Markov 意思決定模型에 의한 最適工程統制에 관한 硏究

A study on the optimal control of bottle-neck process by markovian decision model

The main purpose of this paper study the problem of optimally controlling a discrete-time, bottle-neck process in the production process with countable state space which is subject to one of three control criterions at each observation time interval: produce, inspect while producing, or repair the bottle-neck process. The cost of the bad items to be produced and the inspection and repair costs are assumed dependent on the occurrence state of bad items in the bottle-neck process. When occurrence state of the bad items is followed Markov process at each observation time interval, actions corresponding to the each state can be chosen. Because we must follow any policy in order to choose actions, therefore, we can define a policy to be any rule choosing actions.