Research on the Optimization of Multi-process Production Quality Inspection Strategy Based on Decision Tree and Comprehensive Index Scoring

Haoran Xiong; Lingpeng Zhang; Ruohan Kan

doi:10.54097/8hfd8e62

Authors

Haoran Xiong
Lingpeng Zhang
Ruohan Kan

DOI:

https://doi.org/10.54097/8hfd8e62

Keywords:

Multi-process, production, Decision tree, Comprehensive index scoring, Revenue maximization.

Abstract

This study aims to address the optimization of quality inspection strategies in multi-process production, with the goal of providing enterprises with a scientific decision-making method for quality management. This paper focuses on the complex manufacturing environment with multiple spare parts and multiple processes, and studies how to balance costs and benefits in quality inspection. Firstly, a comprehensive decision-making model including spare part inspection, semi-finished product inspection, finished product inspection, and disassembly decision is established, and cost and revenue functions are constructed. Secondly, the decision tree model is used to analyze the impact of different inspection strategies on costs and benefits, and the optimal strategy is screened through the comprehensive index score. The research results show that the strategy optimization based on the comprehensive index score function can effectively balance costs and benefits. Among the 8192 inspection strategies, the comprehensive score of the selected optimal strategy is 0.246, which significantly improves the revenue and reduces the defective rate. Finally, the feasibility and effectiveness of the model are verified through numerical experiments, and the robustness of the model under different defective rate conditions is analyzed. The model proposed in this paper provides theoretical guidance and practical reference for enterprises to optimize quality inspection strategies and achieve revenue maximization.

Downloads

Download data is not yet available.

References

[1] Guchhait R, Dey B K, Bhuniya S, et al. Investment for process quality improvement and setup cost reduction in an imperfect production process with warranty policy and shortages [J]. RAIRO-Operations Research, 2020, 54 (1): 251-266.

[2] Yang L, Shi Y Q. Multi-process Product Quality Prediction Based on Big Data [J]. Journal of Southwest University of Science and Technology, 2020, 35 (1): 81-89.

[3] Li X, Ban X, Yuan Z, et al. Time series prediction method and application based on deep learning in industrial scenarios [J]. Journal of Engineering Science, 2022, 44 (04): 757-766.

[4] Wang P, Qu H, Zhang Q, et al. Production quality prediction of multistage manufacturing systems using multi-task joint deep learning [J]. Journal of Manufacturing Systems, 2023, 70: 48-68.

[5] Ding J, Guo J. Prediction of Hot Strip Width Based on Principal Component Analysis and Random Forest Algorithm [J]. Journal of Northeastern University (Natural Science), 2021, 42 (09): 1268-1274-1289.

[6] Ge X L, Shi C D, Analysis and Decision Research on the Influence of Defect Rate on Closed-loop Supply Chain [J]. Logistics Sci-Tech, 2020, 43 (10): 137-143.

[7] Zhang Y L, Yan J Y. Fault diagnosis of bearing basedon GDPSO and decision tree [J]. Industrial Engineering Journal, 2021, 24 (6): 41-47.

[8] Kong J, Dai Y, Han M, et al. Nitrated and parent PAHs in the surface water of Lake Taihu, China: Occurrence, distribution, source, and human health risk assessment [J]. Journal of Environmental Sciences, 2021, 102 (04): 159-169.

[9] Filz M A, Bosse J P, Herrmann C. Systematic planning of quality inspection strategies in manufacturing systems [J]. Procedia CIRP, 2021, 104: 1101-1106.

[10] Zahidin M R, Yusof F, Rashid S H A, et al. Research challenges, quality control and monitoring strategy for Wire Arc Additive Manufacturing [J]. Journal of Materials Research and Technology, 2023, 24: 2769-2794.