| The active suspension can generate active control force to improve the suspension characteristics according to different driving conditions,and effectively balance the steering stability and smoothness of the vehicle.However,due to factors such as timevarying parameters,external disturbances,and nonlinearities,the suspension system can not easily meet the robustness requirements of the vehicle when it uses conventional control algorithms.Aiming at the above problems of the active suspension system,this paper designs a sliding mode variable structure controller based on the sliding mode variable structure theory to improve the ride comfort of the vehicle.Aiming at the chattering problem of sliding mode controller,this paper uses the radial basis neural network algorithm to optimize the sliding mode switching switch to achieve the effect of suppressing chattering.In addition,according to the real-time and accurate vehicle vibration state requirements of the control algorithm,the vehicle vibration state observer is designed by Kalman filter algorithm.In this paper,the performance of the proposed controller and algorithm is verified by simulating the designed model.The following is the specific research content and conclusions of this article:(1)The single-wheel road input model is established by the filter white noise method.The coherence of the left and right wheels and the time lag of the front and rear wheels are analyzed.The four-wheel road surface excitation time domain model is established.The seven-degree-of-freedom active suspension dynamics model,the vehicle canopy damping reference model,and the error model based on the two are established to lay the foundation for the subsequent sliding mode controller design.(2)A sliding mode variable controller based on radial basis neural network is designed.The sliding mode controller uses the state variables in the ideal canopy damping suspension model as the tracking target,and controls the active suspension model to improve the ride comfort.By using the radial basis neural network algorithm,the function of fast approximation speed and strong convergence ability is optimized.The sliding mode switching switch is optimized to suppress the buffeting effect of the sliding mode controller,and the robustness of the controller is further improved.(3)The vehicle vibration state observer is designed by Kalman filter algorithm.The simulation proves that the designed state observer can provide real-time and accurate state information for the controller by better estimating the state of the suspension.(4)Simulation modeling and analysis of the established active suspension system controller.The analysis results show that the sliding mode controller based on state observation can improve the ride comfort of the vehicle significantly compared with the passive suspension.The sliding mode controller optimized by the radial basis neural network algorithm has a small improvement in control effect. |
