A Closed-Loop Vehicle Crosswind Stability Dynamic Evaluation System Based on LQR Control Algorithm

Abstract

This invention relates to the field of intelligent driving testing technology and discloses a closed-loop vehicle crosswind stability dynamic evaluation system based on the LQR control algorithm. The system includes a CFD model unit, an active control feedback unit, and an MBD model unit. The CFD model unit acquires the vehicle's motion parameters and outputs aerodynamic load parameters as input to the MBD model unit. The active control feedback unit simulates active control behavior of the vehicle, including acquiring the vehicle's front wheel steering angle parameters through the LQR module and using them as input to the MBD model unit. The MBD model unit uses the front wheel steering angle parameters as control variables to update the aerodynamic load parameters, obtains dynamic response parameters based on the aerodynamic load parameters, and iterates the vehicle's motion parameters based on the dynamic response parameters. This invention effectively ensures that the model does not experience near-uncontrolled lateral movement during solution processing, realistically reproducing actual driving conditions and improving the accuracy and reliability of the evaluation.

Description

Technical Field

[0001] This invention relates to the field of intelligent driving testing technology, specifically to a closed-loop vehicle crosswind stability dynamic evaluation system based on the LQR control algorithm. Background Technology

[0002] Crosswind stability of a vehicle refers to its ability to maintain stable and safe driving when subjected to lateral airflow disturbances while traveling at high speeds, traversing bridges, entering or exiting tunnels, or overtaking other vehicles. This performance directly affects vehicle driving safety, especially in strong crosswind conditions, where vehicles with poor crosswind stability are prone to instability phenomena such as yaw, skidding, lane departure, and even rollover. Therefore, crosswind stability simulation testing of automobiles has become an important means for the automotive industry to improve vehicle safety performance.

[0003] However, current crosswind stability simulation tests, despite the use of advanced computer simulation software and models, still suffer from low simulation fidelity and inaccurate test results. Firstly, the airflow around a vehicle becomes extremely complex under crosswind conditions, especially under high wind speeds and rapidly changing gusts, where this complexity is further amplified. Traditional simulation methods, such as the Lattice Boltzmann Method (LBM), can simulate this flow to some extent, but their applicable velocity discretization models and boundary condition methods are still very limited, making it difficult to fully reproduce the complex airflow changes in real-world scenarios.

[0004] Secondly, existing simulation tests often overlook the two-way coupling relationship between vehicle attitude and crosswind aerodynamic characteristics. In reality, changes in vehicle attitude can, in turn, affect crosswind aerodynamic characteristics, a point that is not fully reflected in current simulation models. Therefore, simulation results often fail to accurately reflect the dynamic response of vehicles in real crosswind environments, making it difficult for test results to achieve the expected accuracy and reliability. Summary of the Invention

[0005] The present invention aims to provide a closed-loop vehicle crosswind stability dynamic evaluation system based on the LQR control algorithm, in order to solve the problem that the crosswind stability simulation test of vehicles in the prior art has a low degree of reproduction of the actual situation, resulting in insufficient accuracy and reliability of the test results.

[0006] To achieve the above objectives, this invention adopts the following technical solution: a closed-loop vehicle crosswind stability dynamic evaluation system based on the LQR control algorithm. This solution considers the influence of lateral active control behavior in actual operating conditions, uses the LQR algorithm for lateral control, simulates the feedback behavior of actual driving to lateral deviation, realistically reflects the vehicle's motion in crosswind environments, improves the simulation fidelity, and enhances the accuracy of test results. The evaluation system specifically includes a CFD model unit, an active control feedback unit, and an MBD model unit; the CFD model unit is used to acquire the vehicle's motion parameters and outputs aerodynamic load parameters as input to the MBD model unit based on these motion parameters.

[0007] The active control feedback unit is used to simulate the active control behavior of the vehicle, including obtaining the front wheel steering angle parameters of the vehicle through the LQR module based on the vehicle's motion parameters, and using them as input to the MBD model unit;

[0008] The MBD model unit is used to use the front wheel steering angle parameter as a control variable and update the aerodynamic load parameter. Based on the aerodynamic load parameter, the dynamic response parameter is obtained. Based on the dynamic response parameter, the vehicle's motion parameter is iterated to complete the closed-loop evaluation of the vehicle's crosswind stability.

[0009] The principles and advantages of this scheme are:

[0010] The study of crosswind stability of automobiles involves multiple disciplines, including automotive aerodynamics and multibody dynamics. To systematically evaluate a vehicle's crosswind stability, it is necessary to consider the coupled effects of aerodynamic disturbances, vehicle path, and changes in vehicle attitude, simulating the dynamic relationship between the six components of aerodynamic forces and vehicle attitude—that is, dynamic evaluation of vehicle crosswind stability. This evaluation method allows for relatively realistic acquisition of flow field information and motion of the vehicle in crosswind environments using numerical simulation methods. For example, invention application CN117829029A discloses a method for testing vehicle crosswind stability using models based on automotive aerodynamics and multibody dynamics.

[0011] However, in strong crosswinds, simulated vehicles often exhibit significant lateral drift due to the large and prolonged lateral velocity calculated, sometimes even veering out of the computational domain, resulting in a state that deviates significantly from reality and leading to inaccurate test results. Analysis revealed that this is because existing technologies neglect a crucial element: the driver's corrective actions and the role of driver assistance systems such as ESP during driving. Even when numerical simulation methods utilize the coupling of vehicle aerodynamics and multibody dynamics to recreate the motion of a vehicle in crosswind conditions without corrective actions, they overlook the active directional control required by real-world drivers when drifting occurs. This limitation results in a low degree of realism in existing evaluation methods, impacting the reliability of the final test results.

[0012] Meanwhile, even though existing technologies take into account the coupling effect of automotive aerodynamics and automotive multibody dynamics, the lack of consideration for active control factors during actual driving and their mutual influence leads to large errors in the established dynamic models and significant deviations in the constructed coupling relationships, resulting in inaccurate evaluation results.

[0013] Therefore, this solution aims to realistically reproduce the motion of a vehicle under crosswind conditions and reflect the impact of active control behavior on crosswind stability when lateral deviation occurs. Considering the feedback effect of active control behavior on crosswind disturbances, a closed-loop dynamic evaluation system for vehicle crosswind stability based on the LQR control algorithm is proposed. Based on the bidirectional dynamic coupling of vehicle aerodynamics and vehicle multibody dynamics (CFD-MBD), a whole-vehicle CFD model and an MBD model are established for coupled calculations. The front wheel steering angle parameters are calculated using the LQR algorithm based on the heading and lateral errors at the current time step, and are applied to the dynamic solution along with the aerodynamic six-component forces. Then, the vehicle attitude parameters of the dynamic response are transferred to the aerodynamic solution for the next time step iteration. The LQR algorithm is calculated synchronously with the bidirectional data exchange, and the state variables of each solution are fed back, thereby achieving a closed-loop evaluation of vehicle crosswind stability.

[0014] Implementing this plan will have the following beneficial effects:

[0015] Compared with existing processing methods, this method is more consistent with the actual situation. In the absence of crosswind stability test conditions, it can realistically reproduce the driving state of a car in complex wind environments. It can effectively ensure that the model does not experience near-uncontrolled lateral movement during the solution process, realistically reproduce actual driving conditions, improve the degree of reproduction, and enhance the accuracy and reliability of the evaluation results.

[0016] Furthermore, the active control feedback unit includes a vehicle attitude parameter module, an LQR calculation module, a structural parameter module, and a path module; the vehicle attitude parameter module is used to obtain the motion parameters of the vehicle at the current time step; the structural parameter module is used to obtain the vehicle's body structural parameters; the path module is used to obtain the vehicle's planned path information; and the LQR calculation module is used to calculate the front wheel steering angle parameters at the current time step based on the motion parameters and structural parameters.

[0017] Furthermore, the CFD model unit includes an update module, an output module, and a decision module; the update module includes a boundary condition submodule and a moving mesh submodule; the boundary condition submodule is used to acquire motion parameters and update the model using the motion parameters as boundary conditions; the moving mesh submodule is used to update the moving mesh model according to the boundary conditions and obtain aerodynamic load parameters through the moving mesh model; the output module is used to output the solved aerodynamic load parameters; and the decision module is used to determine whether the simulation has ended.

[0018] Furthermore, the MBD model unit includes a control quantity module, a motion state module, a response module, and a judgment module; the control quantity module is used to obtain control quantities and update aerodynamic load parameters; the motion state module is used to update the motion state by solving the DAE equation based on the aerodynamic load parameters; the response module is used to obtain dynamic response parameters based on the motion state and output them to the active control feedback unit; the judgment module is used to determine whether the simulation has ended.

[0019] Furthermore, the active control feedback unit also includes an error module and a feedforward interference module; the error module is used to obtain the error parameters between the actual path and the planned path of the vehicle; the feedforward interference module is used to generate interference based on the error parameters, which is used as the feedforward control input in the LQR calculation module for calculation.

[0020] Furthermore, a state equation for the front wheel steering angle is constructed using the LQR calculation module to control the error variation of the front wheel steering angle. The state equation for the front wheel steering angle is expressed using the vehicle dynamics equation in the natural coordinate system as follows:

[0021]

[0022] In the formula, e d , These represent the lateral error and heading error between the planned path and the actual path in the natural coordinate system, respectively. The first derivative represents the lateral error. The first derivative representing the lateral heading error. The second derivative representing the lateral error. δ represents the second derivative of the lateral heading error, m represents the total mass of the vehicle, u represents the longitudinal speed at the center of gravity, δ represents the wheel angle, I represents the moment of inertia, a and b represent the front and rear wheelbases, and K1 and K2 represent the lateral stiffness of the front and rear tires.

[0023] Furthermore, a cost function for reducing errors is constructed based on the state equation of the front wheel steering angle:

[0024]

[0025] In the formula, the error u = δ, Q and R are the weight matrices of the LQR algorithm.

[0026] Furthermore, the interference amount is expressed as:

[0027]

[0028] In the formula, k r The curvature of the path at this point is represented by k3, which is the third element of the K vector calculated by the LQR algorithm, and L represents the wheelbase.

[0029] Furthermore, by constructing the vehicle dynamics equations in the vehicle body coordinate system, the vehicle dynamics equations in the natural coordinate system are obtained. The vehicle dynamics equations in the vehicle body coordinate system are:

[0030]

[0031] In the formula, Indicates lateral acceleration. Indicates yaw acceleration. This represents the longitudinal vehicle speed in the vehicle body coordinate system. This indicates the yaw rate of the car.

[0032] Furthermore, the relationship between the motion parameters in the vehicle body coordinate system and the motion parameters in the natural coordinate system is as follows:

[0033]

[0034] In the formula, This represents the first derivative of the actual steering angle of the front wheel when considering sideslip. Attached Figure Description

[0035] Figure 1 This is a schematic diagram of the structure of a closed-loop vehicle crosswind stability dynamic evaluation system based on the LQR control algorithm of the present invention.

[0036] Figure 2 This is a schematic diagram of the calculation process of a closed-loop vehicle crosswind stability dynamic evaluation system based on the LQR control algorithm of the present invention.

[0037] Figure 3 This is a flowchart of the feedforward control process of a closed-loop vehicle crosswind stability dynamic evaluation system based on the LQR control algorithm of the present invention. Detailed Implementation

[0038] The following detailed description illustrates the specific implementation method:

[0039] Example

[0040] As attached Figure 1As shown, this solution addresses the problems of large lateral displacements and the inability to fully consider driver control and the role of electronic vehicle control systems in dynamic evaluation methods for vehicle crosswind stability during simulation. It considers the feedback effect of active control behavior on crosswind disturbances and uses the LQR algorithm for lateral control to simulate the feedback behavior of actual driving in response to lateral deviation, realistically reflecting the vehicle's motion in crosswind environments. Under this system, the large lateral deviations often observed in numerical simulations are suppressed, and the driver's active correction behavior and the role of electronic assistance systems are also reproduced, improving the accuracy of numerical simulations in reproducing actual conditions. In this embodiment, the closed-loop vehicle crosswind stability dynamic evaluation system mainly includes a CFD model unit, an active control feedback unit, and a MBD model unit.

[0041] The CFD model unit is a partial model of the vehicle's aerodynamics, used to acquire the vehicle's motion parameters and output aerodynamic load parameters as input to the MBD model unit based on these motion parameters. In this embodiment, the aerodynamic load parameters are the six aerodynamic force parameters.

[0042] The active control feedback unit is used to simulate the active control behavior of the vehicle, including obtaining the front wheel steering angle parameters of the vehicle through the LQR module based on the vehicle's motion parameters, and using them as input to the MBD model unit.

[0043] The MBD model unit is a partial model of vehicle aerodynamics. It is used to use the front wheel steering angle parameter as a control variable and update the aerodynamic load parameters, that is, update the aerodynamic six-component force parameters. Based on the aerodynamic load parameters, the dynamic response parameters are obtained. Based on the dynamic response parameters, the vehicle's motion parameters are iterated to complete the closed-loop evaluation of vehicle crosswind stability.

[0044] Combined with appendix Figure 2 As shown, attached Figure 2 This diagram illustrates the computational process of a closed-loop evaluation system based on the LQR control algorithm, built upon a dynamic coupling platform of automotive aerodynamics and multibody dynamics. The aerodynamic component is calculated using the CFD solver Star CCM+, while the multibody dynamics component undergoes motion simulation using the MBD solver Carsim. Real-time computational data for the fluid and multibody systems are acquired and transmitted by scanning a shared database; this part is implemented using Matlab / Simulink software.

[0045] In this embodiment, before coupling, Star CCM+ and Carsim need to solve separately for a period of time. Data scanning and transmission only occur after the flow field and vehicle motion reach a stable state. During this process, the LQR algorithm is calculated along with the data loading process, but it does not produce any actual control effect.

[0046] Specifically, the CFD model unit includes an update module, an output module, and a decision module. The update module comprises a boundary condition submodule and a dynamic mesh submodule. Upon initial coupling, the boundary condition submodule scans and reads motion parameters from a specified shared database using a UDF program. Combined with dynamic mesh modeling technology, it solves for and updates the boundary parameters using the motion parameters of the current time step as boundary conditions. The dynamic mesh submodule updates the dynamic mesh model based on the boundary conditions and obtains aerodynamic load parameters by solving the Navier-Stokes (NS) equations. In Cartesian coordinates, the vector form of the NS equations can be expressed as... Where ρ is the fluid density, p is the pressure, u(u,v,w) is the velocity vector, and F(X,Y,Z) is the body force acting on a unit mass of fluid. The Hamiltonian operator is used, and Δ is used as the Laplace operator; these are existing technologies and will not be elaborated upon here. The output module outputs the solved aerodynamic load parameters, i.e., the aerodynamic six-component force parameters. The decision module determines whether the simulation has ended. If the simulation has ended, it terminates the simulation calculation of the CFD model elements; if it has not ended, it returns to the update module to continue iterating until the simulation results meet the requirements.

[0047] The active control feedback unit includes a vehicle attitude parameter module, an LQR calculation module, a structural parameter module, and a path module. The vehicle attitude parameter module acquires the vehicle's motion parameters for the current time step, serving as real-time data in a shared database. The structural parameter module acquires the vehicle's body structure parameters; in this embodiment, these parameters include front and rear wheel slip angles, vehicle center of gravity position, total vehicle mass, and front and rear wheelbase. The path module acquires the vehicle's planned path information and the actual driving path information for the current time step. The LQR calculation module calculates the front wheel steering angle parameters for the current time step based on the motion and structural parameters.

[0048] The active control feedback unit also includes an error module and a feedforward disturbance module. The error module is used to obtain the error parameters between the actual path and the planned path of the vehicle. The feedforward disturbance module is used to generate disturbance quantities based on the error parameters, which are then used as feedforward control inputs in the LQR calculation module.

[0049] In this embodiment, the LQR control algorithm in the LQR calculation module is based on a linear two-degree-of-freedom vehicle model. However, since this scheme considers the active control behavior during actual driving, assumptions and simplifications have been made regarding vehicle operation, retaining only the lateral and yaw degrees of freedom. Furthermore, the tire lateral stiffness is within the linear range. According to Newton's second law, the linear two-degree-of-freedom vehicle model satisfies:

[0050]

[0051] In the formula, F Y FY1 F Y2 The lateral force on the entire vehicle and its front and rear wheels is represented by δ, the wheel rotation angle is represented by m, and the total mass is represented by a. y M represents lateral acceleration. Z The yaw moment is represented by 'a' and 'b', the front and rear wheelbases are represented by 'a' and 'b', and the moment of inertia is represented by 'I'. Represents angular acceleration.

[0052] In this embodiment, to better reflect the research object with two degrees of freedom of lateral and yaw, equation (1) is transformed to obtain the following formula:

[0053]

[0054] In the formula, K1 and K2 represent the lateral stiffness of the front and rear tires, and α1 and α2 represent the lateral slip angles of the front and rear wheels, respectively. Under the assumption of a small lateral slip angle in the bicycle model, the lateral slip angles of the front and rear wheels are generally no greater than 4° to 5°.

[0055] Due to the limitations of the bicycle model's degrees of freedom, in this embodiment, the control object of the LQR algorithm can only be the wheel angle. The state variables on which the control is based are parameters that reflect the longitudinal and lateral motion. The state equation of the control system needs to be obtained by transforming the vehicle dynamics equation in the geodetic coordinate system using a linear two-degree-of-freedom model, i.e., equation (2). Meanwhile, in this embodiment, the control approach is to minimize the error between the actual path and the planned path. Therefore, for ease of calculation, in this embodiment, the vehicle dynamics equation in the geodetic coordinate system is obtained by transforming the vehicle dynamics equation in the natural coordinate system. This is because the natural coordinate system avoids the deviation between the longitudinal velocity direction and the horizontal axis of the vehicle coordinate system caused by the centroid side slip angle, facilitating the decoupling of longitudinal and lateral motion and simplifying the analysis process. In this embodiment, for the front and rear wheel side slip angles and lateral acceleration a... y and yaw acceleration The parameters in the vehicle body coordinate system can be represented as follows:

[0056]

[0057] In the formula, The longitudinal vehicle speed is represented in the vehicle body coordinate system, where u and υ represent the longitudinal and lateral vehicle speeds at the center of mass, respectively. This represents the yaw rate of the vehicle. Substituting the parameter relationships in equation (3) into equation (2), we can obtain the vehicle dynamics equation in the vehicle coordinate system:

[0058]

[0059] In the formula, Indicates lateral acceleration. Indicates yaw acceleration. This represents the longitudinal vehicle speed in the vehicle body coordinate system. This indicates the yaw rate of the car.

[0060] The motion parameters in the vehicle body coordinate system and the motion parameters in the natural coordinate system have the following relationship:

[0061]

[0062] In the formula, e d , These represent the lateral error and heading error between the planned path and the actual path in the natural coordinate system, respectively. The first derivative represents the lateral error. The first derivative representing the lateral heading error. The second derivative representing the lateral error. The second derivative θ represents the lateral heading error. r Indicates the heading angle. It represents the first derivative of the heading angle.

[0063] Substituting the parameter relationships in equation (5) into equation (4), we can obtain the vehicle dynamics equations in the natural coordinate system:

[0064]

[0065] That is, the state equation (6) of the controller is obtained, which is also the constraint condition of the LQR algorithm. The state variables are the lateral and heading errors in the natural coordinate system and their first derivatives, and the control variable is the front wheel angle. The state equation of the front wheel angle is constructed through the LQR calculation module to control the error change of the front wheel angle, so as to minimize the change of the control variable to ensure comfort. rr Minimize control objectives.

[0066] Therefore, based on the state equation of the front wheel steering angle, the error parameters between the actual path and the planned path of the vehicle are obtained through the error module, and a cost function for reducing the error is constructed:

[0067]

[0068] In the formula, the error u = δ, Q and R are the weight matrices of the LQR algorithm. Therefore, the following can be obtained: Figure 3 The control block diagram under the LQR algorithm shown shows that outside the system control block diagram, the feedforward disturbance module generates disturbance based on the changes in error parameters and structural parameters, forming feedforward control that participates in the front wheel steering angle calculation of the LQR calculation module, preventing the occurrence of "algebraic loop" problems, thereby realizing the simulation of active control behavior.

[0069] In this embodiment, the interference amount δ fRepresented as:

[0070]

[0071] In the formula, k r The curvature of the path at this point is represented by k3, which is the third element of the K vector calculated by the LQR algorithm, and L represents the wheelbase. The front wheel steering angle parameters are then obtained and used as input to the MBD model unit.

[0072] The MBD model unit includes a control quantity module, a motion state module, a response module, and a judgment module. The control quantity module is used to obtain control quantities and update aerodynamic load parameters. In this embodiment, it is scanned and read by Matlab / Simulink, and then the front wheel steering angle parameters calculated based on the current state are sent to Carsim through the interface program.

[0073] The motion state module is used to solve the DAE equations based on aerodynamic load parameters and update the vehicle's motion state, writing the data into the shared database. The response module obtains dynamic response parameters based on the motion state and outputs them to the vehicle attitude parameter module in the active control feedback unit to update the motion parameters. Finally, by outputting time-series aerodynamic six-component forces and vehicle motion parameters, a closed-loop dynamic evaluation of the vehicle's crosswind aerodynamic characteristics and stability is achieved. The judgment module determines whether the simulation has ended. If the simulation has ended, the simulation calculation of the MBD model unit is terminated; if it has not ended, it returns to the update module to continue iterating until the simulation results meet the requirements.

[0074] In this embodiment, a closed-loop system for evaluating vehicle crosswind stability is proposed based on the CFD-MBD dynamic coupling method. Currently, in the study of vehicle crosswind aerodynamic characteristics or crosswind stability, the dynamic coupling method has solved the problem of the separation between aerodynamics and multibody dynamics caused by unidirectional coupling, and has realized a highly realistic evaluation of vehicle crosswind stability without a feedback mechanism. However, the simple bidirectional coupling method often shows a large lateral deviation in numerical simulations that does not conform to reality because it does not consider the active correction behavior that exists during actual driving, which greatly restricts the evaluation effect of the bidirectional coupling mechanism.

[0075] This solution proposes a closed-loop evaluation method that can simulate the active correction behavior of the driver or electronic assistance system under crosswind conditions. It effectively ensures that the model does not experience near-uncontrolled lateral movement during the solution process, realistically reproducing actual driving conditions. Compared with existing methods, it better matches real-world scenarios and, in the absence of crosswind stability testing conditions, accurately reproduces the vehicle's driving state in complex wind environments, quickly and effectively guiding the positive development of vehicle aerodynamics and chassis.

[0076] The above descriptions are merely embodiments of the present invention, and common knowledge such as specific technical solutions and / or characteristics are not described in detail here. It should be noted that those skilled in the art can make various modifications and improvements without departing from the technical solutions of the present invention, and these should also be considered within the scope of protection of the present invention. These modifications and improvements will not affect the effectiveness of the implementation of the present invention or the practicality of the patent. The scope of protection claimed in this application should be determined by the content of its claims, and the specific embodiments described in the specification can be used to interpret the content of the claims.

Claims

1. A closed-loop vehicle crosswind stability dynamic evaluation system based on LQR control algorithm, characterized in that: It includes a CFD model unit, an active control feedback unit, and an MBD model unit; the CFD model unit is used to acquire the vehicle's motion parameters and output aerodynamic load parameters as input to the MBD model unit based on the motion parameters; The active control feedback unit is used to simulate the active control behavior of the vehicle, including obtaining the front wheel steering angle parameters of the vehicle through the LQR module based on the vehicle's motion parameters, and using them as input to the MBD model unit; The MBD model unit is used to use the front wheel steering angle parameter as a control variable and update the aerodynamic load parameter. Based on the aerodynamic load parameter, the dynamic response parameter is obtained. Based on the dynamic response parameter, the vehicle's motion parameter is iterated to complete the closed-loop evaluation of the vehicle's crosswind stability.

2. The closed-loop vehicle crosswind stability dynamic evaluation system based on LQR control algorithm according to claim 1, characterized in that: The active control feedback unit includes a vehicle attitude parameter module, an LQR calculation module, a structural parameter module, and a path module. The vehicle attitude parameter module is used to obtain the motion parameters of the vehicle at the current time step. The structural parameter module is used to obtain the vehicle's body structural parameters. The path module is used to obtain the vehicle's planned path information. The LQR calculation module is used to calculate the front wheel steering angle parameters at the current time step based on the motion parameters and structural parameters.

3. The closed-loop vehicle crosswind stability dynamic evaluation system based on LQR control algorithm according to claim 1, characterized in that: The CFD model unit includes an update module, an output module, and a decision module; the update module includes a boundary condition submodule and a moving mesh submodule; the boundary condition submodule is used to obtain motion parameters and update the model using the motion parameters as boundary conditions; the moving mesh submodule is used to update the moving mesh model according to the boundary conditions and obtain aerodynamic load parameters through the moving mesh model; the output module is used to output the solved aerodynamic load parameters; the decision module is used to determine whether the simulation has ended.

4. A closed-loop vehicle crosswind stability dynamic evaluation system based on LQR control algorithm according to claim 1, characterized in that: The MBD model unit includes a control quantity module, a motion state module, a response module, and a judgment module. The control quantity module is used to obtain control quantities and update aerodynamic load parameters. The motion state module is used to update the motion state by solving the DAE equation based on the aerodynamic load parameters. The response module is used to obtain dynamic response parameters based on the motion state and output them to the active control feedback unit. The judgment module is used to determine whether the simulation has ended.

5. A closed-loop vehicle crosswind stability dynamic evaluation system based on LQR control algorithm according to claim 2, characterized in that: The active control feedback unit further includes an error module and a feedforward interference module; the error module is used to obtain the error parameters between the actual path and the planned path of the vehicle; the feedforward interference module is used to generate interference based on the error parameters and structural parameters, which is used as the feedforward control input in the LQR calculation module for calculation.

6. A closed-loop vehicle crosswind stability dynamic evaluation system based on LQR control algorithm according to claim 5, characterized in that, The state equation for the front wheel steering angle is constructed using the LQR calculation module to control the error variation of the front wheel steering angle. The state equation for the front wheel steering angle is expressed using the vehicle dynamics equation in the natural coordinate system as follows: In the formula, e d , These represent the lateral error and heading error between the planned path and the actual path in the natural coordinate system, respectively. The first derivative represents the lateral error. The first derivative representing the lateral heading error, The second derivative representing the lateral error. δ represents the second derivative of the lateral heading error, m represents the total mass of the vehicle, u represents the longitudinal speed at the center of gravity, δ represents the wheel angle, I represents the moment of inertia, a and b represent the front and rear wheelbases, and K1 and K2 represent the lateral stiffness of the front and rear tires.

7. A closed-loop vehicle crosswind stability dynamic evaluation system based on LQR control algorithm according to claim 6, characterized in that, A cost function for reducing errors is constructed based on the state equation of the front wheel steering angle: In the formula, the error u = δ, Q and R are the weight matrices of the LQR algorithm.

8. A closed-loop vehicle crosswind stability dynamic evaluation system based on LQR control algorithm according to claim 7, characterized in that, The interference amount is expressed as: In the formula, k r The curvature of the path at this point is represented by k3, which is the third element of the K vector calculated by the LQR algorithm, and L represents the wheelbase.

9. A closed-loop vehicle crosswind stability dynamic evaluation system based on LQR control algorithm according to claim 6, characterized in that: The vehicle dynamics equations in the natural coordinate system are obtained by constructing the vehicle dynamics equations in the vehicle body coordinate system. The vehicle dynamics equations in the vehicle body coordinate system are as follows: In the formula, Indicates lateral acceleration. Indicates yaw acceleration. This represents the longitudinal vehicle speed in the vehicle body coordinate system. This indicates the yaw rate of the car.

10. A closed-loop vehicle crosswind stability dynamic evaluation system based on LQR control algorithm according to claim 9, characterized in that: The relationship between the motion parameters in the vehicle body coordinate system and the motion parameters in the natural coordinate system is as follows: In the formula, This represents the first derivative of the actual steering angle of the front wheel when considering sideslip.

Images