Anti-interference Design Of EMAT Reception System And Research On Noise Reduction Algorithm

Electromagnetic acoustic detection technology has many advantages such asbeing free of couplant and heat-resistant, which has been widely used in non-destructive testing fields. However, electromagnetic acoustic transducers (EMATs)have low energy converting efficiency and are sensitive to noise, so that they aresusceptible to external environment. Because of the influence of electromagneticinterference (EMI) in train wheel treads flaw detection system based on EMATs,there are false judgements detection results. In order to improve the anti-interference capability of train wheel tread flaw detection system and increasedetection accuracy, this paper extensively reaserches the anti-interference design ofEMAT receiving system and noise suppressing algorithms.Aiming at EMI problem in wheel tread detection system, this paper analyzes themechanism of EMI. Then the program of anti-interference design for EMATreceiving system and noise suppressing algorithms are put forward.This paper proposes a new coil structure of receiving EMAT in allusion to thenear-field EMI problem. This method includes changing the structure of traditionalcoiling mode, and design parameters of coil and magnet. It is proved bytheoretically analysis that this method could overcome the disadvantage oftraditional receiving EMAT. Experiment shows that according to near-field EMI, thenew pattern coil increases the signal-to-noise ratio (SNR) to5.6times.Aiming at the far-field EMI problem of EMAT, this paper proposes a designmethod of electromagnetic shielding solutions. By means of finite element software,this paper simulates the improving effect of SNR when using different shieldingsulutions. Experiment indicates that according to far-field EMI, the selected shieldsolution impoves the SNR to300%.This paper puts forward a combination of wavelet denoising algorithm,envelope extraction algorithm and defect echo signal identification algorithmaiming at the problem of previous algorithm’s weakness at suppressing gread nosie.The algorithm filtering high frequency nosie firstly and save the low frequencycharacteristic of great noise, then pick out noise according to the difference betweenultrasound signal and noise; according to the propagation characteristics in thewheel tread of ultrasound, a defect detection algorithm is proposed. Experimentindicates that the method of anti-interference design and noise suppressingalgorithm proposed in this paper can improve the detection accuracy from74%to96%.