Power System Fault Diagnosis Methods With Temporal Constraints

Although the theory and technology of modern power system has been evolving, but the fault occurrence of electric power system still can’t avoid. Using the remote sensing, remote communication infomation during the electric power system equiment failure to estimate the fault sections, the malfunctioned protective relays and/or circuit breakers as well as the distorted/missing alarms, fault diagonis can shorten the time of fault handling, prevent accident expanding, accelerate system recovery. How to make full use of the rich and useful temporal information of alarm messages becomes an important problem to be solved.In this context, based on the time series data mining technique, a fault diagnosis method for power systems is proposed with temporal information of alarm messages taken into account. With the temporal information of alarm messages fully utilized, the proposed method can not only estimate the fault sections, but also identify the malfunctioned protective relays and/or circuit breakers as well as the distorted/missing alarms, even for the situations with complicated or successive faults.The existing power system fault diagnosis models employing multi-faceted monitoring and control information mostly implement information fusion in the decision-making level, and may lead to false diagnosis results when the received information is conflict or incomplete. The temporal information of alarms is only employed for preliminary screening, and the relevance between timestamps and information accuracy is not fully taken into consideration. Given this background, based on existing fuzzy Petri net models for power system fault diagnosis, an improved model for this purpose is presented employing electrical data and temporal constraints, and could accommodate the sequence of events (SOE) information from SCAD A, electric parameters as well as their temporal features from the wide area measurement system (WAMS). In the proposed fault diagnosis model, a multi-source information fusion technique is employed to integrate, analyze and process information from multiple sources. The confidence levels of events are evaluated by the relevant temporal information. The redundancy of multiple sources of information is used to verify the correctness of the received information and to estimate some important but missed information, and hence the accuracy and reliability of fault diagnosis results could be significantly improved.Finally, the structure of Petri net model for power system fault diagnosis and automatic generation algorithm is proposed. Also, a fault diagnosis system based on Petri net model is developed to demonstrate the correctness and efficiency of the fault diagnosis method.