中文 | English

J. Hefei Univ. Tech. (Nat. Sci.)

合肥工业大学学报(自然科学版)

JOURNAL OF HEFEI UNIVERSITY OF TECHNOLOGY (NATURAL SCIENCE)

Competent Authorities: Ministry of Education of the People's Republic of China

Sponsor by: Hefei University of Technology

ISSN 1000-1646, CN 34-1001/N

Research on quality control method of clutch manufacturing process based on random forest

Download Full Text

基于随机森林的离合器制造过程质量控制方法研究

Journal Information

J. Hefei Univ. Tech. (Nat. Sci.), Nov.2023, Vol.46 No.11: 1441-1446, 1500

DOI: 10.3969/j.issn.1003-5060.2023.11.001

Authors

SHEN Weilei, ZHANG Xinyang, YANG Xuechun

(School of Mechanical Engineering, Hefei University of Technology, Hefei 230009, China)

Abstract and Keywords

Abstract: As the core tool of statistical process control (SPC), the traditional control chart has limitations in terms of the recognition of systematic factors that affect the manufacturing process. Aiming to solve the problem of insufficient systematic factors recognition of the traditional control chart, this paper proposes a control chart pattern recognition model based on random forest. The grid search method is used to optimize the parameters, and a model based on random forest algorithm flow and control chart pattern recognition is established to identify the systematic factors that affect the process out of control. Taking the automobile clutch as an example, the pattern recognition algorithm based on random forest is applied to the clutch manufacturing process, and then compared with the support vector machine (SVM), logistic regression (LR) and other machine learning methods to verify the feasibility and effectiveness of the random forest model for the control chart pattern recognition.

Keywords: statistical process control (SPC); random forest; clutch; pattern recognition; machine learning

作者信息

沈维蕾，张鑫洋，杨雪春

(合肥工业大学机械工程学院, 安徽合肥 230009)

摘要和关键词

摘要: 传统控制图作为统计过程控制(statistical process control, SPC)的核心工具, 对影响制造过程的系统性因素识别效率较低。文章针对传统控制图对系统性因素识别不充分的问题, 提出基于随机森林的控制图模式识别模型, 利用网格搜索法进行参数优化, 建立基于随机森林算法流程和控制图模式识别模型以识别影响过程失控的系统性因素; 以汽车离合器为例, 将基于随机森林的模式识别算法应用到离合器制造过程中, 并与支持向量机(support vector machine, SVM)、逻辑回归(logistic regression, LR)等机器学习算法相比较, 结果验证了随机森林模型对控制图模式识别的可行性和有效性。

关键词: 统计过程控制(SPC)；随机森林；离合器；模式识别；机器学习

Funding

国家自然科学基金资助项目（51975003）

Next

Back to Contents