Research On Correlation Factors Of Motor Quality Based On Jackknife Bayes Method

With the rise of industrial "4.0" and "manufacturing leader" and the rapid development of cloud storage technology,the data of the manufacturing industry is becoming more perfect.The manufacturing industry is urgently required to explore the way to improve the quality of the product from mass data.The traditional methods of quality analysis qualitatively describe the quality impact factors based on the electrical design,mechanical structure and manufacturing process.These methods cannot meet the high standards for the manufacturing industry in the new era.From the point of data mining and knowledge discovery,this paper aims at mining the potential quality knowledge rules from mass detection data.Research on association rules is one of the hotspots in data mining.It mainly studies the relationship between data sets.Although the classical Apriori algorithm can effectively search for the frequent sets,it read the database over and over.And the efficiency of operation is very low.The improved algorithms,such as FP-Tree and FUP,occupy so much memory space and continue to access disk,and the efficiency is still low.The method based on bayesian estimation is of high efficiency,but its accuracy is poor.This paper focuses on the above problems and proposes a method of association rules based on jackknife bayesian estimation.Firstly,calculating the optimal probability distribution function based on jackknife bayesian estimation.Then searching for 1-term frequent sets.Still combined with the core idea of Apriori algorithm,generating quantitative association rules.And incremental updating algorithm is proposed for continuous incremental update of product data in practical applications.Combined with the knowledge of conjugate prior distribution,the initial sample distribution is regarded as a prior information,and the new data is regarded as the sample.And so the variables distribution parameters is renewed to generate quantitative association rules efficiently in real time.Finally,the effectiveness and superiority of the method are verified by simulation.And it is applied to the correlation analysis about the detected data.Ultimately,it catches the relationship between the motor production index,the assembly index and the electrical index.