摘要
针对存在设置偏差且生产阶段有限的多变量过程,研究了调整费用恒定情形下的使过程质量损失最小的边界调整策略.在建立过程状态空间方程的基础上,利用贝叶斯推断和动态规划给出了调整策略中随生产阶段变化的边界.通过算例解释了边界调整策略的实现方法,并通过仿真将调整策略与其他两种质量控制策略进行了比较分析,仿真结果表明调整策略能够更好的减少过程的总体质量损失.
Aiming at the finite-horizon multivariate process with setup error,the deadband adjustment scheme to minimize the total process quality loss for the situation of adjustment with fixed cost is developed. Based on the state-space process-control model,the time-varying deadband form of this adjustment scheme is derived by using Bayesian inference and dynamic programming.A simulation case is presented to illustrate the implement method of the optimal adjustment policy.Furthermore,the deadband adjustment scheme is compared with other two quality control policies by simulations,the results show that the scheme proposed in this paper can better reduce the total quality loss.
出处
《系统工程理论与实践》
EI
CSSCI
CSCD
北大核心
2010年第3期538-542,共5页
Systems Engineering-Theory & Practice
基金
国家自然科学基金(70672088
70931002)
关键词
多变量过程
经济设计
统计过程控制
边界调整
贝叶斯推断
multivariate process
economic design
statistical process control
deadband adjustment
Bayesian inference
