Predicting the contact wire wear of a railway system using ANN

The aim of this paper is to examine a method that predicts contact wire wear by using the artificial neural network (ANN) approach. A prediction model using relevant input parameters is proposed and illustrated in this paper. The model is to be built around an ANN approach to search and determine the internal relationship between measured physical parameters as inputs and the wire wear as output. Parameters that have significant impact on the wear of contact wire in a railway system are selected to 'educate' a wear model. The parameters used in this paper include the height, stagger, and span length of contact wire in the Overhead Conductor System (OCS), the speed, traction voltage and mode (motoring, coasting and braking) of the train, and the contact force between the contact wire and the pantograph current collecting strip. With the data obtained by automated measuring equipment and construction records, the input-output pairs in the dataset are used to train, validate and test an ANN model. Upon successful training, the ANN model is used to predict and characterise the wire wear.

Media Info

  • Pagination: 9p. ; PDF
  • Monograph Title: Rail - the core of integrated transport: CORE 2012: conference on railway engineering, 7-10 September 2012, Perth, Western Australia

Subject/Index Terms

Filing Info

  • Accession Number: 01532165
  • Record Type: Publication
  • Source Agency: ARRB
  • Files: ATRI
  • Created Date: Jul 29 2014 11:58AM