Empirical mode decomposition applied to fluid leak detection and isolation in process engineering

In this paper, a new algorithm based on signal processing theory is proposed for fluid leak detection and isolation in process engineering. The developed algorithm is sensitive to minor leaks and robust to hydraulic and thermal disturbances and transient operating modes. The proposed approach is based on Empirical Mode Decomposition (EMD) of the sensor measurements and the cross-correlation function. An adaptive threshold, based on stationary analysis of the measurement is calculated, providing a best robustness to false alarm. Experimental results obtained on different processes show the efficiency of the proposed approach.