燕山大学学报2017，Vol.41Issue(1)：27-31,5.DOI:10.3969/j.issn.1007-791X.2017.01.004
递归对角神经网络算法在汽车主动悬架控制系统中的研究
Recurrent diagonal neural network algorithm study on vehicle active suspension control system
摘要
Abstract
A 7-DOF (degree of freedom) vehicle suspension model is established based on the vehicle system dynamics theory in or-der to improve the control effect of vehicle active suspension.The overall scheme design of active suspension control system is car-ried out based on solenoid valve shock absorber technical scheme. The active suspension controller is constructed based on the DRNN ( Diagonal Recurrent Neural Network) algorithm and a neural network is trained using genetic algorithm. The method has a self-study, internal feedback function.The results of real time Hardware-in-the-Loop simulation of Simulink and dSPACE show that the algorithm can obviously improve the control effect of vehicle active suspension by the simulation analysis of vehicle body vertical acceleration, tire dynamic displacement and the vibration analysis of vehicle seat under the effect of intermittent bumpy road excita-tion.DRNN algorithm is better to improve the riding comfort and handling stability.关键词
递归对角神经网络/主动悬架/遗传算法/电磁阀式减振器Key words
diagonal recurrent neural network/active suspension/genetic algorithm/vibration damper of electromagnetic valve分类
信息技术与安全科学引用本文复制引用
吕科,杨正才,赵宝..递归对角神经网络算法在汽车主动悬架控制系统中的研究[J].燕山大学学报,2017,41(1):27-31,5.基金项目
2016十堰市科学技术研究与开发项目(16K46) （16K46）
汽车动力传动与电子控制湖北省重点实验室(湖北汽车工业学院)基金项目(ZDK1201303) （湖北汽车工业学院）
