基于模糊神经网络的列车垂向悬挂控制器研究

Research on Vertical Train Suspension Controller Based on Fuzzy Neural Network

目前列车悬挂控制器参数大多基于专家经验知识提出,人为主观因素强,悬挂系统不能很好地根据实际情况自适应调整。为抑制列车振动提高其运行平稳性,结合模糊神经网络的自学习能力和自整定功能,设计基于Takagi-Sugeno模型的模糊神经网络控制器,将控制器模型与悬挂模型结合实现悬挂系统自适应调整。以德国轨道谱为轮对外部激励,仿真生成加速度时域曲线与功率谱密度曲线,分析结果表明:相比于被动悬挂系统与模糊天棚控制系统,运用模糊神经网络控制可明显抑制车体垂向振动,提高运行平稳性,对控制器设计具有参考价值。

At present,most of the parameters of train suspension controller are based on expert experience and knowledge.Because of the strong subjective factors,the suspension system can’t adapt to the actual situation.In order to suppress train vibration and improve its running stability,a fuzzy neural network controller based on Takagi-Sugeno model was designed in combination with the selflearning ability and self-tuning function of the fuzzy neural network,and the controller model was combined with the suspension model to realize the adaptive adjustment of the suspension system.The acceleration time domain curve and power spectrum density curve were generated by simulation with the German track spectrum as the external excitation.The analysis results show that compared with the passive suspension system and fuzzy sky-hook control system,the fuzzy neural network control can obviously suppress the lateral vibration of the train body,improve the train running stability,and has reference value for future controller design.