Model for Selecting Optimal Dispatching Rules Based Real-time Optimize Job Shop Scheduling Problem

Abstract

Owing to the randomness of the job release time, it is not possible to obtain all job information in real time during the operation of a manufacturing system. Generating a suitable scheduling strategy at the correct moment is the focus of addressing this disturbance. In this study, the flow time of the job in the manufacturing system was used as a criterion for evaluating the performance of the scheduling strategy. Subsequently, a model was constructed for selecting the optimal dispatching rule (DR) to actively change the scheduling strategies during the production process. The constructed model for selecting the optimal DR included an initial model for selecting the optimal DR and an evaluation model. The initial model for selecting the optimal DR outputs a DR with better performance based on the attributes of the job to be scheduled in the manufacturing system. Meanwhile, the evaluation model is responsible for evaluating the DR output of the initial model for selecting the optimal DR and determining whether the DR needs to be updated; the update process is realized based on simulation technology. Following experimental verification, the constructed model could generate scheduling strategies with superior performance in real-time and realize the update of the historical database. The results of this study will be of reference significance for solving the disturbance problem encountered by the manufacturing system in real time.