Research On Life Evaluation Of High Voltage Circuit Breaker Based On Big Data Technology

With the construction of UHV project and the proposal of energy Internet in China,the importance of health management for massive power equipment is increasingly prominent.As the core control and protection equipment in the power system,high-voltage circuit breaker is particularly important to ensure its reliable and stable operation state,which is essential for its condition monitoring and maintenance.However,the deviation of operation and maintenance time node exists in the condition based maintenance of high-voltage circuit breaker.In order to determine the best period of maintenance and reduce the failure rate of circuit breaker and the accident rate of power grid in a large range,this paper studies the law of degradation process in the life cycle of high-voltage circuit breaker,predicts and evaluates the mechanical life and electrical life of high-voltage circuit breaker,so as to provide closure for the maintenance strategy of high-voltage circuit breaker Rational advice.Firstly,this paper analyzes the working principle of hydraulic operating mechanism and arc extinguishing chamber of high-voltage circuit breaker,determines the influencing factors and evaluation criteria of mechanical life and electrical life of high-voltage circuit breaker,and provides theoretical basis for the construction of mechanical life prediction model and calculation module of electrical life.In addition,the Hadoop + Spark big data platform and its ecological components are built,the distributed file system HDFS is used to store data,the Spark is used to calculate data,and the comprehensive application in data collection and integration,data storage and management,data analysis and mining,and module development and design is carried out,which provides the platform foundation for the realization of life prediction and evaluation of high-voltage circuit breakers.In this paper,decision tree algorithm is used to classify and predict the health status of high-voltage circuit breakers,generalized linear regression algorithm is used to predict the mechanical life of high-voltage circuit breakers,and the training model is based on the mechanical life test data of high-voltage circuit breakers,which is verified and evaluated.The results show that the prediction accuracy of the decision tree classification model is about 0.97,while the improved tree model is a random forest algorithm model with AUC value of 0.97,which indicates that the classification model has a high fitting quality.The root mean square error of the generalized linear regression model is 52.96,R2 value is 0.92,which indicates that the deviation between the predicted value and the actual value of the model is small and the fitting effect is good.In addition,this paper proposes some methods to improve and optimize the algorithm model and spark components.Finally,in the aspect of electrical life analysis and evaluation of high-voltage circuit breakers,based on the relative electrical wear and relative electrical life method,this paper designs and develops the relative electrical life analysis and evaluation module,which mainly includes the source code programming of calculation formula,the packaging and release of the relative electrical life module,and develops the whole module as a third-party library to realize the call on Hadoop + Spark big data platform,so as to Realize the analysis and calibration function of relative electrical life of high voltage circuit breaker.