Steer-by-wire hierarchical control system and method

Abstract

The application discloses a kind of drive-by-wire steering hierarchical control system and method, the system mainly includes upper H2 online optimization self-disturbance control, lower double sliding mode controller, corner distribution module, variable transmission ratio module and vehicle reference model;Upper controller is estimated vehicle yaw angular velocity and system collective disturbance in real time by configuring second-order linear extended state observer, H2 online optimization module minimizes with tracking error and control energy quadratic performance index as target, real-time dynamic optimization self-disturbance control law proportion gain;Corner distribution module is based on speed self-adaptive and executes multi-section steering geometry mapping;Lower controller adopts equal-speed approach law and introduces boundary layer saturation function to suppress chattering, realizes independent closed-loop tracking control to left and right steering wheels.The application considers the robustness, response speed and anti-chattering ability under different speed conditions, significantly improves the comprehensive control performance of drive-by-wire steering system.

Description

Technical Field

[0001] This invention belongs to the field of vehicle steer-by-wire technology, specifically relating to a hierarchical control system and method for steer-by-wire that combines active disturbance rejection control (ADRC), H2 online optimization, and dual sliding mode control. Background Technology

[0002] Steer-by-wire (SbW) systems eliminate the mechanical connection between the steering wheel and the steering wheels, achieving steering control through electronic signals. They offer advantages such as fast response, flexible deployment, and ease of integration with advanced driver assistance functions, making them a hot research topic in the field of intelligent vehicles. However, the control performance of a steer-by-wire system directly affects the vehicle's handling stability, safety, and comfort, and its controller design faces numerous technical challenges.

[0003] First, steer-by-wire systems involve complex nonlinear characteristics, such as the nonlinearity of tire sideslip, friction and clearance of the steering actuator, and changes in vehicle mass and center of gravity. Simultaneously, parameters such as road adhesion coefficient and vehicle speed change in real time during vehicle operation, leading to significant uncertainties and parameter perturbations in the system model. While traditional PID control methods are simple in structure, they lack robustness under parameter perturbations and external disturbances, making it difficult to guarantee stability and tracking accuracy across all operating conditions. Yaw rate is a key state variable characterizing vehicle steering response, and its accurate tracking is crucial for achieving the desired steering characteristics. However, in actual driving, vehicles are affected by various external disturbances such as crosswinds, uneven road surfaces, and changes in tire force. These disturbances directly affect yaw motion, causing the actual yaw rate to deviate from the ideal value. Active disturbance rejection control (ADRC) estimates and compensates for total disturbances through a linear extended state observer (ESO), offering unique advantages in disturbance suppression. However, traditional ADRC gain parameters are mostly fixed values, making it difficult to adapt to continuous changes in operating conditions such as vehicle speed, and parameter tuning relies on experience, limiting further improvements in its control performance.

[0004] Secondly, sliding mode control (SMC) is widely used in nonlinear system control due to its strong robustness to parameter perturbations and external disturbances. However, the sign function in traditional sliding mode control leads to high-frequency switching of the control quantity, resulting in chattering, which not only affects control accuracy but may also excite unmodeled dynamics and even damage the actuator. To suppress chattering, researchers have proposed methods such as boundary layer saturation functions and higher-order sliding modes, but balancing chattering suppression and tracking accuracy remains a design challenge. Furthermore, existing steer-by-wire systems mostly use overall front wheel angle control, failing to fully utilize the potential of independent left and right wheel drive. Achieving independent control of the left and right wheels can not only improve the flexibility of steering response but also assist yaw stability control through differential torque, but this places higher demands on the real-time performance and coordination of the controller. Meanwhile, the steering geometry of traditional vehicles (such as Ackermann steering) is usually optimized for specific operating conditions during design, making it difficult to maintain ideal tire side slip characteristics across the entire vehicle speed range. At low speeds, pure Ackerman geometry helps reduce tire wear; at medium speeds, parallel steering improves response consistency; at high speeds, anti-Ackerman geometry increases tire lateral force reserve and enhances stability. How to smoothly transition steering geometry when vehicle speed changes continuously, and optimize tire lateral slip characteristics, is a problem that steer-by-wire systems urgently need to solve.

[0005] In summary, existing steer-by-wire control methods still have shortcomings in areas such as nonlinear disturbance suppression, parameter adaptation, chatter suppression, steering geometry adjustment, and delay compensation. Therefore, there is an urgent need for a hierarchical steer-by-wire control strategy that can effectively suppress disturbances, adapt to changes in operating conditions, and achieve independent and precise tracking of the left and right wheels. Summary of the Invention

[0006] The present invention aims to address the shortcomings of the prior art by providing a layered control system and method for steer-by-wire, which can balance robustness, response speed and anti-vibration capability under different vehicle speed conditions, thereby significantly improving the overall control performance of the steer-by-wire system.

[0007] To achieve its objectives, the present invention employs the following technical solution: The present invention provides a hierarchical control system for steer-by-wire, comprising: a state reference module, an upper-level controller, a steering angle allocation module, and a lower-level execution controller; The state reference module acquires the steering wheel angle signal and vehicle speed signal, and generates the desired front wheel angle and the ideal yaw rate characterizing the ideal steering state of the vehicle based on the preset vehicle reference model and transmission ratio mapping relationship, while extracting the actual yaw rate of the vehicle target model. The upper-level controller uses state and disturbance observation logic to calculate the system lumped disturbance in real time, which includes dynamic perturbations inside the vehicle and external environmental disturbances. At the same time, based on the tracking error between the actual yaw rate and the ideal yaw rate and the dynamically updated control gain, the lumped disturbance is fed forward to compensate for the lumped disturbance and outputs the front wheel compensation steering angle for yaw stability control. Based on the current vehicle speed, the steering angle allocation module uses a multi-segment steering geometry mapping strategy to decouple the desired front wheel steering angle from the total target steering angle composed of the front wheel compensation steering angle, and then obtains the target steering angle of the left front wheel and the target steering angle of the right front wheel. The lower-level execution controller includes a physically decoupled left-wheel tracking controller and a right-wheel tracking controller, which respectively receive the target steering angle of the left front wheel and the target steering angle of the right front wheel, and calculate the target control torque required to drive the left steering actuator and the right steering actuator respectively through a nonlinear robust closed-loop tracking algorithm, so as to achieve independent disturbance rejection control of the steering angles of the left and right front wheels.

[0008] The steer-by-wire layered control system described in this invention is characterized in that the transmission ratio mapping relationship in the state reference module is constructed based on the principle of constant steady-state gain of vehicle yaw rate, and the transmission ratio from the steering wheel angle to the desired front wheel angle is set as a piecewise mapping function that includes the current vehicle speed, vehicle wheelbase, and vehicle understeer stability factor; and when the current vehicle speed exceeds the safety limit, the output transmission ratio is truncated and limited.

[0009] Furthermore, the upper-level controller is an active disturbance rejection controller, wherein the state and disturbance observation logic is implemented by configuring a second-order linear extended state observer. The second-order linear extended state observer expands the system lumped disturbance into a new state variable of the system and estimates it in real time, and then cancels it in reverse in the virtual control law, thereby approximately decoupling the nonlinear vehicle yaw dynamic system into an integral series system.

[0010] Furthermore, the dynamically updated control gain in the upper-level controller is generated through an online optimization algorithm, including: establishing a quadratic performance index functional with the tracking error of yaw rate and control energy input as joint evaluation objectives, and obtaining the optimal proportional feedback gain that minimizes the performance index functional by solving the algebraic Riccati equation in real time; wherein, the weight coefficient used to penalize the tracking error in the quadratic performance index functional is configured as a nonlinear variable that dynamically amplifies with increasing vehicle speed.

[0011] Furthermore, both the left wheel tracking controller and the right wheel tracking controller are sliding mode controllers. The nonlinear robust closed-loop tracking algorithm is constructed based on the sliding surface, which includes the difference between the actual front wheel steering angle and the target steering angle, as well as the rate of change of the steering angle error. The nonlinear robust closed-loop tracking algorithm uses a control law containing feedforward compensation terms and approach terms to drive the actual front wheel steering angle to converge toward the target steering angle.

[0012] Furthermore, the nonlinear robust closed-loop tracking algorithm introduces a smoothing anti-shaking mechanism, including: introducing a boundary layer saturation function of preset thickness into the approach term of the sliding mode controller to replace the standard sliding mode sign function; when the system state trajectory is outside the boundary layer, the sliding mode controller outputs the target control torque with the maximum amplitude limit; when the system state trajectory enters the boundary layer, the target control torque is continuously linearly interpolated within the boundary layer before being output to the steering motor.

[0013] Furthermore, the multi-segment steering geometry mapping strategy of the steering angle allocation module is configured as follows: First, a first speed threshold, a second speed threshold, and a third speed threshold are preset in ascending order; in the range where the current vehicle speed is lower than the first speed threshold, the inner and outer wheel angles are allocated according to the pure Ackerman steering geometry principle; in the range where the current vehicle speed is between the first speed threshold and the second speed threshold, the Ackerman steering angle and the parallel steering angle are linearly interpolated and fused according to the current vehicle speed; in the range where the current vehicle speed is between the second speed threshold and the third speed threshold, the parallel steering angle and the anti-Ackerman steering angle are linearly interpolated and fused according to the current vehicle speed; in the range where the current vehicle speed is higher than the third speed threshold, the inner and outer wheel angles are allocated according to the anti-Ackerman steering geometry principle.

[0014] The steer-by-wire control method of the system described in this invention is characterized by including the following steps: S1: Collects steering wheel angle signals and vehicle speed signals from external inputs; S2: Based on the collected steering wheel angle signal and vehicle speed signal, calculate the desired front wheel angle of the vehicle and input it into the reference model for processing to generate the ideal yaw rate; S3: The error between the ideal yaw rate and the actual yaw rate is input into the upper controller in real time, and the system lumped disturbance composed of external disturbance and internal parameter perturbation is estimated by using a second-order linear extended state observer; then, after combining the dynamic gain based on the online solution of vehicle speed to perform disturbance feedforward compensation on the system lumped disturbance, the front wheel compensation steering angle is output. S4: Superimpose the desired front wheel steering angle and the front wheel compensation steering angle, determine the preset threshold range of the current vehicle speed, and calculate the dynamic weighting factor, thereby decoupling and mapping the total steering angle composed of the desired front wheel steering angle and the front wheel compensation steering angle into independent left and right target steering angles that conform to the characteristics of low-speed Ackerman, medium-speed parallel, or high-speed anti-Ackerman. S5: The left and right lower-level controllers receive the independent target turning angles of their respective sides, calculate the turning angle tracking error and its derivative, and drive the system state to slide towards the preset sliding surface to achieve convergence of the actual turning angle of the front wheels to the target turning angle, and output continuous motor drive torque commands to independently complete the vehicle steering action.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. Improved anti-interference capability and adaptability under all working conditions: Through the upper-level controller, the system lumped disturbance caused by changes in road conditions and perturbations of vehicle internal parameters is calculated and fed forward in real time; at the same time, combined with dynamically updated control gain, the yaw rate tracking performance and control energy consumption under different vehicle speed conditions are taken into account, breaking the limitation of poor adaptability of traditional fixed gain controllers and significantly enhancing the vehicle's handling stability. 2. Balancing the robustness of independent control with the anti-jitter protection of the actuator: The lower-level independent actuator controller adopts a nonlinear robust closed-loop tracking algorithm to control the left and right wheel steering actuators separately. This not only fully leverages the flexibility of the independent decoupled control of the left and right wheels in the steer-by-wire system and improves the system's robustness under nonlinear disturbances, but also effectively filters out high-frequency switching jitter caused by discrete control commands by introducing a smooth anti-jitter mechanism, avoiding unmodeled dynamic excitation and extending the service life of the lower-level steering actuator. 3. Optimization of tire mechanical characteristics across the entire speed range: An angle distribution module and a multi-segment steering geometry mapping strategy are introduced. At low speeds, pure Ackerman steering geometry is used to reduce tire wear; at medium speeds, a transition to parallel steering is used to improve steering response consistency; and at high speeds, anti-Ackerman steering geometry is used to increase the lateral force reserve of the outer tires. This strategy achieves a smooth transition in steering geometry when vehicle speed changes continuously, fully utilizing the physical limits of the tire. 4. The layered and decoupled control architecture facilitates software deployment and OTA upgrades: The overall architecture adopts a layered structure of "state reference - upper-level stabilization - angle allocation - lower-level execution," completely decoupling the macroscopic decision-making of vehicle attitude from the microscopic execution of the underlying motors. Each module has clear functional boundaries and independent logic, greatly reducing the code coupling in the vehicle electronic control unit (ECU), which facilitates modular compilation of algorithms, independent bench calibration, and subsequent remote online upgrades (OTA). Attached Figure Description

[0016] Figure 1This is a control flowchart of a steer-by-wire layered control system according to the present invention. Detailed Implementation

[0017] In this embodiment, a hierarchical steer-by-wire control system based on H2 online optimization, active disturbance rejection, and dual sliding mode is established. This system achieves high-precision yaw rate tracking and independent control of the left and right wheels through an upper-level disturbance observation and stability controller and a lower-level independent execution controller. It also considers the smooth transition of steering geometry to improve steering stability and path tracking capability. Specifically, the system includes: a state reference module, an upper-level controller, a steering angle allocation module, and a lower-level execution controller. The state reference module acquires the steering wheel angle signal and vehicle speed signal, and generates the desired front wheel steering angle and the ideal yaw rate characterizing the ideal steering state of the vehicle based on the preset vehicle reference model and transmission ratio mapping relationship. At the same time, it extracts the actual yaw rate of the vehicle target model.

[0018] In practice, the transmission ratio mapping relationship is constructed based on the principle of constant steady-state gain of vehicle yaw rate. The transmission ratio from steering wheel angle to desired front wheel angle is set as a piecewise mapping function that includes current vehicle speed, vehicle wheelbase, and vehicle understeer stability factor. When the current vehicle speed exceeds the safety limit, the output transmission ratio is truncated and limited.

[0019] The desired front wheel steering angle information is typically obtained based on the driver's input steering wheel angle signal and vehicle speed information. The driver's input steering wheel angle signal and vehicle speed signal are fed into the variable gear ratio module to calculate the ideal front wheel steering angle, enabling the steering gear ratio to continuously change with vehicle speed, thus allowing the vehicle to obtain optimal steering response characteristics under different driving conditions.

[0020] Steering ratio is defined using equation (1). : (1) In equation (1), Steering wheel angle; This refers to the steering angle of the front wheels.

[0021] In the design of traditional mechanical steering systems, the transmission ratio is usually a fixed value; however, because the mechanical connection is eliminated, the transmission ratio can be freely designed to change with vehicle speed, thereby achieving an ideal steering feel that is sensitive at low speeds and stable at high speeds.

[0022] The variable gear ratio design is based on the principle of constant yaw rate gain in vehicle dynamics theory. According to a linear two-degree-of-freedom vehicle model, the steady-state yaw rate gain under front wheel steering angle input is: (2) In equation (2), For vehicle speed; This refers to the vehicle's wheelbase. This refers to the yaw rate; This is the stability factor. With a fixed front wheel steering angle, the yaw rate changes non-linearly with vehicle speed.

[0023] To ensure the driver experiences decoupling between steering response and vehicle speed, i.e., to achieve a constant gain from steering wheel angle to yaw rate, the gear ratio needs to satisfy the following: (3) In equation (3), It is a constant.

[0024] From this, the relationship between the ideal transmission ratio and vehicle speed can be derived: (4) In equation (4), The transmission ratio varies with vehicle speed; This is the gain coefficient, used to adjust the overall steering sensitivity.

[0025] Stability factor It is an important parameter characterizing the understeer or oversteer characteristics of a vehicle, and the calculation formula is: (5) In equation (5), For the overall vehicle weight; This refers to the vehicle's wheelbase. This is the distance from the center of gravity to the front axle. This is the distance from the center of mass to the rear axle; For front axle lateral stiffness; This refers to the rear axle lateral stiffness.

[0026] Considering practical driving needs and engineering requirements, the gear ratio design is divided into three segments. At low speeds, the driver requires a large steering input; a smaller gear ratio allows the front wheel angle to change rapidly with the steering wheel, improving steering sensitivity and reducing steering wheel rotation angle. At medium speeds, the vehicle's steering gear ratio needs to satisfy the principle of constant yaw rate gain, ensuring a linear and predictable steering response for the driver. As vehicle speed increases, the gear ratio gradually increases, and the front wheel angle corresponding to the steering wheel angle decreases, reflecting the characteristic of "high-speed stability." At high speeds, excessive steering response reduces driving stability; therefore, the maximum gear ratio is limited to restrict the range of front wheel angle changes, improving safety at high speeds.

[0027] The transmission ratio is defined as follows: (6) In equation (6), For vehicle speed; , These are the speed thresholds; , This is the transmission ratio limit.

[0028] By inputting the steering wheel angle and vehicle speed information into the variable gear ratio module, the desired front wheel angle can be obtained. .

[0029] The variable gear ratio module and the generated ideal front wheel steering angle are input into a linear two-degree-of-freedom reference vehicle model, which is used to generate the ideal yaw rate. This serves as the tracking target for the upper-level controller. The equations of motion for the linear two-degree-of-freedom model are: (7) In equation (7), For the overall vehicle weight; For vehicles to bypass Moment of inertia of the shaft; It is lateral acceleration; Longitudinal velocity; This refers to the yaw rate; This is the yaw acceleration; The lateral force acting on the front wheel; This refers to the lateral force acting on the rear wheel.

[0030] Take the state vector as , For system input, The state-space equations for the linear two-degree-of-freedom model can be obtained as follows: (8) In equation (8), The system state vector The derivative of ; A is the system state matrix; B is the system input control matrix.

[0031] The upper-level controller uses state and disturbance observation logic to calculate the system lumped disturbance in real time, which includes dynamic perturbations inside the vehicle and external environmental disturbances. At the same time, based on the tracking error between the actual yaw rate and the ideal yaw rate and the dynamically updated control gain, it performs feedforward compensation on the lumped disturbance and outputs the front wheel compensation steering angle for yaw stability control. In specific implementation, the upper-level controller is an active disturbance rejection controller. The state and disturbance observation logic is implemented by configuring a second-order linear extended state observer. The second-order linear extended state observer expands the system's lumped disturbance into a new state variable of the system and estimates it in real time. It then cancels it out in the virtual control law, thereby approximately decoupling the nonlinear vehicle yaw dynamic system into an integral series system.

[0032] The vehicle's yaw motion system is described as follows: (9) In equation (9), This is the yaw acceleration; For control input, namely the front wheel compensated steering angle; This refers to the lumped disturbance that includes both unmodeled dynamic disturbances of the system and external disturbances; This represents the model gain.

[0033] Total disturbance Expand into new state variables, let , Then the state equation of the system is: (10) In equation (10), , Two new state variables are defined, representing the actual yaw rate and the lumped disturbance, respectively; This is an estimate of the yaw rate. The derivative of the estimated yaw rate; This is an estimate of the total disturbance; The derivative of the estimated total disturbance; , These are the state correction gain and the disturbance correction gain, respectively.

[0034] The observer gain is configured using the bandwidth method, and the observer poles are placed at... If: (11) In equation (11), The observer bandwidth determines the observer's response speed.

[0035] To achieve digital control, the Euler method is used to discretize the observer, and the sampling time is set to be... Then the discretized observer is: (12) In equation (12), for The estimated yaw rate at time t, for The estimated value of the total disturbance at time 1. for Front wheel compensation steering angle at any moment for Output estimation error at time 10:00 for The actual yaw rate at that moment; for The estimated yaw rate at time t, for The total disturbance estimate at time t.

[0036] The observer's initial state is set to To prevent observer divergence caused by integral saturation, [the following is done]: The estimated value is subjected to amplitude limiting.

[0037] Model gain This reflects the degree to which the control input affects the rate of change of yaw rate. Definition for: (13) The system after ESO observer compensation can be approximated as a single integrator: (14) In equation (14), This refers to the yaw rate tracking error. This is the derivative of the yaw rate tracking error; This is a virtual control variable.

[0038] In specific implementation, the dynamically updated control gain in the upper-level controller is generated through an online optimization algorithm, including: establishing a quadratic performance index functional with the tracking error of yaw rate and control energy input as joint evaluation objectives, and obtaining the optimal proportional feedback gain that minimizes the performance index functional by solving the algebraic Riccati equation in real time; wherein, the weight coefficient used to penalize the tracking error in the quadratic performance index functional is configured as a nonlinear variable that dynamically amplifies with increasing vehicle speed.

[0039] Define the performance index of the quadratic form as : (15) in, These are the error weighting coefficients. This is the control quantity weighting coefficient. This indicator reflects the comprehensive optimization of tracking accuracy and control energy.

[0040] For the above single integrator system, the Riccati equation is solved according to the theory of linear quadratic regulators (LQR): (15) In equation (15), This represents the state matrix of a single-integral error system. This represents the transpose of the state matrix of a single-integral error system. ; This represents the input matrix of the single-integral error system. This represents the transpose of the input matrix of a single-integral error system. ; Represents the error state weight matrix. ; This represents the control energy input weight matrix. This represents the inverse of the control energy input weight matrix. , ; This represents the positive definite solution to the algebraic Riccati equation; the solution yields... Therefore, the optimal gain for: (16) To enable the controller's performance to automatically adjust with changes in vehicle speed, the error weights are designed as a function of vehicle speed. : (17) In equation (17), These are the base weight values ​​to ensure basic tracking performance at low speeds; This is the vehicle speed gain coefficient, which controls the rate at which the vehicle speed increases. This is the speed threshold; below this value, the base value remains unchanged. The speed is the vehicle speed.

[0041] Based on the above design, design the optimal gain. The calculation formula is: (18) In equation (18), This is the proportional gain that varies with vehicle speed.

[0042] The yaw rate estimated by the ESO observer and H2 online optimized proportional gain Substituting these values ​​into the controller's control law, we obtain the virtual control quantity. : (19) In equation (19), The desired yaw rate.

[0043] Total perturbation estimated using ESO observer For virtual control variables Compensation is performed to obtain the front wheel compensated steering angle. : (20) In equation (20), This refers to the front wheel compensation angle. To prevent excessive front wheel compensation angle from causing the tires to enter the non-linear region or the actuators to saturate, the front wheel compensation angle needs to be limited. (twenty one) In equation (21), This is the actual front wheel compensation steering angle output.

[0044] The steering angle allocation module, based on the current vehicle speed, uses a multi-segment steering geometry mapping strategy to determine the total target steering angle, which consists of the desired front wheel steering angle and the front wheel compensation steering angle. After decoupling, the target steering angles of the left front wheel and the right front wheel are obtained.

[0045] For a given total target turning angle In this case, first calculate the theoretical value of the inner and outer wheel rotation angles under pure Ackermann geometry based on this value. , : (twenty two) In equation (22), , The steering angle of the inner and outer wheels under strict Ackermann steering geometry; Wheelbase (center distance of kingpin); This refers to the wheelbase.

[0046] The wheel angle under parallel steering is : (twenty three) When a vehicle turns, the turning angles of the inner and outer wheels differ due to their different turning radii. The required steering geometry varies at different vehicle speeds. At low speeds, tire lateral forces are relatively small, and the primary focus is on tire wear and steering ease, making pure Ackermann geometry more suitable. At medium speeds, a balance between consistent response and tire wear is needed, and parallel steering provides a good linear response. At high speeds, the demand for tire lateral forces increases, and inverse Ackermann geometry allows the outer tires to achieve a larger slip angle, thereby increasing lateral force reserve and improving steering stability.

[0047] Under pure Ackermann geometry, the steering angles of the inner and outer sides of the steering wheel satisfy the following relationship: (twenty four) In equation (24), The turning angle of the inner wheel; This refers to the outer wheel's steering angle. Pure Ackerman geometry minimizes tire wear and is suitable for low-speed conditions.

[0048] Parallel steering geometry, meaning the steering angles of the inner and outer wheels are equal: (25) This geometry improves the consistency of steering response at moderate vehicle speeds.

[0049] Inverse Ackermann geometry, where the outer wheel's turning angle is greater than the inner wheel's turning angle: (26) This geometry increases the lateral force reserve of the outer tires, improving stability at high speeds.

[0050] In specific implementation, the multi-segment steering geometry mapping strategy of the corner allocation module is configured as follows: firstly, a first speed threshold, a second speed threshold, and a third speed threshold are preset in sequence; To achieve a smooth transition in vehicle speed adaptation, three speed thresholds are defined. , , , representing low speed, medium speed, and high speed thresholds, respectively. This embodiment employs a speed-adaptive steering geometry scheduling strategy to smoothly transition between pure Ackerman, parallel steering, and anti-Ackerman.

[0051] (1) Low-speed zone, ; Within the range where the current vehicle speed is below the first speed threshold, the steering angles of the inner and outer wheels are allocated based on the pure Ackerman steering geometry principle; At this point, a pure Ackermann steering geometry is used: (27) (2) Medium and low speed transition zone, ; When the current vehicle speed is between the first speed threshold and the second threshold, linear interpolation is performed on the Ackerman steering angle and the parallel steering angle based on the current vehicle speed. At this point, a linear transition occurs between Ackermann geometry and parallel turning: (28) (3) Medium-to-high speed transition zone ; When the current vehicle speed is between the second and third speed thresholds, the parallel steering angle and the anti-Ackermann steering angle are linearly interpolated and fused based on the current vehicle speed. At this point, a linear transition occurs between parallel turning and anti-Ackermann geometry: (29) (4) High-speed zone, ; In the range where the current vehicle speed is higher than the third speed threshold, the steering angles of the inner and outer wheels are allocated according to the anti-Ackerman steering geometry principle.

[0052] At this point, anti-Ackermann geometry is used entirely: (30) Based on the aforementioned weighting coefficients, calculate the amplitude of the fused inner and outer wheel rotation angles: (31) In equations (27), (28), (29), (30) and (31), , , These are the weighting coefficients for Ackermann steering, parallel steering, and anti-Ackermann steering, respectively.

[0053] By inputting the desired front wheel steering angle and vehicle speed information into the steering angle allocation module, the desired values ​​of the inner and outer steering front wheel angles can be obtained. , .

[0054] The lower-level execution controller includes a physically decoupled left-wheel tracking controller and a right-wheel tracking controller, which respectively receive the target steering angle of the left front wheel and the target steering angle of the right front wheel, and calculate the target control torque required to drive the left steering actuator and the right steering actuator respectively through a nonlinear robust closed-loop tracking algorithm, so as to achieve independent disturbance rejection control of the steering angles of the left and right front wheels.

[0055] In practice, both the left-wheel tracking controller and the right-wheel tracking controller are sliding mode controllers. The nonlinear robust closed-loop tracking algorithm is based on the sliding surface, which includes the difference between the actual front wheel steering angle and the target steering angle, as well as the rate of change of the steering angle error. The nonlinear robust closed-loop tracking algorithm uses a control law that includes feedforward compensation terms and approach terms to drive the actual front wheel steering angle to converge toward the target steering angle.

[0056] The nonlinear robust closed-loop tracking algorithm introduces a smoothing anti-shaking mechanism, including: introducing a boundary layer saturation function with a preset thickness in the approach term of the sliding mode controller to replace the standard sliding mode sign function; when the system state trajectory is outside the boundary layer, the sliding mode controller outputs the target control torque with the maximum amplitude limit; when the system state trajectory enters the boundary layer, the target control torque is continuously linearly interpolated within the boundary layer before being output to the steering motor.

[0057] The sliding mode controller is the lower-level execution controller of the steer-by-wire system of this invention. It consists of two independent controllers with identical structures, which control the steering motors of the left and right front wheels respectively, so that the actual turning angle of the left and right front wheels accurately tracks the target turning angle output by the turning angle distribution module.

[0058] The steering actuator can be simplified as a gear and rack system driven by an electric motor, and its dynamic model can be expressed as: (32) In equation (32), The equivalent moment of inertia of the motor rotor and load referred to the motor shaft; This refers to the actual angular velocity of the steering actuator; This refers to the actual angular acceleration of the steering actuator; This refers to the output torque of the motor; The damping coefficient; This includes lumped disturbances that incorporate friction, load torque, and unmodeled dynamics.

[0059] Define the corner tracking error and its derivative for: (33) In equation (33), The actual turning angle of the steering actuator.

[0060] Design the sliding surface of the controller. for: (34) In equation (34), The parameters of the sliding surface determine the convergence rate of the error.

[0061] To ensure that the system state approaches the sliding surface from any initial position, a reaching law needs to be designed. This embodiment uses a constant velocity reaching law: (35) In equation (35), The rate of change of the sliding surface; The approach rate is determined by the approach law gain; Boundary layer thickness; Let saturation function be defined as: (36) In equation (36), Let be the independent variable of the saturation function; in the calculation of the actual sliding mode reaching law, .

[0062] Replacing the sign function in traditional sliding mode control with a saturation function is a key measure to achieve smooth anti-shaking. In the boundary layer... Inside, the control quantity changes continuously to avoid high-frequency switching; outside the boundary layer, the control quantity maintains maximum output to ensure that the system state quickly approaches the sliding surface.

[0063] Based on the definition of a sliding surface, take its derivative: (37) In equation (37), This is the second derivative of the corner tracking error.

[0064] Substituting the system model and deriving the torque command output by the sliding mode controller, It can be represented as: (38) To prevent motor overload and actuator damage, the control torque is adjusted. Limiting the amplitude: (39) In equation (39), , These represent the maximum and minimum torque output by the motor, respectively.

[0065] The output motor torque is reduced and amplified by a reducer, and then the steering wheel is controlled by a gear and rack mechanism to achieve precise tracking of the desired steering angle of the left and right front wheels.

[0066] In this embodiment, a steer-by-wire control method based on the above system includes the following steps: S1: Collects steering wheel angle signals and vehicle speed signals from external inputs; S2: Based on the collected steering wheel angle signal and vehicle speed signal, calculate the desired front wheel angle of the vehicle and input it into the reference model for processing to generate the ideal yaw rate; S3: The error between the ideal yaw rate and the actual yaw rate is input into the upper controller in real time, and the system lumped disturbance composed of external disturbance and internal parameter perturbation is estimated by using a second-order linear extended state observer; then, after combining the dynamic gain based on the online solution of vehicle speed to perform disturbance feedforward compensation on the system lumped disturbance, the front wheel compensation steering angle is output. S4: Superimpose the desired front wheel steering angle and the front wheel compensation steering angle, determine the preset threshold range of the current vehicle speed, and calculate the dynamic weighting factor, thereby decoupling and mapping the total steering angle composed of the desired front wheel steering angle and the front wheel compensation steering angle into independent left and right target steering angles that conform to the characteristics of low-speed Ackerman, medium-speed parallel, or high-speed anti-Ackerman. S5: The left and right lower-level controllers receive the independent target turning angles of their respective sides, calculate the turning angle tracking error and its derivative, and drive the system state to slide towards the preset sliding surface to achieve convergence of the actual turning angle of the front wheels to the target turning angle, and output continuous motor drive torque commands to independently complete the vehicle steering action.

Claims

1. A layered control system for steer-by-wire, characterized in that, include: The system consists of a state reference module, an upper-level controller, a corner allocation module, and a lower-level execution controller. The state reference module acquires the steering wheel angle signal and vehicle speed signal, and generates the desired front wheel angle and the ideal yaw rate characterizing the ideal steering state of the vehicle based on the preset vehicle reference model and transmission ratio mapping relationship, while extracting the actual yaw rate of the vehicle target model. The upper-level controller uses state and disturbance observation logic to calculate the system lumped disturbance in real time, which includes dynamic perturbations inside the vehicle and external environmental disturbances. At the same time, based on the tracking error between the actual yaw rate and the ideal yaw rate and the dynamically updated control gain, the lumped disturbance is fed forward to compensate for the lumped disturbance and outputs the front wheel compensation steering angle for yaw stability control. Based on the current vehicle speed, the steering angle allocation module uses a multi-segment steering geometry mapping strategy to decouple the desired front wheel steering angle from the total target steering angle composed of the front wheel compensation steering angle, and then obtains the target steering angle of the left front wheel and the target steering angle of the right front wheel. The lower-level execution controller includes a physically decoupled left-wheel tracking controller and a right-wheel tracking controller, which respectively receive the target steering angle of the left front wheel and the target steering angle of the right front wheel, and calculate the target control torque required to drive the left steering actuator and the right steering actuator respectively through a nonlinear robust closed-loop tracking algorithm, so as to achieve independent disturbance rejection control of the steering angles of the left and right front wheels.

2. The steer-by-wire layered control system according to claim 1, characterized in that, The transmission ratio mapping relationship in the state reference module is constructed based on the principle of constant steady-state gain of vehicle yaw rate. The transmission ratio from the steering wheel angle to the desired front wheel angle is set as a piecewise mapping function that includes the current vehicle speed, vehicle wheelbase, and vehicle understeer stability factor. When the current vehicle speed exceeds the safety limit, the output transmission ratio is truncated and limited.

3. The steer-by-wire layered control system according to claim 1, characterized in that, The upper-level controller is an active disturbance rejection controller. The state and disturbance observation logic is implemented by configuring a second-order linear extended state observer. The second-order linear extended state observer expands the system lumped disturbance into a new state variable of the system and estimates it in real time. It then cancels it out in the virtual control law, thereby approximately decoupling the nonlinear vehicle yaw dynamic system into an integral series system.

4. The steer-by-wire layered control system according to claim 2, characterized in that, The dynamically updated control gain in the upper-level controller is generated through an online optimization algorithm, including: establishing a quadratic performance index functional with the tracking error of yaw rate and control energy input as joint evaluation objectives, and obtaining the optimal proportional feedback gain that minimizes the performance index functional by solving the algebraic Riccati equation in real time; wherein, the weight coefficient used to penalize the tracking error in the quadratic performance index functional is configured as a nonlinear variable that dynamically amplifies with increasing vehicle speed.

5. The steer-by-wire layered control system according to claim 1, characterized in that, Both the left-wheel tracking controller and the right-wheel tracking controller are sliding mode controllers. The nonlinear robust closed-loop tracking algorithm is based on a sliding surface, which includes the difference between the actual front wheel steering angle and the target steering angle, as well as the rate of change of the steering angle error. The nonlinear robust closed-loop tracking algorithm uses a control law containing feedforward compensation terms and approach terms to drive the actual front wheel steering angle to converge toward the target steering angle.

6. The steer-by-wire layered control system according to claim 5, characterized in that, The nonlinear robust closed-loop tracking algorithm introduces a smoothing and anti-shaking mechanism, including: introducing a boundary layer saturation function of preset thickness into the approach term of the sliding mode controller to replace the standard sliding mode sign function; when the system state trajectory is outside the boundary layer, the sliding mode controller outputs the target control torque with the maximum amplitude limit; when the system state trajectory enters the boundary layer, the target control torque is continuously linearly interpolated within the boundary layer before being output to the steering motor.

7. The steer-by-wire layered control system according to claim 1, characterized in that, The multi-segment steering geometry mapping strategy of the steering angle allocation module is configured as follows: First, a first speed threshold, a second speed threshold, and a third speed threshold are preset in ascending order; in the range where the current vehicle speed is lower than the first speed threshold, the inner and outer wheel angles are allocated according to the pure Ackerman steering geometry principle; in the range where the current vehicle speed is between the first speed threshold and the second speed threshold, the Ackerman steering angle and the parallel steering angle are linearly interpolated and fused according to the current vehicle speed; in the range where the current vehicle speed is between the second speed threshold and the third speed threshold, the parallel steering angle and the anti-Ackerman steering angle are linearly interpolated and fused according to the current vehicle speed; in the range where the current vehicle speed is higher than the third speed threshold, the inner and outer wheel angles are allocated according to the anti-Ackerman steering geometry principle.

8. A steer-by-wire control method based on the system described in any one of claims 1 to 7, characterized in that, Includes the following steps: S1: Collects steering wheel angle signals and vehicle speed signals from external inputs; S2: Based on the collected steering wheel angle signal and vehicle speed signal, calculate the desired front wheel angle of the vehicle and input it into the reference model for processing to generate the ideal yaw rate; S3: The error between the ideal yaw rate and the actual yaw rate is input into the upper controller in real time, and the system lumped disturbance composed of external disturbance and internal parameter perturbation is estimated by using a second-order linear extended state observer; then, after combining the dynamic gain based on the online solution of vehicle speed to perform disturbance feedforward compensation on the system lumped disturbance, the front wheel compensation steering angle is output. S4: Superimpose the desired front wheel steering angle and the front wheel compensation steering angle, determine the preset threshold range of the current vehicle speed, and calculate the dynamic weighting factor, thereby decoupling and mapping the total steering angle composed of the desired front wheel steering angle and the front wheel compensation steering angle into independent left and right target steering angles that conform to the characteristics of low-speed Ackerman, medium-speed parallel, or high-speed anti-Ackerman. S5: The left and right lower-level controllers receive the independent target turning angles of their respective sides, calculate the turning angle tracking error and its derivative, and drive the system state to slide towards the preset sliding surface to achieve convergence of the actual turning angle of the front wheels to the target turning angle, and output continuous motor drive torque commands to independently complete the vehicle steering action.

Images