Rail Ultrasonic Flaw Signal Recognition And The Design Of The Detection Device

Railway is an important traffic tool in our daily life.Once an accident occurs,it will cause serious loss of lives and property.Therefore,to ensure the safe operation of the railway has always been an important task for most countries.Periodic inspection of the rail has an important practical significance for avoiding the occurrence of major accidents.In order to obtain the high detection speed,a small rail detection device is designed in this paper.The highest detection of the device is 10 km/h.In order to normally detect,this paper designed a detection device which must meet the following three requirements:to ensure a certain flaw detection speed,the effective coupling between probe and the rail and the high-speed acquisition of defect signal.Ultrasonic signals often contain a lot of noise,how to extract defects in the background noise has been a major problem in ultrasonic signal processing.Separation spectrum technology can improve the signal to noise ratio significantly,so it has played an important role for a long time in ultrasonic signal processing.However,separation spectrum technology is related to the number of filter banks,the size of bandwidth,the frequency interval and so on.So it has not been widely used in engineering area.Based on the idea of split-spectrum technology,the method of variable bandwidth time-frequency analysis is established.Because the method of combining minimum value and polarity threshold has the Maximum signal to noise ratio in variable bandwidth time-frequency analysis,it is chosen to reconstruct the signal.Based on the characteristic of the waveform and frequency of the defect signal,the model of the defect signals is constructed,and correlation filtering was used in processing the sampling ultrasonic echo.In the rail flaw detection,the frequency of the defect echo will be concentrated in the vicinity of the probe center frequency.However,the frequency of the noise is random,so the energy of the noise is often lower than the energy of the defect signal.The method of the pseudo Wigner-Ville distribution will be used to eliminate the clutter.In this paper,the above-mentioned three methods are combined together to form a three steps of time-frequency analysis to process the ultrasonic signals,which can improve the defect detection rate and signal to noise ratio at the same time.In this paper,we separately use ultrasonic 0° probe,37° probe and 70° probe to detect the crack in the different parts of rail defect hole and crack detection.And the actual ultrasonic signal is processed and analyzed by the three steps of time frequency analysis method.After the signal processing,the holes defects and crack flaws can be quickly identified.