Font Size: a A A

Research On Contact Wire Wear Predicting Model By Using MATLAB Based On Artificial Neural Network

Posted on:2019-12-28Degree:MasterType:Thesis
Country:ChinaCandidate:Z H ChengFull Text:PDF
GTID:2382330545465394Subject:Power system and its automation
Abstract/Summary:PDF Full Text Request
Traction power system with two kinds of power supply system of DC and AC,DC power supply for city rail transit system,but the catenary power supply principle of electrified railway and AC power supply similar with urban rail.Due to its advantages such as high speed,high transportation efficiency and low environmental pollution,high-speed railway EMU train has become the main transportation means of transportation transport industry abroad today.Catenary is responsible for the electric power obtained from the traction substation directly to the use of electric locomotives,the quality and status of the catenary network directly affects the transport capacity of electrified railway.Therefore,it is necessary to grasp the working condition of the contact line in time.However,the contact wire wear test is a test item with the most strength,the longest research time and the smaller harvest in the measurement of the geometric parameters of the contact wire.Therefore,this paper assumes a multi-input,single-output mathematical model which is artificial neural network model,looking for the link between the height of the wire and other relatively easy to measure the acquisition of parameters and contact wire wear,and then calculate the evolution predict the amount of wear to reduce the workload of wear measurement.The method used in this paper is BP neural network,which is the back propagation network in artificial neural network(ANN).Artificial neural network in the prediction can automatically simulate the natural relationship between the various wear factors,reduce human intervention and improve the accuracy of the final contact line wear prediction rate.In this paper,MATLAB neural network toolbox to achieve the BP network model of the contact wire wear of the fitting process and prediction(simulation),and conducted on-site simulation.The results show that the calculated values can be basically consistent with the measured values based on the model simulation.The experimental errors are approximately between 10% and can be used to predict the contact wire wear in the field.Because the two kinds of power supply system in the construction of the similar contact network,also can be electrified railway the research method is applied to AC power system.
Keywords/Search Tags:contact wire wear, artificial neural network, prediction
PDF Full Text Request
Related items