基于RBF神经网络的ABS滑模变结构控制研究

Study on ABS sliding mode variable structure control based on RBF neural network

针对汽车制动过程中防抱死制动系统(ABS)具有的非线性、时变性和不确定性,设计了以最佳滑移率为目标的滑模变结构控制器,并且采用径向基神经网络(RBF)实时调整滑模变结构控制器参数,以削弱常规滑模变结构控制的抖振现象。利用MATLAB/Simulink仿真平台搭建单轮车辆制动模型,并进行ABS控制策略的仿真实验。仿真结果表明:在指定路面上制动时,基于RBF神经网络的滑模变结构控制策略能够有效削弱常规滑模变结构控制输出的高频抖振,并能使车辆具有良好的制动效果。

Aiming at the nonlinearity,time-varying and uncertainty of anti-lock braking system(ABS)in the process of automobile braking,a sliding mode variable structure controller with optimal slip rate is designed.The parameters of sliding mode variable structure controller are adjusted in real time using radial basis function(RBF)neural network to weaken the chattering phenomenon of conventional sliding mode variable structure control.The MATLAB/Simulink simulation platform was used to build a single-wheel vehicle braking model,and the simulation experiment of ABS control strategy was carried out.The simulation results show that the sliding mode variable structure control strategy based on RBF neural network can effectively weaken the high frequency buffeting of the output of conventional sliding mode variable structure control,and enable the vehicle a good braking effect when braking on specified road surface.