Abstract: In order to solve the problem of unified management of multi-source and heterogeneous data faced by the protection setting verification and dynamic diagnosis of electrical system in the centralized control operation of thermal power plants, this paper constructs the method framework of protection setting online verification and dynamic diagnosis of electrical system in thermal power plants based on multi-source data monitoring, and proposes the unified Kalman filter Bayesian fusion algorithm and support vector machine dynamic diagnosis model, which improves the consistency of protection setting online auxiliary verification and conventional offline verification results and the correctness of fault analysis, and can assist in guiding the maintenance of electrical equipment in thermal power plants.