首页 | 本学科首页   官方微博 | 高级检索  
     

基于非线性神经网络的摆式列车检测系统动态补偿方法研究
引用本文:王雪梅,倪文波,李芾. 基于非线性神经网络的摆式列车检测系统动态补偿方法研究[J]. 宇航计测技术, 2004, 24(4): 54-57
作者姓名:王雪梅  倪文波  李芾
作者单位:1. 西南交通大学机械工程学院,成都,610031
2. 西南交通大学牵引动力国家重点实验室,成都,610031
基金项目:铁道部科技研究开发计划项目 ( 99J45 -E),教育部博士点基金项目 ( 2 0 0 2 0 613 0 16)
摘    要:
针对零极点匹配动态补偿方法的不足 ,提出基于非线性神经网络的摆式列车检测系统动态补偿方法。根据陀螺平台实测超高数据 ,采用LM算法对所设计的非线性神经网络进行训练。仿真结果表明 ,所设计的神经网络动态补偿器能有效改善陀螺平台系统的动态特性 ,满足摆式列车倾摆控制实时性的要求。

关 键 词:神经网络  动态补偿  点匹配  仿真结果  实时性  检测系统  非线性  摆式列车  倾摆控制  超高  Tilting train Neural network Dynamic compensation Single-axis gyroscope platform
文章编号:1000-7202(2004)04-0054-04
修稿时间:2003-12-15

Research on Dynamic Compensation Method of Tilting Train Measurement System Based on Non-Linear Neural Network
WANG Xue-mei NI Wen-bo LI Fu. Research on Dynamic Compensation Method of Tilting Train Measurement System Based on Non-Linear Neural Network[J]. Journal of Astronautic Metrology and Measurement, 2004, 24(4): 54-57
Authors:WANG Xue-mei NI Wen-bo LI Fu
Affiliation:WANG Xue-mei1 NI Wen-bo2 LI Fu2
Abstract:
In view of the disadvantages of zero-pole matching dynamic compensation method, the dynamic compensation method of tilting train measurement system based on non-linear neural network is put forward. According to the practical superelevation data measured with single-axis gyroscope platform, the designed non-linear neural network is trained with LM algorithm. Simulation shows that dynamic characters of single-axis gyroscope platform can be effectively compensated for with the designed non-linear neural network, meeting the needs of real time controlling of tilting train.
Keywords:Tilting train Neural network Dynamic compensation Single-axis gyroscope platform
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号