Two stage hierarchical strategy based stochastic job shop scheduling with random processing times

In the real-life manufacturing environment, even the schedule with optimal efficiency measure would deteriorate due to uncertainty. This problem could be optimized when stability measure is employed as an important evaluation criterion of the candidate schedules. This work studies the stochastic job shop scheduling problem subjected to the disturbances of random processing times. To address the disturbances, a proactive scheduling strategy is adopted, which include both the efficiency measure and stability measure. A two stage hierarchical strategy based on the ordinal optimization genetic algorithm (OOGA) is designed to solve the stochastic optimization problem. In the first stage, the stochastic dominance based on expectation is applied to optimize efficiency measure; in the second stage, the variance of is applied as the basis of the stability measure. It aims to provide a satisfied near optimal schedule for the decision makers by considering both efficiency and stability simultaneously. According to the preliminary simulation results, the significance of studying the stability performance of the schedule subjected to random processing times is discussed, and the feasibility of the proposed hierarchical strategy and the algorithm is proved.