摘要
通过检测坠砣位移、摇杆倾角以及环境参数可以得到接触网张力传递系统的运行状态,而风荷载会导致棘轮补偿器坠砣位移的变化。为准确提取接触网张力传递系统中的故障信息,需对风荷载作用导致的坠砣位移分量进行去除。考虑到风荷载分量是弱信号,提出基于虚拟观测的FastICA的去除方法,并针对ICA的不确定性,提出SS-MDP法消除分离信号的不确定性,并通过仿真验证该方法的可行性。仿真结果证明,该方法可实现坠砣位移信号中风荷载分量的去除。去除风荷载分量后的信号可用于接触网张力传递系统的故障状态的提取。
The operation state of catenary tension transmit system can be got by detecting the balance weight displacement,swing lever inclination and environmental parameters.In practice,windloads always exist in nature,which will cause changes of balance weight displacements of the tension wheel equipment.The component of balance weight displacement affected by windloads should be removed so as to extract accurate failure information of the catenary tension transmit system.Considering that the windload component is a weak signal,the FastICA removal method based on virtual observation was proposed.In view of the uncertainty of ICA,the SS-MDP method was used to eliminate the uncertainty of separated signals,the feasibility of this method was verified by simulation.With the proposed method the windload component was removed from the balance weight displacement signal.The signal without the windload component can be used to extract failure states of the catenary tension transmit system.
出处
《铁道学报》
EI
CAS
CSCD
北大核心
2014年第7期44-50,共7页
Journal of the China Railway Society
基金
北京交通大学校基本科研业务费(E11JB00110)
关键词
虚拟观测
FASTICA
风荷载
坠砣位移
virtual observation
FastICA
windload
balance weight displacement
