Research on railroad turnout faults based on CART classification tree algorithm

As an important part of railroad equipment, turnout is the weakest part of the track structure and its quality directly affects the safety and quality of railroad transportation. In order to solve the shortcomings that maintenance personnel can only use professional testing instruments for testing and maintenance manually and lack of systematic theoretical knowledge and limitations when barriers occur, the key features of turnout failure can be extracted by using the turnout action curve and transition time provided by the monitoring system to build a turnout failure identification model to automatically determine whether the turnout is in a fault state, which can effectively manage the turnout. It can effectively manage turnout equipment, predict turnout faults, and improve equipment monitoring efficiency. The experimental results show that the accuracy of fault detection is as high as 96.86%.