针对多维数据的贝叶斯更新难题,首先利用变量转换将多变量的贝叶斯更新公式转变为单变量更新公式.在对过程状态进行重新划分的基础上,利用马尔科夫链分别构造了不同情形下的贝叶斯过程状态转移矩阵,区分了扰动发生的不同时刻对于成本函数的影响.利用迭代的思想构造了过程的成本函数,在此基础上最终构造了面向小批量生产过程的多变量控制图优化模型.仿真结果表明,相比于已有的多变量控制图,本文设计的控制图可以有效地降低过程成本,因此具有更大的实用价值.
Dealing with the Bayesian updation of multivariate data, we convert the multivariate Bayesian updation into a univariate one using variable conversion. Based on new partition of process state, process transition matrices under different scenarios were constructed employing Markov chain method. Thus, the influence of the disturbance occurrence time on process cost was taken into consideration. A process cost function was constructed iteratively, and finally a multivariate control chart optimization model for short-run production was formulated. Simulation results show that, compared with the existing multivariate control chart, the control chart proposed in this paper can reduce the process cost effectively, therefore has greater application value.