Motor train unit train running fault information detection method and system

Abstract

The embodiment of the invention discloses a motor train unit running fault information detection method and system. The method comprises the steps of inputting a preset number of images of the same type of motor train unit trains with fault information to a deep learning framework and training the images to obtain a fault recognition model; acquiring identity information and the images of the motor train unit trains; according to the identity information and the images of the motor train unit trains and the rule that the same trains are at the same position, classifying the images; inputting the images to the fault recognition model of the same type as the images and judging whether the images contain the fault information or not. The recognizability of images of the fault information is effectively improved, the false alarm rate is lowered, the accuracy is improved, image watching operating personnel can also get rid of the fate that a lot of image data is manually detected, the working intensity is lowered, the occurrence of missing detection is avoided, and the running safety of the motor train unit trains is guaranteed.

Description

Technical field [0001] The present invention relates to the technical field of train failure information detection, in particular to a method and system for detecting train operation failure information of an EMU. Background technique [0002] With the development of China's high-speed railways, there are more and more EMU trains. The detection of train operation fault information of EMUs is an important means to ensure the safety of EMU trains. The detection of train operation fault information of EMUs cannot rely on manual online detection, and can only use high-tech detection equipment to detect trains. [0003] The traditional EMU train operation fault information detection technology scheme mainly uses the railside equipment to collect the car body image of the running EMU train, transfer the image data to the data service device of the railside detection station, and use the image comparison on the identification service device The algorithm performs fault information identi...

AI Technical Summary

Problems solved by technology

[0004] However, the current detection system has the problems of high false alarm rate of fault information and low accuracy rate. The operator still cannot get rid of the fate of manually detecting a large amount of image data. The working hours are long, the number of images is large, and the work intensity is high. In case of missing inspection, it is difficult to guarantee the operation safety of EMU trains

Images