A Solution to Single-Machine Inverse Job-Shop Scheduling Problem

Abstract

Concerning the inverse job-shop scheduling problem (JSP), this paper proposes a hybrid solution based on genetic algorithm (GA) and improved particle swarm optimization (PSO), with the aim to minimize the parameter adjustment. The solution was presented as a block coding plan with decimal mechanism, under which both processes and parameters can be optimized simultaneously. To enhance the local search ability of the proposed algorithm, four neighbourhood structures were designed, and an adaptive selection mechanism was created to select the most suitable neighbourhood. Finally, the proposed algorithm was proved valid through discrete event simulation (DES) and comparison with other algorithms.

A Multi-Objective Optimization Model Based on Non-Dominated Sorting Genetic Algorithm

“…The completion time and cost are the top concerns of the scheduler. Below is the mathematical description of the JSP [19].…”

Optimization of welding job-shop scheduling problem under variable workstation constraint: an industrial application with Arena simulation based genetic algorithm

“…Liao and Lin [19] studied on optimization of jobshop supply chain scheduling problem using PSO. Wang et al [20] used hybrid GA-PSO for inverse JSSP and performed discrete event simulation. Zhang et al [21] used PSO and ANN together for JSSP.…”

“…Some researchers [24] used fuzzy processing times, however in this study GA aided Arena simulation is used and probability distributions of processing times are suitable for this software. GA MR-FJSSP Makespan Deng et al [6] Bee evolutionary guiding -NSGA MO-FJSSP Total workload of all machines, makespan, workload of the most loaded machine Ocaktan et al [7] GA, Arena simulation Stochastic JSSP Makespan Zhang et al [8] Multi-population GA FJSSP Makespan, load of each machine, load of all machines Hu et al [9] Improved CSA JSSP Makespan, Jiang et al [10] GWO with double-searching mode [20] Hybrid GA-PSO, discrete event simulation…”