Fault Diagnosis Model And Recovery Strategy For Intelligent Substation

With the social and economic development, people and enterprises demand highly service of the power grid increasingly. Therefor, more and more researcher pay attention on the reliability of substation which play a important role in the smart gird operation normally. Beside, the management and maintenance of substations also become more important. However, due to the human factor, environment, device itself, substations suffer different type of disturbance during their operation, which result to a variety of types of failures, even lead to serious incident. So the substations fault handing is a important task of establishing a high-performance grid infrastructure.The manual fault handling is hard to meet the real-time requirement, because of the unpredictability and irregularity. Beside, security devices produce a lot of information when a fault occur which lead to getting useful law in these information become much trouble. Thus, a real-time, automatic fault handling model can provide useful information support for substations managers who can talk action to make system quickly return to work after it suffer a failure, and reducing the economic losses.At present, much researcher have proposed a variety knowledge of detail program of substations fault handling, they don't describe the process of fault handing in the perspective of the system ontology. This paper divides the process into three continuous and associated parts, fault alarm signal operation, fault diagnosis, fault recovery, and theirs relations. This specific work as follows:1) Describe the substation fault handling from the system level, and divide the process into three relate parts, alarm signal sample quality improvement stage, substation fault diagnosis stage, substation fault recovery stage. And describes this three stage content and their relationship in detail, these step form a close-loop system.2) Because of false positives or redundancy of alarm signals, we analysis the generation and propagation of alarm signals, and improve the accuracy and the streamlining of the sample in this two stages. Adopt signal shaking mechanism in the sample acquisition, add filters and integration mechanism in the transmission process, so we can greatly improve the quality of the signal samples, which can provide a high quality knownledge to fault-handing, then ensure the effectiveness of followed algorithm.3) After receiving the alarm signal sample, we use special method to handle the content of alarm packet, in order to obtain the type of fault, the segment which fault occur, the information of fault devices. In this process, we use fault tree method to model system and then use Petri Net knowledge to diagnose. We used this algorithm in a 220KV Bus in Donghulong Substation locked in JieXiu City, then we compared the calculation with expert inference, and the consequence verify the effectiveness of our algorithm.4) Decision-makers based on the diagnostic result and monitoring information to establish a recovery mechanism of substation. In this paper, we used the Petri Net knowledge to model the recovery process and get the operation of equipment and task order form, we also implicated this recovery algorithm in Donghulong interlligent substation, and verify the accruacy and real-time of our algorithm...