Research On Behavior Law Of High Voltage Circuit Breaker

The rapid development of smart grids and the continuous expansion of power grids have placed higher demands on the safe and reliable operation of power equipment.As the most basic power equipment,high-voltage circuit breakers are closely linked to the safe and stable operation of the entire power grid.If the abnormal operation of the high-voltage circuit breaker can be detected in time or the running trend of the high-voltage circuit breaker is mastered,it will have a positive effect on the state maintenance of the high-voltage circuit breaker and the stable operation of the entire power system.With the development of information technology,computer technology and artificial intelligence technology,large-scale and abundant types of power system state data have been produced.Digitalization and informationization have become the key features of the future power grid.Data analysis techniques are also constantly improving.The research method based on data mining analysis is different from the traditional research method based on mathematical model.The analysis of large amount of data can be independent of the model and the hypothesis.The analysis of the correlation of state data can be independent of the model and the hypothesis.As long as there is correlation between the data,the new model,new knowledge,and even new laws that cannot be found by the traditional methods can be obtained through analysis and calculation.At present,many data analysis methods have been applied to various fields of power equipment state assessment.Power equipment fault prediction based on data mining analysis and power equipment state anomaly detection based on statistical analysis are typical application scenarios for power equipment state data analysis.Applying it to high voltage circuit breakers will be a meaningful exploration.There are many types of power equipment faults,but there are few fault samples.It is difficult to establish an accurate anomaly detection model and set an abnormality detection threshold under the lack of fault samples.Imprecise Probability(IP)is an effective method for processing non-complete sample information.It is more comprehensive and reasonable to describe the possibility of events in the form of probability interval under the lack of fault samples.The Bayesian Network(BN)can predict the failure probability of power equipment by mining the correlation between equipment state and state data.Based on this,this paper combines the imprecise Dirichlet model(IDM)with Bayesian network to predict the mechanical failure probability of high voltage circuit breakers under the lack of fault samples.This method provides a new idea for the reliability evaluation and state maintenance of the,mechanical mechanism.Also,the high-voltage circuit breaker will be affected by bad working conditions and its state will change abnormally.If the abnonnal operating state is not found in time,it will eventually turn into a fault state,affecting the safe and stable operation of the power system.Based on this,this paper uses statistical methods to detect abnormality of state data.The abnormal detection result can not only provide reference for maintenance decision,but also analyze the cause of the abnormal value,according to the different forms of abnormal detection results.