Research On Decision Support System Of Power Grid Fault Information Analysis In Jinhua Region

Power grid failure occurs, how to rapidly and accurately determine the cause of the problem, removal of fault equipment and to recover the normal operation of power grids, to improve the reliability of power supply is of great significance and sustainability. Along with the development of the automation system, substation all information to the dispatching control center have been collected. In the event of power grid failure, a large number of fault alarm information up to the control center. Especially in bad weather, frequent accidents, a huge amount to upload information constantly brush screen, control personnel it is difficult to capture the precise and in the first time a decision-making reference information, increased the complexity of the judgment and handling of accidents, to control the working brings many problems. Therefore, developing a grid fault information analysis and decision system is very necessary. Auxiliary control personnel in accident cases to fault location quickly and accurately and provide decision-making processing opinion, greatly shorten the processing time of the accident, quickly restore power grid safe and stable operation.This article first to power system fault diagnosis research at home and abroad as a starting point, combing the different methods of fault diagnosis theory knowledge. After analyzing the advantages and disadvantages of different methods were compared, finally chose the application of combining expert system and artificial neural network method. Artificial neural network system based on the numerical computation of reasoning is introduced into the expert system, and make the expert system reasoning and artificial neural network self-learning and the complementary advantages of fast operation ability, suitable for the construction of jinhua power grid fault information analysis and decision system.Based on the grid fault information analysis and decision system is proposed in the design of system function structure and the overall structure framework, and the knowledge acquisition and reasoning machine, explanation mechanism, knowledge base and so on has carried on the analysis and research, puts forward the design idea and relevant requirements. Paper presents a method of knowledge acquisition based on artificial neural network, the reasoning of positive and negative two-way reasoning strategy, explanation mechanism adopts the method of path tracking. Knowledge base with relational database establishment and maintenance, simple questions by the expert system through explicit knowledge base for processing, complex problems by the trained artificial neural network system after learning through the tacit knowledge base for processing. And carried out in front of the knowledge base maintenance scheduling console window information optimization work, the information standard naming, latency, shielding, directly put in storage, reduce the amount of power grid fault diagnosis system information processing. This paper focuses on the construction of the neural network model, select the 110kV lines of 220kV Huajin substation as part of the object in accordance with the corresponding test, prototype sample input neural network to learn the value of training repeatedly and correct connection weights and threshold parameters and target output value, the output to the experts in the field of target value. The test basically meet the needs of power grid fault information analysis and decision support functions, to achieve the intended objectives.This paper mainly analyses the grid fault information decision system on the structure and function of the study, proved that the system based on expert system and artificial neural network of power grid fault information diagnosis method is effective. This paper carried out some useful exploration and attempt, related work remains to be further in-depth study and practice.