Vehicle control method and device, electronic equipment, storage medium and program product

Abstract

The invention relates to the technical field of vehicle control, in particular to a vehicle control method and device, electronic equipment, a storage medium and a program product. The method comprises the following steps: predicting a first predicted side slip angle corresponding to a current moment of a target vehicle based on a first preset method; predicting a second predicted side slip angle corresponding to the current moment of the target vehicle based on a second preset method; fusing the first predicted side slip angle and the second predicted side slip angle to obtain a final predicted side slip angle corresponding to the current moment of the target vehicle; calculating a target additional yawing moment corresponding to the current moment of the target vehicle based on the final predicted side slip angle; and controlling the target vehicle based on the target additional yaw moment. Sideslip, excessive steering or insufficient steering of the vehicle can be rapidly inhibited, and the lateral stability, the driving safety and the control smoothness of the vehicle under complex working conditions are remarkably improved. Therefore, the high-precision and high-robustness requirements of vehicle lateral stability control under complex roads and extreme working conditions are met.

Description

Technical Field

[0001] This invention relates to the field of vehicle control technology, specifically to vehicle control methods, devices, electronic equipment, storage media, and program products. Background Technology

[0002] With the rapid development of the automotive industry and the increasing complexity of road environments, vehicles are prone to oversteer, sideslip, and other dangerous phenomena under extreme conditions such as emergency avoidance, seriously threatening the lives of drivers and passengers. Therefore, vehicle lateral stability control has become a key research focus and core direction in the field of automotive active safety. Among the many vehicle body state variables characterizing lateral stability, yaw rate and sideslip angle are the two most critical parameters. They can accurately reflect the vehicle's lateral motion attitude and stable state, providing core inputs for control algorithm design. Yaw rate can be directly acquired by onboard inertial measurement unit (IMU) sensors, while sideslip angle is difficult to measure directly due to its special physical characteristics, requiring the design of a special estimation algorithm. Its estimation performance directly determines the effectiveness of lateral stability control.

[0003] Currently, among the centroid sideslip angle estimation algorithms, those based on kinematic and dynamic models are the most widely used, but both have significant limitations, making it difficult to balance accuracy and robustness under all operating conditions. Therefore, they cannot meet the high-precision and high-robustness requirements for vehicle lateral stability control under complex road conditions and extreme operating conditions, necessitating the development of corresponding solutions. Summary of the Invention

[0004] This invention provides a vehicle control method, device, electronic device, storage medium, and program product to solve the problem that existing technologies cannot meet the high precision and high robustness requirements of vehicle lateral stability control under complex roads and extreme working conditions.

[0005] In a first aspect, the present invention provides a vehicle control method, the method comprising: predicting a first predicted centroid sideslip angle of a target vehicle at the current moment based on a first preset method; predicting a second predicted centroid sideslip angle of the target vehicle at the current moment based on a second preset method; fusing the first predicted centroid sideslip angle and the second predicted centroid sideslip angle to obtain a final predicted centroid sideslip angle of the target vehicle at the current moment; calculating a target additional yaw moment of the target vehicle at the current moment based on the final predicted centroid sideslip angle; and controlling the target vehicle based on the target additional yaw moment.

[0006] The vehicle control method provided in this application predicts the first predicted centroid sideslip angle of the target vehicle at the current moment based on a first preset method. It then predicts the second predicted centroid sideslip angle based on a second preset method. The first and second predicted centroid sideslip angles are fused to obtain the final predicted centroid sideslip angle of the target vehicle at the current moment. This achieves a complementary advantage of high precision in linear conditions and high robustness in nonlinear conditions, ensuring stable and accurate output of the centroid sideslip angle across the entire operating range and improving the overall reliability of state estimation. Based on the final predicted centroid sideslip angle, the target additional yaw moment of the target vehicle at the current moment is calculated. Using a more reliable vehicle state as input, the calculation of the additional yaw moment is more accurate and better reflects the vehicle's actual stability requirements, avoiding over- or under-control due to estimation errors. The target vehicle is controlled based on the target additional yaw moment. This can quickly suppress vehicle sideslip, oversteering, or understeering, significantly improving the vehicle's lateral stability, driving safety, and handling smoothness under complex operating conditions. This meets the requirements for high precision and high robustness in vehicle lateral stability control under complex road conditions and extreme working conditions.

[0007] In one optional implementation, predicting the first predicted centroid sideslip angle of the target vehicle at the current moment based on a first preset method includes: constructing an initial state equation based on a two-degree-of-freedom vehicle dynamics model, using the centroid sideslip angle and yaw rate of the target vehicle as state variables; wherein the centroid sideslip angle is the core variable to be estimated; constructing an initial observation equation based on a two-degree-of-freedom vehicle dynamics model, using the lateral acceleration and yaw rate of the target vehicle as observations; and calculating the first predicted centroid sideslip angle based on the initial state equation and the initial observation equation.

[0008] The vehicle control method provided in this application uses the sideslip angle and yaw rate of the target vehicle as state variables and constructs an initial state equation based on a two-degree-of-freedom vehicle dynamics model. The model has clear physical meaning, low computational cost, and can accurately describe the lateral dynamic characteristics of the vehicle, providing a reliable model foundation for high-precision prediction of the sideslip angle. Using the lateral acceleration and yaw rate of the target vehicle as observations, an initial observation equation is constructed based on the two-degree-of-freedom vehicle dynamics model. The observations can be directly measured by an IMU, eliminating the need for additional sensors, achieving low-cost, high-real-time state observation, and effectively correcting prediction biases. Based on the initial state equation and the initial observation equation, the first predicted sideslip angle is calculated. This achieves closed-loop correction between prediction and observation, providing high estimation accuracy and no accumulated error under normal driving conditions, thus providing high-quality state information for vehicle stability control.

[0009] In one optional implementation, the first predicted centroid sideslip angle is calculated based on the initial state equation and the initial observation equation, including: performing a first-order Taylor expansion of the initial state equation and the initial observation equation using the forward Euler method to obtain a discretized state equation and a discretized observation equation; initializing the discretized state equation and the discretized observation equation; and predicting the first predicted centroid sideslip angle based on the initialized discretized state equation and the discretized observation equation.

[0010] The vehicle control method provided in this application uses the forward Euler method to perform a first-order Taylor expansion of the initial state equation and the initial observation equation, resulting in discretized state equations and discretized observation equations. This method is suitable for discrete computation by the onboard controller, with low computational complexity, high real-time performance, and ensures the engineering feasibility of vehicle state prediction. The discretized state equations and discretized observation equations are initialized, providing a stable initial starting point for state estimation, avoiding algorithm divergence, and improving the convergence speed and stability of the estimation process. Based on the initialized discretized state equations and discretized observation equations, the first predicted centroid sideslip angle is predicted. This achieves closed-loop prediction of the dynamic model, exhibiting high prediction accuracy and no integral accumulation error under normal operating conditions, providing reliable state input for vehicle stability control.

[0011] In one optional implementation, the first predicted centroid sideslip angle is predicted based on the initialized discretized state equation and discretized observation equation, including: predicting the first initial centroid sideslip angle and prior error covariance matrix corresponding to the current time based on the initialized discretized state equation and discretized observation equation; calculating the Kalman gain based on the prior error covariance matrix, the observation equation Jacobian matrix corresponding to the observation equation, and the noise covariance; and correcting the first initial centroid sideslip angle based on the Kalman gain to obtain the first predicted centroid sideslip angle.

[0012] The vehicle control method provided in this application predicts the first initial centroid sideslip angle and the prior error covariance matrix at the current moment based on the initialized discretized state equation and discretized observation equation. It can perform state prediction based on the dynamic model and quantify the prediction error, providing a basis for subsequent accurate correction. Based on the prior error covariance matrix, the Jacobian matrix of the observation equation corresponding to the observation equation, and the noise covariance, the Kalman gain is calculated to achieve optimal weighting of prediction and observation information, obtaining optimal observation correction weights under different noise environments. Based on the Kalman gain, the first initial centroid sideslip angle is corrected to obtain the first predicted centroid sideslip angle. This forms a prediction-correction closed loop, effectively suppressing noise and model errors, resulting in higher and more stable centroid sideslip angle estimation accuracy.

[0013] In one optional implementation, predicting the second predicted centroid sideslip angle of the target vehicle at the current moment based on a second preset method includes: obtaining the kinematic equations corresponding to the target vehicle; deriving the kinematic differential equations of the centroid sideslip angle of the target vehicle based on the kinematic equations; discretizing the kinematic differential equations using the forward Euler method to obtain discrete kinematic equations; and calculating the second predicted centroid sideslip angle based on the discrete kinematic equations.

[0014] The vehicle control method provided in this application obtains the kinematic equations corresponding to the target vehicle. It relies only on geometric and kinematic relationships, without depending on variable dynamic parameters such as tires and stiffness, making it applicable to a wider range of working conditions. Based on the kinematic equations, the kinematic differential equations of the center-of-gravity sideslip angle corresponding to the target vehicle are derived. A direct relationship between the center-of-gravity sideslip angle and measurable signals is established, resulting in clear physics, a simple model, and strong robustness. The forward Euler method is used to discretize the kinematic differential equations, obtaining discrete kinematic equations; this requires minimal computation, offers high real-time performance, and is suitable for online operation of onboard controllers. Based on the discrete kinematic equations, the second predicted center-of-gravity sideslip angle is calculated. It remains stable and reliable under extreme working conditions and nonlinear scenarios such as road surface changes, and will not fail due to model inaccuracies.

[0015] In one optional implementation, the first predicted centroid sideslip angle and the second predicted centroid sideslip angle are fused to obtain the final predicted centroid sideslip angle corresponding to the target vehicle at the current moment, including: obtaining the lateral acceleration corresponding to the target vehicle; determining the weights corresponding to the first predicted centroid sideslip angle and the second predicted centroid sideslip angle respectively based on the lateral acceleration; and fusing the first predicted centroid sideslip angle and the second predicted centroid sideslip angle based on the weights corresponding to the first predicted centroid sideslip angle and the second predicted centroid sideslip angle respectively to obtain the final predicted centroid sideslip angle corresponding to the target vehicle at the current moment.

[0016] The vehicle control method provided in this application obtains the lateral acceleration corresponding to the target vehicle. Based on physically meaningful operating condition parameters, it can accurately identify the vehicle's driving conditions, providing a reliable basis for subsequent weight allocation. According to the lateral acceleration, the weights corresponding to the first and second predicted centroid sideslip angles are determined respectively; the weights are dynamically and adaptively allocated according to the operating conditions, balancing the accuracy of linear operating conditions and the robustness of nonlinear operating conditions, avoiding estimation bias caused by fixed weights. Based on the weights corresponding to the first and second predicted centroid sideslip angles respectively, the first and second predicted centroid sideslip angles are fused to obtain the final predicted centroid sideslip angle corresponding to the target vehicle at the current moment. This achieves complementary advantages of dynamic and kinematic methods, obtaining stable, accurate, and robust centroid sideslip angle estimation results across the entire operating condition range.

[0017] In one optional implementation, the target additional yaw moment corresponding to the target vehicle at the current moment is calculated based on the final predicted sideslip angle, including: obtaining the target sideslip angle and target yaw rate of the target vehicle at the current moment; obtaining the actual yaw rate of the target vehicle at the current moment; calculating the sideslip angle difference between the final predicted sideslip angle and the target sideslip angle, and the yaw rate difference between the actual yaw rate and the target yaw rate; constructing an initial sliding surface corresponding to a preset sliding mode control algorithm based on the sideslip angle difference and the yaw rate difference; calculating an initial exponential convergence rate based on the initial sliding surface; obtaining a preset sliding surface width corresponding to the initial sliding surface; correcting the initial exponential convergence rate based on the preset sliding surface width to obtain a target exponential convergence rate; and calculating the target additional yaw moment corresponding to the target vehicle at the current moment based on the target exponential convergence rate.

[0018] The vehicle control method provided in this application obtains the target centroid sideslip angle and target yaw rate of the target vehicle at the current moment; it also obtains the actual yaw rate of the target vehicle at the current moment. The method calculates the difference between the predicted centroid sideslip angle and the target centroid sideslip angle, as well as the difference between the actual yaw rate and the target yaw rate. This accurately quantifies the deviation between the actual and ideal states of the vehicle, intuitively reflecting the degree of stability and facilitating subsequent control correction. Based on the difference in centroid sideslip angle and yaw rate, an initial sliding surface corresponding to a preset sliding mode control algorithm is constructed. This simultaneously constrains the centroid sideslip angle and yaw rate, taking into account both lateral and yaw stability, resulting in a more comprehensive control target. Based on the initial sliding surface, an initial exponential approach rate is calculated. This allows the target vehicle state to quickly approach the sliding surface, improving control response speed and ensuring convergence efficiency. The preset sliding surface width corresponding to the initial sliding surface is obtained, providing a quantitative boundary basis for suppressing chattering, making the correction process more stable and controllable. Based on the preset sliding surface width, the initial exponential convergence rate is corrected to obtain the target exponential convergence rate. This effectively reduces sliding mode control chattering, improves control smoothness, and maintains robustness. Based on the target exponential convergence rate, the target additional yaw moment corresponding to the target vehicle at the current moment is calculated. This achieves rapid convergence and smooth output, improving vehicle driving stability and ride comfort.

[0019] In one optional implementation, the target additional yaw moment corresponding to the target vehicle at the current moment is calculated based on the target exponential approximation rate, including: calculating the basic expression corresponding to the target additional yaw moment based on the target exponential approximation rate; the basic expression includes unknown true parameters; the unknown true parameters represent the tire lateral stiffness as a function of the distance from the center of gravity to the axle; calculating the parameter adaptive law corresponding to the unknown true parameters; substituting the parameter adaptive law into the basic expression corresponding to the target additional yaw moment to calculate the target additional yaw moment.

[0020] The vehicle control method provided in this application calculates the basic expression corresponding to the target additional yaw moment based on the target exponential approach rate. It fully preserves the vehicle's dynamic characteristics while simplifying the model structure, providing a foundation for subsequent parameter adaptive compensation. It calculates the parameter adaptive law corresponding to unknown real parameters. This enables online real-time estimation and compensation of time-varying parameters such as tire lateral stiffness, significantly improving the robustness of the control system to changes in road surface and load. Substituting the parameter adaptive law into the basic expression corresponding to the target additional yaw moment, the target additional yaw moment is calculated. Eliminating the influence of unknown parameters makes the control commands more precise and stable, reliably improving vehicle lateral stability under all operating conditions.

[0021] In one optional implementation, calculating the parameter adaptation law corresponding to the unknown true parameters includes: obtaining the parameter estimate corresponding to the unknown true parameters; calculating the parameter estimation error based on the unknown true parameters and the parameter estimate; constructing the parameter adaptation law corresponding to the parameter estimation error; constructing a Lyapunov function based on the initial sliding surface and the parameter estimation error; verifying whether the parameter adaptation law meets the stability requirements of the target vehicle based on the Lyapunov function; if the parameter adaptation law meets the stability requirements of the target vehicle, then the parameter adaptation law corresponding to the unknown true parameters is obtained.

[0022] The vehicle control method provided in this application obtains parameter estimates corresponding to unknown true parameters. This provides iteratively updatable basic variables for real-time compensation of parameter uncertainty. Based on the unknown true parameters and parameter estimates, the parameter estimation error is calculated. The deviation between the true parameters and the estimated values ​​is quantified, providing a clear basis for adaptive adjustment. A parameter adaptive law corresponding to the parameter estimation error is constructed. Online adaptive parameter correction is achieved, enabling automatic adaptation to changes in parameters such as tires and road surface. A Lyapunov function is constructed based on the initial sliding surface and the parameter estimation error. This provides a rigorous theoretical basis for determining system stability, ensuring the convergence of the control algorithm. The Lyapunov function is used to verify whether the parameter adaptive law meets the stability requirements of the target vehicle. Theoretically, this ensures the algorithm's stability and prevents divergence, improving the reliability of the control system. If the parameter adaptive law meets the stability requirements of the target vehicle, the parameter adaptive law corresponding to the unknown true parameters is obtained. A theoretically rigorous and engineering-reliable adaptive strategy is obtained, providing stable support for the final control law.

[0023] In a second aspect, the present invention provides a vehicle control device, the device comprising: The first prediction module is used to predict the first predicted centroid sideslip angle of the target vehicle at the current moment based on a first preset method. The second prediction module is used to predict the second predicted centroid sideslip angle of the target vehicle at the current moment based on the second preset method. The fusion module is used to fuse the first predicted centroid sideslip angle and the second predicted centroid sideslip angle to obtain the final predicted centroid sideslip angle of the target vehicle at the current moment. The calculation module is used to calculate the target additional yaw moment of the target vehicle at the current moment based on the final predicted centroid sideslip angle; The control module is used to control the target vehicle based on the target additional yaw moment.

[0024] The vehicle control device provided in this application predicts the first predicted centroid sideslip angle of the target vehicle at the current moment based on a first preset method. It also predicts the second predicted centroid sideslip angle of the target vehicle at the current moment based on a second preset method. The first and second predicted centroid sideslip angles are fused to obtain the final predicted centroid sideslip angle of the target vehicle at the current moment. This achieves a complementary advantage of high precision in linear conditions and high robustness in nonlinear conditions, ensuring stable and accurate output of the centroid sideslip angle across the entire operating range and improving the overall reliability of state estimation. Based on the final predicted centroid sideslip angle, the target additional yaw moment of the target vehicle at the current moment is calculated. Using a more reliable vehicle state as input, the calculation of the additional yaw moment is more accurate and better reflects the actual stability requirements of the vehicle, avoiding over- or under-control due to estimation errors. Based on the target additional yaw moment, the target vehicle is controlled. This can quickly suppress vehicle sideslip, oversteering, or understeering, significantly improving the lateral stability, driving safety, and handling smoothness of the vehicle under complex operating conditions. This meets the requirements for high precision and high robustness in vehicle lateral stability control under complex road conditions and extreme working conditions.

[0025] Thirdly, the present invention provides an electronic device, comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the computer instructions to perform the vehicle control method of the first aspect or any corresponding embodiment described above.

[0026] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to perform the vehicle control method of the first aspect or any corresponding embodiment thereof.

[0027] Fifthly, the present invention provides a computer program product, including computer instructions for causing a computer to execute the vehicle control method of the first aspect or any corresponding embodiment thereof. Attached Figure Description

[0028] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0029] Figure 1 This is a schematic flowchart of a first embodiment of a vehicle control method according to the present invention; Figure 2 This is a schematic diagram of a second process for a vehicle control method according to an embodiment of the present invention; Figure 3 This is a schematic diagram of a vehicle model according to an embodiment of the present invention; Figure 4 This is a structural block diagram of a vehicle control device according to an embodiment of the present invention; Figure 5 This is a schematic diagram of the hardware structure of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0030] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0031] It is understood that before using the technical solutions disclosed in the various embodiments of the present invention, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in the present invention and their authorization should be obtained in accordance with relevant laws and regulations through appropriate means.

[0032] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "multiple" means two or more, unless otherwise explicitly specified.

[0033] According to an embodiment of the present invention, a vehicle control method embodiment is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer target vehicle, such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0034] This embodiment provides a vehicle control method that can be used in electronic devices within a target vehicle. Figure 1 This is a flowchart of a vehicle control method according to an embodiment of the present invention, such as... Figure 1 As shown, the process includes the following steps: Step S101: Predict the first predicted centroid sideslip angle of the target vehicle at the current moment based on the first preset method.

[0035] Specifically, the electronic device can predict the first predicted centroid sideslip angle of the target vehicle at the current moment based on a first preset method.

[0036] The first preset method can be a two-degree-of-freedom vehicle dynamics model estimation algorithm based on extended Kalman filter (EKF), a dynamics model estimation algorithm based on unscented Kalman filter (UKF), or a dynamics model estimation algorithm based on model predictive control (MPC). This application embodiment does not specifically limit the first preset method.

[0037] This step will be explained in detail below.

[0038] Step S102: Predict the second predicted centroid sideslip angle of the target vehicle at the current moment based on the second preset method.

[0039] Specifically, the electronic device can predict the second predicted centroid sideslip angle of the target vehicle at the current moment based on the second preset method.

[0040] The second preset method can be a two-degree-of-freedom vehicle dynamics model estimation algorithm based on extended Kalman filter (EKF), a dynamics model estimation algorithm based on unscented Kalman filter (UKF), or a dynamics model estimation algorithm based on model predictive control (MPC). This application embodiment does not specifically limit the second preset method. The first preset method and the second preset method are different methods.

[0041] This step will be explained in detail below.

[0042] Step S103: The first predicted centroid sideslip angle and the second predicted centroid sideslip angle are fused to obtain the final predicted centroid sideslip angle corresponding to the target vehicle at the current moment.

[0043] Specifically, the electronic device can fuse the first predicted centroid sideslip angle and the second predicted centroid sideslip angle to obtain the final predicted centroid sideslip angle corresponding to the target vehicle at the current moment.

[0044] This step will be explained in detail below.

[0045] Step S104: Based on the final predicted centroid sideslip angle, calculate the target additional yaw moment corresponding to the target vehicle at the current moment.

[0046] Specifically, the electronic device can use a preset calculation method to calculate the target additional yaw moment corresponding to the target vehicle at the current moment based on the final predicted centroid sideslip angle.

[0047] This step will be explained in detail below.

[0048] Step S105: Control the target vehicle based on the target additional yaw moment.

[0049] Specifically, electronic devices can convert the calculated target additional yaw moment into specific control actions for the target vehicle, and achieve closed-loop control of the target vehicle's lateral stability through the adjustment of the actuator.

[0050] The vehicle control method provided in this application predicts the first predicted centroid sideslip angle of the target vehicle at the current moment based on a first preset method. It then predicts the second predicted centroid sideslip angle based on a second preset method. The first and second predicted centroid sideslip angles are fused to obtain the final predicted centroid sideslip angle of the target vehicle at the current moment. This achieves a complementary advantage of high precision in linear conditions and high robustness in nonlinear conditions, ensuring stable and accurate output of the centroid sideslip angle across the entire operating range and improving the overall reliability of state estimation. Based on the final predicted centroid sideslip angle, the target additional yaw moment of the target vehicle at the current moment is calculated. Using a more reliable vehicle state as input, the calculation of the additional yaw moment is more accurate and better reflects the vehicle's actual stability requirements, avoiding over- or under-control due to estimation errors. The target vehicle is controlled based on the target additional yaw moment. This can quickly suppress vehicle sideslip, oversteering, or understeering, significantly improving the vehicle's lateral stability, driving safety, and handling smoothness under complex operating conditions. This meets the requirements for high precision and high robustness in vehicle lateral stability control under complex road conditions and extreme working conditions.

[0051] This embodiment provides a vehicle control method that can be used in electronic devices within a target vehicle. Figure 2 This is a flowchart of a vehicle control method according to an embodiment of the present invention, such as... Figure 2 As shown, the process includes the following steps: Step S201: Predict the first predicted centroid sideslip angle of the target vehicle at the current moment based on the first preset method.

[0052] Specifically, step S201 above may include the following steps: Step S2011: Using the sideslip angle and yaw rate of the target vehicle as state variables, construct the initial state equations based on the two-degree-of-freedom vehicle dynamics model.

[0053] Among them, the centroid sideslip angle is the core quantity to be estimated.

[0054] Specifically, the electronic device can select the sideslip angle of the target vehicle's center of gravity. and yaw rate As the target vehicle state quantity Front wheel steering angle Input quantity The sideslip angle is the core quantity to be estimated, and the yaw rate is an auxiliary state quantity. Together, they characterize the vehicle's lateral dynamics state and are the only two state quantities describing the vehicle's lateral motion and yaw motion around the z-axis in the two-degree-of-freedom vehicle dynamics model. The two-degree-of-freedom vehicle dynamics model is a simplified lateral dynamics model, assuming the vehicle only performs lateral translation and yaw motion around the z-axis, neglecting secondary factors such as suspension deformation, tire sideslip stiffness nonlinearity, and longitudinal motion coupling, thus conforming to the dynamic characteristics of a vehicle under normal linear driving conditions.

[0055] Based on the force analysis of a two-degree-of-freedom vehicle dynamics model and Newton's second law, the electronic device derives the first-order differential relationships of the sideslip angle and yaw rate. It incorporates the inherent parameters of the target vehicle (vehicle mass, distance from the center of mass to the front and rear axles, front and rear axle sideslip stiffness, and moment of inertia about the z-axis) and external inputs (front wheel steering angle). Simultaneously, it introduces Gaussian white noise (representing modeling errors, minor road surface disturbances, and other internal uncertainties of the target vehicle) to ultimately form an initial state equation in the continuous domain. The core function of this initial state equation is to describe the changes in the sideslip angle and yaw rate over time, reflecting the intrinsic evolutionary characteristics of the vehicle's lateral dynamics.

[0056] In step S2012, the initial observation equations are constructed based on the two-degree-of-freedom vehicle dynamics model, using the lateral acceleration and yaw rate of the target vehicle as observations.

[0057] Specifically, the electronic device can select the lateral acceleration of the target vehicle. and yaw rate As a target vehicle observation Both can be measured directly and in real time by the vehicle's IMU (Inertial Measurement Unit) sensors, making the data easy to acquire and highly reliable. The yaw rate is both a state variable and an observable, forming a self-checking mechanism for the state variable and improving estimation accuracy.

[0058] Electronic devices are based on the same two-degree-of-freedom vehicle dynamics model. The functional relationship between the observed quantities (lateral acceleration, yaw rate) and the state quantities (center of mass sideslip angle, yaw rate) is derived to ensure the dynamic consistency between the observation equation and the state equation and to avoid estimation errors caused by model inconsistency.

[0059] The electronic device establishes explicit expressions for lateral acceleration and yaw rate with respect to the centroid sideslip angle and yaw rate through dynamic derivation. Simultaneously, it introduces target vehicle measurement noise in the form of Gaussian white noise (characterizing external uncertainties such as IMU sensor measurement errors and signal transmission errors), ultimately forming the initial observation equation in the continuous domain. The core function of this equation is to bridge the gap between state variables and observed variables, allowing the centroid sideslip angle, which cannot be directly measured, to be indirectly estimated through measurable observations.

[0060] For example, the initial state equation and the initial observation equation are shown below: (1) In the formula, , , , , and The target vehicle's process noise and measurement noise are defined as Gaussian white noise, respectively. , , Let the process noise covariance of the target vehicle be denoted as . The covariance matrix of the measurement noise; For the overall vehicle quality, For longitudinal velocity, The distance from the center of gravity to the front axle. The distance from the center of gravity to the rear axle. For front axle lateral stiffness, For rear axle lateral stiffness, Let be the vehicle's moment of inertia about the z-axis.

[0061] Step S2013: Calculate the first predicted centroid sideslip angle based on the initial state equation and the initial observation equation.

[0062] Specifically, step S2013 above may include the following steps: Step a1: Perform a first-order Taylor expansion on the initial state equation and initial observation equation based on the forward Euler method to obtain the discretized state equation and discretized observation equation.

[0063] Specifically, the electronic device can discretize the initial state equation of the continuous domain using a sampling step size T (typically 0.01~0.05s in engineering, which is the fixed sampling interval for vehicle sensors / controllers) as the time increment. Simultaneously, it linearizes the nonlinear terms through a first-order Taylor expansion, introducing the Jacobian matrix of the initial state equation (representing the rate of change of state variables), ultimately obtaining the discretized state equation. This equation clarifies the discrete recursive relationship between the state variables (centroid sideslip angle, yaw rate) at the current time k and the state variables and input variables (front wheel steering angle) at the previous time k-1, incorporating the vehicle's inherent parameters and the target vehicle's process noise.

[0064] Furthermore, electronic devices can also employ a sampling step size T consistent with the state equation to discretize the initial observation equation in the continuous domain using the forward Euler method. Simultaneously, a first-order Taylor expansion is used for linearization, introducing the Jacobian matrix of the observation equation (representing the linear mapping relationship between the observed quantities and the state variables), ultimately yielding the discretized observation equation. This equation clarifies the discrete mapping relationship between the observed quantities (lateral acceleration, yaw rate) at the current time k and the state variables at the previous time k, incorporating measurement noise.

[0065] The discretized state equations and observation equations are normalized into the standard matrix form of the extended Kalman filter, defining core matrices such as the state transition matrix, input matrix, observation matrix, and identity matrix, providing a standardized mathematical form for subsequent initialization and iterative calculations.

[0066] For example, the discretized state equation and the discretized observation equation are shown below: (2) In the formula, , It is the identity matrix. , , T is the sampling step size.

[0067] Step a2: Initialize the discretized state equation and the discretized observation equation.

[0068] Specifically, an initial state vector is set. This vector contains the initial values ​​of the sideslip angle and yaw rate. It is assumed that the target vehicle is initially traveling at a constant speed in a straight line (no sideslip, no yaw), therefore the initial value of the sideslip angle β = 0 and the initial value of the yaw rate γ = 0 are taken as follows: .

[0069] Set the initial estimation error covariance matrix This matrix represents the degree of error in the initial state estimate. To ensure the reliability of the initial estimate, a small diagonal matrix is ​​used (e.g., P0=diag(0.01,0.01)). The diagonal elements are extremely small positive numbers, indicating that the error of the initial state estimate is extremely low and there is no obvious deviation.

[0070] The electronic device can also set the target vehicle process noise covariance matrix Q and the measurement noise covariance matrix R. Q represents the statistical characteristics of the target vehicle process noise (modeling error, road surface micro-disturbance) in the initial state equation, and R represents the statistical characteristics of the measurement noise (IMU sensor measurement error, signal transmission error) in the initial observation equation. Both are calibrated as fixed diagonal matrices based on actual vehicle tests, providing a basis for subsequent error calculation.

[0071] The electronic device substitutes the above initial values ​​into the discretized state equation and the discretized observation equation to complete the initialization of the equation, transforming the equation from mathematical form into a practical equation that can be directly substituted into the calculation.

[0072] Step a3: Based on the initialized discretized state equation and discretized observation equation, the first predicted centroid sideslip angle is predicted.

[0073] Specifically, step a3 above may include the following steps: Step a31: Based on the initialized discretized state equation and discretized observation equation, predict the first initial centroid side slip angle and prior error covariance matrix corresponding to the current time.

[0074] Specifically, the electronic device can estimate the optimal state at time k-1. (Including the optimal values ​​of the centroid sideslip angle and yaw rate at time k-1) Substitute these values ​​into the initialized discretized state equations, and combine them with the current input (front wheel steering angle δ). f Based on the target vehicle's process noise, the predicted prior state value at time k is calculated. Extracting the centroid sideslip angle component from this prior state prediction value yields the first initial centroid sideslip angle at the current moment. This value is a pure model prediction result without incorporating measured data.

[0075] Next, the electronic device can combine the state equation Jacobian matrix Φ corresponding to the state equation and the optimal estimation error covariance matrix P at time k-1. k-1 The target vehicle process noise covariance matrix Q is calculated using the prior error covariance calculation formula of the extended Kalman filter, yielding the prior error covariance matrix at time k. This matrix is ​​a symmetric positive definite matrix, representing the error magnitude and error coupling relationship of the prior prediction values ​​of the first initial centroid sideslip angle and yaw rate, providing a core error basis for subsequent Kalman gain calculations.

[0076] If it is the first sampling time of the filtering algorithm, then the initialization state value of step a2 will be... The initial covariance matrix P0 is substituted into the calculation as a parameter at time k-1; if it is a subsequent sampling time, the optimal value at time k-1 obtained in the previous iteration step a33 is substituted into the calculation to achieve continuous iteration.

[0077] For example, the predicted prior state value at the current time obtained based on the discretized state equation. and prior error covariance matrix As shown below: (3) In the formula, Let be the state transition matrix of the target vehicle. , Let be the Jacobian matrix of the state equation. .

[0078] Step a32: Calculate the Kalman gain based on the prior error covariance matrix, the observation equation Jacobian matrix corresponding to the observation equation, and the noise covariance.

[0079] Specifically, the electronic device will calculate the prior error covariance matrix. The observation equations, including the Jacobian matrix H (obtained through linearization in step a1, representing the linear mapping between the observed quantity and the state quantity) and the measurement noise covariance matrix R (initialized in step a2, representing the sensor measurement error), are substituted into the Kalman gain standard calculation formula for the extended Kalman filter: The Kalman gain matrix K at time k is obtained through matrix operations. k .

[0080] If the sensor measurement accuracy is high (R is small), the Kalman gain will be too large, and subsequent corrections will focus more on the IMU's measured observations; if the prior prediction model accuracy is high ( If the value is small, the Kalman gain will be too small, and subsequent corrections will focus more on the prior prediction value of step a31 to achieve optimal fusion with weighted fusion as needed.

[0081] Step a33: Based on the Kalman gain, the first initial centroid sideslip angle is corrected to obtain the first predicted centroid sideslip angle.

[0082] Specifically, the electronic device can calculate the prior state prediction value at time k. Substituting the initialized discretized observation equations, the predicted observation values ​​at time k are calculated. Simultaneously, the actual value z of the observation at time k measured by the IMU sensor is acquired. k (Including measured values ​​of lateral acceleration and yaw rate), calculate the deviation between the two. This bias characterizes the difference between the model's predictions and actual measurements.

[0083] Electronic devices can correlate the Kalman gain matrix Kk with the observation bias. Multiply to obtain the state correction, then add this correction to the prior state prediction at time k. The optimal state estimate at time k is obtained. .

[0084] Then, the electronic device obtains the optimal state estimate from time k. In the process, the centroid sideslip angle component is extracted, which is the first predicted centroid sideslip angle of the target vehicle at the current moment. This value is the optimal estimate that combines model prediction and measured data, and its accuracy is much higher than the first initial centroid sideslip angle predicted by pure model.

[0085] To ensure the accuracy of error calculation in the next iteration, the electronic device can use the Kalman gain matrix K. k Prior error covariance matrix The optimal estimation error covariance matrix P at time k is obtained by updating the Jacobian matrix H of the observation equation. k This is used as the input parameter for step a31 at time k+1 to achieve continuous iteration of the filtering algorithm.

[0086] For example, an electronic device can calculate the Kalman gain matrix based on the following formula, and calculate the first predicted centroid sideslip angle and the optimal estimation error covariance matrix based on the Kalman gain matrix.

[0087] (4) In the formula, Let Jacobian matrix be the prediction equation. .

[0088] Step S202: Predict the second predicted centroid sideslip angle of the target vehicle at the current moment based on the second preset method.

[0089] Specifically, step S202 above may include the following steps: Step S2021: Obtain the kinematic equations corresponding to the target vehicle.

[0090] Specifically, the kinematic equations of the target vehicle are derived based on rigid body planar kinematics, which only describe the geometric relationship between the linear motion (lateral and longitudinal) of the vehicle's center of mass and its angular motion (yaw) around the z-axis, without involving dynamic factors such as forces and tire characteristics.

[0091] Electronic devices can acquire the kinematic equations of the target vehicle input by the user, construct the kinematic equations of the target vehicle, or extract them from a vehicle kinematic model library without additional calibration.

[0092] The kinematic equations corresponding to the target vehicle are: ;in, The lateral velocity of the vehicle's center of mass (the velocity component perpendicular to the vehicle's longitudinal axis). β is the longitudinal velocity of the vehicle's center of mass (the velocity component along the vehicle's longitudinal axis, which can be measured by wheel speed sensors / GPS); β is the sideslip angle of the center of mass (a core quantity to be estimated, defined as the angle between the vehicle's longitudinal axis and the velocity vector of the center of mass); γ is the yaw rate of the vehicle (the rotational angular velocity about the z-axis, measured by an IMU sensor); L is the wheelbase of the vehicle (the distance from the center of mass to the front axle + the distance from the center of mass to the rear axle, an inherent parameter of the vehicle, which is known).

[0093] Step S2022: Based on the kinematic equations, derive the kinematic differential equations of the sideslip angle of the target vehicle's center of gravity.

[0094] Specifically, electronic devices can estimate the sideslip angle of the center of mass based on the kinematic integral method, and derive the kinematic differential equation of the sideslip angle of the center of mass based on the kinematic equation.

[0095] For example, the derivation process is as follows: (5) Step S2023: Discretize the kinematic differential equations using the forward Euler method to obtain discrete kinematic equations.

[0096] Specifically, the electronic device can define the sampling step size. The forward Euler method is used to discretize the kinematic differential equations.

[0097] Specifically, the forward Euler method is a first-order numerical integration method. Its core principle is to approximate the current state value using the rate of change from the previous time step. The formula is: Where: T: sampling step size (generally taken as 0.01~0.05s in engineering, to match the sampling frequency of vehicle sensors / controllers) The state value at time k-1 (here, the centroid sideslip angle β) k-1 ). The rate of change of state at time k-1 (here, the rate of change of the centroid sideslip angle) ); The state value at time k (the centroid sideslip angle β to be determined) k ).

[0098] Then, the electronic device substitutes the kinematic differential equations into the forward Euler formula to obtain the discrete kinematic equations.

[0099] For example, the discrete kinematic equations are: (6) Step S2024: Based on the discrete kinematic equations, the second predicted centroid sideslip angle is calculated.

[0100] Specifically, the electronic device sets the initial value of the integral, assuming that the target vehicle is initially in a state of uniform straight-line driving (without side slip or yaw). Therefore, the initial value of the centroid side slip angle at time 0 is: β0=0. This initial value is the starting point of the first calculation, without the dependence of actual measured data, and conforms to the normal operating conditions of vehicle startup.

[0101] Then, taking a sampling step size T=0.01s as an example, the calculation is performed time-by-time in chronological order: time k=1 (the first sampling point). Initial value β0=0, measured value a at time 0. y,0 (lateral acceleration), v x,0 (Longitudinal velocity), r0 (yaw rate). Time k=2 (second sampling point). Input: Result β1 from the previous time step, measured value a from the first time step. y,1 v x,1 , r1. Repeat the above logic, taking "the estimated centroid sideslip angle of the previous moment + the sensor's measured value at the current moment" as input, and substituting it into the discrete equation to calculate β at the current moment. k .

[0102] Step S203: The first predicted centroid sideslip angle and the second predicted centroid sideslip angle are fused to obtain the final predicted centroid sideslip angle corresponding to the target vehicle at the current moment.

[0103] Specifically, step S203 above may include the following steps: Step S2031: Obtain the lateral acceleration corresponding to the target vehicle.

[0104] Specifically, electronic devices can collect the lateral acceleration of the target vehicle in real time through the IMU (Inertial Measurement Unit) sensor on the target vehicle. The sampling frequency is consistent with the sampling step size for estimating the center of mass sideslip angle (e.g., 100Hz, corresponding to a step size of 0.01s), and the data is a continuous numerical signal.

[0105] Among them, the lateral acceleration a y It is the acceleration of the target vehicle along the direction perpendicular to the longitudinal axis, and its absolute value directly reflects the lateral dynamic characteristics of the vehicle: |a y |small (e.g., |a) y |<0.5m / s 2 The vehicle is in a linear driving condition (uniform speed, straight line, slight steering), the tire sideslip characteristics are linear, and the dynamic model (first predicted value) has high accuracy; |a y | large (e.g., | a)y |≥0.5m / s 2 When a vehicle is in a nonlinear driving condition (large-angle steering, emergency avoidance), the tire side slip characteristics are nonlinear, the dynamic model is prone to distortion, and the kinematic model (second prediction value) is more reliable.

[0106] Step S2032: Based on the lateral acceleration, determine the weights corresponding to the first predicted centroid sideslip angle and the second predicted centroid sideslip angle, respectively.

[0107] Specifically, the electronic device can calculate the absolute value of the lateral acceleration corresponding to the lateral acceleration, and obtain the initial weights corresponding to the first and second predicted centroid sideslip angles, respectively. Then, the absolute value of the lateral acceleration of the electronic device is compared with a preset threshold value for the absolute value of lateral acceleration. If the absolute value of the lateral acceleration is less than the preset threshold value (|a|), the electronic device will calculate the absolute value of the lateral acceleration. y |<0.5m / s 2 If the initial weight w1 corresponding to the first predicted centroid sideslip angle is multiplied by a first coefficient greater than 1, and the initial weight w2 corresponding to the second predicted centroid sideslip angle is multiplied by a second coefficient less than 1, the high-precision estimation result of the dynamic method is preferred.

[0108] If the absolute value of lateral acceleration is greater than or equal to the preset absolute value threshold of lateral acceleration (e.g., |a|), y |≥0.5m / s 2 The initial weight w1 corresponding to the first predicted centroid sideslip angle is multiplied by a second coefficient less than 1, and the initial weight w2 corresponding to the second predicted centroid sideslip angle is multiplied by a first coefficient greater than 1. The robust estimation result of the kinematic method is preferred.

[0109] The adjusted weights satisfy w1+w2=1 (the weights sum to 1, ensuring that the physical meaning of the values ​​remains unchanged after fusion), and 0≤w1,w2≤1 (the weights are non-negative, to avoid reverse correction).

[0110] For example, the electronic device can employ an exponentially smoothed weighting function (avoiding abrupt changes in the fusion value caused by step weights) to calculate the weights at the current moment, using the absolute value of the lateral acceleration |ay,k| as input. The weights w1,k for the first predicted centroid sideslip angle are: Where α is the weight attenuation coefficient (accurate on a real vehicle, e.g., α = 0.8s). 2 / m 2 The larger |ay,k| is, the smaller w1,k is, which conforms to the weight decay law of linear and nonlinear working conditions; the weight w2,k of the second predicted centroid sideslip angle: w2,k=1 w1,k satisfies the constraint that the sum of the weights is 1. The larger |ay,k| is, the larger w2,k is, which adapts to the weight enhancement requirements of nonlinear working conditions.

[0111] For example, in step S201, the first predicted centroid sideslip angle is estimated based on the extended Kalman filter algorithm. And the second predicted centroid sideslip angle estimated based on the kinematic integral method in step S202 As input, the absolute value of the lateral acceleration characterizes the nonlinearity of the target vehicle, and the determination is made. Weighting coefficients : (7) In the formula, , and for Follow The calibration coefficients that vary.

[0112] Step S2033: Based on the weights corresponding to the first predicted centroid sideslip angle and the second predicted centroid sideslip angle respectively, the first predicted centroid sideslip angle and the second predicted centroid sideslip angle are fused to obtain the final predicted centroid sideslip angle corresponding to the target vehicle at the current time.

[0113] Specifically, the electronic device fuses the first predicted centroid sideslip angle and the second predicted centroid sideslip angle with their respective adjusted weights to obtain the final predicted centroid sideslip angle of the target vehicle at the current moment.

[0114] For example, ;in, The adjusted weights are those corresponding to the first predicted centroid sideslip angle. The adjusted weights are those corresponding to the second predicted centroid sideslip angle. The first predicted centroid sideslip angle, This is the second predicted centroid sideslip angle.

[0115] Step S204: Based on the final predicted centroid sideslip angle, calculate the target additional yaw moment corresponding to the target vehicle at the current moment.

[0116] Specifically, step S204 above may include the following steps: Step S2041: Obtain the target centroid sideslip angle and target yaw rate of the target vehicle at the current moment.

[0117] Specifically, the electronic device can calculate the target centroid sideslip angle and target yaw rate of the target vehicle at the current moment based on the historical state data corresponding to the target vehicle.

[0118] Electronic devices can also derive the target centroid sideslip angle and target yaw rate of the target vehicle at the current moment based on a two-degree-of-freedom vehicle dynamics model.

[0119] For example, the target yaw rate is derived from a two-degree-of-freedom vehicle dynamics model, as shown in the formula: Where K is the vehicle stability factor ( ), m is the total vehicle mass, L f L is the distance from the center of gravity to the front axle. r C is the distance from the center of mass to the rear axle. cf For the front axle tire lateral stiffness, C cr For the rear axle tire lateral stiffness, The front wheel steering angle (measured), v x L is the longitudinal velocity (measured), and L is the wheelbase (known); the target centroid sideslip angle is derived from a linear two-degree-of-freedom model, and the formula is... , or simplified to .

[0120] Step S2042: Obtain the actual yaw rate of the target vehicle at the current moment.

[0121] Specifically, electronic devices can collect the actual yaw rate of the target vehicle at the current moment in real time through the IMU (Inertial Measurement Unit) sensor on the vehicle. The sampling frequency is consistent with the frequency of target value calculation and centroid sideslip angle estimation (e.g., 100Hz).

[0122] Step S2043: Calculate the difference between the predicted center of mass sideslip angle and the target center of mass sideslip angle, and the difference between the actual yaw rate and the target yaw rate.

[0123] Specifically, the electronic device calculates the difference between the predicted centroid sideslip angle and the target centroid sideslip angle. The formula is: ;in, To ultimately predict the centroid sideslip angle, For the target centroid sideslip angle, This represents the difference in the centroid sideslip angle.

[0124] The electronic device calculates the difference in yaw rate between the actual yaw rate and the target yaw rate. The formula is: ;in, This is the actual yaw rate. The target yaw rate.

[0125] Step S2044: Based on the difference in the center of mass side slip angle and the difference in the yaw rate, construct the initial sliding surface corresponding to the preset sliding mode control algorithm.

[0126] Specifically, the sliding surface is the "reference trajectory" of sliding mode control. The design goal is to make the target vehicle state (deviation) quickly approach and move along the sliding surface, eventually converging to 0. The center of gravity sideslip angle deviation and yaw rate deviation are selected as the core variables, and weighting factors are introduced to balance the control priority of the two.

[0127] Sliding surface for: (8) Where s is the sliding surface function value (core control variable); This is the weighting factor for the yaw rate deviation. (Originally obtained through calibration, set to 10 in simulation). and These are the target's yaw rate and the target's center of mass sideslip angle, respectively. Electronic equipment can control the target to "make s→0".

[0128] Step S2045: Based on the initial sliding surface, calculate the target additional yaw moment corresponding to the target vehicle at the current moment.

[0129] Specifically, step S2045 above may include the following steps: Step b1: Calculate the initial exponential convergence rate based on the initial sliding surface.

[0130] Specifically, the initial exponential approach rate allows the target vehicle state to quickly approach the sliding surface, and the formula is: ;in, The time derivative of the sliding surface function; To approach velocity ( >0, the larger the value, the faster it approaches the sliding surface); Exponential approach term adjustment parameter ( >0, adjusts the convergence speed of the rapid arrival phase); sgn(s): sign function (1 when s>0, -1 when s<0, 0 when s=0), ensures that the target vehicle moves toward the sliding surface.

[0131] Step b2: Obtain the preset sliding surface width corresponding to the initial sliding surface.

[0132] Specifically, the electronic device can receive the preset sliding surface width corresponding to the initial sliding surface input by the user. It can also receive the preset sliding surface width corresponding to the initial sliding surface sent by other devices. Furthermore, the preset sliding surface width can be set according to the current state of the target vehicle. .

[0133] This application does not specifically limit the method by which the electronic device obtains the preset sliding surface width corresponding to the initial sliding surface.

[0134] Step b3: Based on the preset sliding surface width, the initial exponential convergence rate is corrected to obtain the target exponential convergence rate.

[0135] Specifically, the electronic device can construct a saturation function sat(s) based on a preset sliding surface width. Then, the initial exponential approximation rate is corrected based on the saturation function sat(s) to obtain the target exponential approximation rate.

[0136] For example, the corrected target exponential convergence rate is: (9) Step b4: Based on the target exponential convergence rate, calculate the target additional yaw moment corresponding to the target vehicle at the current moment.

[0137] Specifically, step b4 above may include the following steps: Step b41: Based on the target exponential convergence rate, calculate the basic expression corresponding to the target additional yaw moment.

[0138] The basic expression contains unknown real parameters; these unknown real parameters represent the tire's lateral stiffness as a function of the distance from the center of mass to the shaft.

[0139] Specifically, the unknown real parameter is the tire lateral stiffness (front and rear axle tire lateral stiffness C). cf C cr ) and the distance from the center of mass to the front and rear axes (L) f L r The composite function of ), i.e., θ=f(C cf C cr ,L f ,L r The reason why this unknown real parameter is an unknown quantity is that tire lateral stiffness changes in real time with road surface adhesion coefficient, tire wear, and load transfer, and cannot be accurately characterized by a fixed value. It is the most important source of uncertainty in vehicle lateral control parameters.

[0140] Specifically, the electronic device can calculate the first-order time derivative of the initial sliding surface s, and combine it with the derivatives of the centroid sideslip angle deviation and the yaw rate deviation to obtain the sliding surface derivative. Converted to the rate of change of centroid side slip angle yaw acceleration The form is then used. Then, the target exponential convergence rate is substituted into the sliding surface derivative. The yaw acceleration is obtained. The constraint expression (forced yaw motion converges according to the approach law of sliding mode control).

[0141] Then, the above yaw acceleration constraint expression is substituted into the vehicle yaw motion dynamics equation, and the known terms related to tire lateral stiffness and center of gravity wheelbase are integrated into a functional form of the unknown true parameter θ. Finally, the integrated yaw motion equation is rearranged and simplified to obtain the expression for the target additional yaw moment Mz, which is the basic expression for the target additional yaw moment.

[0142] For example, see Figure 3 Combining the vehicle yaw motion equation shown in equation (10), the direct yaw moment required for vehicle lateral stability control is obtained. Expression: The calculation process can be shown in the following formula: (10) Step b42: Calculate the parameter adaptation law corresponding to the unknown true parameters.

[0143] Specifically, step b42 above may include the following steps: Step b421: Obtain the parameter estimates corresponding to the unknown true parameters.

[0144] Specifically, electronic devices can define parameter estimates corresponding to unknown true parameters. . Unknown true parameters The real-time estimate is a variable that updates dynamically over time. Its initial value can be set as the nominal value of the parameter (the composite value of tire lateral stiffness and wheelbase calibrated by vehicle design parameters, such as...). , , (This is the nominal value of tire lateral stiffness) to ensure the reasonableness of the initial estimate.

[0145] rate of change over time The final implementation is determined by the parameter adaptation law in the subsequent design. For unknown true parameters asymptotic tracking (i.e., over time, Gradually converged to (the true value).

[0146] Step b422: Based on the unknown true parameters and parameter estimates, the parameter estimation error is calculated.

[0147] Specifically, electronic devices can obtain the parameter estimation error by subtracting the parameter estimate from the unknown true parameter, using the following formula: . This indicates that there is a deviation between the current parameter estimate and the true value. This indicates that the estimated value perfectly tracks the true value. The design goal of the adaptive law is to ensure that... It converges rapidly to 0 over time, achieving accurate parameter estimation.

[0148] Electronic devices can calculate the time first derivative of the parameter estimation error, and thus obtain... Due to unknown true parameters The rate of change is much slower than the control sampling rate (tire side stiffness is a slow time-varying parameter), and in engineering, it can be approximated as... Therefore, it is simplified to This relationship provides a crucial basis for subsequent differentiation and stability verification of Lyapunov functions.

[0149] Step b423: Construct the parameter adaptive law corresponding to the parameter estimation error.

[0150] Specifically, the adaptive law needs to satisfy the following: "The larger the sliding surface s, the faster the parameter update speed; the parameter estimation error..." "The larger the deviation, the greater the parameter update magnitude," which links the parameter estimation update pattern with the deviation state of vehicle lateral stability control, thus achieving "the greater the deviation, the more timely the compensation."

[0151] Electronic devices can incorporate the robust design requirements of sliding mode control. In engineering, a proportional adaptive law is used as the initial form, with the core being the time derivative of the parameter estimates. The parameter update rate is proportional to the known function terms related to the sliding surface s and the parameters.

[0152] For example, the parameter adaptive law is as follows: (11) In the formula, For adaptive gain, .

[0153] This adaptive law is in a form to be verified. It only guarantees that parameter updates are linked with the control state and has not yet been verified for stability. It needs to be verified by subsequent Lyapunov functions to see if it meets the vehicle stability requirements. Once the verification is successful, it will become the final parameter adaptive law.

[0154] Step b424: Based on the initial sliding surface and parameter estimation error, construct the Lyapunov function.

[0155] Specifically, the electronic device can combine the state variables of sliding mode control (sliding surface s, representing the vehicle's lateral stability deviation) and the state variables of parameter adaptation (parameter estimation error). The parameter estimation bias is represented by a quadratic positive definite function, which is constructed to ensure the positive definiteness of the function. At the same time, the function includes all key deviation variables of the target vehicle, so as to realize the overall stability verification of the "control state + parameter estimation state".

[0156] The specific form of the Lyapunov function: (12) Step b425 verifies whether the parameter adaptive law meets the stability requirements of the target vehicle based on the Lyapunov function.

[0157] Specifically, The "total energy" of the target vehicle (characterized by a Lyapunov function, including control error energy and parameter estimation error energy) is monotonically decreasing or remains constant, eventually converging to a minimum of 0, corresponding to the target vehicle state s=0 and the parameter estimation error. This means that the vehicle's lateral stability deviation converges, the parameter estimation is error-free, and the target vehicle achieves asymptotic stability.

[0158] Specifically, the electronic device can take the time first derivative of the constructed Lyapunov function V. Then, substitute the derivative relationship between the target exponential convergence rate and the parameter estimation error. ,Will Transform into a form containing s, , In the form of.

[0159] Then, verify the simplified version. Does it meet the requirements? If simplified If the initial adaptive law satisfies the Lyapunov stability criterion, then return to step b423 to adjust the form of the adaptive law (such as modifying the correlation method of the adaptive gain or adjusting the known function terms) and re-verify until the requirements are met.

[0160] Only when When the initial adaptive law is always true, it satisfies the stability requirements of the target vehicle and can guarantee the overall stability of the parameter estimation process and the vehicle lateral control process. If it is not true, the adaptive law will cause the parameter estimation to diverge and the control quantity to become unstable, making it unusable.

[0161] Step b426: If the parameter adaptive law satisfies the stability requirements of the target vehicle, then the parameter adaptive law corresponding to the unknown real parameters is obtained.

[0162] Specifically, if the verification result of step b425 meets the vehicle stability requirements, then the initial adaptive law designed in step b423 is used as the unknown true parameter. The corresponding final parameter adaptive law. This final adaptive law can calculate parameter estimates in real time. The update rate enables the handling of unknown real parameters. Asymptotic tracking.

[0163] The electronic device combines the final parameter adaptive law with the vehicle's sampling step size, and discretizes it using the forward Euler method. (where T is the sampling step size), which is converted into a discretized form that can be executed by the vehicle controller, ensuring the real-time performance of parameter estimation.

[0164] Step b43: Substitute the parameter adaptive law into the basic expression corresponding to the target additional yaw moment to calculate the target additional yaw moment.

[0165] Specifically, the unknown true parameters in the fundamental expression of torque. From parameter estimates Substitute ( It is updated in real time by the parameter adaptive law, and can accurately track The true value), while in the adaptive law and Substitute the relevant terms into the basic expression to eliminate all unknowns.

[0166] Finally, the electronic device will estimate the parameters. Substitute the basic torque expression and replace all unknown real parameters. The expression is then finalized and simplified to obtain the final calculation result of the target additional yaw moment.

[0167] For example, the electronic device can substitute the above formula (11) into formula (10) to obtain the target additional yaw moment control law as follows: .

[0168] Step S205: Control the target vehicle based on the target additional yaw moment.

[0169] Please refer to the description of step S105 for details on this step, which will not be repeated here.

[0170] The vehicle control method provided in this application uses the sideslip angle and yaw rate of the target vehicle as state variables and constructs an initial state equation based on a two-degree-of-freedom vehicle dynamics model. The model has clear physical meaning, low computational cost, and can accurately describe the lateral dynamic characteristics of the vehicle, providing a reliable model foundation for high-precision prediction of the sideslip angle. Using the lateral acceleration and yaw rate of the target vehicle as observations, an initial observation equation is constructed based on the two-degree-of-freedom vehicle dynamics model. The observations can be directly measured by an IMU, eliminating the need for additional sensors, achieving low-cost, high-real-time state observation, and effectively correcting prediction biases. A first-order Taylor expansion is performed on the initial state equation and initial observation equation using the forward Euler method to obtain discretized state equations and discretized observation equations. This is suitable for discrete computation by the onboard controller, with low computational cost and high real-time performance, ensuring the engineering feasibility of vehicle state prediction. The discretized state equations and discretized observation equations are initialized, providing a stable initial starting point for state estimation, avoiding algorithm divergence, and improving the convergence speed and stability of the estimation process. Based on the initialized discretized state equation and discretized observation equation, the first initial centroid sideslip angle and prior error covariance matrix corresponding to the current time are predicted. This enables state prediction based on the dynamic model, while simultaneously quantifying the prediction error, providing a basis for subsequent accurate correction. Based on the prior error covariance matrix, the Jacobian matrix of the observation equation corresponding to the observation equation, and the noise covariance, the Kalman gain is calculated to achieve optimal weighting of prediction and observation information, obtaining optimal observation correction weights under different noise environments. Based on the Kalman gain, the first initial centroid sideslip angle is corrected to obtain the first predicted centroid sideslip angle. This forms a prediction-correction closed loop, effectively suppressing noise and model errors, resulting in higher and more stable centroid sideslip angle estimation accuracy.

[0171] Then, the kinematic equations corresponding to the target vehicle are obtained. Relying only on geometric and kinematic relationships, and not on variable dynamic parameters such as tires and stiffness, it is applicable to a wider range of operating conditions. Based on the kinematic equations, the kinematic differential equations of the center-of-gravity sideslip angle corresponding to the target vehicle are derived. A direct relationship between the center-of-gravity sideslip angle and measurable signals is established, which is physically clear, the model is simple, and it has strong robustness. The forward Euler method is used to discretize the kinematic differential equations, resulting in discrete kinematic equations; the computational load is minimal, and the real-time performance is high, making it suitable for online operation of the vehicle controller. Based on the discrete kinematic equations, the second predicted center-of-gravity sideslip angle is calculated. It remains stable and reliable under extreme operating conditions and nonlinear scenarios such as road surface changes, and will not fail due to model inaccuracy. The lateral acceleration corresponding to the target vehicle is obtained. Based on physically meaningful operating condition characterization parameters, the vehicle's driving conditions can be accurately identified, providing a reliable basis for subsequent weight allocation. Based on lateral acceleration, the weights corresponding to the first and second predicted sideslip angles are determined. This allows for dynamic adaptive allocation of weights according to different operating conditions, balancing accuracy under linear conditions with robustness under nonlinear conditions, and avoiding estimation bias caused by fixed weights. Based on the weights corresponding to the first and second predicted sideslip angles, they are fused to obtain the final predicted sideslip angle of the target vehicle at the current moment. This achieves a complementary approach between dynamic and kinematic methods, obtaining stable, accurate, and robust sideslip angle estimation results across the entire operating range.

[0172] Next, the target centroid sideslip angle and target yaw rate for the target vehicle at the current moment are obtained; the actual yaw rate for the target vehicle at the current moment is also obtained. The difference between the predicted centroid sideslip angle and the target centroid sideslip angle, and the difference between the actual yaw rate and the target yaw rate are calculated. This accurately quantifies the deviation between the actual and ideal states of the vehicle, intuitively reflects the degree of stability, and facilitates subsequent control correction. Based on the difference in centroid sideslip angle and yaw rate, an initial sliding surface corresponding to the preset sliding mode control algorithm is constructed. Simultaneously constraining the centroid sideslip angle and yaw rate, both lateral and yaw stability are considered, resulting in more comprehensive control. Based on the initial sliding surface, the initial exponential approach rate is calculated. This allows the target vehicle state to quickly approach the sliding surface, improving control response speed and ensuring convergence efficiency. The preset sliding surface width corresponding to the initial sliding surface is obtained, providing a quantitative boundary basis for suppressing chattering, making the correction process more stable and controllable. Based on the preset sliding surface width, the initial exponential approach rate is corrected to obtain the target exponential approach rate. This method effectively reduces chattering in sliding mode control, improves control smoothness, and maintains robustness. Based on the target exponential convergence rate, the fundamental expression corresponding to the target additional yaw moment is calculated; the vehicle dynamics characteristics are fully preserved while the model structure is simplified, providing a foundation for subsequent adaptive parameter compensation. Parameter estimates corresponding to unknown true parameters are obtained. This provides iteratively updatable basic variables for real-time compensation of parameter uncertainties. Based on the unknown true parameters and parameter estimates, the parameter estimation error is calculated. The deviation between the true parameters and the estimated values ​​is quantified, providing a clear basis for adaptive adjustment. An adaptive parameter law corresponding to the parameter estimation error is constructed. Online adaptive parameter correction is achieved, automatically adapting to changes in tire, road surface, and other parameters. A Lyapunov function is constructed based on the initial sliding surface and parameter estimation error. This provides a rigorous theoretical basis for system stability, ensuring the convergence of the control algorithm. The adaptive parameter law is verified based on the Lyapunov function to ensure the stability requirements of the target vehicle. This theoretically ensures the algorithm's stability and prevents divergence, improving the reliability of the control system. If the adaptive parameter law meets the target vehicle stability requirements, the adaptive parameter law corresponding to the unknown true parameters is obtained. A theoretically rigorous and engineering-reliable adaptive strategy is obtained, providing stable support for the final control law. By substituting the adaptive law of parameters into the basic expression corresponding to the target additional yaw moment, the target additional yaw moment is calculated. This eliminates the influence of unknown parameters, making control commands more precise and stable, and reliably improving the vehicle's lateral stability under all operating conditions.

[0173] This embodiment also provides a vehicle control device for implementing the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0174] This embodiment provides a vehicle control device, such as... Figure 4 As shown, it includes: The first prediction module 401 is used to predict the first predicted centroid sideslip angle of the target vehicle at the current moment based on a first preset method. The second prediction module 402 is used to predict the second predicted centroid sideslip angle of the target vehicle at the current moment based on the second preset method. The fusion module 403 is used to fuse the first predicted centroid sideslip angle and the second predicted centroid sideslip angle to obtain the final predicted centroid sideslip angle corresponding to the target vehicle at the current moment. Calculation module 404 is used to calculate the target additional yaw moment of the target vehicle at the current moment based on the final predicted centroid sideslip angle; Control module 405 is used to control the target vehicle based on the target additional yaw moment.

[0175] In some optional implementations, the first prediction module 401 is specifically used to construct an initial state equation based on a two-degree-of-freedom vehicle dynamics model, using the sideslip angle and yaw rate of the target vehicle as state variables; wherein, the sideslip angle is the core variable to be estimated; using the lateral acceleration and yaw rate of the target vehicle as observations, an initial observation equation is constructed based on a two-degree-of-freedom vehicle dynamics model; and the first predicted sideslip angle is calculated based on the initial state equation and the initial observation equation.

[0176] In some optional implementations, the first prediction module 401 is specifically used to perform a first-order Taylor expansion of the initial state equation and the initial observation equation based on the forward Euler method to obtain the discretized state equation and the discretized observation equation; initialize the discretized state equation and the discretized observation equation; and predict the first predicted centroid sideslip angle based on the initialized discretized state equation and the discretized observation equation.

[0177] In some optional implementations, the first prediction module 401 is specifically used to predict the first initial centroid sideslip angle and the prior error covariance matrix corresponding to the current time based on the initialized discretized state equation and discretized observation equation; calculate the Kalman gain based on the prior error covariance matrix, the observation equation Jacobian matrix corresponding to the observation equation, and the noise covariance; and correct the first initial centroid sideslip angle based on the Kalman gain to obtain the first predicted centroid sideslip angle.

[0178] In some optional implementations, the second prediction module 402 is specifically used to obtain the kinematic equations corresponding to the target vehicle; derive the kinematic differential equations of the centroid sideslip angle corresponding to the target vehicle based on the kinematic equations; discretize the kinematic differential equations using the forward Euler method to obtain discrete kinematic equations; and calculate the second predicted centroid sideslip angle based on the discrete kinematic equations.

[0179] In some optional implementations, the fusion module 403 is specifically used to obtain the lateral acceleration corresponding to the target vehicle; determine the weights corresponding to the first predicted centroid sideslip angle and the second predicted centroid sideslip angle respectively based on the lateral acceleration; and fuse the first predicted centroid sideslip angle and the second predicted centroid sideslip angle based on the weights corresponding to the first predicted centroid sideslip angle and the second predicted centroid sideslip angle respectively to obtain the final predicted centroid sideslip angle corresponding to the target vehicle at the current moment.

[0180] In some optional implementations, the calculation module 404 is specifically used to obtain the target centroid sideslip angle and target yaw rate of the target vehicle at the current moment; obtain the actual yaw rate of the target vehicle at the current moment; calculate the difference between the predicted centroid sideslip angle and the target centroid sideslip angle, and the difference between the actual yaw rate and the target yaw rate; construct an initial sliding surface corresponding to a preset sliding mode control algorithm based on the difference between the centroid sideslip angle and the difference between the yaw rates; and calculate the target additional yaw moment of the target vehicle at the current moment based on the initial sliding surface.

[0181] In some optional implementations, the calculation module 404 is specifically used to calculate the initial exponential approach rate based on the initial sliding surface; obtain the preset sliding surface width corresponding to the initial sliding surface; correct the initial exponential approach rate based on the preset sliding surface width to obtain the target exponential approach rate; and calculate the target additional yaw moment corresponding to the target vehicle at the current moment based on the target exponential approach rate.

[0182] In some optional implementations, the calculation module 404 is specifically used to calculate the basic expression corresponding to the target additional yaw moment based on the target exponential approach rate; the basic expression includes unknown true parameters; the unknown true parameters characterize the tire lateral stiffness as a function of the distance from the center of mass to the axle; calculate the parameter adaptive law corresponding to the unknown true parameters; substitute the parameter adaptive law into the basic expression corresponding to the target additional yaw moment to calculate the target additional yaw moment.

[0183] In some optional implementations, the calculation module 404 is specifically used to obtain the parameter estimates corresponding to the unknown true parameters; calculate the parameter estimation error based on the unknown true parameters and the parameter estimates; construct the parameter adaptive law corresponding to the parameter estimation error; construct the Lyapunov function based on the initial sliding surface and the parameter estimation error; verify whether the parameter adaptive law meets the stability requirements of the target vehicle based on the Lyapunov function; if the parameter adaptive law meets the stability requirements of the target vehicle, then the parameter adaptive law corresponding to the unknown true parameters is obtained.

[0184] The vehicle control device provided in this embodiment of the invention can execute the vehicle control method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects for executing the method. Further functional descriptions of the various modules and units described above are the same as in the corresponding embodiments described above, and will not be repeated here.

[0185] Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention.

[0186] The following is a detailed reference. Figure 5 The diagram illustrates a structural schematic suitable for implementing an electronic device according to embodiments of the present invention. The electronic device may include a processor (e.g., a central processing unit, graphics processor, etc.) 01, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 02 or a program loaded from a memory 08 into a random access memory (RAM) 03. The RAM 03 also stores various programs and data required for the operation of the electronic device. The processor 01, ROM 02, and RAM 03 are interconnected via a bus 04. An input / output (I / O) interface 05 is also connected to the bus 04.

[0187] Typically, the following devices can be connected to I / O interface 05: input devices 06 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 07 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; memory devices 08 including, for example, magnetic tapes, hard disks, etc.; and communication devices 09. Communication device 09 allows electronic devices to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 5Electronic devices with various devices are shown, but it should be understood that it is not required to implement or have all of the devices shown, and more or fewer devices may be implemented or have instead.

[0188] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication device 09, or installed from memory 08, or installed from ROM 02. When the computer program is executed by processor 01, it performs the functions defined in the vehicle control method of the embodiments of the present invention.

[0189] Figure 5 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments of the present invention.

[0190] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code. When the software or computer code is accessed and executed by the computer, processor, or hardware, the vehicle control method shown in the above embodiments is implemented.

[0191] A portion of this invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the invention through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.

[0192] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A vehicle control method, characterized in that, The method includes: The first predicted centroid sideslip angle of the target vehicle at the current moment is predicted based on the first preset method; The second predicted centroid sideslip angle of the target vehicle at the current moment is predicted based on the second preset method; The first predicted centroid sideslip angle and the second predicted centroid sideslip angle are fused to obtain the final predicted centroid sideslip angle of the target vehicle at the current moment; Based on the final predicted centroid sideslip angle, calculate the target additional yaw moment corresponding to the target vehicle at the current moment; The target vehicle is controlled based on the target additional yaw moment.

2. The method according to claim 1, characterized in that, The first predicted centroid sideslip angle of the target vehicle at the current moment, predicted based on the first preset method, includes: Using the sideslip angle and yaw rate of the target vehicle as state variables, an initial state equation is constructed based on a two-degree-of-freedom vehicle dynamics model; wherein, the sideslip angle is the core variable to be estimated. Using the lateral acceleration and yaw rate of the target vehicle as observations, an initial observation equation is constructed based on a two-degree-of-freedom vehicle dynamics model; Based on the initial state equation and the initial observation equation, the first predicted centroid sideslip angle is calculated.

3. The method according to claim 2, characterized in that, The calculation of the first predicted centroid sideslip angle based on the initial state equation and the initial observation equation includes: Based on the forward Euler method, a first-order Taylor expansion is performed on the initial state equation and the initial observation equation to obtain the discretized state equation and the discretized observation equation: The discretized state equation and the discretized observation equation are initialized; Based on the initialized discretized state equation and discretized observation equation, the first predicted centroid sideslip angle is predicted.

4. The method according to claim 3, characterized in that, The prediction of the first predicted centroid sideslip angle based on the initialized discretized state equation and discretized observation equation includes: Based on the initialized discretized state equation and discretized observation equation, the first initial centroid side slip angle and prior error covariance matrix corresponding to the current time are predicted; The Kalman gain is calculated based on the prior error covariance matrix, the observation equation Jacobian matrix corresponding to the observation equation, and the noise covariance. Based on the Kalman gain, the first initial centroid sideslip angle is corrected to obtain the first predicted centroid sideslip angle.

5. The method according to claim 1, characterized in that, The prediction of the second predicted centroid sideslip angle of the target vehicle at the current moment based on the second preset method includes: Obtain the kinematic equations corresponding to the target vehicle; Based on the kinematic equations, the kinematic differential equation of the sideslip angle of the center of mass of the target vehicle is derived. The kinematic differential equations are discretized using the forward Euler method to obtain discrete kinematic equations; Based on the discrete kinematic equations, the second predicted centroid sideslip angle is calculated.

6. The method according to claim 1, characterized in that, The step of fusing the first predicted centroid sideslip angle and the second predicted centroid sideslip angle to obtain the final predicted centroid sideslip angle of the target vehicle at the current moment includes: Obtain the lateral acceleration corresponding to the target vehicle; Based on the lateral acceleration, determine the weights corresponding to the first predicted centroid sideslip angle and the second predicted centroid sideslip angle, respectively; Based on the weights corresponding to the first and second predicted centroid sideslip angles, the first and second predicted centroid sideslip angles are fused to obtain the final predicted centroid sideslip angle corresponding to the target vehicle at the current moment.

7. The method according to claim 1, characterized in that, The step of calculating the target additional yaw moment corresponding to the target vehicle at the current moment based on the final predicted centroid sideslip angle includes: Obtain the target centroid sideslip angle and target yaw rate of the target vehicle at the current moment; Obtain the actual yaw rate of the target vehicle at the current moment; Calculate the difference between the final predicted centroid sideslip angle and the target centroid sideslip angle, and the difference between the actual yaw rate and the target yaw rate; Based on the difference in the centroid side slip angle and the difference in the yaw rate, an initial sliding surface corresponding to the preset sliding mode control algorithm is constructed. Based on the initial sliding surface, calculate the initial exponential convergence rate; Obtain the preset sliding surface width corresponding to the initial sliding surface; Based on the preset sliding surface width, the initial exponential convergence rate is corrected to obtain the target exponential convergence rate; Based on the target exponential convergence rate, the target additional yaw moment corresponding to the target vehicle at the current moment is calculated.

8. The method according to claim 7, characterized in that, The step of calculating the target additional yaw moment corresponding to the target vehicle at the current moment based on the target exponential convergence rate includes: The basic expression corresponding to the target additional yaw moment is calculated based on the target exponential convergence rate; the basic expression contains unknown real parameters; the unknown real parameters characterize the tire lateral stiffness as a function of the distance from the center of mass to the axis; Calculate the parameter adaptive law corresponding to the unknown true parameters; The target additional yaw moment is calculated by substituting the parameter adaptive law into the basic expression corresponding to the target additional yaw moment.

9. The method according to claim 8, characterized in that, The calculation of the parameter adaptive law corresponding to the unknown true parameters includes: Obtain the parameter estimates corresponding to the unknown true parameters; Based on the unknown true parameters and the estimated parameter values, the parameter estimation error is calculated; Construct the parameter adaptive law corresponding to the parameter estimation error; Based on the initial sliding surface and the parameter estimation error, a Lyapunov function is constructed; Verify whether the parameter adaptive law satisfies the target vehicle stability requirements based on the Lyapunov function; If the parameter adaptive law satisfies the stability requirements of the target vehicle, then the parameter adaptive law corresponding to the unknown true parameters is obtained.

10. A vehicle control device, characterized in that, The device includes: The first prediction module is used to predict the first predicted centroid sideslip angle of the target vehicle at the current moment based on a first preset method. The second prediction module is used to predict the second predicted centroid sideslip angle of the target vehicle at the current moment based on the second preset method. The fusion module is used to fuse the first predicted centroid sideslip angle and the second predicted centroid sideslip angle to obtain the final predicted centroid sideslip angle of the target vehicle at the current moment. The calculation module is used to calculate the target additional yaw moment corresponding to the target vehicle at the current moment based on the final predicted centroid sideslip angle; The control module is used to control the target vehicle based on the target additional yaw moment.

Images