Leak detection for long transportation pipeline using a state coupling analysis of pump units

Leak detection for long transportation pipeline with a large economic and environmental impact has been an area of intensive research for more than five decades. This paper presents a novel pipeline leak detection scheme based on a state coupling analysis (SCA). Instead of monitoring the pipeline and pump units separately, SCA introduces a new detecting method of analyzing data in a coupling running condition. A novel capture method for abnormal pressure based on logical reasoning algorithm is proposed. Hamming approach degree arithmetic is applied to calculate the matching mode identifying the state of units. SCA is used to reduce the rate of false alarm and detect the leak with a high detecting sensitivity for long transportation pipeline. An on-line software system based on SCA is utilized to achieve superior accuracy and implementation. An industrial case study for coupling system pipeline leak detection is used as an example to validate the effectiveness of the proposed method.

► Logical reasoning algorithm is used to capture the abnormal pressure.
► For the pipeline leak, a state coupling analysis (SCA) is used to obtain a high detection sensitivity but low false alarm.
► Hamming approach degree arithmetic is applied to calculate the matching mode identifying the state of pump units.
► The effectiveness of the proposed approach is verified.