神经网络模糊PID控制半主动悬架系统

Research on Semi-active Suspension System Based on Neural Network Fuzzy PID Control

针对半主动悬架系统随路面激励变化产生的振动与冲击,设计一种基于BP(Back Propagation)神经网络模糊PID(Proportional Integral Derivative)控制的半主动悬架控制器,以提升悬架系统的控制精度,改善汽车平顺性。以悬架系统的偏差与偏差变化率为输入参数,利用模糊逻辑规则对输入参数进行模糊化和归一化处理,处理结果作为BP神经网络的输入,通过神经网络在线调整加权系数,实现PID控制参数的优化。为了验证控制效果,对比了BP神经网络模糊PID控制半主动悬架系统、模糊PID控制半主动悬架系统和被动悬架系统的性能,BP神经网络模糊PID控制策略的半主动悬架系统具有更好的控制效果,能够显著改善不同路况下车辆悬架的性能。

Aiming at the vibration and impact of semi-active suspension system with the change of road excitation,a controller based on BP(back propagation)neural network fuzzy PID(proportional integral derivative)control is designed and applied to semi-active suspension system to improve the control accuracy of suspension system and vehicle ride comfort.Taking the deviation and deviation change rate of the suspension system as the input parameters,the fuzzy logic rules are used to fuzzify and normalize the input parameters,and the results above as BP neural network input are used to adjust the weighting coefficient,so as to optimize the PID control parameters.In order to verify the control effect,the performances of semi-active suspension system controlled by BP neural network fuzzy PID and fuzzy PID and passive suspension are compared.The results show that the semi-active suspension system controlled by BP neural network fuzzy PID has better optimal control effect,which can significantly improve the performance of vehicle suspension under different road conditions.