Ultrasonic signal compressive sampling and its application in pipeline flaw inner inspection

Abstract

Ultrasonic inner testing method can be used to detect almost every kind of flaws in oil and gas transmission pipeline. Due to the large number of transducers on the testing pig, a huge amount of sampling data has been produced which makes data storage and real-time processing impossible. For the purpose of reducing sampling data without loss of resolution, a compressive sampling method applied to the front end of sensor array is introduced. The paper firstly focuses on the sparsity of flaw ultrasonic echo in pipeline which is the precondition of the method. Then, compressive sampling technique based on multi-channel analog-to-information convertor (AIC) is designed to achieve low sampling rate data acquisition. The experimental results show that the flaw echoes can be accurately reconstructed with only 22% of sampling data compared with traditional sampling method. Furthermore, the flaw dual identification can be directly performed without signal reconstruction which greatly improves the time efficiency.