Study On THDS Hot Box Comprehensive Prediction Of Qinghai-Tibet Railway

The burning and cutting box of railway freight car is accident-prone, it is also one of the major content of safety work for railway vehicles in china. Currently, the evident problems of hot box alarm system of Qinghai-Tibet Railway Golmud-Lhasa section is difficult to formulate the hot detection model and unable to automatically identify the waveform, and render the redemptions remains low. Because the false positives and rejection of THDS, it brought a great disturbance to the railway transportation. With the increasing of transportation density, the contradiction between the safety and transportation efficiency is becoming sharply.Therefore, aiming at these specific issues this article makes a study on how to improve the hot box redemptions. The main research of this article are as follows:Firstly, it establishes the infrared intelligent recognition model, and makes the waveform recognition. Secondly, it establishes hot box integrated alarm model, and makes a comprehensive evaluation of 5T hot box alarm.Thirdly, deploys the software of the 5T comprehensive early-warning system, to provide real-time monitoring and information query services.According to the statistical results analysis of the new methods, under the premise of better than artificial recognition, the new wave of intelligent recognition achieves automatic recognition and reached more than 80%. At the same time, bearing decompose of fault sample without missing and the strong thermal alarm redemptions reached 64.29% after optimization.All of this shows that the methods can improve the accuracy of hot box estimation, and achieves the purpose of increasing redemptions.