基于BP神经网络的铁路货车横向稳定性评估  

Study of Lateral Stability Assessment of Railway Freight Wagon based on BP Neural Network

作　　者：辛民 江亚男[2] 

机构地区：[1]神华铁路货车运输有限责任公司,北京100011 [2]西南交通大学牵引动力国家重点实验室,四川成都610031

出　　处：《大连交通大学学报》2017年第4期60-63,共4页Journal of Dalian Jiaotong University

摘　　要：探讨了BP神经网络技术在铁路货车横向稳定性评估中的应用:首先通过货车动力学模型的单参数影响分析筛选出了车辆横向稳定性关键影响因素集;然后采用MATLAB的Newff函数设计因素集与临界速度的映射关系神经网络模型,通过对比不同设计参数下预测值与期望值的误差确定合理的神经网络模型参数;最后,通过预测结果与实际值的对比分析和相关性分析验证了该映射关系模型的有效性.根据该映射关系可以实现基于部件状态的货车横向稳定性评估.The application of BP neural network in lateral stability assessment of railway freight wagon is investigated. Firstly the dynamic model of a vehicle is built, and the key factor set is obtained through the influence analysis of single parameter. According to Newff function in MATLAB, the mapping relationship model between the factors and critical velocity is designed. Comparing the prediction value of BP neural network with the actual one, the effectiveness of the mapping relationship model is verified. The lateral stability assessment of the vehicle based on the state of components is realized by the mapping relationship model.

关 键 词：铁路货车 检修制度 状态修 神经网络 稳定性评估 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]