Active Fault Diagnosis Based on Dual-Observer for Leakage Detection and Estimation With Adaptive Auxiliary Signal Mechanism in Industrial Gas Networks

For industrial gas networks, it is essential to provide accurate leakage fault detection and estimation to ensure the safety and stability of its delivery process. The topology of industrial gas networks has become more complex due to many demand-side users and the large load fluctuations. A unified fault diagnosis framework for the overall networks is urgently needed to realize leakage fault detection and estimation even under complex working conditions. This article proposes a novel active fault diagnosis (AFD) framework based on a dual-observer that includes an unknown input observer (UIO) and fault estimation observer (FEO), to achieve active fault detection and fault estimation simultaneously. In this framework, the UIO is designed for enhanced fault performance by decoupling the partial disturbances and estimating the fault size accurately through the FEO, in which an auxiliary variable is utilized to establish the relation between the fault and the system state. Furthermore, an adaptive auxiliary signal mechanism is designed, which adjusts the time of auxiliary signal injection into the system in real time according to the system operation state. To verify the effectiveness of the proposed method, a single pipeline system and a real industrial gas network with 21 nodes are employed. The simulation results demonstrate that the proposed method significantly improves the ability of gas pipeline leakage fault detection and fault estimation accuracy.