Research and Application of Association Rule Mining Algorithm for Electric Power Equipment Manufacturing

The electric power equipment (EPE) manufacturing industry is currently facing challenges related to technological dependency and insufficient automation, limiting improvements of the production efficiency and product quality. The objective of this paper is to propose an improved association rule mining algorithm to extract the potential relationships among production elements of EPE manufacturing. First, the Random Forest algorithm (RF) is employed for feature dimensionality reduction to enhance computational efficiency. Additionally, the k-means clustering algorithm is used for data discretization to better handle continuous and multi-modal data. Finally, a support penalty for low-frequency itemsets is introduced to improve the algorithm's adaptability and accuracy in EPE manufacturing environments. Experimental results indicate that the improved association rule mining algorithm outperforms traditional algorithms in rule generation efficiency, rule quality, and computational time. Furthermore, the algorithm could be used to analyze and identify defects and faults, etc., in EPE production lines to improve production efficiency and product quality.