Based on the criteria for electrical data，protective relays and circuit breakers，a weighted fuzzy Petri net based model with time-delay constraints for power system fault diagnosis employing information from multiple sources is presented，in which the temporal features of the three criteria are taken into account.First，the candidate fault components are identified by analyzing the power network topologies before and after the occurrence of a fault or faults.Then，the action paths can be found by the backward reasoning routing，thus the fault diagnosing model based on the weighted fuzzy Petri net is developed to identify the fault components.The electrical data and their temporal features are exploited in the development of the Petri net as well.The time-delay constraints between faulty components and operating protective relays as well as between operating protective relays and tripped circuit breakers，and the malfunctions of protective relays and circuit breakers，can be accommodated.A matrix operation based method is employed for fault diagnosis reasoning.The backward reasoning is next used to identify the malfunctioning protective relays and/or circuit breakers.Finally，the IEEE39-bus power system and an actual fault scenario of Xingren substation in Guizhou Province are used for demonstrating the feasibility and efficiency of the fault diagnosis model and method developed.