基于第一类错误的常规控制图检验模式优化  

Optimization of the Conventional Control Chart Inspection Model Based on Type I Errors

作　　者：李天淋 王天时 LI Tianin;WANG Tianshi(Beijing Sankuai Online Technology Co.,Ltd.,Bejing 100102;Beijing JiaoTong University,Beijing 100102)

机构地区：[1]北京三快在线科技有限公司,北京100102 [2]北京交通大学,北京100102

出　　处：《中国质量》2025年第3期123-127,共5页China Quality

摘　　要：本文针对质量管理工具常规控制图中的八条检验规则计算产生第一类错误的概率,发现原规则存在差异较大、不平衡的问题;运用蒙特卡洛法、概率论等计算不同数值假设下的第一类错误概率,提出和推荐了一组更平衡的取值,帮助企业降低第一类错误概率;面对常规控制图应用过程中可能产生的各种问题,提出了基于警报验证结果调节检验严苛程度的方法,动态平衡第一类错误和第二类错误,帮助企业降低损失。This article calculates the probability of Type I errors for the eight inspection rules in conventional control charts of quality management tools,and finds that there are significant differences and imbalances in the original rules.By employing Monte Carlo methods and probability theory,the article calculates the probabilities of Type I errors under different numerical assumptions and proposes a set of more balanced values,helping enterprises reduce the probability of Type I errors.In response to various issues that may arise in the application of conventional control charts,the article suggests a method to adjust the stringency of inspections based on the results of alarm verification,dynamically balancing Type I and Type Il errors,and assisting enterprises in reducing losses.

关 键 词：质量管理 常规控制图 控制图检验模式 蒙特卡洛法 

分 类 号：F27[经济管理—企业管理]