摘要

By taking full use of the temporal information of alarm message for quick and accurate diagnosis is very important. This paper presents an effective method for fault diagnosis, based on the time constraint fuzzy directed graph (TCFDG), in order to solve the cases of incomplete alarm information. An improved recognition algorithm is proposed which can effectively recognize misreports, missing messages and timing inconsistencies etc. And false alarms are filtered to overcome the deficiencies of incomplete alarm information, and to provide protection for the precise positioning of faults. Then, a reasoning method based on TCFDG is proposed, and use min operator and multiplication operator to determine the fault source, and use incidence matrix reduction method, reducing the dimension of the matrix, improved operational efficiency. Finally, a typical power system compared with some other typical methods was shown for demonstrating the feasibility and efficiency of the proposed method.