BI_Apriori Algorithm: Research and Application Based on Battery Production Data
Ning Zhou, Mengshi Qiao, Jianxin Zhou · 2019
As the problem that the traditional Apriori algorithm has low mining efficiency and generates a large number of irrelevant rules, combined with the characteristics of the process parameters of the battery industry automated production process, an association rule algorithm for battery automation industrial production is proposed--BI_Apriori algorithm. The performance of the algorithm is optimized in the following two aspects: (1) The production process data is stored in the form of a matrix; (2) Delete the item set of the candidate set that does not contain the battery qualification rate parameter (zYsNG) in the connection step. The simulation results show that compared with the traditional Apriori algorithm and Apriori-Tid algorithm, BI_Apriori algorithm has better performance in mining efficiency, generation rule number and generation rule correlation. The algorithm can mine the relationship effectively between process parameters and product qualification rate in the battery industry automation process.