Research On Adaptive Neural Network H_∞ Servo Control Method Of Electronic Brake Booster

Electronic Brake Booster(EBB)is a new type of executive structure of Brake-By-Wire system applied to new energy vehicles.It is one of the important components of Electro-Hydraulic Brake system(EHB),and plays an increasingly important role in brake energy recovery,advanced driving assistance and other functions.In this paper,the Adaptive Neural Network H_∞(ANN-H_∞)control strategy design of EBB system is taken as the theme,and a series of researches are carried out in combination with the proposal of EBB structure scheme and comparative simulation analysis.The main contents of this paper are as follows:(1)This paper analyzes the research status of EBB’s structural scheme and control strategy at home and abroad.Aiming at the mechanical structure of EBB,the self-designed EBB structure scheme is proposed,which has the characteristics of simple structure,light weight and simple implementation of control strategy.The working principle and mathematical model of the main components of EBB system are introduced in detail.Considering the factors of"Coulomb+viscous"friction,a simulation model of EBB system is established on MATLAB/Simulink.(2)A fuzzy PID Three Loop control strategy for servo drive of EBB system is proposed.A fuzzy controller is designed on the position loop to adjust the parameters online,so as to improve the anti-interference ability of the control strategy.In view of the difficulty of establishing fuzzy rules base,Particle Swarm Optimization(PSO)algorithm is applied to optimize the rules.Based on the control model of EBB system,the simulation analysis of emergency braking and general braking is carried out.(3)In view of the situation that the fuzzy PID Three Loop control strategy will produce the current fluctuation and the moment impact under the emergency braking condition,an Adaptive Neural Network H_∞control strategy based on the H_∞control is proposed to approach the nonlinear part and the uncertain factors of EBB system through the Radial Neural Network.In this paper,the stability and convergence of ANN-H_∞control strategy are proved in the sense of the Lyapunov Theory.Under the same condition as Three Loop control strategy,the ANN-H_∞control strategy is modeled and simulated.The results show that the ANN-H_∞controller is superior to the Three Loop control strategy in the aspects of rapidity,static error and stability.The simulation results of robustness verification show that the ANN-H_∞control strategy can still complete the EBB servo action in the absence of C-phase current,which significantly improves the anti-interference ability of the controller.(4)TMS320F28335 chip is used as the controller to realize the data interaction between the host computer and the controller through the serial port.Combined with the MATLAB/Simulink environment to realize the automatic code generation and download,a semi physical simulation platform is built.The ANN-H_∞control is verified by the Processor In the Loop test.The test results show that the ANN-H_∞control satisfies the numerical equivalence with the Model In the Loop,and the execution time of the control algorithm in the processor is less than 0.1 ms,which verifies the feasibility of the control algorithm at the processor level.