Decision fusion framework in diagnostic and prognostic assessment of long-distance oil pipeline leakage and damage based on Dempster–Shafer theory

Abstract

A decision fusion algorithm in diagnostic and prognostic assessment of long-distance oil pipeline leakage and damage based on Dempster–Shafer (D–S) theory is proposed in this paper. To monitor the leakage and external damage, a new monitoring system is constructed with an optical fiber cable installed along the pipeline as a distributed optical fiber sensor (DOFS) and a pair of pressure meters installed at the pipeline inlet and outlet. When leaking, the oil stream disturbs the cable and a negative-pressure wave originates which propagates to pipe inlet and outlet. The probability of leaking is evaluated according to the correlation coefficient of negative-pressure waves acquired at the inlet and outlet, and the location is pinpointed using the negative-pressure method. The optical fiber cable is sensitive to soakage of oil products and mechanical deformation/vibration caused by leaking, tampering, or mechanical impacting. An optical time domain reflectometry (OTDR) is employed to locate the additional attenuation and an optical power meter is employed to record the transmitted optical power. By using the D-S rule, features generated from the waveforms of optical power and negative-pressures are fused to make a decision about whether and where there is leakage and/or external damage.