Application of support vector machine learning to leak detection and location in pipelines

Abstract

Leak detection in oil pipelines is important for safe operation of pipelines. Negative Pressure Wave (NPW) in pressure curve can be an indication of leakage of a pipeline. In this paper, we propose to use the support vector machine (SVM) learning to detect NPW in pressure curves. In the approach, NPWs detection is formulated as a supervised-learning problem and the method of SVM is employed to develop the detection algorithm. The proposed method is evaluated using a database of 1500 pressure curves containing 500 NPWs. Experimental results demonstrate that, when compared to the Wavelet based methods, the proposed SVM framework offers the better performance.