A four-wheel steering vehicle steering lock cross system fault-tolerant control method

Abstract

The application is suitable for the field of intelligent driving technology, and provides a four-wheel steering vehicle steering stuck cross system fault-tolerant control method, which comprises the following steps: step one: vehicle dynamics reconstruction, taking the stuck steering shaft rotation angle as a state quantity, and taking the healthy steering shaft and the brake braking force as control quantities, so as to realize dynamic reconstruction of the vehicle dynamics model; step two: error calculation, calculating the lateral position error and the heading error of the current position of the vehicle and the expected trajectory point according to the Frenet formula; step three: expected lateral velocity and yaw rate calculation, combining the vehicle state to calculate the lateral velocity and the yaw angular velocity required for the vehicle to track the expected trajectory; and step four: fault-tolerant control quantity calculation, distributing the healthy steering shaft rotation angle and the left and right side brake braking force according to the expected lateral velocity, the expected yaw angular velocity and the vehicle state to realize fault-tolerant control, which has the advantages of good fault tolerance performance and attitude stability control ability.

Description

Technical Field

[0001] This invention relates to the field of intelligent driving technology, specifically to a cross-system fault-tolerant control method for steering jamming in a four-wheel steering vehicle. Background Technology

[0002] As the core execution component of an automated hierarchical framework, the motion control system directly determines whether an autonomous vehicle can reliably and stably follow the desired trajectory, making it crucial for ensuring vehicle safety. Autonomous vehicles incorporate an increasing number of sensors, controllers, and actuators, leading to a corresponding increase in the probability of malfunctions. Based on the location of the malfunction, these can be categorized as sensor failures, ECU failures, communication failures, and actuator failures. This study designs a fault-tolerant control system for a multi-axle steering autonomous vehicle under the condition of steering actuator jamming. When the steering system of a multi-axle steering vehicle jams, the vehicle cannot obtain sufficient lateral force to perform steering actions, losing some or all of its steering ability. Under the action of unexpected lateral forces, the vehicle may deviate from the expected trajectory, potentially causing traffic accidents. Therefore, to ensure safety, a fault-tolerant control function needs to be designed to ensure that the vehicle maintains basic trajectory tracking ability even when the steering system is jammed.

[0003] Current research on fault-tolerant control focuses on steering system jamming faults under high-speed conditions, and most existing control algorithms are designed for specific fault conditions, lacking system analysis, modeling, and fault-tolerant algorithm design for multi-axle steering system fault types. Furthermore, most current research simply designs fault-tolerant functions from the perspective of algorithm constraints, without proposing adaptive fault-tolerant control methods based on dynamic characteristic analysis under fault conditions. Summary of the Invention

[0004] In view of the shortcomings of the existing technology, the purpose of this invention is to provide a cross-system fault-tolerant control method for steering jamming in four-wheel steering vehicles, so as to solve the problems in the background technology mentioned above.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A cross-system fault-tolerant control method for steering jamming in a four-wheel steering vehicle includes the following steps:

[0007] Step 1: Vehicle dynamics reconstruction, using the stuck steering shaft angle as the state variable and the healthy steering shaft and brake force as the control variables, to achieve dynamic reconstruction of the vehicle dynamics model;

[0008] Step 2: Error Calculation. Calculate the lateral position error between the vehicle's current position and the desired trajectory point using the Frenet formula. and heading error ;

[0009] Step 3: Calculate the desired lateral velocity and yaw rate. Design a fault-tolerant control algorithm based on a hierarchical collaborative framework. The upper-level control algorithm calculates the lateral velocity and yaw rate required for the vehicle to track the desired trajectory based on the lateral error and heading error of the vehicle trajectory tracking, combined with the vehicle state.

[0010] Step 4: Calculate the fault-tolerant control quantity. Design a fault-tolerant control algorithm based on a hierarchical collaborative framework. The lower-level control algorithm uses a dynamic model that is dynamically reconstructed based on the fault type of the vehicle steering system. It allocates the healthy steering shaft angle and the braking force of the left and right brakes according to the desired lateral speed, desired yaw rate and vehicle state to achieve fault-tolerant control.

[0011] As a further aspect of the present invention, the vehicle dynamics model is as follows:

[0012] ;

[0013] in A For the vehicle state matrix; C f For the lateral stiffness of the vehicle's front wheels; C r The rear wheel lateral stiffness of the vehicle; a This is the distance from the front axle to the center of gravity of the vehicle. b This is the distance from the rear axle to the center of gravity of the vehicle. For vehicle quality; The vehicle's current longitudinal speed; Let be the moment of inertia about the z-axis; B is the control matrix. C This is the output matrix; D is the feedforward matrix; I represents the identity matrix; 0 represents the zero matrix; w This refers to the distance between the left and right wheels of the vehicle.

[0014] As a further aspect of the present invention, in step one, when the current axis is stuck, the system state... ,in It is the vehicle's current lateral speed. It is the vehicle's current yaw rate. This is the front wheel steering angle, which is a constant here; the model outputs the control quantity. ,in For the rear wheel steering angle, This represents the difference in braking force between the left and right brakes. T The matrix transpose symbol is used to reconstruct the state matrix. for:

[0015] ;

[0016] in For the vehicle's front axle sideslip angle; reconstruct the control matrix. for:

[0017] ;

[0018] in w The distance between the left and right wheels of the vehicle;

[0019] When the rear axle is stuck, the system status is as follows: ,in This refers to the rear wheel steering angle, which is a constant here; the model outputs the control quantity. Reconstructing the state equation for:

[0020] ;

[0021] in Rear wheel sideslip angle; reconstruct control matrix for:

[0022] ;

[0023] When both the front and rear axles are stuck, and They are all constants. Reconstruct the state matrix for:

[0024] ;

[0025] Reconstruct the control matrix for:

[0026] .

[0027] As a further aspect of the present invention, the error calculation formula in step two is as follows:

[0028] ;

[0029] in The derivative of the heading error; This is the derivative of the lateral position error; This refers to the vehicle's heading angle; The heading angle is used as a reference trajectory point; The desired yaw rate; Let be the trajectory curvature; let Let the vehicle position vector be... The reference trajectory point position vector; The lateral velocity unit vector of the reference trajectory point; The vehicle's current longitudinal speed; This represents the vehicle's current lateral speed.

[0030] As a further embodiment of the present invention, the model state equation in step three can be expressed as:

[0031] ;

[0032] in The derivative of the state vector of the upper-level control algorithm; This is the observation vector for the upper-level control algorithm; This represents the state vector of the upper-level control algorithm. The upper-level control algorithm controls the output vector. The trajectory curvature is a constant here; the matrices in the state equations are:

[0033] ;

[0034] in and Let these be the state matrix, control matrix, output matrix, and feedforward matrix of the upper-level control algorithm, respectively. Based on model predictive control, the upper-level control algorithm is designed, and its cost function is defined as follows:

[0035] ;

[0036] in This is the cost function of the upper-level control algorithm; j For the first j One prediction step; N For prediction in the time domain; For the first One time step; and This is the weight matrix for the state vector and the control vector.

[0037] As a further aspect of the present invention, the constraint condition of the cost function of the upper-level control algorithm is:

[0038] .

[0039] As a further aspect of the present invention, the state equation of the lower-level control algorithm in step four is as follows:

[0040] ;

[0041] in The derivative of the state vector of the lower-level control algorithm; For the observation vector of the lower-level control algorithm; This represents the state vector of the lower-level control algorithm. The lower-level control algorithm controls the output vector; and These are the state matrix, control matrix, output matrix, and feedforward matrix of the lower-level control algorithm, respectively.

[0042] The lower-level control algorithm is designed based on the model predictive control algorithm. The cost function of the lower-level control algorithm is defined as follows:

[0043] ;

[0044] in The cost function for the lower-level control algorithm; The lower-level control algorithm controls the output change. It is a relaxation factor; H It is a Hessian matrix; g This is the gradient vector.

[0045] As a further aspect of the present invention, the constraint condition of the cost function of the lower-level control algorithm is:

[0046] .

[0047] Compared with the prior art, the beneficial effects of the present invention are:

[0048] By incorporating fault information from the vehicle's steering actuators, a vehicle dynamics model adaptable to various steering fault conditions was reconstructed. Corresponding state equations and control input matrices were designed for three typical fault scenarios: front axle jamming, rear axle jamming, and simultaneous front and rear axle jamming. This provides model support for subsequent fault-tolerant control, significantly improving the control system's adaptability and stability to actuator faults. Based on this, a trajectory tracking error modeling method based on heading and lateral error calculations was proposed. This method accurately captures the deviation between the vehicle's actual trajectory and the reference trajectory. The desired lateral velocity and yaw rate are generated through an upper-level MPC optimization control algorithm, achieving dynamic compensation for tracking errors. Furthermore, a lower-level MPC fault-tolerant control strategy was constructed by combining the vehicle's current state and actuator health status. The fault model is discretized using the midpoint Euler or forward Euler method, and the algorithm's continuity and robustness are enhanced by introducing control increments and relaxation factors. This controller, while ensuring trajectory tracking accuracy, possesses excellent fault tolerance and attitude stability control capabilities, making it particularly suitable for stable driving scenarios involving vehicles with four-wheel steering capabilities when partial steering axle failure occurs. Attached Figure Description

[0049] Figure 1 This is a flowchart of the present invention.

[0050] Figure 2 The results are for HIL test scenario 1; (a) trajectory comparison of HIL test scenario 1; (b) tracking error of HIL test scenario 1; (c) front and rear axle rotation angle comparison of HIL test scenario 1; (d) left and right braking force comparison of HIL test scenario 1.

[0051] Figure 3 The results are for HIL test scenario 2; (a) trajectory comparison of HIL test scenario 2; (b) tracking error of HIL test scenario 2; (c) front and rear axle rotation angle comparison of HIL test scenario 2; (d) left and right braking force comparison of HIL test scenario 2.

[0052] Figure 4 The results are for HIL test scenario 3; (a) trajectory comparison of HIL test scenario 3; (b) tracking error of HIL test scenario 3; (c) front and rear axle rotation angle comparison of HIL test scenario 3; (d) left and right braking force comparison of HIL test scenario 3.

[0053] Figure 5 This is a test site for actual vehicles.

[0054] Figure 6 The results are for the actual vehicle test scenario 1; (a) trajectory comparison of actual vehicle test scenario 1; (b) tracking error of actual vehicle test scenario 1; (c) front and rear axle rotation angle comparison of actual vehicle test scenario 1; (d) left and right braking force comparison of actual vehicle test scenario 1.

[0055] Figure 7 The results are for real vehicle test scenario 2; (a) trajectory comparison of real vehicle test scenario 2; (b) tracking error of real vehicle test scenario 2; (c) front and rear axle rotation angle comparison of real vehicle test scenario 2; (d) left and right braking force comparison of real vehicle test scenario 2.

[0056] Figure 8 The results are for real vehicle test scenario 3; (a) trajectory comparison of real vehicle test scenario 3; (b) tracking error of real vehicle test scenario 3; (c) front and rear axle rotation angle comparison of real vehicle test scenario 3; (d) left and right braking force comparison of real vehicle test scenario 3. Detailed Implementation

[0057] In order to provide a clearer understanding of the technical features, objectives and beneficial effects of the present invention, the technical solution of the present invention will now be described in detail below, but it should not be construed as limiting the scope of implementation of the present invention.

[0058] This invention provides a cross-system fault-tolerant control method for steering jamming in four-wheel steering vehicles, comprising the following steps (see...). Figure 1 ):

[0059] Step 1: Vehicle dynamics reconstruction. The angle of the stuck steering shaft is used as the state variable, and the healthy steering shaft (i.e., the steering shaft that is not stuck) and the braking force of the brake are used as the control variables to realize the dynamic reconstruction of the vehicle dynamics model.

[0060] Step 2: Error Calculation. Calculate the lateral position error between the vehicle's current position and the desired trajectory point using the Frenet formula. and heading error ;

[0061] Step 3: Calculate the desired lateral velocity and yaw rate. Design a fault-tolerant control algorithm based on a hierarchical collaborative framework. The upper-level control algorithm calculates the lateral velocity and yaw rate required for the vehicle to track the desired trajectory based on the lateral error and heading error of the vehicle trajectory tracking, combined with the vehicle state.

[0062] Step 4: Calculate the fault-tolerant control quantity. Design a fault-tolerant control algorithm based on a hierarchical collaborative framework. The lower-level control algorithm uses a dynamic model that is dynamically reconstructed based on the fault type of the vehicle steering system. It allocates the healthy steering shaft angle and the braking force of the left and right brakes according to the desired lateral speed, desired yaw rate and vehicle state to achieve fault-tolerant control.

[0063] In this invention, step one: vehicle dynamics reconstruction;

[0064] When a vehicle with four-wheel steering capability experiences partial steering shaft jamming, the vehicle dynamics model can be reconstructed based on the fault type. During model reconstruction, the angle of the jammed steering shaft is used as the state variable, while the healthy steering shaft and brake force are used as control variables, achieving dynamic reconstruction of the vehicle dynamics model. The vehicle dynamics model can be represented as:

[0065] ;

[0066] in for System status at any time; for Vehicle state observation vector at any given time; for Vehicle state vector at any given time; for The timing algorithm outputs the control matrix;

[0067] ;

[0068] in A For the vehicle state matrix; C f For the lateral stiffness of the vehicle's front wheels; C r The rear wheel lateral stiffness of the vehicle; a This is the distance from the front axle to the center of gravity of the vehicle. b This is the distance from the rear axle to the center of gravity of the vehicle. For vehicle quality; The vehicle's current longitudinal speed; Let be the moment of inertia about the z-axis; B is the control matrix.C This is the output matrix; D is the feedforward matrix; I represents the identity matrix; 0 represents the zero matrix; w This refers to the distance between the left and right wheels of the vehicle.

[0069] In the event of partial steering shaft failure, the dynamic model can be reconstructed into different forms depending on the type of steering shaft failure. In this invention, the failure types are classified as front axle failure, rear axle failure, and both front and rear axle failure.

[0070] When the current axis is stuck, the system status is as follows ,in It is the vehicle's current lateral speed. It is the vehicle's current yaw rate. This is the front wheel steering angle, which is a constant here; the model outputs the control quantity. ,in For the rear wheel steering angle, This represents the difference in braking force between the left and right brakes. T The matrix transpose symbol is used to reconstruct the state matrix. for:

[0071] ;

[0072] in For the vehicle's front axle sideslip angle; reconstruct the control matrix. for:

[0073] ;

[0074] in w The distance between the left and right wheels of the vehicle;

[0075] When the rear axle is stuck, the system status is as follows: ,in This refers to the rear wheel steering angle, which is a constant here; the model outputs the control quantity. Reconstructing the state equation for:

[0076] ;

[0077] in Rear wheel sideslip angle; reconstruct control matrix for:

[0078] ;

[0079] When both the front and rear axles are stuck, and They are all constants. Reconstruct the state matrix for:

[0080] ;

[0081] Reconstruct the control matrix for:

[0082] .

[0083] Step 2: Error Calculation;

[0084] Before performing trajectory tracking, it is necessary to first calculate the lateral position error between the vehicle's current position and the desired trajectory point. and heading error ;make Let the vehicle position vector be... For the reference trajectory point position vector, and These are the unit vectors of lateral velocity for the vehicle and the reference trajectory point, respectively. and These are the unit vectors of longitudinal velocity for the vehicle and the reference trajectory point, respectively. and These are the vehicle's heading angle and the heading angle of the reference trajectory point, respectively. Heading error. It can be calculated using the following formula:

[0085] ;

[0086] make , representing the lateral position error. According to vector geometric relationships, we have:

[0087] ;

[0088] Therefore, we can conclude that;

[0089] ;

[0090] in The vehicle position vector; The phasor of the reference trajectory point position;

[0091] Differentiating both sides of the lateral position error, we get:

[0092] ;

[0093] in This is the derivative of the lateral position error; , The derivative of the vehicle position vector. The vehicle velocity vector; The derivative of the position vector of the reference trajectory point. , For reference speed, we can obtain:

[0094] ;

[0095] in The derivative of the unit normal vector of the reference trajectory point can be calculated using Frenet's formula:

[0096] ;

[0097] ;

[0098] in s The distance along the trajectory; For trajectory curvature; For reference heading angle; This refers to the vehicle's heading angle; The sideslip angle is the angle at the vehicle's center of gravity. , The vehicle's current longitudinal speed; Let the vehicle's current lateral speed be denoted as:

[0099] ;

[0100] In summary, the vehicle tracking error is as follows:

[0101] ;

[0102] in The derivative of the heading error; This is the derivative of the lateral position error; The desired yaw rate; Let be the trajectory curvature.

[0103] Step 3: Calculate the desired lateral velocity and yaw velocity;

[0104] This invention designs a fault-tolerant control algorithm based on a hierarchical collaborative framework. The upper-level control algorithm calculates the lateral velocity and yaw rate required for the vehicle to track the desired trajectory based on the lateral and heading errors of the vehicle trajectory tracking, combined with the vehicle state. This invention designs the upper-level control algorithm based on Model Predictive Control (MPC).

[0105] The model's state equations can be expressed as:

[0106] ;

[0107] in The derivative of the state vector of the upper-level control algorithm; This is the observation vector for the upper-level control algorithm; This represents the state vector of the upper-level control algorithm. The upper-level control algorithm controls the output vector. The trajectory curvature is a constant here; the matrices in the state equations are:

[0108] ;

[0109] in and Let these be the state matrix, control matrix, output matrix, and feedforward matrix of the upper-level control algorithm, respectively. Based on model predictive control, the upper-level control algorithm is designed, and its cost function is defined as follows:

[0110] ;

[0111] in This is the cost function of the upper-level control algorithm; j For the first j One prediction step; N For prediction in the time domain; For the first One time step; and This is the weight matrix for the state vector and control vector. They are as follows:

[0112] ;

[0113] in, Weights for lateral position error; Weights for heading error; Lateral velocity weights; Weighted by yaw rate; For trajectory curvature weights; diag ( ) is the constructor for a diagonal matrix.

[0114] The constraints on the cost function of the upper-level control algorithm are:

[0115] .

[0116] The upper-level control algorithm obtains the desired lateral velocity and yaw rate through quadratic programming iteration optimization, which are then used as the optimization targets of the lower-level control algorithm.

[0117] Step 4: Calculate the fault-tolerant control quantity;

[0118] This invention designs a fault-tolerant control algorithm based on a hierarchical collaborative framework. The lower-level control algorithm uses a dynamic model reconstructed from the vehicle steering system fault types. It allocates the healthy steering shaft angle and braking forces of the left and right brakes based on the desired lateral velocity, desired yaw rate, and vehicle state to achieve fault-tolerant control. This invention designs the lower-level control algorithm based on MPC and reconstructs the vehicle state equation based on actuator fault states, with the following form:

[0119] ;

[0120] in The derivative of the state vector of the lower-level control algorithm; For the observation vector of the lower-level control algorithm; This represents the state vector of the lower-level control algorithm. The lower-level control algorithm controls the output vector; and These are the state matrix, control matrix, output matrix, and feedforward matrix of the lower-level control algorithm, respectively.

[0121] As described in step one, when different actuators fail... There are different expressions. The state matrix is ​​discretized using the midpoint Euler method. The discrete matrix can be obtained Discretize the state matrix using the forward Euler method. The discrete matrix can be obtained To ensure the continuity of the algorithm's output control quantity, Control vector at time Decomposed into Control vector at time add Control vector increment at time 1 ,Right now ,Will Merged into the state vector, the state vector becomes:

[0122] ;

[0123] in This is the merged state vector; for The state vector of the lower-level control algorithm at time t;

[0124] The algorithm optimization objective becomes The new state-space expression is as follows:

[0125] ;

[0126] in for The state vector at any given time; for The state vector at any given time; for The increment of the control vector determined at any given time; for The system output vector at each time step; and They are respectively The state transition matrix, control input matrix, and observation matrix at each time step are represented as follows:

[0127] ;

[0128] in for Time step The augmented state matrix at time step; for Time step The augmented control matrix at time step; for Time step The augmented observation matrix at time points; It is the identity matrix. This is a zero matrix. To ensure the tracking accuracy and stability of the fault-tolerant control algorithm, the cost function is defined as:

[0129] ;

[0130] in The cost function for the lower-level control algorithm; N For prediction in the time domain; i For the first i One forecast period; For the current moment Predicted future moments The system status; For a moment Reference state; For at any time Planning for the future Incremental control input; The state vector weight matrix; The control vector weight matrix; The weights are relaxation factor weights; Let be the relaxation factor. The system's predicted output expression in the prediction time domain is as follows:

[0131] ;

[0132] in For the system from the current moment Predicted future output sequence; This is the state observation matrix; This is a state estimate; For dynamic control matrix; To control the input increment vector, it is represented as follows:

[0133] ;

[0134] ;

[0135] in At time respectively Predicted The system output at that moment; For time-varying output matrices; Time-varying state transition matrices at different times; They are respectively Time calculation The system control vector output at time t; The output matrix maps the system state to the output. The input matrix describes how the control input affects the state changes; These are the state transition matrices at different times, used to describe the evolution of the system state over time.

[0136] The cost function can be expressed as:

[0137] ;

[0138] in To control the transpose of the input increment vector; To control the transpose of dynamic matrices; To output the weight matrix; For dynamic control matrix; To control the input weight matrix; To control the input increment vector; This is the state observation matrix; This is the merged state vector; The reference trajectory vector;

[0139] Will Place The last term yields:

[0140] ;

[0141] make:

[0142] ;

[0143] Finally, the cost function can be written as:

[0144] ;

[0145] in The cost function for the lower-level control algorithm; The input increment vector is used to control the lower-level control algorithm; It is a relaxation factor; H It is a Hessian matrix;g This is the gradient vector.

[0146] This is a standard quadratic programming problem, which can be solved using a quadratic programming solver. The constraints of the quadratic programming problem can be expressed as:

[0147] .

[0148] 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.

[0149] The specific implementation of the present invention will be described in detail below with reference to specific embodiments.

[0150] Example 1: Hardware-in-the-loop (HIL) and real-vehicle testing;

[0151] HIL testing scenarios include, but are not limited to:

[0152] Table 1 HIL Test Scenarios

[0153]

[0154] The test results for HIL test scenario 1 are as follows: Figure 2 As shown in (a)-(d), the test results for HIL test scenario 2 are as follows: Figure 3 As shown in (a)-(d), the test results for HIL test scenario 3 are as follows: Figure 4 As shown in (a)-(d). Figure 2 The HIL test results show that the fault-tolerant control algorithm proposed in this invention can reasonably allocate the rear axle rotation angle and the difference in braking force between the left and right sides of the vehicle when the front axle is stuck. It can control the lateral error of the vehicle within 0.2m and enable the vehicle to continue to drive safely along the desired trajectory under this condition. It has the fault-tolerant control function under the condition of front axle stuck. Figure 3 The HIL test results show that the fault-tolerant control algorithm proposed in this invention can reasonably allocate the front axle steering angle and the difference in braking force between the left and right sides of the vehicle when the rear axle is stuck. It can control the lateral error of the vehicle within 0.2m, so that the vehicle can continue to drive safely along the desired trajectory under this condition. It has the fault-tolerant control function under the condition of rear axle stuck. Figure 4The HIL test results show that the fault-tolerant control algorithm proposed in this invention can reasonably allocate the difference in braking force between the left and right sides of the vehicle when both the front and rear axles are locked, and can control the lateral error of the vehicle within 0.3m, enabling the vehicle to continue to drive safely along the desired trajectory under this condition. It possesses fault-tolerant control functionality under conditions where both the front and rear axles are locked. In summary, the method proposed in this invention can dynamically allocate the unlocked actuators to maintain high-precision trajectory tracking capability under fault conditions including but not limited to front axle lockup, rear axle lockup, and both front and rear axles lockup, allowing the vehicle to continue driving safely.

[0155] The actual vehicle testing site for this invention is as follows: Figure 5 As shown, real-vehicle test scenarios include, but are not limited to:

[0156] Table 2 Real-vehicle test scenarios

[0157]

[0158] The test results for real vehicle test scenario 1 are as follows: Figure 6 As shown in (a)-(d), the test results for real vehicle test scenario 2 are as follows: Figure 7 As shown in (a)-(d), the test results for real vehicle test scenario 3 are as follows: Figure 8 As shown in (a)-(d). Figure 6 The real-vehicle test results show that the fault-tolerant control algorithm proposed in this invention can reasonably allocate the rear axle rotation angle and the difference in braking force between the left and right sides of the vehicle when the front axle is stuck. It can control the lateral error of the vehicle within 0.3m, so that the vehicle can continue to drive safely along the desired trajectory under this condition. It has the fault-tolerant control function under the condition of front axle stuck. Figure 7 The real-vehicle test results show that the fault-tolerant control algorithm proposed in this invention can reasonably allocate the front axle steering angle and the difference in braking force between the left and right sides of the vehicle when the rear axle is stuck. It can control the lateral error of the vehicle within 0.2m, so that the vehicle can continue to drive safely along the desired trajectory under this condition. It has the fault-tolerant control function under the condition of rear axle stuck. Figure 8Real-vehicle test results show that the fault-tolerant control algorithm proposed in this invention can reasonably allocate the braking force difference between the left and right sides of the vehicle when both the front and rear axles are stuck, controlling the lateral error of the vehicle to within 0.35m. This allows the vehicle to continue driving safely along the desired trajectory under these conditions, demonstrating fault-tolerant control functionality even when both the front and rear axles are stuck. In summary, the method proposed in this invention can effectively control faults in test scenarios including, but not limited to, front axle stuck, rear axle stuck, and both front and rear axles stuck, achieving fault-tolerant control under fault conditions. Under actuator fault conditions, the failure of conventional motion control algorithms is due to the fact that the vehicle dynamics model used deviates completely from the actual vehicle dynamics characteristics. All predictions based on this model are erroneous, thus causing the algorithm to fail. This invention proposes a model dynamic reconstruction algorithm to address this situation. When a fault occurs, the dynamics model is reconstructed, eliminating the difference between the model and the actual dynamics characteristics, ensuring the accuracy of the prediction results of the model predictive control algorithm. Therefore, it can still maintain a certain level of tracking accuracy under fault conditions, thereby improving vehicle safety.

[0159] The above are merely preferred embodiments of the present invention. It should be noted that those skilled in the art can make several modifications and improvements without departing from the concept of the present invention, and these should also be considered within the scope of protection of the present invention. These modifications and improvements will not affect the effectiveness of the implementation of the present invention or the practicality of the patent.

Claims

1. A cross-system fault-tolerant control method for steering jamming in a four-wheel steering vehicle, characterized in that, Includes the following steps: Step 1: Vehicle dynamics reconstruction, using the stuck steering shaft angle as the state variable and the healthy steering shaft and brake force as the control variables, to achieve dynamic reconstruction of the vehicle dynamics model; Step 2: Error Calculation. Calculate the lateral position error between the vehicle's current position and the desired trajectory point using the Frenet formula. and heading error ; Step 3: Calculate the desired lateral velocity and yaw rate. Design a fault-tolerant control algorithm based on a hierarchical collaborative framework. The upper-level control algorithm calculates the lateral velocity and yaw rate required for the vehicle to track the desired trajectory based on the lateral error and heading error of the vehicle trajectory tracking, combined with the vehicle state. Step 4: Calculate the fault-tolerant control quantity. Design a fault-tolerant control algorithm based on a hierarchical collaborative framework. The lower-level control algorithm uses a dynamic model that is dynamically reconstructed based on the fault type of the vehicle steering system. It allocates the healthy steering shaft angle and the braking force of the left and right brakes according to the desired lateral speed, desired yaw rate and vehicle state to achieve fault-tolerant control. The vehicle dynamics model is as follows: ； in A For the vehicle state matrix; C f For the lateral stiffness of the vehicle's front wheels; C r The rear wheel lateral stiffness of the vehicle; a This is the distance from the front axle to the center of gravity of the vehicle. b This is the distance from the rear axle to the center of gravity of the vehicle. For vehicle quality; The vehicle's current longitudinal speed; Let be the moment of inertia about the z-axis; B is the control matrix; C This is the output matrix; D is the feedforward matrix; I represents the identity matrix; 0 represents the zero matrix; w The distance between the left and right wheels of the vehicle; In step one, when the current axis is stuck, the system state is... ,in It is the vehicle's current lateral speed. It is the vehicle's current yaw rate. This is the front wheel steering angle, which is a constant here; the model outputs the control quantity. ,in For the rear wheel steering angle, This represents the difference in braking force between the left and right brakes. T The matrix transpose symbol is used to reconstruct the state matrix. for: ； in For the vehicle's front axle sideslip angle; reconstruct the control matrix. for: ； in w The distance between the left and right wheels of the vehicle; When the rear axle is stuck, the system status is as follows: ,in This refers to the rear wheel steering angle, which is a constant here; the model outputs the control quantity. Reconstructing the state equation for: ； in Rear wheel sideslip angle; reconstruct control matrix for: ； When both the front and rear axles are stuck, and They are all constants. Reconstruct the state matrix for: ； Reconstruct the control matrix for: 。 2. The cross-system fault-tolerant control method for steering jamming in a four-wheel steering vehicle according to claim 1, characterized in that, The error calculation formula in step two is as follows: ； in The derivative of the heading error; This is the derivative of the lateral position error; This refers to the vehicle's heading angle; The heading angle is used as a reference trajectory point; The desired yaw rate; Let be the trajectory curvature; let Let the vehicle position vector be... The reference trajectory point position vector; The lateral velocity unit vector of the reference trajectory point; The vehicle's current longitudinal speed; This represents the vehicle's current lateral speed.

3. The cross-system fault-tolerant control method for steering jamming in a four-wheel steering vehicle according to claim 2, characterized in that, The model state equation in step three can be expressed as: ； in The derivative of the state vector of the upper-level control algorithm; This is the observation vector for the upper-level control algorithm; This represents the state vector of the upper-level control algorithm. The upper-level control algorithm controls the output vector. The trajectory curvature is a constant here; the matrices in the state equations are: ； in and Let these be the state matrix, control matrix, output matrix, and feedforward matrix of the upper-level control algorithm, respectively. Based on model predictive control, the upper-level control algorithm is designed, and its cost function is defined as follows: ； in This is the cost function of the upper-level control algorithm; j For the first j One prediction step; N For prediction in the time domain; For the first One time step; and This is the weight matrix for the state vector and the control vector.

4. The cross-system fault-tolerant control method for steering jamming in a four-wheel steering vehicle according to claim 3, characterized in that, The constraints on the cost function of the upper-level control algorithm are as follows: 。 5. The cross-system fault-tolerant control method for steering jamming in a four-wheel steering vehicle according to claim 4, characterized in that, The state equation of the lower-level control algorithm in step four is as follows: ； in The derivative of the state vector of the lower-level control algorithm; For the observation vector of the lower-level control algorithm; This represents the state vector of the lower-level control algorithm. The lower-level control algorithm controls the output vector; and These are the state matrix, control matrix, output matrix, and feedforward matrix of the lower-level control algorithm, respectively. The lower-level control algorithm is designed based on the model predictive control algorithm. The cost function of the lower-level control algorithm is defined as follows: ； in The cost function for the lower-level control algorithm; The lower-level control algorithm controls the output change. It is a relaxation factor; H It is a Hessian matrix; g This is the gradient vector.

6. The cross-system fault-tolerant control method for steering jamming in a four-wheel steering vehicle according to claim 5, characterized in that, The constraints on the cost function of the lower-level control algorithm are as follows: 。

Images