive power steering caused by motor failure or performance degradation, ensuring the continuity and safety of vehicle steering and enhancing the driver's control experience.

[0057] 3. This torque-offset-based electric power steering fault diagnosis and adaptive system adaptively adjusts the power assist and return torque ratio. This system can smoothly respond to driver operations even in the event of minor faults or performance fluctuations, avoiding abrupt changes and improving overall driving comfort. It also utilizes existing sensor data and return torque models for diagnosis and control, eliminating the need for new, complex testing equipment and facilitating system integration and cost control.

[0058] 4. This torque-offset-based electric power steering fault diagnosis and adaptive system has self-learning and dynamic adaptability. It can combine historical data and multi-cycle statistics to achieve dynamic correction and self-learning of the return torque model, further reducing false alarms and improving the stability and robustness of system diagnosis. When a serious fault is detected, the system can automatically switch to protection mode or redundant control path to ensure basic steering power function and enhance vehicle driving safety. At the same time, the diagnostic results and status information are transmitted to other vehicle control units or remote management platforms via the CAN bus communication interface, facilitating real-time monitoring and maintenance management. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 Shown is a schematic diagram of the overall external structure of the present invention. DETAILED DESCRIPTION

[0060] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0061] Example:

[0062] Figure 1As shown, the electric power steering fault diagnosis and adaptive system based on torque offset includes a data acquisition unit, a model building unit, a torque offset evaluation unit, a fault judgment unit and a power output adjustment unit, wherein:

[0063] The data acquisition unit is used to obtain real-time sensor data during the vehicle's driving process through the sensor group, eliminate invalid data in the real-time sensor data, classify it, and send it to the torque offset evaluation unit and the power output adjustment unit according to the classification results;

[0064] The sensor group includes a steering wheel torque sensor, a steering wheel angle sensor and a vehicle speed sensor, wherein:

[0065] The steering wheel torque sensor is used to collect the steering wheel torque signal applied by the driver in real time;

[0066] The steering wheel angle sensor is used to obtain the steering wheel rotation angle;

[0067] The vehicle speed sensor is used to obtain the current speed of the vehicle;

[0068] The steering wheel torque signal is sent to the torque offset evaluation unit, and the steering wheel rotation angle and the current speed of the vehicle are sent to the power output adjustment unit.

[0069] The model building unit is used to obtain real-time sensor data and build a self-centering assist torque model based on vehicle dynamics and empirical adjustments. The real-time sensor data is used as training samples to train the self-centering assist torque model. The ideal self-centering assist torque that should be output under ideal conditions is obtained by table lookup or function fitting. The model parameters are then optimized to obtain the self-centering assist torque model.

[0070] The specific process of obtaining the return-to-center assist torque model is as follows:

[0071] S101. Modeling is performed based on the tire self-aligning torque, where the tire self-aligning torque is the core input of the self-aligning assist. The specific expression is as follows:

[0072] Talign = -D*sin(C*arctan(B*α-E*α-arctan(B*α))), where Talign is the tire aligning torque and α is the tire slip angle, calculated based on the steering wheel rotation angle and the current vehicle speed. B, C, D, and E are preset formula parameters whose specific values ​​are calibrated through tire testing.

[0073] S102. Construct a steering wheel return dynamics equation based on the steering system dynamics. The specific expression is as follows: Where θ is the steering wheel rotation angle, J is the moment of inertia, B is the damping coefficient, K is the torsional stiffness, Tdriver is the driver input torque, which is 0 when returning to the center position, Tassist is the motor assist torque, and Tfric is the friction torque;

[0074] S103. Set up rule-based assistance strategy:

[0075] When the vehicle speed is less than or equal to the preset speed threshold, that is, under low-speed conditions, when the steering wheel rotation angle θ is greater than the preset angle threshold, and the rotation angular velocity When the direction is consistent with the return direction, power assistance is provided, and the motor power assistance torque Tassist is expressed as follows: Where kp and kd are PD control gains respectively, Tcomp is the torque to compensate for friction and damping, and its specific value is a constant set according to the vehicle speed;

[0076] When the vehicle speed is greater than the preset speed threshold, that is, under high-speed conditions, damping control is used to suppress excessive return to the positive position. The motor assist torque Tassist is expressed as follows:

[0077] S104: Acquire key parameters, including the speed threshold Vthreshold, the angle threshold θthreshold, and the PD control gain. Adjust the key parameters through actual vehicle testing. The specific adjustment method is as follows:

[0078] After fixing the vehicle speed, gradually increase kp until the return time meets the standard;

[0079] Adjust kd to suppress oscillations;

[0080] Repeatedly test different vehicle current speeds and steering wheel rotation angles to optimize the parameter mapping table;

[0081] S105. Calibrate the Coulomb friction Tcoulomb and the viscous friction Tviscous through bench tests, and establish a friction torque model expressed as follows:

[0082] S106 , obtaining multiple sets of real-time sensor data and integrating them into training samples, and obtaining the ideal self-centering assist torque that should be output under the ideal state through table lookup or function fitting, and optimizing the model parameters to obtain a self-centering assist torque model.

[0083] The torque deviation evaluation unit is used to obtain real-time sensor data, specifically the steering wheel torque signal, calculate the real-time output torque of the EPS system based on the steering wheel torque signal, calculate the difference between the real-time output torque and the ideal return-to-center assist torque, and simultaneously obtain the aging coefficient of the EPS system, comprehensively calculate the deviation coefficient of the EPS system, and send it to the fault judgment unit;

[0084] The fault judgment unit is used to obtain the deviation coefficient of the EPS system and obtain a preset fault classification interval to evaluate the deviation coefficient, so as to generate a fault judgment result and send it to the power output adjustment unit;

[0085] The specific process of generating fault judgment results is as follows:

[0086] S201 , obtaining the real-time output torque Tactual and the ideal return-to-center assist torque Tassist of the EPS system, and calculating the difference between the real-time output torque and the ideal return-to-center assist torque: ΔT = |Tactual-Tassist|;

[0087] S202: Obtain an aging coefficient η of the EPS system. The aging coefficient η is calculated based on the historical driving distance of the vehicle and the equipment life using a collective exponential decay model. The deviation coefficient of the EPS system is calculated according to the following formula: Where e and d are preset weight factors, and the deviation coefficient is used to evaluate the difference between the actual braking state of the EPS system and the ideal braking state;

[0088] S203: Obtain a preset fault classification interval (R1, R2, R3, R4). If the deviation coefficient of the EPS system is less than or equal to R1, it indicates that the EPS system is in a normal state.

[0089] If the deviation coefficient of the EPS system is greater than R1 and less than R2, it means that the EPS system is slightly deviated at this time;

[0090] If the deviation coefficient of the EPS system is greater than R2 and less than R3, it means that the EPS system is in a moderate fault at this time;

[0091] If the deviation coefficient of the EPS system is greater than R3 and less than R4, it means that the EPS system is in a serious fault;

[0092] If the deviation coefficient of the EPS system is greater than or equal to R4, it means that the EPS system is in an emergency failure.

[0093] The power output adjustment unit is used to obtain the fault judgment result and trigger a preset hierarchical response mechanism. The hierarchical response mechanism analyzes the fault judgment result and adapts the corresponding power output strategy according to the fault level. The main controller command is generated according to the power output strategy and sent to the EPS motor and driver. The EPS motor and driver output power torque according to the main controller command;

[0094] The hierarchical response mechanism includes a hierarchical sensing module, a policy storage module, and a policy matching module, wherein:

[0095] The hierarchical sensing module is used to obtain the fault judgment results, analyze the fault judgment results, and generate the fault level feature Ai, where i = 1, 2, 3, 4;

[0096] The strategy storage module is used to preset the corresponding power output strategy according to the degree of torque deviation during vehicle operation, as follows:

[0097] If the fault level characteristic quantity is A1, corresponding to a mild offset, the standard power assist mapping is used;

[0098] If the fault level characteristic is A2, corresponding to a moderate fault, the power assist amplitude is reduced by 5-10%, the proportion of the return torque is increased, and the directional sense consistency is ensured;

[0099] If the fault level characteristic is A3, it corresponds to a serious fault, the maximum power output is limited, and an abnormal warning signal is sent through the instrument;

[0100] If the fault level characteristic is A4, corresponding to an emergency fault, the main power assist motor is turned off, the mechanical backup is enabled, and the redundant power assist strategy is engaged;

[0101] The strategy matching module is used to find the corresponding power-assistance output strategy according to the fault level characteristic quantity, and generate the main controller instruction according to the power-assistance output strategy.

[0102] The actual power assist back-calculation unit is used to obtain the feedback current of the EPS motor and driver, calculate the actual execution torque based on the feedback current, calculate the difference between the actual execution torque and the desired return power assist torque, and perform secondary optimization of the return power assist torque model based on the difference to obtain an accurate return power assist torque model.

[0103] The specific process of calculating the difference between the actual execution torque and the assist torque is as follows:

[0104] S301. Obtain the feedback current I of the EPS motor and driver, and calculate the actual execution torque according to the following formula: Tmodel = kl*I, where kl is a preset proportional coefficient;

[0105] S302 : Obtain the ideal self-aligning assist torque Tassist and calculate the difference: ΔT′=|Tmodel-Tassist|.

[0106] It also includes an alarm module for obtaining abnormal warning signals and feeding back abnormal conditions to the driver.

[0107] The present invention constructs a self-centering assist torque model based on the vehicle dynamics law and experience adjustment, and optimizes the model parameters to obtain the self-centering assist torque model. At the same time, the real-time output torque of the EPS system is calculated according to the steering wheel torque signal, and the deviation coefficient of the EPS system is comprehensively calculated. The deviation coefficient is evaluated according to the preset fault classification interval, and a multi-level fault judgment mechanism is established according to its change law to realize dynamic health evaluation of the EPS system motor performance. As the basis for adjusting the control strategy, the output assist value can be dynamically adjusted, the self-centering control can be strengthened, or it can be switched to the safety mode, thereby improving the redundant control capability and driving safety.

[0108] The size of the interval and threshold is set to facilitate comparison. The size of the threshold depends on the amount of sample data and the number of bases set by technical personnel in this field for each set of sample data; as long as it does not affect the proportional relationship between the parameter and the quantized value.

[0109] The above formulas are all dimensionless and numerical calculations. The formulas are obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters in the formulas are set by those skilled in the art according to actual conditions.

[0110] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. Electric power steering fault diagnosis and adaptive system based on torque offset, characterized by: It includes a data acquisition unit, a model building unit, a torque offset evaluation unit, a fault judgment unit and a power output adjustment unit, wherein: The data acquisition unit is used to obtain real-time sensor data during the vehicle's driving process through the sensor group, eliminate invalid data in the real-time sensor data, classify it, and send it to the torque offset evaluation unit and the power output adjustment unit according to the classification results; The model building unit is used to obtain real-time sensor data and build a self-centering assist torque model based on vehicle dynamics and empirical adjustments. The ideal self-centering assist torque that should be output under ideal conditions is obtained by table lookup or function fitting, and the model parameters are optimized to obtain the self-centering assist torque model. The torque deviation evaluation unit is used to obtain real-time sensor data, specifically the steering wheel torque signal, calculate the real-time output torque of the EPS system based on the steering wheel torque signal, calculate the difference between the real-time output torque and the ideal return-to-center assist torque, and simultaneously obtain the aging coefficient of the EPS system, comprehensively calculate the deviation coefficient of the EPS system, and send it to the fault judgment unit; The fault judgment unit is used to obtain the deviation coefficient of the EPS system and obtain a preset fault classification interval to evaluate the deviation coefficient, so as to generate a fault judgment result and send it to the power output adjustment unit; The power output adjustment unit is used to obtain the fault judgment result and trigger a preset hierarchical response mechanism. The hierarchical response mechanism analyzes the fault judgment result and adapts the corresponding power output strategy according to the fault level. The main controller command is generated according to the power output strategy and sent to the EPS motor and driver. The EPS motor and driver output power torque according to the main controller command; The actual power assist back-calculation unit is used to obtain the feedback current of the EPS motor and driver, calculate the actual execution torque based on the feedback current, calculate the difference between the actual execution torque and the desired return power assist torque, and perform secondary optimization of the return power assist torque model based on the difference to obtain an accurate return power assist torque model.

2. The electric power steering fault diagnosis and adaptive system based on torque offset according to claim 1, characterized in that: The sensor group includes a steering wheel torque sensor, a steering wheel angle sensor and a vehicle speed sensor, wherein: The steering wheel torque sensor is used to collect the steering wheel torque signal applied by the driver in real time; The steering wheel angle sensor is used to obtain the steering wheel rotation angle; The vehicle speed sensor is used to obtain the current speed of the vehicle; The steering wheel torque signal is sent to the torque offset evaluation unit, and the steering wheel rotation angle and the current speed of the vehicle are sent to the power output adjustment unit.

3. The electric power steering fault diagnosis and adaptive system based on torque offset according to claim 1, characterized in that: The specific process of obtaining the return-to-center assist torque model is as follows: S101. Modeling is performed based on the tire self-aligning torque, where the tire self-aligning torque is the core input of the self-aligning assist. The specific expression is as follows: Talign = -D*sin(C*arctan(B*α-E*α-arctan(B*α))), where Talign is the tire aligning torque and α is the tire slip angle, calculated based on the steering wheel rotation angle and the current vehicle speed. B, C, D, and E are preset formula parameters whose specific values ​​are calibrated through tire testing. S102. Construct a steering wheel return dynamics equation based on the steering system dynamics. The specific expression is as follows: Where θ is the steering wheel rotation angle, J is the moment of inertia, B is the damping coefficient, K is the torsional stiffness, Tdriver is the driver input torque, which is 0 when returning to the center position, Tassist is the motor assist torque, and Tfric is the friction torque; S103. Set up rule-based assistance strategy: When the vehicle speed is less than or equal to the preset speed threshold, that is, under low-speed conditions, when the steering wheel rotation angle θ is greater than the preset angle threshold, and the rotation angular velocity When the direction is consistent with the return direction, power assistance is provided, and the motor power assistance torque Tassist is expressed as follows: Where kp and kd are PD control gains respectively, Tcomp is the torque to compensate for friction and damping, and its specific value is a constant set according to the vehicle speed; When the vehicle speed is greater than the preset speed threshold, that is, under high-speed conditions, damping control is used to suppress excessive return to the positive position. The motor assist torque Tassist is expressed as follows: S104: Acquire key parameters, including the speed threshold Vthreshold, the angle threshold θthreshold, and the PD control gain. Adjust the key parameters through actual vehicle testing. The specific adjustment method is as follows: After fixing the vehicle speed, gradually increase kp until the return time meets the standard; Adjust kd to suppress oscillations; Repeatedly test different vehicle current speeds and steering wheel rotation angles to optimize the parameter mapping table; S105. Calibrate the Coulomb friction Tcoulomb and the viscous friction Tviscous through bench tests, and establish a friction torque model expressed as follows: S106 , obtaining multiple sets of real-time sensor data and integrating them into training samples, and obtaining the ideal self-centering assist torque that should be output under the ideal state through table lookup or function fitting, and optimizing the model parameters to obtain a self-centering assist torque model.

4. The electric power steering fault diagnosis and adaptive system based on torque offset according to claim 1, characterized in that: The specific process of generating fault judgment results is as follows: S201 , obtaining the real-time output torque Tactual and the ideal return-to-center assist torque Tassist of the EPS system, and calculating the difference between the real-time output torque and the ideal return-to-center assist torque: ΔT = |Tactual-Tassist|; S202: Obtain an aging coefficient η of the EPS system. The aging coefficient η is calculated based on the historical driving distance of the vehicle and the equipment life using a collective exponential decay model. The deviation coefficient of the EPS system is calculated according to the following formula: Where e and d are preset weight factors, and the deviation coefficient is used to evaluate the difference between the actual braking state of the EPS system and the ideal braking state; S203: Obtain a preset fault classification interval (R1, R2, R3, R4). If the deviation coefficient of the EPS system is less than or equal to R1, it indicates that the EPS system is in a normal state. If the deviation coefficient of the EPS system is greater than R1 and less than R2, it means that the EPS system is slightly deviated at this time; If the deviation coefficient of the EPS system is greater than R2 and less than R3, it means that the EPS system is in a moderate fault at this time; If the deviation coefficient of the EPS system is greater than R3 and less than R4, it means that the EPS system is in a serious fault; If the deviation coefficient of the EPS system is greater than or equal to R4, it means that the EPS system is in an emergency failure.

5. The electric power steering fault diagnosis and adaptive system based on torque offset according to claim 1, characterized in that: The hierarchical response mechanism includes a hierarchical sensing module, a policy storage module, and a policy matching module, wherein: The hierarchical sensing module is used to obtain the fault judgment results, analyze the fault judgment results, and generate the fault level feature Ai, where i = 1, 2, 3, 4; The strategy storage module is used to preset the corresponding power output strategy according to the degree of torque deviation during vehicle operation, as follows: If the fault level characteristic quantity is A1, corresponding to a mild offset, the standard power assist mapping is used; If the fault level characteristic is A2, corresponding to a moderate fault, the power assist amplitude is reduced by 5-10%, the proportion of the return torque is increased, and the directional sense consistency is ensured; If the fault level characteristic is A3, it corresponds to a serious fault, the maximum power output is limited, and an abnormal warning signal is sent through the instrument; If the fault level characteristic is A4, corresponding to an emergency fault, the main power assist motor is turned off, the mechanical backup is enabled, and the redundant power assist strategy is engaged; The strategy matching module is used to find the corresponding power-assistance output strategy according to the fault level characteristic quantity, and generate the main controller instruction according to the power-assistance output strategy.

6. The electric power steering fault diagnosis and adaptive system based on torque offset according to claim 1, characterized in that: The specific process of calculating the difference between the actual execution torque and the assist torque is as follows: S301. Obtain the feedback current I of the EPS motor and driver, and calculate the actual execution torque according to the following formula: Tmodel = kl*I, where kl is a preset proportional coefficient; S302 : Obtain the ideal self-aligning assist torque Tassist and calculate the difference: ΔT′=|Tmodel-Tassist|.

Images