A Segmented Adjustment Least Mean Square Deconvolution Rail Crack Signal Restoration Method

Abstract

The invention discloses an improved segmented adjustment least mean square deconvolution-based steel rail crack signal recovery method, and aims at exciting acoustic emission signals having same characteristics of steel rail cracking processes through an acoustic emission probe. The method comprises the steps of: 1, inputting a known pulse signal from the acoustic emission probe into a steel rail;2, carrying out FFT transformation on an output and an input, and dividing the spectrums of the two to obtain an initial value of a system characteristic; 3, carrying out least mean square filteringon the system characteristic; 4, repeating the step 3 until a reduction error is less than a preset value, and outputting a system pulse response; 5, carrying out deconvolution via a crack signal andthe system pulse response to obtain an inverted signal; and 6, setting a weight according to attenuation degrees of the steel rail at different frequency segments, and weighting and superposing time domain components of the inverted signal to obtain a final input signal. According to the method, a segmented adjustment least mean square deconvolution method is used to invert emission signals of cracks in steel rails, so that the crack source simulation methods in the existing acoustic emission test are improved, and the method is safer and easier to implement and has strong theoretical and engineering practical significance.

Description

technical field [0001] The invention relates to a method in the field of rail crack signal detection, in particular to an improved segmented adjustment minimum mean square deconvolution rail crack signal recovery method. Background technique [0002] With the "Fuxing" put into operation, my country's high-speed rail construction has entered the peak period of large-scale construction, and it is also the peak period when a large number of new lines are put into operation. With the increase of the speed of passenger trains and the increase of the axle load of freight trains in my country's railways, the rails are subject to alternating bending stress and transverse stress during service, resulting in increased wear, plastic deformation and fatigue damage, and the performance of rails will deteriorate rapidly. However, a large number of rails are in the state of overload operation and overdue service. If the rails break and fail, it is likely to cause a major rail breaking acci...

AI Technical Summary

Problems solved by technology

[0004] Due to the passive characteristics of the acoustic emission detection technology, the on-site detection of the reliability of the formed acoustic emission equipment needs to artificially impose a simulated crack source on the rail. Methods such as scribing cracks or applying large hydraulic pressure to the rail specimen until the rail breaks have disadvantages such as a large difference from the actual crack signal characteristics, high cost, and difficulty in implementation.

Secondly, due to the defects in the frequency characteristics of the sensor itself, the acoustic emission signals received in the current research are all distorted by the characteristics of the sensor to varying degrees, which is not conducive to the research and restoration of the characteristics of the acoustic emission source

Images