Reliability Evaluation Of Power Grid Based On Inexact Probability

In the context of the continuous transformation of China's power grid structure to UHV,large capacity,long-distance and regional networking,and the continuous improvement of people's production and living standards,all walks of life in society have higher requirements for the reliability of power system operation.Power grid reliability assessment is an important link to measure the overall operating status of the system and check the system reliability level.Its evaluation results will directly affect the planning,design,and operation of the power grid.However,with the wide application of new equipment and new technologies in the power grid,many new components lack historical fault statistics.Under small sample conditions,limited statistical data should make limited estimates of component reliability parameters.The paper is based on improved The accurate Dirichlet model(IDM)uses limited data to perform inaccurate interval probability estimation on the reliability parameters of power components,and completes grid reliability assessment and non-coherent component identification based on inexact probabilities.main tasks as follows:(1)Discusses the difference and correlation between the precise point probability and the imprecise interval probability,and studies the influence of the value of the hyperparameter s on the accuracy and credibility of the imprecise probability estimation in the imprecise Dirichlet model,in order to balance Due to the contradiction between the accuracy and credibility of the estimation results,an improved inaccurate Dirichlet model based on dynamic hyperparameter s is proposed.(2)Introduces the non-coherent characteristic of reliability in power systems,clarifies the two types of characteristics and the defects of traditional identification methods,based on inexact interval probability and improved particle swarm optimization algorithm,based on the expected power shortage EENS index A new method for quickly identifying the non-coherent components of system reliability is proposed,and the effectiveness of the method is verified with a 15-node test system.(3)Based on the non-precision interval probability,considering the non-coherentcharacteristics of reliability,the improved sequential Monte-Carlo simulation method was used to complete the grid reliability evaluation of the IEEE-RTS79 test system,and the non-precision reliability index interval was calculated.value.And a new algorithm for identifying non-coherent components of reliability is used to obtain the association table of non-coherent components of the system.On this basis,the component importance analysis and maintenance decision evaluation of the test system were carried out,and the difference between the component importance ranking table and the maintenance decision evaluation before and after considering the non-coherent characteristics of reliability was compared.