The Failure Rate Level Evaluation Of Intelligent Electricity Meters In A Typical Environment Based On Apriori Association Rules

This dissertation proposes a method of association rules based on Apriori algorithm analysis,to improve the algorithm’s support-confidence measure IS the framework,apply it to the watt-hour meter failure rate analysis,based on MATLAB simulation platform,to build a typical environment influence factors-watt-hour meter grade evaluation model of failure rate,failure rate analysis for typical environmental impact factors and watt-hour meter level,the relation between to select intelligent watt-hour meter manufacturers in a typical environment and provide a reference basis.Firstly,this dissertation introduces the research background and significance of failure rate rating of intelligent electricity meters in typical environments,and provides a theoretical basis for the necessity of building a failure rate rating system for electricity meters.The failure mechanism of intelligent electricity meters in typical environment is studied and the current situation of failure rate evaluation of electricity meters in foreign countries is expounded,which lays a theoretical foundation for the design method of failure rate rating evaluation system of electricity meters in typical environment.Secondly,the existing assessment methods of failure rate grade are introduced,and the applicable conditions and advantages and disadvantages of each method are compared and analyzed.This dissertation focuses on the basic principle of association rule analysis and summarizes its steps.Are introduced in detail,and various mining algorithm of association rules analysis,analysis of the existing advantages and disadvantages and applicability of association rule algorithm,an Apriori algorithm is proposed,b ased on the intelligent watt-hour meter raw data grouped into experimental group and the control group and to discretization of data,the typical environment failure rate and the factors affecting the grading,in order to achieve the purpose of reducing no ise value,the influence of invalid values.The framework of support degree and confidence degree is constructed to explore the deep connection between typical environmental factors and failure rate.Rule pruning can filter strong association rules and reduc e the running cost of analyzing data.According to the actual needs,the threshold values of minimum support degree and minimum confidence degree can be adjusted flexibly,and the association rules can be screened to achieve the artificial control of the ru nning times of the algorithm.Then,the Apriori algorithm simulation model was built and an example was analyzed to obtain effective association rules.The analysis results obtained typical environmental influencing factors and failure efficiency.The resul ts were compared with the control group data to verify the feasibility of the algorithm.Finally,the limitations of Apriori algorithm are studied.In order to obtain more comprehensive and accurate results,the improvement scheme of Apriori algorithm is given.Firstly,it introduces the other objective measures commonly used in association rules,studies the nature of objective measures and analyzes their applicability.Selection IS to measure the added support-confidence framework of original Apriori algorithm,improved Apriori algorithm simulation model was constructed,the factors affecting the draw IS measured as the standard of arrange table-failure rate level IS measured,and the results of correlation analysis,contrast did not improve the results of the algorithm,to verify the superiority of the improved algorithm.Simulation and evaluation results show that the Apriori association rules-based intelligent electricity meter failure rate evaluation method proposed in this dissertation can analyze the internal relationship between typical environmental factors and failure rate of intelligent electricity meter.Compared with other evaluation methods,it has the characteristics of low computing cost,high mining accuracy and high computing efficiency.