Data-Driven Fault Detection in Gas Regulators Using Empirical Mode Decomposition

This paper proposes a data-driven approach for fault detection in gas regulators using empirical mode decomposition (EMD). The structure of direct gas regulators is first explained. Next, the paper introduces data-driven residual generators and presents an optimization problem for designing the parameter matrix of residuals. Based on this, a set of EMD-based evaluation function is developed, and the detection decision is constructed. Finally, the effectiveness of this approach is verified using experimental data from an experimental platform for gas transmission and distribution network.