A hierarchical collaborative fault-tolerant control method and system for corner module vehicle single-wheel steering failure

Abstract

This invention discloses a hierarchical collaborative fault-tolerant control method and system for single-wheel steering failure in corner module vehicles. The method includes: acquiring vehicle state and environmental information; calculating a dynamic adaptive residual threshold in real time; triggering failure detection when both wheel angle residual and yaw rate residual exceed the threshold; diagnosing failure types from steering jamming, steering loosening, performance loss, and angle limitation; constructing a two-degree-of-freedom reference model; adjusting model parameters using pre-failure trend extrapolation and failure type equivalence; calculating a safety reference value based on the road adhesion coefficient; and generating a smooth transition reference yaw rate via a first-order exponential transition; constructing a nonlinear predictive control model; using the smooth transition reference as the tracking target; designing a cost function including state error, control input, increment, and relaxation variables; applying additional constraints for different failure types; and performing rolling optimization to solve for the optimal wheel angle and torque and collaboratively allocating them. This invention achieves refined failure diagnosis and hierarchical classification.

Description

Technical Field

[0001] This invention relates to the field of vehicle active safety and chassis cooperative control technology, specifically to a hierarchical cooperative fault-tolerant control method and system for single-wheel steering failure in corner module vehicles. Background Technology

[0002] With the development of steer-by-wire chassis technology, independent steering systems are increasingly widely used in vehicles. However, due to factors such as hardware aging and external impacts, the risk of single-wheel steering failure still exists. Many existing fault-tolerant control strategies directly switch control objectives or reconstruct control laws, leading to abrupt changes in yaw rate and lateral acceleration, which can cause significant discomfort to the driver. Even when considering post-failure stability, a single additional yaw moment compensation method is typically used, lacking precise identification of the specific form of single-wheel failure. This can easily cause sudden changes in the vehicle's dynamic response at the moment of failure, leading to driver panic and misoperation; moreover, a single compensation method cannot fully utilize the frequency domain response characteristics of the multi-actuator vehicles in corner-module vehicles, resulting in insufficient fault-tolerant ride comfort and ultimate safety. Therefore, a steering fault-tolerant control method is needed that can cope with different failure types and achieve a smooth transition while balancing transient ride comfort and steady-state tracking capability. Summary of the Invention

[0003] The purpose of this invention is to provide a hierarchical collaborative fault-tolerant control method for single-wheel steering failure of corner module vehicles, so as to solve the problems of insufficient vehicle stability caused by the inability of existing technologies to accurately diagnose and implement hierarchical collaborative fault-tolerant control for multiple failure types of single-wheel steering of corner module vehicles in real time.

[0004] To achieve the above objectives, the technical solution provided by this invention is: a hierarchical collaborative fault-tolerant control method for single-wheel steering failure in corner module vehicles, comprising the following steps: S1: Acquire vehicle status and environmental information, calculate dynamic adaptive residual threshold in real time, and calculate wheel angle residual and yaw rate residual; when the wheel angle residual is greater than the dynamic angle residual threshold and the yaw rate residual is greater than the dynamic yaw rate residual threshold, trigger failure classification detection; otherwise, maintain monitoring status. S2: After triggering the failure classification detection, based on the actuator current, actual angle change rate, high-frequency variance of the actual angle, and the relationship between the target angle change rate and the preset threshold, the current failure type is diagnosed from the set of faults including steering jam, steering loosening, performance loss, and angle limitation; wherein, the high-frequency variance characterizes the high-frequency oscillation characteristics of the actual angle. S3: Construct a two-degree-of-freedom reference model, calculate the trend reference yaw rate before the failure and the diagnosed failure type based on the yaw rate within a preset time period before the failure occurs, make equivalent adjustments to the parameters of the two-degree-of-freedom reference model, and calculate the yaw rate threshold in combination with the road adhesion coefficient as a safe reference value for yaw rate after the failure; calculate the smooth transition reference yaw rate through the smooth transition formula. S4: Construct a nonlinear predictive control model with the smooth transition reference yaw rate as the tracking target, and construct a cost function that includes state error, control input, control input increment, and slack variables; based on the diagnosed failure type, apply additional constraints corresponding to the failure type to the nonlinear predictive control model; solve the cost function using rolling optimization, obtain the optimal control input, and coordinately distribute it to each actuator, wherein the optimal control input includes the wheel rotation angle and torque.

[0005] To optimize the above technical solution, the specific measures also include: In step S1, the real-time calculation of the dynamic adaptive residual threshold and the calculation of the wheel steering angle residual and yaw rate residual are specifically as follows: The dynamic adaptive residual threshold is calculated in real time using the following formula:

[0006]

[0007] In the formula, The dynamic angle residual threshold, The dynamic yaw rate residual threshold. and These are the basic thresholds for the rotational residual and the yaw rate residual, respectively. The road surface adhesion coefficient, For longitudinal vehicle speed, Indicates reference speed. The actuator load gain coefficient, The road surface adhesion gain coefficient is positive. The tire lateral force saturation function takes into account the adhesion limit.

[0008] Furthermore, the wheel angle residual is calculated in real time. and yaw rate residual :

[0009]

[0010] in, To estimate the wheel angle; This is the actual turning angle of the wheel. To estimate the yaw rate, This represents the actual yaw rate.

[0011] In step S2, the diagnosis of the current failure type from the set of faults including steering jamming, steering loosening, loss of efficiency, and limited steering angle specifically involves: When the actuator current remains abnormal and the actual rate of change of steering angle approaches zero, the current failure type is diagnosed as steering jam. When the actuator current drops and the actual rate of change of rotation exceeds the upper limit, the current failure type is diagnosed as steering loosening; When the high-frequency variance of the actual rotation angle exceeds the preset threshold, the current failure type is diagnosed as performance loss. Furthermore, when the target turning angle continues to increase while the actual turning angle stagnates and the residual difference increases abnormally, the current failure type is diagnosed as angle-limited.

[0012] In step S3, the expression for the two-degree-of-freedom reference model is:

[0013]

[0014] in, For the overall vehicle quality, and These are the lateral stiffnesses of the front and rear axles, respectively. Wheelbase This is the distance from the front axle to the center of gravity. This is the distance from the rear axle to the center of mass. Let be the moment of inertia of the vehicle about the z-axis. and These are the front wheel steering angle input and the rear wheel steering angle input, respectively.

[0015] Furthermore, in step S3, the equivalent adjustment of the parameters of the two-degree-of-freedom reference model includes: correcting the tire lateral stiffness according to the failure type, adjusting the equivalent wheelbase, and changing the steering weights of the front and rear axles.

[0016] In step S3, the smooth transition reference yaw rate is calculated using the smooth transition formula, which is as follows:

[0017] in, For a smooth transition, refer to the yaw rate. t The time elapsed since the start of the transition process. For the transition velocity coefficient, The yaw rate is used as a reference for the trend before the fault. This is a safe reference value after a yaw rate failure.

[0018] In step S4, the cost function expression is:

[0019] in, To predict the step size, To control the step size, For reference output; Output as the target. For control input; To control the input increment; The state weight matrix is... To control the input weight matrix, To control the input increment weight matrix, For extremely large slack variable weights, These are slack variables; For the current control cycle, i To predict the future relative number of steps within the step size.

[0020] Furthermore, in step S4, the additional constraints include: locking the corresponding wheel angle according to the steering jam failure type, limiting the angle change rate according to the steering loosening failure type, reducing the upper limit of the corresponding actuator output according to the efficiency loss failure type, and limiting the angle range according to the angle limitation failure type.

[0021] As another important technical solution, the present invention also provides a hierarchical cooperative fault-tolerant control system for single-wheel steering failure of corner module vehicles, comprising: The failure detection trigger module is used to acquire vehicle status and environmental information, calculate the dynamic adaptive residual threshold in real time, and calculate the wheel angle residual and yaw rate residual. When the wheel angle residual is greater than the dynamic angle residual threshold and the yaw rate residual is greater than the dynamic yaw rate residual threshold, the failure classification detection is triggered; otherwise, the monitoring state is maintained. The failure type diagnosis module is used to diagnose the current failure type from a set of faults including steering jamming, steering loosening, performance loss, and steering angle limitation after triggering failure classification detection, based on the actuator current, actual angle change rate, high-frequency variance of actual angle, and the relationship between the target angle change rate and a preset threshold; wherein, the high-frequency variance characterizes the high-frequency oscillation characteristics of the actual angle. The smooth transition reference generation module is used to construct a two-degree-of-freedom reference model. Based on the yaw rate within a preset time period before the failure, it calculates the pre-failure trend reference yaw rate and the diagnosed failure type, performs equivalent adjustments to the parameters of the two-degree-of-freedom reference model, and calculates the yaw rate threshold in conjunction with the road adhesion coefficient as a safe reference value for the yaw rate after the failure. The smooth transition reference yaw rate is calculated using the smooth transition formula. The predictive control fault-tolerant allocation module is used to construct a nonlinear predictive control model. Taking the smooth transition reference yaw rate as the tracking target, it constructs a cost function that includes state error, control input, control input increment, and slack variables. Based on the diagnosed failure type, it applies additional constraints corresponding to the failure type to the nonlinear predictive control model. It then performs rolling optimization to solve the cost function, obtains the optimal control input, and allocates it to each actuator.

[0022] The present invention also proposes an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements a hierarchical collaborative fault-tolerant control method for single-wheel steering failure of a corner module vehicle as described above.

[0023] The present invention also proposes a computer-readable storage medium storing a computer program that enables a computer to execute a hierarchical collaborative fault-tolerant control method for single-wheel steering failure of a corner module vehicle as described above.

[0024] Compared with the prior art, the beneficial effects of the present invention are: This invention constructs a dynamic adaptive residual threshold and combines it with multi-dimensional features such as actuator current, rotation angle change rate, and high-frequency variance to accurately diagnose specific failure types. This refined diagnosis provides a reliable basis for subsequent differentiated fault-tolerant control and avoids control mismatch problems caused by single fault mode processing.

[0025] The invention employs a two-degree-of-freedom reference model with pre-fault trend extrapolation and equivalent adjustment of failure type in the upper layer to generate a safety reference value. It also achieves smooth switching through a first-order exponential transition formula, effectively eliminating the step jump in yaw rate and lateral acceleration caused by sudden changes in reference value in traditional fault-tolerant control. This significantly reduces driver discomfort and improves transient driving smoothness.

[0026] The lower layer of this invention is based on nonlinear model predictive control, with smooth transition yaw rate as the tracking target. It constructs a cost function that includes state error, control input, increment and relaxation variables. By setting the weights of steering angle and torque differently, it achieves priority coordination between low-frequency steering compensation and high-frequency yaw moment compensation. Corresponding additional constraints are applied for different failure types to ensure the feasibility of the solution. It makes full use of the independent steering / driving capability of four wheels of the corner module vehicle, while taking into account the transient smoothness of failure and the steady-state trajectory tracking / stability limit.

[0027] The dynamic residual threshold of this invention can adapt to changes in road surface adhesion coefficient and longitudinal vehicle speed, avoiding false triggering or missed triggering diagnosis under low adhesion or high speed conditions; the smooth transition coefficient is dynamically adjusted with vehicle speed and road surface condition to ensure safe take-off under various conditions; the slack variable ensures that the optimization problem always has a solution under extreme conditions, preventing control system failure due to constraint conflicts. Attached Figure Description

[0028] Figure 1 This is a standard flowchart of the single-wheel steering failure hierarchical collaborative fault-tolerant control method of the present invention. Detailed Implementation

[0029] The present invention will be further described in detail below through specific embodiments, but it should not be construed as limiting the scope of the subject matter of the present invention to the following embodiments. All technologies implemented based on the above content of the present invention fall within the scope of the present invention.

[0030] In one embodiment, this invention proposes a hierarchical collaborative fault-tolerant control method for single-wheel steering failure in corner module vehicles, the flowchart of which is shown below. Figure 1 As shown, the entire method includes the following steps: S1: Acquire vehicle status and environmental information, calculate dynamic adaptive residual threshold in real time, and calculate wheel angle residual and yaw rate residual; when the wheel angle residual is greater than the dynamic angle residual threshold and the yaw rate residual is greater than the dynamic yaw rate residual threshold, trigger failure classification detection; otherwise, maintain monitoring status. The dynamic adaptive residual threshold is calculated in real time using the following formula:

[0031]

[0032] In the formula, The dynamic angle residual threshold, The dynamic yaw rate residual threshold. and These are the basic thresholds for the rotational residual and the yaw rate residual, respectively. The road surface adhesion coefficient, For longitudinal vehicle speed, Indicates reference speed. For the target turning angle of the wheel, The actuator load gain coefficient, The road surface adhesion gain coefficient is positive. The positive speed regulation gain coefficient ranges from 0.1 to 0.5. The positive pavement adhesion influence factor ranges from 1.5 to 3.0. The expression for the tire lateral force saturation function that takes into account the adhesion limit is:

[0033] in, A positive saturation shape adjustment coefficient. This is a preset minimum value used to prevent errors caused by dividing by zero.

[0034] Real-time calculation of wheel angle residuals and yaw rate residual :

[0035]

[0036] in, To estimate the wheel angle, the steering ratio of the steering system under fault-free conditions is calculated in real time by combining the driver's steering wheel angle input and the inertial filtering circuit that takes into account response delay. This is the actual turning angle of the wheel; To estimate the yaw rate, parameters such as the current longitudinal vehicle speed and the driver's steering wheel angle are input into the vehicle reference dynamics model for real-time calculation. This represents the actual yaw rate.

[0037] If the wheels simultaneously satisfy:

[0038] This triggers a refined failure classification detection.

[0039] S2: After triggering the failure classification detection, based on the actuator current, actual angle change rate, high-frequency variance of the actual angle, and the relationship between the target angle change rate and the preset threshold, the current failure type is diagnosed from the set of faults including steering jam, steering loosening, performance loss, and angle limitation; wherein, the high-frequency variance characterizes the high-frequency oscillation characteristics of the actual angle. If the vehicle meets the following conditions:

[0040] The current failure type is diagnosed as steering jam, which means that the steering actuator is mechanically jammed and the wheel is fixed at a non-desired turning angle. Not only can it not provide the expected steering force, but it will also generate continuous and strong interfering lateral force and yaw moment.

[0041] If the vehicle meets the following conditions:

[0042] The current failure type is diagnosed as steering detachment, meaning that the wheel loses its connection with the vehicle in the steering dimension, is not controlled by the steering actuator, and follows the vehicle's direction of travel under the action of the return torque, providing almost no lateral force and generating no active interference.

[0043] In some implementations, the current failure type is diagnosed as steering loosening, and the power supply to the steering motor of the faulty wheel is cut off.

[0044] If the vehicle meets the following conditions:

[0045] The current failure type is diagnosed as performance loss, i.e., actuator aging or a sharp increase in friction, leading to a decrease in steering gain or severe response lag. After determining this fault, the mean of the actual rate of change of steering angle at each sampling point within the determination time window is calculated. The mean of the rate of change of the target angle Calculate the efficiency reduction ratio coefficient This is used for subsequent calculations of the yaw rate as a safety reference value after a fault. The expression is:

[0046] If the actuator current and wheel angle residual are normal within a certain target turning angle range, but the actual turning angle remains stagnant at a certain value as the target turning angle continues to increase. place, and If the value increases, it is determined to be a limited steering angle, meaning the steering mechanism operates normally within a specific range, and the steering angle is limited to that range. Record this value. The value is one of the boundaries of the restricted interval.

[0047] in, For the measured actuator current, The preset jamming current threshold, This represents the actual rate of change of rotation angle. The preset threshold for the rate of change of wheel angle is close to 0. The preset release current threshold, for high frequency variance, The preset high-frequency variance threshold, The preset redundancy coefficient, This represents the actual rate of change of rotation angle. Let be the expected rate of change of the rotation angle.

[0048] S3: Construct a two-degree-of-freedom reference model, calculate the trend reference yaw rate before the failure and the diagnosed failure type based on the yaw rate within a preset time period before the failure occurs, make equivalent adjustments to the parameters of the two-degree-of-freedom reference model, and calculate the yaw rate threshold in combination with the road adhesion coefficient as a safe reference value for yaw rate after the failure; calculate the smooth transition reference yaw rate through the smooth transition formula. In some embodiments, the upper layer of the hierarchical control strategy of the present invention is based on the recorded data before the fault occurs. yaw rate over time To simulate the dynamic trend of the car before the failure, a second-order Taylor expansion is used to extrapolate these data. The extrapolation formula is as follows:

[0049] In the formula, The yaw rate is used as a reference for the trend before the fault. The time when the fault occurred yaw rate at that time and These are the first and second derivatives of the yaw rate at the moment the fault occurs, respectively.

[0050] The expression for the two-degree-of-freedom reference model is:

[0051]

[0052] in, For the overall vehicle quality, and These are the lateral stiffnesses of the front and rear axles, respectively. Wheelbase This is the distance from the front axle to the center of gravity. This is the distance from the rear axle to the center of mass. Let be the moment of inertia of the vehicle about the z-axis. and These are the front wheel angle input and the rear wheel angle input, respectively. When the left and right wheel angles are different, the angle input here refers to the average of the left and right wheel angles.

[0053] In some implementations, when the centroid sideslip angle The value is 0, so the base yaw rate based on the driver's front wheel steering angle input is obtained when there is no fault:

[0054] in, The equivalent vehicle stability factor is expressed as:

[0055] When steering lock-up occurs, the stuck wheel cannot respond to driver commands, but its stuck steering angle remains unchanged. This will generate continuous lateral force disturbance. At this time, the equivalent stiffness remains unchanged, but the input changes.

[0056] In some implementations, the total front axle input consists of the healthy wheel input and the stuck wheel interference, used to calculate the yaw rate during a steering lockup fault. :

[0057] in, For healthy rotation angles, This refers to the stuck angle of the wheel.

[0058] When a steering malfunction occurs, the faulty wheel loses its steering constraint and, under the influence of the return torque, becomes a free-following state. Its tire slip angle approaches zero and it loses its ability to generate lateral force. In the two-degree-of-freedom reference model, the equivalent slip stiffness parameter corresponding to the faulty wheel is directly set to zero to decouple the mathematical influence of the faulty wheel on the vehicle's lateral dynamics objective.

[0059] The yaw rate at the time of steering disengagement was calculated. :

[0060] In some implementations, when a performance failure occurs, the wheels still retain some steering ability, but the steering gain is reduced, resulting in incomplete steering. The performance degradation ratio coefficient provided by the failure refinement diagnosis module is then utilized. The front axle lateral stiffness is effectively reduced as follows:

[0061] In some implementations, the yaw rate at the time of performance loss failure is calculated. :

[0062] Yaw rate during cornering failure It is a piecewise function, and its expression is:

[0063] in, and These are the maximum and minimum values ​​of the restricted wheel rotation angle, respectively.

[0064] Obtain the road surface adhesion coefficient Calculate the yaw rate threshold that the road surface can provide. :

[0065] in, The preset safety margin coefficient is preferably taken as follows: The value is 0.85~0.9, where g is the acceleration due to gravity.

[0066] The safe reference value after a yaw rate failure is:

[0067] The formula for calculating the smooth transition reference state is:

[0068] in, For a smooth transition, refer to the yaw rate. t The time elapsed since the start of the transition process. This is the transition speed coefficient, which can be dynamically selected based on vehicle speed and road surface adhesion coefficient. For high-speed vehicles or low-adhesion roads, a smaller value can be appropriately chosen. To extend the transition time and ensure stability; under low-speed, high-adhesion road conditions, increase To improve the efficiency of takeover.

[0069] S4: Construct a nonlinear predictive control model with the smooth transition reference yaw rate as the tracking target, and construct a cost function that includes state error, control input, control input increment, and slack variables; based on the diagnosed failure type, apply additional constraints corresponding to the failure type to the nonlinear predictive control model; solve the cost function using rolling optimization, obtain the optimal control input, and coordinately distribute it to each actuator, wherein the optimal control input includes the wheel rotation angle and torque.

[0070] The hierarchical control strategy of this invention employs a multi-actuator cooperative control based on a nonlinear predictive control model at the lower level. Define state variables for:

[0071] Define control input for:

[0072] Combining the nonlinear tire model and the vehicle dynamics differential equations, we obtain:

[0073]

[0074] in, , , and Each wheel has an independent steering angle for the front left, front right, rear left, and rear right wheels. , , and The independent torques are for the front left, front right, rear left, and rear right wheels. The function is derived from the nonlinear tire model and the differential equations of vehicle dynamics. For the output matrix, To control the cycle; The cost function expression is:

[0075] in, To predict the step size, To control the step size, For reference output; Output as the target. To control the input increment; The state weight matrix is... To control the input weight matrix, The weights are for extremely large slack variables.

[0076] To prevent the control algorithm from becoming unsolvable due to constraint conflicts and causing the system to freeze even in extreme cases, slack variables are used to address constraints beyond objective hard constraints such as mechanical limits and actuator performance limitations. To soften it, the expression is: .

[0077] By setting the control input increment weight matrix This design ensures that the penalty weight for changes in four-wheel torque is much greater than that for changes in wheel steering angle, thus achieving a priority division of dynamic response. Under low-frequency steady-state conditions, the system prioritizes rear-wheel steering for compensation. Only under sudden large deviations or high-frequency transient disturbances does it instantaneously activate the four-wheel differential brake / drive actuators for high-frequency yaw torque compensation.

[0078] Under different fault modes, the control input and the change in control input There are different additional constraints: When the steering wheel is stuck:

[0079] When the steering wheel comes loose:

[0080] When a steering wheel becomes disengaged, the direction of the braking torque on the disengaged wheel is determined by the actual following angle of the steering wheel. Decide:

[0081] When performance is lost:

[0082] When cornering is restricted:

[0083] In the formula, and These are the minimum and maximum steering angles of the faulty wheel, respectively. The jamming angle obtained by the failure refinement diagnosis module; and These are the minimum and maximum rotation angles of the faulty wheel, respectively. The maximum braking torque of the faulty wheel. and These are the maximum and minimum values ​​of the rate of change of the rotation angle of the faulty wheel, respectively.

[0084] During normal driving, and It should be subject to constraints such as actuator performance, mechanical structure, and road surface adhesion limits. For example, the road surface adhesion limit constraint according to the friction circle theory is:

[0085] in, and respectively wheels The longitudinal and lateral forces, It is a wheel The vertical load, .

[0086] In the event of a failure, additional constraints are added based on the specific failure mode.

[0087] In some implementations, for the four steering faults mentioned in this invention, except for the steering loosening fault where the torque is additionally limited to prevent the positive torque from causing the shimmy effect, only the steering angle is subject to additional constraints, while the original constraints on the magnitude of the wheel torque are retained.

[0088] In each control cycle Within this framework, the solver, under the conditions of satisfying the aforementioned comprehensive constraints, seeks to optimize the cost function through rolling optimization. The optimal control increment sequence that reaches its minimum value is taken as the first element of the sequence as the actual execution increment for the current cycle. The optimal input is defined as the optimal wheel angle and torque. The expression for the solution process is as follows:

[0089]

[0090] In the formula, To predict the optimal control increment sequence obtained by solving within the step size. To provide a solution operator that minimizes the cost function, This represents the comprehensive set of constraints imposed on the system, including the dynamic model, tire friction circle, relaxation variables, and the aforementioned additional constraints for different failure modes. For the optimal control increment sequence The first control increment element in the process. This is the optimal target input vector that should be allocated to each actuator in the current control cycle; the solved optimal value is then input to the controller of the corresponding actuator.

[0091] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Any simple modifications, equivalent substitutions, and improvements made by those skilled in the art to the above embodiments without departing from the scope of the technical solution of the present invention, based on the technical essence of the present invention, shall still fall within the protection scope of the technical solution of the present invention.

Claims

1. A hierarchical collaborative fault-tolerant control method for single-wheel steering failure in corner module vehicles, characterized in that, Includes the following steps: S1: Acquire vehicle status and environmental information, calculate dynamic adaptive residual threshold in real time, and calculate wheel angle residual and yaw rate residual; when the wheel angle residual is greater than the dynamic angle residual threshold and the yaw rate residual is greater than the dynamic yaw rate residual threshold, trigger failure classification detection; otherwise, maintain monitoring status. S2: After triggering the failure classification detection, based on the actuator current, actual angle change rate, high-frequency variance of the actual angle, and the relationship between the target angle change rate and the preset threshold, the current failure type is diagnosed from the set of faults including steering jam, steering loosening, performance loss, and angle limitation; wherein, the high-frequency variance characterizes the high-frequency oscillation characteristics of the actual angle. S3: Construct a two-degree-of-freedom reference model, calculate the trend reference yaw rate before the failure and the diagnosed failure type based on the yaw rate within a preset time period before the failure occurs, make equivalent adjustments to the parameters of the two-degree-of-freedom reference model, and calculate the yaw rate threshold in combination with the road adhesion coefficient as a safe reference value for yaw rate after the failure; calculate the smooth transition reference yaw rate through the smooth transition formula. S4: Construct a nonlinear predictive control model with the smooth transition reference yaw rate as the tracking target, and construct a cost function that includes state error, control input, control input increment, and slack variables; based on the diagnosed failure type, apply additional constraints corresponding to the failure type to the nonlinear predictive control model; solve the cost function using rolling optimization, obtain the optimal control input, and coordinately distribute it to each actuator, wherein the optimal control input includes the wheel rotation angle and torque.

2. The hierarchical collaborative fault-tolerant control method for single-wheel steering failure of a corner module vehicle according to claim 1, characterized in that: In step S2, the current failure type is diagnosed from a set of faults including steering jamming, steering loosening, loss of efficiency, and limited steering angle. The specific diagnostic method is as follows: When the actuator current remains abnormal and the actual rate of change of steering angle approaches zero, the current failure type is diagnosed as steering jam. When the actuator current drops and the actual rate of change of rotation exceeds the upper limit, the current failure type is diagnosed as steering loosening; When the high-frequency variance of the actual rotation angle exceeds the preset threshold, the current failure type is diagnosed as performance loss. When the target turning angle continues to increase while the actual turning angle stagnates and the residual difference increases abnormally, the current failure type is diagnosed as angle-limited.

3. The hierarchical collaborative fault-tolerant control method for single-wheel steering failure of a corner module vehicle according to claim 1, characterized in that: In step S3, the expression for the two-degree-of-freedom reference model is: in, For the overall vehicle quality, and These are the lateral stiffnesses of the front and rear axles, respectively. Wheelbase This is the distance from the front axle to the center of gravity. This is the distance from the rear axle to the center of mass. Let be the moment of inertia of the vehicle about the z-axis. and These are the front wheel steering angle input and the rear wheel steering angle input, respectively. This refers to the longitudinal vehicle speed.

4. The hierarchical collaborative fault-tolerant control method for single-wheel steering failure of a corner module vehicle according to claim 3, characterized in that: In step S3, the equivalent adjustment of the parameters of the two-degree-of-freedom reference model includes: correcting the tire lateral stiffness according to the failure type, adjusting the equivalent wheelbase, and changing the steering weights of the front and rear axles.

5. The hierarchical collaborative fault-tolerant control method for single-wheel steering failure of a corner module vehicle according to claim 1, characterized in that: In step S3, the smooth transition reference yaw rate is calculated using the smooth transition formula, which is as follows: in, For a smooth transition, refer to the yaw rate. t The time elapsed since the start of the transition process. For the transition velocity coefficient, The yaw rate is used as a reference for the trend before the fault. This is a safe reference value after a yaw rate failure.

6. The hierarchical collaborative fault-tolerant control method for single-wheel steering failure of a corner module vehicle according to claim 1, characterized in that: In step S4, the cost function expression is: in, To predict the step size, To control the step size, For reference output, Output as the target. For control input; To control the input increment; The state weight matrix is... To control the input weight matrix, To control the input increment weight matrix, For slack variable weights, These are slack variables; For the current control cycle, i To predict the future relative number of steps within the step size.

7. The hierarchical collaborative fault-tolerant control method for single-wheel steering failure of a corner module vehicle according to claim 1, characterized in that: In step S4, the additional constraints include: locking the corresponding wheel angle according to the steering jam failure type, limiting the angle change rate according to the steering loosening failure type, reducing the upper limit of the corresponding actuator output according to the efficiency loss failure type, and limiting the angle range according to the angle limitation failure type.

8. A hierarchical cooperative fault-tolerant control system for single-wheel steering failure in corner module vehicles, characterized in that, include: The failure detection trigger module is used to acquire vehicle status and environmental information, calculate the dynamic adaptive residual threshold in real time, and calculate the wheel angle residual and yaw rate residual. When the wheel angle residual is greater than the dynamic angle residual threshold and the yaw rate residual is greater than the dynamic yaw rate residual threshold, failure classification detection is triggered; otherwise, the monitoring status is maintained. The failure type diagnosis module is used to diagnose the current failure type from a set of faults including steering jamming, steering loosening, performance loss, and steering angle limitation after triggering failure classification detection, based on the actuator current, actual angle change rate, high-frequency variance of actual angle, and the relationship between the target angle change rate and a preset threshold; wherein, the high-frequency variance characterizes the high-frequency oscillation characteristics of the actual angle. The smooth transition reference generation module is used to construct a two-degree-of-freedom reference model. Based on the yaw rate within a preset time period before the failure, it calculates the pre-failure trend reference yaw rate and the diagnosed failure type, performs equivalent adjustments to the parameters of the two-degree-of-freedom reference model, and calculates the yaw rate threshold in conjunction with the road adhesion coefficient as a safe reference value for the yaw rate after the failure. The smooth transition reference yaw rate is calculated using the smooth transition formula. The predictive control fault-tolerant allocation module is used to construct a nonlinear predictive control model. Taking the smooth transition reference yaw rate as the tracking target, it constructs a cost function that includes state error, control input, control input increment, and slack variables. Based on the diagnosed failure type, it applies additional constraints corresponding to the failure type to the nonlinear predictive control model. It then performs rolling optimization to solve the cost function, obtains the optimal control input, and allocates it to each actuator.

9. An electronic device, characterized in that, include: The system includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements a hierarchical collaborative fault-tolerant control method for single-wheel steering failure of a corner module vehicle as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: The computer program causes the computer to execute a hierarchical collaborative fault-tolerant control method for single-wheel steering failure of a corner module vehicle as described in any one of claims 1 to 7.

Images