| 基于LM神经网络模型的机车牵引力和制动力的计算 |
| |
| 引用本文： | 刘典政,冯晓云.基于LM神经网络模型的机车牵引力和制动力的计算[J].机车电传动,2007(3):20-23. |
| |
| 作者姓名： | 刘典政 冯晓云 |
| |
| 作者单位： | 西南交通大学,电气工程学院,四川,成都,610031 |
| |
| 摘 要： | 针对传统机车牵引力、制动力计算的不足,提出了基于LM(Levenberg-Marquardt)算法的前向多层神经网络模型方法,阐述了该神经网络的结构设计,利用LM算法使得在网络结构最小、训练步长最短的情况下实现精度最优.仿真与试验的结果表明,利用该神经网络模型能较好地计算机车的牵引力与制动力,其精确性、快速性和抗干扰性都优于其他计算方法.
|
| 关 键 词： | LM算法 牵引力 制动力 神经网络 机车 神经网络 网络模型 机车牵引力 制动力 力的计算 Neural Network Model Based Force Braking Locomotive 计算方法 抗干扰性 快速性 地计算 结果 仿真与试验 最优 精度 情况 训练步长 |
| 文章编号： | 1000-128X（2007）03-0020-04 |
| 修稿时间： | 2006-10-20 |
Calculation of Locomotive Tractive and Braking Force Based on LM Neural Network Model |
| |
| LIU Dian-zheng,FENG Xiao-yun.Calculation of Locomotive Tractive and Braking Force Based on LM Neural Network Model[J].Electric Drive For Locomotive,2007(3):20-23. |
| |
| Authors: | LIU Dian-zheng FENG Xiao-yun |
| |
| Institution: | School of Electrical Engineering, Southwest Jiaotong University, Chengdu, Sichuan 610031, China |
| |
| Abstract: | Aiming at the disadvantages of traditional calculations of locomotive tractive and braking force,the method of forward multi-layer neural network model is proposed based on Levenberg-Marquardt(LM)algorithm.The structure design of the neural network is elaborated.With LM algorithm,the optimum accuracy could be realized while the network structure is smallest and the training epochs are shortest.The simulation and test results show that with the neural network model,the locomotive tractive and braking forces could be calculated,with higher accuracy,speed and stronger anti-interference capacity against other algorithms. |
| |
| Keywords: | Levenberg-Marquardt algorithm traction force braking force neural networks locomotive |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
