Study On The Key Technologies Of Automatic Ultrasonic Inspection For Seamless Tubes

In order to meet the demands of on-line automatic detection for seamless tubes and regard improving the ultrasonic testing(UT) reliability,accuracy,and efficiency of seamless tubes as the research objects of this dissertation,some key technologies such as increasing the signal to noise ration(SNR),improving the ultrasonic inspection resolution and intelligent recognition of flaws types were completely studied using the hybridized swarm intelligent optimization theory.Moreover,an automatic ultrasonic inspection system for seamless tubes was developed.The detail work was presented as follows:Chapter 1,the urgency concerned with speeding up the research on ultrasonic nondestructive testing(UNDT) equipments for seamless tubes was discussed,some typical applications to detect seamless tubes were introduced,and the current research status of automatic ultrasonic inspection as well as their trend was summarized.At the end of this chapter,the key technologies of UNDT for seamless tubes were presented in order to determine the research direction and contents of this dissertation.Chapter 2,a new kind of de-noising technology based on model estimation and hybridized swarm intelligent algorithm was proposed by analyzing the limitations of the typical UT de-noising methods used in the seamless tube on-line inspection.The object signal was extracted successfully from UT echo signal through using the hybridized swarm intelligent algorithm to estimate the UT echo signal model parameters.The experiment results showed that this technique was able to improve SNR greatly and possessed strong generalization.Chapter 3,a new sort of deconvolution technique based on hybridized swarm intelligent algorithm and wavelet transform was developed to improve the UT resolution for seamless tubes.After the position set of ultrasonic reflection coefficient was determined by using wavelet transform,hybridized swarm intelligent algorithm was adapted to solve the ultrasonic reflection coefficient amplitudes corresponding to the position set.Simulation and experiment results showed that the new method owned strong robustness and wide adaptability.Chapter 4,in order to apply the automatic identification technique of flaw types in seamless tube UNDT,particle swarm optimization cluster analysis was taken to divide the eigen vectors so as to fulfill flaw type interpretation automatically on the basis of extracting the eigenvalue of defect echoes by complex wavelet transform.The efficiency and practicality of this method were proved by the experiment results.Chapter 5,some key parts and modules were developed after the general project of the automatic UT system for seamless tubes was designed,and the automatic UT system was integrated.The practical operation results testified that this system was able to meet the requirements of on-line inspection.Chapter 6,the main work and conclusions in this dissertation were summarized and the prospect for further research objects was put forward.