| The railway industry in our country develops fast in the recent years. The electric railway is the most efficient, energy-saving and environment-friendly railway system, so it has been used widely. The pantograph-catenary system is one of the key technologies in the electric railway. But its fault rates are very high because this system is a solo system without any redundancy, the force and power transmission mechanism are complex. It’s a weakness in the electric railway. In order to keep the electric railway operation trustworthy and safe, the pantograph-catenary system’s dynamic characteristics must be assessed effectively to promote its personal performance, and in the other hand, the system must be monitored efficiently to keep the safety. This paper do not use the traditional methods like mean and standard deviation in the time domain. The one dimension physical signals and two dimension image signals of the pantograph-catenary system are analyzed using the modern signal processing technologies in frequency domain or time-frequency domain.This paper is supported by the Scientific Project of Railway Ministry of China (2011J016-B); the National Natural Science Foundation of China(No.51007074); the New Century Excellent Talents in University of Education Ministry of China(No. NECT-08-0825); the PhD Innovation Fund of Southwest Jiaotong University.The main purpose of this research is to assess pantograph-catenary system characteristic signal accurately. Based on the modern signal processing technologies, the main work in the paper includes:the obtain of pantograph-catenary dynamic data-> the pre-process of the data-> the analysis of basic statistic characteristic-> the new assessment method in time domain-> the new assessment method in frequency domain-> the fault image intelligent identification.(1) The pantograph-catenary model automatic creation method using the secondary development technology is proposed to obtain the necessary data for analyzing the characteristic, and this method can promote the efficiency of model creation, improve the reuse rate, reduce the fault rate of creating the model and simplify the comparison between the different models. The model is modified using the falling temperature method and the nonlinear characteristics of material. The suggested choice of finite element length in the model is given. The model is proved reliable via the compare with EN50318and simulation of Siemens.(2) In order to research the influence of the wind to the catenary, the formula of catenary wave speed is modified considering the high tension, catenary stiffness and random wind. The wind tunnel test is performed to measure the lift coefficient and damp coefficient of catenary. The pantograph-catenary dynamic formula is constructed considering the extra wind damp.(3) In order to assess the pantograph-catenary characteristics, the correlation between different signals of system are analyzed using EEMD. The contact force data are pre-processed more accurately according to the correlation. The stationarity and periodicity of contact force data are assessed. A new second order difference standard deviation assessment method is proposed to assess the system characteristics. The rationality of this method is analyzed from the lifetime’s point of view.(4) The pantograph-catenary characteristics showed by spectrum are analyzed by the modern signal spectrum technology. The catenary inherent frequencies are obtained from ARMA model. The advantage of spectrum analysis is showed by analyzing the pantograph-catenary coupling performance using the uplift spectrum, and the bad coupling frequency range can be found besides the basic result. Finally, the quantitative representation of spectrum is researched, and this can provide the reference of grading the pantograph-catenary characteristics.(5) The pantograph-catenary detection image signals are researched by using the modern signal sparse representation technology. The image pre-process, the pantograph crack fault, the insulator damage fault can be identified intelligently using curvelet transform. These algorithms are accurate and fast. Its time cost can be accepted. These algorithms provide the feasible scheme to monitor the pantograph-catenary system. |
PDF Full Text Request