Improved method for fault diagnosis of red band of track circuit

Abstract

The invention discloses an improved method for fault diagnosis of the red band of a track circuit. The method includes the following steps of: 1) qualitative analysis of FTA for establishing a fault tree model according to field failure analysis; 2) determining the diagnostic attribute of a diagnostic model by a minimum cut set and extracting a fault diagnosis rule by analyzing knowledge contained in the fault tree; 3) designing a diagnostic system structure and constructing an improved particle swarm optimization support vector machine (SVM) model according to a FTA analysis result; 4) setting up an improved particle swarm optimization algorithm parameter according to a diagnostic object, and establishing a fault diagnosis model through the SVM model improved by sample learning and training; and 5) inputting actual measured data in real operation into the fault diagnosis model and obtaining a fault diagnosis result by calculation. The method can be used for diagnosing the red band failure of a ZPW-2000 non-insulated frequency shift track circuit, and is high in fault diagnosis accuracy and high in fault sample utilization rate so as to provide a new technical means for the intelligent fault diagnosis of the track circuit.

Description

technical field [0001] The present invention relates to the field of fault processing of high-speed railway signal equipment, and relates to a ZPW-2000 track circuit fault diagnosis method, specifically a ZPW-2000 non-insulated frequency-shifting track circuit red light based on FTA and improved particle swarm optimization support vector machine With fault intelligent diagnosis method. technical background [0002] Due to the special role and status of the track circuit, the performance of the track circuit directly affects the transportation efficiency and driving safety. The failure of the track circuit function will affect the transportation efficiency and may bring disastrous consequences. Therefore, the normal operation of the track circuit must be guaranteed. At present, the mainstream equipment with blocked sections mostly adopts the ZPW-2000 series non-insulated track circuit. The track circuit works outdoors all the year round and is easily affected by the environme...

AI Technical Summary

Problems solved by technology

However, in practical applications, the SVM method still has the problem of how to select parameters to optimize its diagnostic performance. A single SVM method alone cannot meet the requirements of rapid diagnosis and accurate location of fault points in the red light belt fault diagnosis of track circuits.

Images