An Adaptive Separation Algorithm For Direct Wave Detection Of Ultrasonic Rail Flaw Detection Based On Interpolation Cross-correlation

Ultrasonic rail flaw detection is a very important issue,which determines the safety of railway transportation.The Ministry of Railways requires regular inspection and inspection of railways.Especially for high-speed rail,we know that high-speed rails do not operate at night,that is,high-speed rails need to be detected every night,and high-density inspections require high-precision inspections.In the rail flaw detection test,since the detection signal adopts a linear frequency modulation signal,there is a direct wave interference in the analysis of the flaw detection result after the end of the detection.Therefore,it is necessary to study the direct wave suppression method suitable for ultrasonic rail flaw detection.When an adaptive filter is used to perform direct wave suppression on the received signal,the selection of the filter reference signal determines the performance of the adaptive filter algorithm and the effect of direct wave suppression.Since the reference signal is used to fit the direct wave,the direct wave filtering will have a better effect when the reference signal is strongly correlated with the direct wave component of the received signal.Therefore,the focus of this paper is on how to fit the reference signal most directly related to the direct wave component in the received signal.The main contributions are:1.For the problem of using another way to receive the signal analog reference signal,that is,without using prior knowledge simulation,the reflection waveforms in the two-way input of the adaptive filter are too similar;and artificially estimate the direct wave waveform analog reference signal.The problem that the correlation between the analog reference signal and the direct wave is not enough.This paper studies a reference signal selection algorithm based on interpolating cross-correlation.Using the transmitted signal plus interpolating crosscorrelation algorithm,the simulated reference signal effectively solves the problem that the traditional reference signal selection method does not utilize prior knowledge simulation,reference signal and direct wave.Related issues such as not much.2.Unlike the ground penetrating radar direct wave suppression,the ultrasonic signal used by the rail flaw detection is longer than the wavelength of the electromagnetic signal used by the bottom radar.The sample estimation error that is easily introduced when the signal is sampled and an adaptive filter is used to separate the direct wave.In order to solve the above problems,this paper studies a sampling rate traversal method combined with interpolation cross-correlation algorithm,which effectively solves the problem of sampling estimation error introduced by the longer wavelength of the ultrasonic signal,and further improves the accuracy of the direct wave separation.3.In order to simulate the direct wave suppression effect in an actual rail flaw detection scenario simply and conveniently,based on the above method,a set of soft and hard combined detection system based on ultrasonic rail flaw detection was developed and the test of the system was completed.The direct direct wave suppression experiment was carried out for the actual rail flaw detection scene.The reference signal selection method studied in this paper and the direct wave suppression effect under the traditional reference signal selection method were compared,and a good contrast effect was achieved.The feasibility of the method.