A New Method Based on Simulation-Optimization Approach to Find Optimal Solution in Dynamic Job-shop Scheduling Problem with Breakdown and Rework

Abstract

In this paper we propose an integrated algorithm based on combination of a discrete- event simulation and genetic algorithm. The simulation model is considered as a constraint-satisfaction procedure and if the streaming operations are initiated, then the meta-heuristic takes predefined steps to improve the solution. The latter is constructed through an interface, namely control matrix, implemented as interaction between the simulation model and refined solution of meta-heuristic. In run-time, the control matrix is accessed via simulation model for further modifications. The proposed method is implemented on classical job-shop problems with objective of makespan and results are compared with mixed integer programming model. Moreover, the appropriate dispatching priorities are achieved for dynamic job-shop problem minimizing a multi-objective criteria. The results show that simulation-based optimization are highly capable to capture the main characteristics of the shop and produce optimal/near-optimal solutions with highly credibility degree.

Keywords

• Genetic Algorithm
• Simulation based optimization
• Dynamic job-shop scheduling