Nonlinear fitting based intelligent detection method for running exception state of contact network

Abstract

The invention discloses a nonlinear fitting based intelligent detection method for a running exception state of a contact network. The method is implemented by the specific steps of (1) initializing a system; (2) preprocessing an image; (3) performing machine learning and region extraction; (4) performing edge detection; (5) extracting an exception state feature edge and judging whether a pixel value reaches a threshold or not; (6) extracting a to-be-fitted curve and selecting a pixel point; (7) performing nonlinear fitting; (8) calculating whether a coefficient of determination reaches a threshold or not; and (9) manually judging whether an exception exists or not. According to the method, the problem of incapability of accurately detecting and quickly identifying an exception state by an existing safety inspection apparatus is solved; the working efficiency and quality of the contact network are improved; and new and effective scientific technical support is provided for improving information automation level and working efficiency of power supply management.

Description

technical field [0001] The invention relates to the field of computer technology, in particular to a method for intelligent detection of catenary operation abnormal state based on nonlinear fitting in the field of visual computing technology and nonlinear fitting calculation method. The invention organically integrates visual computing, computer network, wired / wireless communication technology, nonlinear fitting method, artificial intelligence and human-computer interaction, and can provide intelligent detection for the power supply operation status of the railway catenary to improve the quality of the catenary. Detect work efficiency and quality, and provide new and effective scientific and technological methods for the improvement of information automation level and work efficiency of power supply management. Background technique [0002] The catenary safety inspection device (2C) plays an important role in detecting the defects of catenary equipment and ensuring the safet...

AI Technical Summary

Problems solved by technology

[0005] In order to overcome the problems existing in the above-mentioned prior art, the present invention proposes an intelligent detection method for catenary abnormal state based on nonlinear fitting. For image information, the brightness and contrast of the image are enhanced by preprocessing the catenary state image. According to the construction characteristics of the catenary in similar railway sections, machine learning technology is used to extract the area of ​​the image captured by the high-speed camera, and focus on the clearer ones. The key component of the catenary, the clearer catenary edge information image is detected by the sobel operator, and the "exclusive OR" operation is performed with the result map of the Hough line detection, and the corrosion operation is performed on the image after the "exclusive or" operation to obtain a more accurate Suspected abnormal state edge map, using nonlinear fitting method to select the best fitting function, judging the coefficient of determination to detect suspected defects, and finally using human-computer interaction to screen catenary state images with real defects, which solves the problem of full-time personnel's observation records Playback and analysis of a large amount of catenary operation image data received, to judge whether there is an abnormal state, time-consuming and labor-intensive, and problems such as equipment defects cannot be found in time

Images