A Simulation Based Genetic Algorithm for Flowshop Scheduling Problem Considering Energy Cost under Uncertainty

A flowshop problem with objective functions of minimizing makespan and energy cost has been investigated. Reducing production costs is one of the goals that industries always have in mind. Increasing public awareness about the energy issues creates a new attitude toward minimizing energy costs. In order to make the problem more compatible with the real-world conditions, the problem is considered under uncertainty. An existing research gap inspired this study. It is assumed that machines can use the three slow, normal and fast speeds to process jobs. At high speeds, consumption rate increases and completion time decreases, and vice versa. The difference in machine processing speeds yields different and contradictory values in the objective functions. Therefore, a method should be proposed in which, in addition to the order of jobs, the speed of machines could be determined. A mathematical model is presented, and then a simulation-based genetic algorithm is used to solve the problem on a large scale. Simulation is used for each evaluation of the objective function in the genetic algorithm to consider the uncertainty of processing times. Due to the stochastic processing time, the expected value model is used to deal with uncertainty. The computational results indicate that the algorithm and approach show a good performance.