Abstract: In order to improve the accuracy and timeliness of the fault diagnosis in nuclear power plant, a event-triggering mechanism regarding the alarm status of operating parameters of the main equipment as the triggering event was proposed for the fault diagnosis in this paper, and a belief rule base of intelligent diagnosis expert system based on event-triggering mechanism was established by using the set of information of the relevant operation parameters in nuclear power plant. The belief rule base was built by combining the failure mechanism model with the nuclear power plant model, namely extracting the symptoms which describe the faults, expressing rules and setting variables of rules under the alarms caused by the faults. The faults were artificially inserted into the nuclear power plant model and diagnosed by the diagnostic system researched above. The diagnostic results show that the type of fault obtained from the diagnostic system matches with that inserted into the nuclear power plant model, which verifies the accuracy of the diagnosis of this system and proves the validity and feasibility of this rule base.