汽车主动悬架系统的神经网络控制算法  

Neural Network Control Algorithm of Automobile Active Suspension System

作　　者：黄剑鸣[1] 富丽娟[1] 

机构地区：[1]重庆工学院重庆汽车学院,重庆400050

出　　处：《重庆工学院学报（自然科学版）》2008年第3期100-103,120,共5页Journal of Chongqing Institute of Technology

摘　　要：通过对汽车主动悬架建立再生神经网络模型,提出了一种神经模糊适应性控制算法,该算法在多层神经网络的基础上,借助一套模糊规则,调节模糊神经控制器参数,通过神经网络建模确定了汽车主动悬架的动态参数,并向神经模糊适应性控制器提供学习信号.对一辆装有磁流变液减振器及基于DSP微处理器模糊神经控制系统的微车,在各种速度与路面条件下进行试验,将控制效果与开环被动悬架系统的进行了比较,结果表明,神经控制算法在减小微车振动方面表现出了良好的性能.By building regenerating neural network model for automobile active suspension system, this paper puts forward a neural fuzzy adaptive control algorithm, which employs a set of fuzzy rules to adjust fuzzy neural controller parameters based on multi-layer neural network, establishes neural network model to determine the dynamic parameters of automobile active suspension, and provides learning signals for neural luzzy adaptive controllers. A minicar, equipped with Magneto-rbeological Fluid Shock Absorbers and DSP- based micro-processor fuzzy neural controlling system, is tested at various speeds and on different road conditions. The controlling effects are compared with those of open-loop passive suspension system and the results show that neural control algorithm has excellent performance in reducing minicar＇s vibration.

关 键 词：神经模糊控制器 神经网络 汽车悬架 振动控制 

分 类 号：U463.33[机械工程—车辆工程]