A Bayesian detector-based approach for determining arrival and departure times of negative pressure waves in pipeline leakage localization

Abstract

The pipeline leakage detection and location method based on negative pressure wave (NPW) characteristics are widely used in existing pipe network systems. The accuracy of leakage localization significantly relies on precisely picking up the arrival time of the NPW. A novel method based on Bayesian detection theory is proposed to determine both the arrival and the departure times of the NPW. The use of NPWs as features was theoretically derived based on the transient fluid dynamics of the pipeline, and then the nonstationary NPW was transformed into a piecewise stationary Gaussian process by differencing the pressure time series data. A Bayesian optimal detector was constructed to identify the multiple transition points (corresponding to the arrival and departure times) in the differentiated pressure data to identify the NPW. A series of pipeline leakage tests were carried out to verify the effectiveness of the proposed method. It is demonstrated that the combination of time differencing nonstationary pressure data and employing the Bayesian detector can precisely capture the arrival and departure times of NPWs, enabling more accurate leakage localization in the pipeline.