An integrated Gaussian–Probabilistic–Fuzzy framework for health assessment and remaining useful life prediction of medium-voltage switchgears

Abstract

Reliable operation of Switchgear is crucial for ensuring the safety and stability of modern power systems. However, prolonged exposure to thermal, mechanical, and electrical stresses progressively deteriorates switchgear performance, often resulting in catastrophic failures and costly outages. This paper proposes a novel hybrid framework for comprehensive health assessment and remaining useful life (RUL) prediction of medium-voltage (MV) switchgears by integrating Gaussian normalization, probabilistic modeling, and fuzzy logic evaluation. Initially, all measured indicators are systematically classified into four functional subsystems, electrical, mechanical, insulation, and auxiliary, to enable structured condition evaluation. The Gaussian method normalizes heterogeneous indicator values, the probabilistic approach quantifies the failure likelihood of each indicator, and the fuzzy logic system handles uncertain and linguistic expert information. The resulting subsystem health indices are aggregated to derive a global health index (HI), which is then used to estimate RUL through a (Gaussian, Probability, Fuzzy) GPF-based algorithm. The proposed framework has been validated using real field data collected from two MV substations, and its performance has been benchmarked against existing techniques. Results demonstrate superior accuracy, robustness, and interpretability of the proposed method under uncertainty, providing actionable insights for condition-based maintenance scheduling and risk-informed decision-making in power utilities.