Power System Fault Evolution Modeling And Analysis

With the enactment and promulgation of the “13th Five-Year Plan” for the regulation and operation of the State Grid,it is required to carry out big data analysis based on equipment operation,research equipment monitoring operating state forecasting technology based on big data technology,and establish equipment accounting information database to realize equipment associated analysis of defects,equipment alarms,and accounting records of the ledgers,etc.Finally,a state monitoring and trend forecasting model for monitoring equipment is built.The establishment of the fault model of the power system is of great significance for the accurate fault prediction of the equipment.It is imperative to adapt to the rapid development of the power grid by establishing a systematic fault analysis mechanism,constructing a fault evolution mechanism model,and optimizing the decision-making process and management mode for equipment maintenance.This paper analyzes the actual alarm signals of the power system,and proposes an iterative optimized sequence pattern mining algorithm to generate frequent sequence patterns,and stores the extracted sequence patterns hierarchically in the sequence template database.By introducing the idea of gene alignment in biology into the grid fault sequence model,the sequence model is optimized and adjusted,the homology of the fault sequence is analyzed,the key nodes in the sequence model are identified,and the algorithm for the combination comparison is designed to realize the comparison between the sample data and the template database,and the correct classification assessment of sample data.Finally,the Bayesian network is used to reconstruct the generated sequence model,and the fault model is described by the probability map model.This paper introduces in detail the symbolization of the text signal in the sample data,the selection of the window function parameters,and the characteristic distribution of the sequence of the weighted sequence model library under each iteration.This article also introduces the design of replacement matrix based on grid alarms in the sequence comparison process,judging the categories of input sample sequences and signal state evaluation based on the comparison results,and the initial Bayesian construction through sequence templates and expert knowledge.In the end,the nodeprobability adjustment is implemented according to the subsequent data.In practical applications,an alarm signal mining system based on the power big data platform is established to implement sequence mining,comparison,and early warning modules,which has certain reference significance for the analysis of the system fault evolution mechanism.