Abstract: In recent years, the operating environment of the secondary relay protection system in substations has become increasingly complex, and the fault forms have shown an obvious tendency of concealment. The traditional maintenance mode has deficiencies in risk identification and prevention capabilities. Therefore, focusing on the technical requirements of condition-based maintenance, this paper constructs a full-loop condition monitoring architecture for relay protection, and proposes an equipment health evaluation model integrating multi-source condition information and a fault diagnosis method based on fuzzy logic. Verification and analysis carried out in combination with the engineering application of a 220 kV intelligent substation show that the proposed method can realize early warning of potential hidden dangers of relay protection and hierarchical risk assessment, providing effective technical support for the safe and stable operation of the secondary substation system.