A fuzzy Petri net based approach for fault diagnosis in power systems considering temporal constraints

The fuzzy Petri net is a promising and efficient approach that can tackle the complexities of power system fault diagnosis. In this work, the temporal constraint between event occurrences in power systems is investigated. Then, it is introduced to a fuzzy Petri net (FPN) for fault diagnosis. The temporal attributes are assigned to the propositions in the Petri net, so that temporal information can be taken into account, which makes the true hypothesis distinguishable from the false ones. The modified matrix execution algorithm can enhance computational efficiency, with a “weighted average” operation included to improve the fault-tolerance. The developed model possesses a modular structure, which is easy to adapt to topology changes, and to accommodate modern protection schemes. A preliminary evaluation of the operating performance of protective devices is also carried out after fault section identification. The testing results on the IEEE 14-bus power system and Zhejiang provincial power system in China demonstrate that the developed model is correct and efficient. Compared with three existing fault diagnosis methods, the proposed one has stronger fault-tolerance with lower computational cost, and is suitable for on-line fault diagnosis in large-scale power systems.