Fault-tolerant control method, system and vehicle for steering lockup failure of four-wheel steering vehicle

Abstract

The present application belongs to the field of vehicle control, and particularly relates to a four-wheel steering vehicle steering stick failure fault-tolerant control method, system and vehicle. The method comprises three layers of control strategies. The upper layer models the vehicle as a two-degree-of-freedom single-track vehicle model, generates the expected steering angle of the vehicle capable of tracking the predetermined trajectory by using MPC, and calculates the ideal yaw rate and lateral velocity of the vehicle. The middle layer controller is responsible for receiving the output of the upper layer controller, calculating the control deviation of the vehicle, constructing a three-degree-of-freedom dynamic model and constraint conditions of the vehicle based on the Uka theory, and then solving the control force of each wheel satisfying the constraint. The lower layer optimizer is used to minimize the tire load rate as the optimization objective, consider the constraint condition of the tire force, and redistribute the torque and steering angle of each wheel capable of achieving the control force of each wheel through iterative optimization. The trajectory tracking and fault-tolerant control of the full-line controlled distributed drive vehicle are realized, and the problems of insufficient reliability and scene limitation of the traditional scheme are overcome.

Description

Technical Field

[0001] This invention belongs to the field of vehicle control, specifically relating to a fault-tolerant control method for steering jamming failure in a four-wheel steering vehicle, and its corresponding vehicle control system and fully drive-by-wire distributed drive vehicle. Background Technology

[0002] In fully drive-by-wire distributed electric vehicles, each wheel is equipped with a wheel-side / hub electric drive system, ensuring precise control of torque across all four wheels. When one or more electric drive systems fail, the additional yaw moment caused by unequal driving forces on opposite sides can easily lead to vehicle yaw instability, compromising vehicle safety. To address this issue, torque reconfiguration of multiple drive units can ensure both vehicle power and safety.

[0003] The technical solution disclosed in Chinese invention patent application 202510738744.1 utilizes the differential steering torque of the front axle to assist the faulty steer-by-wire system in smoothly tracking the desired steering wheel angle when the steering actuator motor fails, thereby meeting the vehicle trajectory tracking requirements for steering and improving the stability of vehicle trajectory tracking. The technical solution disclosed in Chinese invention patent application 202410119029.5 utilizes the differential steering characteristic of the front wheels of a distributed drive vehicle to compensate for the front wheel angle when the steering motor completely fails. Simultaneously, when a drive wheel fails severely, the steering motor's ability to actively apply additional steering angle helps the vehicle quickly correct its posture, thus ensuring vehicle safety and stability.

[0004] However, existing control methods for steering fault tolerance primarily rely on tracking and controlling vehicle state parameters to solve for differential torque, thereby achieving lateral stability control when a single wheel fails. Dynamic modeling often references a single-track two-degree-of-freedom model, assuming uniform vehicle speed and neglecting the impact of speed variations on lateral stability under actual operating conditions. Furthermore, existing fault-tolerant control methods offer limited consideration for vehicle control in rear-wheel sticking failure scenarios, failing to address the fault-tolerant control problem in this situation. Therefore, rear-wheel sticking failure can easily lead to vehicle instability, posing a significant safety risk to drivers and passengers. Summary of the Invention

[0005] To address the stability issues of fully steerable distributed drive vehicles under speed variations and rear wheel jamming failures, this invention provides a fault-tolerant control method for steering jamming failures in four-wheel steering vehicles, along with a corresponding vehicle control system and a fully steerable distributed drive vehicle.

[0006] This invention is achieved using the following technical solution: A fault-tolerant control method for steering sticking failure in a four-wheel steering vehicle, comprising: S1: Based on a two-degree-of-freedom monorail vehicle model, a state-space equation representing the vehicle's motion is created using the front wheel rotation angle as input; and a model predictive controller is designed by combining the vehicle's state-space equation with the goal of tracking a preset path.

[0007] S2: Collect real-time vehicle status data, and use the model predictive controller to generate the expected front wheel steering angle for future moments that enable the vehicle to follow the preset path based on the real-time status data; and calculate the ideal lateral speed and ideal yaw rate at the current moment.

[0008] S3: Construct a three-degree-of-freedom dynamic model of the vehicle, calculate the error vector based on real-time state data, and determine the vehicle's constraints by combining ideal lateral velocity and ideal yaw rate. Based on Vucic's theory, transform the dynamic model and constraints into matrix form, and then solve for the wheel control forces that satisfy the constraints.

[0009] S4: Taking the four-wheel steering angle and four-wheel torque as the decision objects, the objective function is defined with minimizing the tire load rate as the optimization objective. The four-wheel output is constrained to meet the wheel control force and the output torque and road adhesion conditions are lower than the preset values. An iterative optimization model is constructed.

[0010] S5: Under non-fault conditions, the iterative optimization model generates optimized results for the steering angles and torques of all four wheels based on the current wheel control forces. Under steering sticking failure conditions, the steering angle of the faulty wheel is fixed at the current value, and the same strategy is used to optimize the torques of all four wheels and the steering angles of the remaining wheels.

[0011] As a further improvement of the present invention, in step S1, the expression for the two-degree-of-freedom monorail vehicle model is: ; In the above formula, It is the centroid sideslip angle; This refers to the sideslip angular velocity; This refers to the yaw rate; This is the yaw acceleration; For heading angle; Y represents the angular velocity of the heading; Y is the lateral displacement in the geodetic coordinate system. X represents the lateral velocity in the geodetic coordinate system; X represents the longitudinal displacement in the geodetic coordinate system. I represents the longitudinal velocity in the geodetic coordinate system. z Let v be the vehicle's moment of inertia about the Z-axis; m be the vehicle's mass; v and v' be the vehicle's rotational inertia. x v y These are the vehicle velocity, longitudinal velocity, and lateral velocity in the vehicle coordinate system, respectively. The steering angle of the front wheels; For the rear wheel steering angle, l f and lr These are the distances from the front axle and rear axle to the vehicle's center of gravity, respectively. and These are the lateral forces on the front and rear wheels of the vehicle, respectively.

[0012] As a further improvement of the present invention, in step S1, the optimization objective of the model prediction controller is... The expression is: ; In the above formula, This represents the output vector predicted at time t+i from time t. Represents the reference output vector at time t+i; N represents the control increment at time t+i; p N represents the length of the prediction time domain; C represents the length of the control time domain; Q represents the weight matrix for tracking accuracy; R represents the weight matrix for control increment; These are the preset weighting coefficients for the relaxation terms; This is the preset relaxation factor.

[0013] As a further improvement of the present invention, in step S2, the ideal lateral velocity and ideal yaw rate The calculation formula is: ; In the above formula, C f Indicates the front wheel lateral stiffness; C r This indicates the rear wheel lateral stiffness; L indicates the vehicle's wheelbase. The feedback gain represents the lateral motion state of the vehicle. The ratio of the front and rear wheel steering angles is represented by ; N represents the system characteristic coefficient. .

[0014] As a further improvement of the present invention, in step S3, the expression of the three-degree-of-freedom dynamic model of the vehicle is: ; In the above formula, Let x be the second derivative of the vehicle's longitudinal displacement, i.e., the longitudinal acceleration. Let y be the second derivative of the vehicle's lateral displacement y, i.e., the lateral acceleration; v represents the vehicle's velocity. and To represent the vehicle yaw angles The first and second derivatives, i.e., yaw rate and yaw acceleration; F x and F y These are the longitudinal force and lateral force of the vehicle, respectively; M z This refers to the yaw moment of the vehicle.

[0015] Based on Vucic's theory, its matrix form is obtained as follows: ; Where M represents the inertia matrix, and These represent the first-order and second-order steering variables of the state vector q, respectively; C represents the Coriolis force matrix; This represents the input control force vector; and satisfies: .

[0016] As a further improvement of the present invention, the formula for calculating the error vector e is as follows: ; In the above formula, q d x represents the desired state vector of the vehicle; d y d and These represent the ideal longitudinal displacement, ideal lateral displacement, and ideal yaw angle of the vehicle, respectively; e1, e2, and e3 represent the longitudinal displacement error, lateral displacement error, and yaw angle error of the vehicle, respectively.

[0017] The constraints are: ; In the above formula, These are the first derivatives of e1, e2, and e3, respectively; The second derivatives of e1, e2, and e3 are respectively; , , , , , These are the coefficients of the error term or the first-order error term, respectively.

[0018] The matrix form of the constraints, derived from Vucic's theory, is as follows: ; in, ; x d y d and The second derivative of .

[0019] As a further improvement of the present invention, in step S3, the control force vector is determined based on the Uka theory. The calculation formula is as follows: ; In the above formula, p1 and p2 represent the nominal force vector and the additional force vector of the robust stability control, respectively; M represents the square root matrix of the moment of inertia; P represents a pre-selected positive definite symmetric constant matrix, which is set here as the identity matrix; This represents the control gain coefficient.

[0020] As a further improvement of the present invention, the objective function of the iterative optimization model in step S4 is... The expression is: ; In the above formula, F txij F represents the longitudinal force of the wheel with index ij. tyij F represents the lateral force of the wheel with index ij. zij This indicates the vertical load on the wheel with index ij; when ij = fl, fr, rl, rr, they represent the left front wheel, right front wheel, left rear wheel, and right rear wheel of the vehicle, respectively. This is the road surface adhesion coefficient.

[0021] As a further improvement of the present invention, the constraint conditions of the iterative optimization model in step S4 are as follows: ; In the above formula, This represents the wheel angle of the wheel with index ij. , , and These represent the wheel angles of the vehicle's left front wheel, right front wheel, left rear wheel, and right rear wheel, respectively; F txfl F txfr F txrl F txrr F represents the longitudinal force on the left front wheel, right front wheel, left rear wheel, and right rear wheel of the vehicle, respectively. tyfl F tyfr F tyrl F tyrr These represent the lateral forces on the left front wheel, right front wheel, left rear wheel, and right rear wheel of the vehicle, respectively; b l It represents half the track width between the front and rear wheels; r ij Indicates the wheel radius; M z T represents the yaw moment of the vehicle. ijmax This represents the upper limit of the output torque of the wheel with index ij; C D A represents the drag coefficient. D Indicates the windward area.

[0022] The present invention also includes a vehicle control system, which includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it adopts the fault-tolerant control method for steering jamming failure of a four-wheel steering vehicle as described above to create an upper-level controller, a middle-level controller, and a lower-level optimizer; thereby realizing the generation and issuance of control commands containing the torque and steering angle of each wheel of the vehicle based on the real-time status data of the vehicle.

[0023] The upper-level controller is based on a two-degree-of-freedom vehicle model, using the front wheel steering angle as input to create a state-space equation representing the vehicle's motion. A model predictive controller is designed based on the vehicle's state-space equation, with the optimization objective of tracking a preset path. Real-time vehicle state data is then collected, and the model predictive controller uses this data to generate the ideal front wheel steering angle for future moments that enable the vehicle to track the preset roadbed; it also calculates the ideal lateral velocity and ideal yaw rate at the current moment.

[0024] The intermediate controller is used to construct a three-degree-of-freedom dynamic model of the vehicle, calculate the error vector based on real-time state data, and determine the vehicle's constraints by combining ideal lateral velocity and ideal yaw rate. Then, based on Vucic's theory, the dynamic model and constraints are transformed into matrix form, and the wheel control forces that satisfy the constraints are solved.

[0025] The lower-level optimizer uses the four-wheel steering angle and four-wheel torque as decision objects, minimizes the tire load rate as the optimization objective, and defines an objective function. The constraints are that the four-wheel output must meet wheel control force requirements and that the output torque and road adhesion conditions are below preset values. An iterative optimization model is constructed. Under non-fault conditions, the iterative optimization model generates the optimized results for each of the four-wheel steering angle and four-wheel torque based on the current wheel control force. Under steering jamming failure conditions, the steering angle of the faulty wheel is fixed at the current value, and the same strategy is used to optimize the four-wheel torque and the steering angles of the remaining wheels.

[0026] The present invention also includes a fully drive-by-wire distributed drive vehicle that employs the vehicle control system described above.

[0027] The technical solution provided by this invention has the following beneficial effects: This invention proposes a novel three-layer control strategy for four-wheel steering vehicles. The upper-layer controller is based on a simplified two-degree-of-freedom single-rail vehicle model and uses the MPC method to efficiently solve for the ideal lateral velocity and ideal yaw rate to achieve trajectory tracking. In the middle-layer controller, this invention uses a method based on Ukkarian theory to efficiently solve for the total vehicle control force; it introduces second-order constraints along with first-order constraints to reduce the vehicle's trajectory tracking error. Furthermore, it introduces a robust stability term control force on top of the nominal control force to ensure the strong robustness of the control method and enhance the stability of the vehicle's trajectory tracking in the event of steering failure. In the lower-layer optimizer, for any wheel steering sticking failure scenario, this invention obtains the longitudinal / lateral forces and additional yaw moment of the vehicle by tracking the desired trajectory, and designs an iterative optimization model to minimize the tire load rate by combining the constraints of the four-wheel tire forces. This enables dynamic reconstruction of the chassis angle and driving torque distribution according to the failure scenario; thus, it improves the stability control effect of four-wheel steering vehicles in failure scenarios while reducing the tire load rate and ensuring vehicle safety. Attached Figure Description

[0028] Figure 1 This is a flowchart of the steps of the fault-tolerant control method for steering jamming failure in a four-wheel steering vehicle provided in Embodiment 1 of the present invention.

[0029] Figure 2 This is a schematic diagram of the two-degree-of-freedom monorail vehicle model established in Embodiment 1 of the present invention.

[0030] Figure 3 This is a schematic diagram of the three-degree-of-freedom dynamic model of the vehicle established in Embodiment 1 of the present invention.

[0031] Figure 4 This is a comparison chart showing the trajectory tracking performance of the control scheme of the present invention and the traditional control scheme without steering fault tolerance in a simulation experiment. Detailed Implementation

[0032] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0033] Example 1

[0034] To address the aforementioned shortcomings of traditional vehicle stability control schemes for fully drive-by-wire distributed drive vehicles, this embodiment provides a fault-tolerant control method for steering failure in four-wheel steering vehicles. This method designs a three-layer control strategy for control scenarios where steering failure occurs at any wheel in a four-wheel steering vehicle. This strategy dynamically allocates the steering angle and driving torque of each wheel in the chassis based on the vehicle's real-time status data, controlling the vehicle to move along a preset trajectory and enhancing vehicle stability and trajectory tracking accuracy during steering failure. In practical applications, the three-layer control strategy provided in this embodiment can be implemented using a vehicle control system comprising an upper-layer controller, a middle-layer controller, and a lower-layer optimizer.

[0035] In detail, such as Figure 1 As shown, the fault-tolerant control method for steering sticking failure in four-wheel steering vehicles provided in this embodiment adopts a three-layer control strategy, which includes the following steps: I. Upper-level controller In this embodiment, the upper-level controller is used to control the vehicle for trajectory tracking. To achieve this goal, the vehicle is modeled as a two-degree-of-freedom single-rail vehicle. Then, a Model Predictive Control (MPC) is used to dynamically plan the optimal control parameters for the vehicle to achieve closed-loop control. Specifically, the upper-level controller first acquires the vehicle's real-time state parameters, then uses the MPC to output the desired turning angle of the vehicle based on the vehicle's reference state parameters required for tracking the predetermined trajectory; finally, it calculates the vehicle's ideal yaw rate and ideal lateral velocity. The detailed control logic of the upper-level controller is as follows: S1: Based on a two-degree-of-freedom monorail vehicle model, a state-space equation representing the vehicle's motion is created using the front wheel rotation angle as input; and a model predictive controller is designed by combining the vehicle's state-space equation with the goal of tracking a preset path.

[0036] Among them, such as Figure 2 As shown in the figure These are the front wheel slip angle and the rear wheel slip angle, respectively. For reference heading angle, These are the front wheel speed and the rear wheel speed, respectively. This represents the longitudinal displacement difference in the geodetic coordinate system. The expression for the two-degree-of-freedom monorail vehicle model established in this embodiment is: ; In the above formula, It is the centroid sideslip angle; This refers to the sideslip angular velocity; This refers to the yaw rate; This is the yaw acceleration; For heading angle; Y represents the angular velocity of the heading; Y is the lateral displacement in the geodetic coordinate system. X represents the lateral velocity in the geodetic coordinate system; X represents the longitudinal displacement in the geodetic coordinate system. I represents the longitudinal velocity in the geodetic coordinate system. z Let v be the vehicle's moment of inertia about the Z-axis; m be the vehicle's mass; v and v' be the vehicle's rotational inertia. x v y These are the vehicle velocity, longitudinal velocity, and lateral velocity in the vehicle coordinate system, respectively. The steering angle of the front wheels; For the rear wheel steering angle, l f and l r These are the distances from the front axle and rear axle to the vehicle's center of gravity, respectively. and These are the lateral forces on the front and rear wheels of the vehicle, respectively.

[0037] To ensure that the rear wheel steering responds to changes in lateral motion and front wheel steering angle, the rear wheel steering angle... It should also satisfy the following formula: ; In the above formula, The feedback gain represents the lateral motion state of the vehicle. This represents the ratio of the steering angles of the front and rear wheels. The formula for calculating them is: ; Among them, C f Indicates the front wheel lateral stiffness; C r This indicates the rear wheel lateral stiffness.

[0038] In the upper-level controller of this embodiment, x is selected as the system state variable. ;u is the control input quantity. y represents the system output. Create a state-space equation representing the vehicle's motion. This state-space equation can be simplified as follows: ; in, Represents the system state variables at time t; The first derivative of the system state variables; C represents the system input at time t; C represents the system output observation matrix.

[0039] Model predictive control (MPC) is one of the most successful advanced control strategies in modern control theory. Its core idea is to use the mathematical model of the system to solve a finite-time optimization problem at each sampling time to obtain the optimal control sequence, but only implement the first control action, and then repeat the process at the next time step. This "rolling time" strategy enables MPC to naturally handle control objects in multivariable systems, complex constraints, and multi-optimization scenarios.

[0040] To better apply the MPC control strategy, this embodiment first linearizes the nonlinear vehicle control system, and then transforms it into the following linear time-varying equation based on the above state-space equation: ; In the above formula, Represents the time-varying state coefficient matrix. The time-varying input coefficient matrix, in this embodiment, satisfies the following equation: .

[0041] Furthermore, by discretizing the linear time-varying equation, we obtain: ; In the above formula, k represents the current sampling time; and These represent the system states at time k and time k+1, respectively. This represents the system input at time k; This represents the output matrix of the discretized system. This represents the discretized system input matrix; in this embodiment, the two satisfy: ; Where T is the sampling period and I represents the identity matrix.

[0042] Reconstruct a new state vector , This leads to the new state-space expression: ; In the above formula, k is the current sampling time, and k+1 is the next sampling time; This represents the control increment at time k-1; This represents the derived output at time k; Represents the state matrix; Indicates the input control matrix; Represents the system output matrix; and satisfies: .

[0043] In this embodiment, in order to enable the vehicle to accurately track the preset path, the optimization objective of the designed model prediction controller is... The expression is: ; In the above formula, This represents the output vector predicted at time t+i from time t. Represents the reference output vector at time t+i; N represents the control increment at time t+i; p N represents the length of the prediction time domain; C represents the length of the control time domain; Q represents the weight matrix for tracking accuracy; R represents the weight matrix for control increment; These are the preset weighting coefficients for the relaxation terms; This is the preset relaxation factor.

[0044] S2: Collect real-time vehicle status data, and use the model predictive controller to generate the expected front wheel steering angle for future moments that enable the vehicle to follow the preset path based on the real-time status data; and calculate the ideal lateral speed and ideal yaw rate at the current moment.

[0045] In the model predictive controller of this embodiment, for any time k, based on the system state at time k and the control increment in the control time domain, the output of the system at each future time in the prediction time domain can be derived as follows: ; Rewriting the above equation in matrix form yields: ; In the above formula, This represents the output quantity in the prediction time domain; This represents the system state matrix in the prediction time domain; This represents the state vector at sampling time k; This represents the control input matrix in the prediction time domain; Represents the control input in the prediction time domain; and satisfies: ; .

[0046] In this embodiment, the optimization objective of the model prediction controller is set as follows: middle R=50000; By solving the objective function in each control time domain, a series of optimal control increments can be obtained. Applying the first control increment at the current moment to the system yields the optimal control quantity at that moment, which is also the optimal front wheel steering angle increment for the next moment. Applying this to the system will create a closed-loop control for optimizing vehicle trajectory.

[0047] Next, this embodiment will introduce The new front wheel steering angle is taken as the expected front wheel steering angle at the current moment, and substituted into the following single-track two-DOF vehicle model with additional rear wheel steering: This yields the ideal lateral velocity of the vehicle. and ideal yaw rate : ; In this embodiment, the ideal lateral velocity and ideal yaw rate The calculation formula is: ; In the above formula, L represents the vehicle's wheelbase; N represents the system characteristic coefficient, which satisfies: .

[0048] II. Middle Layer Controller

[0049] In this embodiment, the middle-level controller receives the ideal yaw rate and ideal lateral velocity output from the upper-level controller and calculates the vehicle's control deviation based on the vehicle's real-time state parameters. Then, it constructs a three-degree-of-freedom dynamic model and constraints for the vehicle based on Udwadia–Kalaba theory. Finally, it solves for the control forces at each wheel that satisfy the constraints, including the additional yaw moment, total longitudinal force, and total lateral force. Specifically, the control logic of the middle-level controller is as follows: S3: Construct a three-degree-of-freedom dynamic model of the vehicle, calculate the error vector based on real-time state data, and determine the vehicle's constraints by combining ideal lateral velocity and ideal yaw rate. Based on Vucic's theory, transform the dynamic model and constraints into matrix form, and then solve for the wheel control forces that satisfy the constraints.

[0050] Under this control objective, it is necessary to construct such as Figure 3 The vehicle's three-degree-of-freedom dynamic model is shown. In addition, to facilitate the solution, this embodiment also needs to convert the vehicle's dynamic model and constraints into matrix form based on Uka theory.

[0051] Specifically, in this embodiment, the expression for the three-degree-of-freedom dynamics model of the vehicle is: ; In the above formula, Let x be the second derivative of the vehicle's longitudinal displacement, i.e., the longitudinal acceleration. Let y be the second derivative of the vehicle's lateral displacement y, i.e., the lateral acceleration. and To represent the vehicle yaw angles The first and second derivatives, i.e., yaw rate and yaw acceleration; F x and F y These are the longitudinal force and lateral force of the vehicle, respectively; M z This refers to the yaw moment of the vehicle.

[0052] Based on Uka theory, the matrix form of the dynamic model can be obtained as follows: ; Where M represents the inertia matrix, and These represent the first-order and second-order steering variables of the state vector q, respectively; C represents the Coriolis force matrix; This represents the input control force vector; and satisfies: .

[0053] To characterize the system's constraints, this embodiment first defines the formula for calculating the error vector e as follows: ; In the above formula, q dx represents the desired state vector of the vehicle; d y d and These represent the ideal longitudinal displacement, ideal lateral displacement, and ideal yaw angle of the vehicle, respectively; e1, e2, and e3 represent the longitudinal displacement error, lateral displacement error, and yaw angle error of the vehicle, respectively.

[0054] Then, based on the system error, the following constraints are set: ; In the above formula, These are the first derivatives of e1, e2, and e3, respectively; The second derivatives of e1, e2, and e3 are respectively; , , , , , These are the coefficients of the error term or the first-order error term, respectively. .

[0055] Furthermore, based on Vucic's theory, the matrix form of the constraints is obtained as follows: ; Where A represents the constraint matrix; b represents the error constraint vector; and the two satisfy: ; In the above formula, x d y d and The second derivative of .

[0056] In summary, by combining the vehicle dynamics model and the matrix form of the constraints under Uka theory, the control force vector can be obtained. The calculation formula is as follows: ; In the above formula, p1 and p2 represent the nominal force vector and the additional force vector of the robust stability control, respectively; M represents the square root matrix of the moment of inertia; P represents a pre-selected positive definite symmetric constant matrix, which is chosen here as the identity matrix; This represents the control gain coefficient.

[0057] III. Lower-level optimizer

[0058] In this embodiment, the lower-level optimizer, based on the control forces output by the middle-level controller, minimizes the tire load rate as the optimization objective, considers tire force constraints, and redistributes the torque and steering angle of each wheel to achieve trajectory tracking and fault-tolerant control of the fully steerable distributed drive vehicle. Specifically, the control logic of the lower-level optimizer is as follows: S4: Taking the four-wheel steering angle and four-wheel torque as the decision objects, the objective function is defined with minimizing the tire load rate as the optimization objective. The four-wheel output is constrained to meet the wheel control force and the output torque and road adhesion conditions are lower than the preset values. An iterative optimization model is constructed.

[0059] Specifically, in this embodiment, the objective function of the iterative optimization model is... The expression is: ; In the above formula, F txij F represents the longitudinal force of the wheel with index ij. tyij F represents the lateral force of the wheel with index ij. zij This indicates the vertical load on the wheel with index ij; when ij = fl, fr, rl, rr, they represent the left front wheel, right front wheel, left rear wheel, and right rear wheel of the vehicle, respectively. This is the road surface adhesion coefficient.

[0060] Considering that the wheel angles and torques allocated by the lower-level optimizer not only require control of the longitudinal force, lateral force, and yaw moment issued by the middle-level controller, but also need to be limited by the maximum output torque of the hub motor and the road surface adhesion conditions, this embodiment sets the constraints of the iterative optimization model as follows: ; In the above formula, This represents the wheel angle of the wheel with index ij. , , and These represent the wheel angles of the vehicle's left front wheel, right front wheel, left rear wheel, and right rear wheel, respectively; F txfl F txfr F txrl F txrr F represents the longitudinal force on the left front wheel, right front wheel, left rear wheel, and right rear wheel of the vehicle, respectively. tyfl F tyfr F tyrl F tyrr These represent the lateral forces on the left front wheel, right front wheel, left rear wheel, and right rear wheel of the vehicle, respectively. It represents half the track width between the front and rear wheels; r ij Indicates the wheel radius; T ijmax This represents the upper limit of the output torque of the wheel with index ij; C D A represents the drag coefficient. D Indicates the windward area.

[0061] S5: Under non-fault conditions, the iterative optimization model generates optimized results for the steering angles and torques of all four wheels based on the current wheel control forces. Under steering sticking failure conditions, the steering angle of the faulty wheel is fixed at the current value, and the same strategy is used to optimize the torques of all four wheels and the steering angles of the remaining wheels.

[0062] Based on the above iterative optimization model, this embodiment can iteratively optimize the optimal control force output by the middle-level controller for each round according to the objective function. Specifically, firstly, by The torques of the four tires are obtained, and the slip angles of each tire are obtained by combining the following inverse model of tire slip characteristics. : ; In the above formula, This indicates the lateral stiffness of the corresponding wheel; Indicates the longitudinal load attenuation coefficient; Represents the equivalent characteristic coefficient of a tire; and satisfies: , Where p represents the empirical fitting constant, p=2.9; Based on the relationship between wheel steering angle and tire slip angle, the wheel steering angle of each wheel can be expressed as:

[0063] In the above formula, These represent the sideslip angles of the vehicle's left front wheel, right front wheel, left rear wheel, and right rear wheel, respectively.

[0064] Get the wheel angles of the four wheels back, Then calculate the torque T of the four wheels. ij , Then according to and T ij Joint control of the vehicle.

[0065] In practical applications, vehicles need to flexibly adopt differentiated wheel steering angle and torque distribution strategies based on fault conditions: Assuming a vehicle failure, during steering, the right rear wheel may experience steering sticking at a certain moment due to a steering motor malfunction, i.e., after a certain moment t. The constant value is used to simulate failure scenarios and represents the output of the lower-level optimizer. The value should be constant at time t, meaning that the steering angle inputted by the lower-level optimizer to the vehicle model after time t is always the same as the steering angle at time t. In this case, the lower-level optimizer needs to redistribute the torque output of each wheel based on the new constraint that the steering angle of the faulty wheel is fixed. At this time, and At time t, the torque and angle of the remaining wheels are redistributed according to the new constraints. This optimizes the torque and angle so that they can reduce the impact of the failed wheel and allow the vehicle to continue traveling along the predetermined trajectory. They can also minimize the tire load rate, thereby improving tire life and vehicle safety.

[0066] Example 2

[0067] To better apply the solution in Embodiment 1, this embodiment further provides a vehicle control system, which includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it adopts the fault-tolerant control method for steering jamming failure of a four-wheel steering vehicle as described above to create an upper-level controller, a middle-level controller, and a lower-level optimizer; thereby realizing the generation and issuance of control commands containing the torque and steering angle of each wheel of the vehicle based on the real-time status data of the vehicle.

[0068] The upper-level controller is based on a two-degree-of-freedom monorail vehicle model, using the front wheel angle as input to create a state-space equation representing the vehicle's motion. A model predictive controller is designed based on the vehicle's state-space equation, with the optimization objective of tracking a preset path. Real-time vehicle state data is then collected, and the model predictive controller generates the ideal front wheel angle for future moments that enables the vehicle to track the preset roadbed, based on this data. The ideal lateral velocity and ideal yaw rate are also calculated for the current moment.

[0069] The intermediate controller is used to construct a three-degree-of-freedom dynamic model of the vehicle, calculate the error vector based on real-time state data, and determine the vehicle's constraints by combining ideal lateral velocity and ideal yaw rate. Then, based on Vucic's theory, the dynamic model and constraints are transformed into matrix form, and the wheel control forces that satisfy the constraints are solved.

[0070] The lower-level optimizer uses the four-wheel steering angle and four-wheel torque as decision objects, minimizes the tire load rate as the optimization objective, and defines an objective function. The constraints are that the four-wheel output must meet wheel control force requirements and that the output torque and road adhesion conditions are below preset values. An iterative optimization model is constructed. Under non-fault conditions, the iterative optimization model generates the optimized results for each of the four-wheel steering angle and four-wheel torque based on the current wheel control force. Under steering jamming failure conditions, the steering angle of the faulty wheel is fixed at the current value, and the same strategy is used to optimize the four-wheel torque and the steering angles of the remaining wheels.

[0071] Furthermore, this embodiment also provides a fully steerable distributed drive vehicle integrated with the aforementioned vehicle control system. This vehicle can perform autonomous driving according to a preset vehicle trajectory. In the event of wheel steering failure, it can also flexibly and adaptively redistribute the steering angle and torque of each wheel, thereby reducing the impact of the failed wheel and allowing the vehicle to continue traveling along the predetermined trajectory. In addition, the new vehicle stability control strategy can minimize tire load, improving tire lifespan and vehicle driving safety. Compared to existing control strategies, the technical solution of this invention offers superior stability, reliability, and safety, especially overcoming driving safety issues in scenarios involving wheel steering failure.

[0072] Simulation Experiment

[0073] To verify the performance of the fault-tolerant control method for steering sticking failure in four-wheel steering vehicles provided by this invention, technicians conducted simulations and tests on the relevant schemes.

[0074] The simulated vehicle parameters in this experiment include: m = 1609 kg, I z =1765kg·m 2 ;l f =1.15m, l r =1.35m; b l =0.76m, C f =-75000 N / rad; C r =-80000N / rad, drag coefficient C D =0.28, windward area A D =2.15m 2 T ijmax =400 N·m; =0.85. Some control parameters of the intermediate controller. =55s -1 , =38s -2 ; =55s -1 ; =55s -2 ; =55s -1 ; =55s -2 .

[0075] The simulation conditions for this experiment are as follows: when the car travels along a predetermined trajectory (i.e., the double lane change condition), the coefficient of adhesion of the vehicle on the road surface is adjusted according to the predetermined trajectory. The working condition was simulated on a high adhesion road surface with a coefficient of friction of 0.85, and the right rear wheel failed and jammed at time t = 2s.

[0076] Under the above simulation conditions, the trajectory tracking performance of the control scheme provided by this invention and the traditional control scheme without steering fault tolerance in the simulation scenario is compared. The obtained trajectory tracking performance is as follows: Figure 4 As shown in the figure, analysis of the data reveals that the control scheme of this invention still exhibits excellent trajectory tracking performance even when rear wheel jamming occurs, enabling fault-tolerant control of four-wheel steering vehicles.

[0077] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A fault-tolerant control method for steering sticking failure in a four-wheel steering vehicle, characterized in that, It includes: S1: Based on a two-degree-of-freedom monorail vehicle model, a state-space equation representing the vehicle's motion is created using the front wheel rotation angle as input. A model predictive controller is designed based on the vehicle's state-space equations, with the goal of tracking a preset path. S2: Collect real-time vehicle status data, and use the model prediction controller to generate the expected front wheel steering angle for future moments that enable the vehicle to follow the preset path based on the real-time status data; and calculate the ideal lateral speed and ideal yaw rate at the current moment. S3: Construct a three-degree-of-freedom dynamic model of the vehicle, calculate the error vector based on real-time state data, and determine the vehicle's constraints by combining ideal lateral velocity and ideal yaw rate; based on Vucic's theory, transform the dynamic model and constraints into matrix form, and then solve for the wheel control force that can satisfy the constraints. S4: Taking the four-wheel steering angle and four-wheel torque as the decision objects, the objective function is defined with minimizing the tire load rate as the optimization objective. The four-wheel output is constrained to meet the wheel control force and the output torque and road adhesion conditions are lower than the preset values. An iterative optimization model is constructed. S5: In a non-faulty state, the optimization model generates the optimized results of the four wheel angles and four wheel torques based on the wheel control force at the current moment through iterative optimization; under the condition of steering jam failure, the wheel angle of the faulty side is fixed at the current value, and the four wheel torques and the other wheel angles are optimized using the same strategy.

2. The fault-tolerant control method for steering jamming failure in a four-wheel steering vehicle as described in claim 1, characterized in that: In step S1, the expression for the two-degree-of-freedom monorail vehicle model is: ； In the above formula, It is the centroid sideslip angle; This refers to the sideslip angular velocity; This refers to the yaw rate; This is the yaw acceleration; For heading angle; Y represents the angular velocity of the heading; Y is the lateral displacement in the geodetic coordinate system. X represents the lateral velocity in the geodetic coordinate system; X represents the longitudinal displacement in the geodetic coordinate system. I represents the longitudinal velocity in the geodetic coordinate system. z Let v be the vehicle's moment of inertia about the Z-axis; m be the vehicle's mass; v and v' be the vehicle's rotational inertia. x v y These are the vehicle velocity, longitudinal velocity, and lateral velocity in the vehicle coordinate system, respectively. The steering angle of the front wheels; For the rear wheel steering angle, l f and l r These are the distances from the front axle and rear axle to the vehicle's center of gravity, respectively. and These are the lateral forces on the front and rear wheels of the vehicle, respectively.

3. The fault-tolerant control method for steering jamming failure in a four-wheel steering vehicle as described in claim 2, characterized in that: In step S1, the optimization objective of the model predictive controller is... The expression is: ； In the above formula, This represents the output vector predicted at time t+i from time t. Represents the reference output vector at time t+i; N represents the control increment at time t+i; p N represents the length of the prediction time domain; C represents the length of the control time domain; Q represents the weight matrix for tracking accuracy; R represents the weight matrix for control increment; These are the preset weighting coefficients for the relaxation terms; This is the preset relaxation factor.

4. The fault-tolerant control method for steering jamming failure in a four-wheel steering vehicle as described in claim 1, characterized in that: In step S2, the ideal lateral velocity and ideal yaw rate The calculation formula is: ； In the above formula, C f Indicates the front wheel lateral stiffness; C r This indicates the rear wheel lateral stiffness; L indicates the vehicle's wheelbase. The feedback gain represents the lateral motion state of the vehicle. The ratio of the front and rear wheel steering angles is represented by ; N represents the system characteristic coefficient. .

5. The fault-tolerant control method for steering sticking failure in a four-wheel steering vehicle as described in claim 4, characterized in that, In step S3, the expression for the vehicle's three-degree-of-freedom dynamics model is: ； In the above formula, Let x be the second derivative of the vehicle's longitudinal displacement, i.e., the longitudinal acceleration. Let y be the second derivative of the vehicle's lateral displacement y, i.e., the lateral acceleration; v represents the vehicle's velocity. and To represent the vehicle yaw angles The first and second derivatives, i.e., yaw rate and yaw acceleration; F x and F y These are the longitudinal force and lateral force of the vehicle, respectively; M z The yaw moment of the vehicle; Based on Vucic's theory, its matrix form is obtained as follows: ; Where M represents the inertia matrix, and These represent the first-order and second-order steering variables of the state vector q, respectively; C represents the Coriolis force matrix; This represents the input control force vector; and satisfies: ; And / or, the formula for calculating the error vector e is: ； In the above formula, q d x represents the desired state vector of the vehicle; d y d and These represent the ideal longitudinal displacement, ideal lateral displacement, and ideal yaw angle of the vehicle, respectively; e1, e2, and e3 represent the longitudinal displacement error, lateral displacement error, and yaw angle error of the vehicle, respectively. The constraints are: ； In the above formula, These are the first derivatives of e1, e2, and e3, respectively; The second derivatives of e1, e2, and e3 are respectively; , , , , , These are the coefficients of the error term or the first-order error term, respectively. The matrix form of the constraints, derived from Vucic's theory, is as follows: ; in, ; x d y d and The second derivative of .

6. The fault-tolerant control method for steering jamming failure in a four-wheel steering vehicle as described in claim 5, characterized in that: In step S3, based on Vucic's theory, the control force vector The calculation formula is as follows: ； In the above formula, p1 and p2 represent the nominal force vector and the additional force vector of the robust stability control, respectively; M represents the square root matrix of the moment of inertia; P represents a pre-selected positive definite symmetric constant matrix. This represents the control gain coefficient.

7. The fault-tolerant control method for steering sticking failure in a four-wheel steering vehicle as described in claim 6, characterized in that: In step S4, the objective function of the iterative optimization model is... The expression is: ； In the above formula, F txij F represents the longitudinal force of the wheel with index ij. tyij F represents the lateral force of the wheel with index ij. zij This indicates the vertical load on the wheel with index ij; when ij = fl, fr, rl, rr, they represent the left front wheel, right front wheel, left rear wheel, and right rear wheel of the vehicle, respectively. This is the road surface adhesion coefficient.

8. The fault-tolerant control method for steering jamming failure in a four-wheel steering vehicle as described in claim 7, characterized in that: The constraints of the iterative optimization model in step S4 are: ； In the above formula, This represents the wheel angle of the wheel with index ij. and These represent the wheel angles of the vehicle's left front wheel, right front wheel, left rear wheel, and right rear wheel, respectively; F txfl F txfr F txrl F txrr F represents the longitudinal force on the left front wheel, right front wheel, left rear wheel, and right rear wheel of the vehicle, respectively. tyfl F tyfr F tyrl F tyrr These represent the lateral forces on the left front wheel, right front wheel, left rear wheel, and right rear wheel of the vehicle, respectively; b l It represents half the track width between the front and rear wheels; r ij Indicates the wheel radius; M z T represents the yaw moment of the vehicle. ijmax This represents the upper limit of the output torque of the wheel with index ij; C D A represents the drag coefficient. D Indicates the windward area.

9. A vehicle control system, characterized in that: It includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it adopts the fault-tolerant control method for steering jamming failure of a four-wheel steering vehicle as described in any one of claims 1-8 to create an upper-level controller, a middle-level controller, and a lower-level optimizer; thereby realizing the generation and issuance of control commands containing the torque and steering angle of each wheel of the vehicle based on the real-time status data of the vehicle. The upper-level controller is based on a two-degree-of-freedom vehicle model. It creates a state-space equation representing the vehicle's motion with the front wheel angle as input. It also designs a model predictive controller based on the vehicle's state-space equation, with the goal of tracking a preset path. The controller collects real-time vehicle state data and uses the model predictive controller to generate an ideal front wheel angle for future moments that enables the vehicle to track the preset roadbed. It also calculates the ideal lateral velocity and ideal yaw rate at the current moment. The intermediate controller is used to construct a three-degree-of-freedom dynamic model of the vehicle, calculate the error vector based on real-time state data, and determine the vehicle's constraints by combining ideal lateral velocity and ideal yaw rate; based on Vucic's theory, the dynamic model and constraints are transformed into matrix form, and then the wheel control force that can satisfy the constraints is solved. The lower-level optimizer uses the four-wheel steering angle and four-wheel torque as decision objects, minimizes the tire load rate as the optimization objective, and defines an objective function. The four-wheel output is constrained to meet the wheel control force and the output torque and road adhesion conditions are lower than preset values. An iterative optimization model is constructed. Under non-fault conditions, the iterative optimization model generates the optimization results of each four-wheel steering angle and four-wheel torque based on the wheel control force at the current moment. Under the condition of steering jam failure, the steering angle of the faulty wheel is fixed at the current value, and the same strategy is used to optimize the four-wheel torque and the steering angle of the other wheels.

10. A fully drive-by-wire distributed drive vehicle, characterized in that: It employs the vehicle control system as described in claim 9.

Images