Monitoring method for performance analysis and comparison on steel rail welding seam data set based on MDCD

Abstract

The invention discloses a monitoring method for performance analysis and comparison on a steel rail welding seam data set based on MDCD. The monitoring method comprises the following steps that S1, analyzing a Lamb wave structure in a steel rail welding seam; S2, monitoring Lamb wave data characteristics on the steel rail weld crack damage; S3, building and training an MDCD model for health monitoring of the steel rail welding seam structure; S4, evaluating the model training effects of the two stages, comparing the performance reduction rate of the second-stage model relative to the first-stage model, and comparing the execution effects of the two-stage training models on different tasks with a current advanced and popular deep learning algorithm and a traditional digital signal processing algorithm. According to the invention, crack damage on the weld structure can be effectively detected, and the Lamb wave can be ensured to monitor the weld structure within a certain distance and range due to low propagation damage.

Description

technical field [0001] The invention relates to the technical field of non-destructive testing, in particular to a monitoring method based on MDCD performance analysis and comparison on rail weld data sets. Background technique [0002] As a part of the rail that undertakes the continuity of two modules, the health status of rail welds is closely related to the safety of railway transportation. During the production process of railway transportation, due to the irregular shape of the welded joints of the rails, when the train passes by, the force generated by the wheel-rail relationship cannot be evenly transmitted and dispersed at the welded joints, resulting in some parts of the welded joints of the rails being stressed. serious. In the case of passing high-frequency trains for a long time, the vibration fatigue caused by the wheel-rail relationship will directly lead to fatigue crack damage in the severely stressed parts of the rail welds. As the crack damage deepens, ra...

AI Technical Summary

Problems solved by technology

[0005] In order to ensure the detectability of rail weld cracks, conventional non-destructive ultrasonic detection methods need to use flaw detectors, flaw detectors and other equipment to continuously slide on the rails to achieve crack damage. Therefore, for rail welds with fixed positions and a small range, it is necessary The detection of cracks can only be achieved by continuously sliding on it. The mobile detection of the ultrasonic probe prevents it from solidifying around the rail weld to achieve continuous structural health monitoring of the rail weld. Each inspection consumes manpower and material resources

Images