Pipe Corrosion Cracking With Ultrasonic Detection Signal Processing Research

The pipeline as a kind of safe and economical carrying means of transporting a mass of energy sources, such as oil, gas and so forth is applied widely in every place of the world. For the sake of guaranteeing their safe movement and prolonging their life, the in-service pipelines should be inspected periodically in order to discover the problem and adopt the measure timely. Based on the need for pipeline inspection of local petroleum operation department, the thesis carries through the design and development of long-distance pipeline corrosion cracking ultrasonic inspection signal processing. Aiming at the difficulty of data processing, use the wavelet analysis method, select the appropriate small Porgy and wavelet level, to resolve the signal denoising and other issues. To further enhance the ability to crack signal recognition, did a great deal of research work, And in accordance with ultrasound signal contained in the crack signal of the energy value of different, Extracted crack signal whether the signal characteristics of. The wavelet analysis of the package for the characteristics of entropy, Use of SVM select the appropriate function for automatic identification Category, has achieved good results.By the theories argument and experiment measurement, In this paper, the use of ultrasound pipeline corrosion cracking in the detection signal processing methods to effectively deal with some of the crack signal, Implementation of the noise signal. Signal and noise characteristics of value can be extracted using SVM to achieve pipeline corrosion cracking signals in the classification.