A nonanticipatory policy for stochastic seru scheduling problems

This paper addresses the scheduling problem in seru production system (SPS) from stochastic dynamic optimization perspective, in which seru is one of successful new-type manufacturing modes arising from Japanese production practice. To minimize the weighted sum of completion times (TWCT), stochastic seru scheduling problems can be solved in polynomial time is showed, and a nonanticipatory scheduling policy pi is provided by means of a time-indexed linear programming relaxation. The performance guarantee depending on the squared coefficient of the processing time's variation is showed, and the upper bound of expected completion time of job is presented. Numerical examples are provided finally, the difference between deterministic and stochastic seru scheduling problems, along with the illustration of the nonanticipatory policy pi for stochastic seru scheduling problems are demonstrated. The results indicate that the stochastic seru scheduling problem is more complicated than the deterministic one, and the decision can be obtained according to pi based on the information up to now and a priori knowledge.