Multi-start Tabu Agents-Based Model for the Dual-Resource Constrained Flexible Job Shop Scheduling Problem

Typically, processing jobs on a production floor require both machines and human resources. However, most classical scheduling problems ignore possible constraints caused by the availability of workers and treat machines only as limited resource. This paper presents a Multi-Start Tabu Agents-based Model (MuSTAM) for Dual-Resource Constrained Flexible Job shop Scheduling Problem (DRCFJSP). It considers a set of initial solutions running in parallel using the intensification technique. It has a single objective which is to minimize the maximum completion time (makespan) due to its importance in research workshops. The proposed model consists of two classes of agents: MainAgent and TabuAgents. The MainAgent receives inputs, generates the initial population, creates TabuAgents based on the number of solutions in the initial population PopSize, launches the system and finally displays the best solution. Each TabuAgent takes a solution from the created initial population and applies Tabu Search using the technique of concentrated intensification to neighborhood search. TabuAgents cooperate and communicate between them in order to improve the search quality. In experimental phase, numerical tests are performed to evaluate our MuSTAM model compared to ITS based on FJSPW benchmark instances of Gong. The obtained results show the efficiency of the Multi-Start Tabu Agents-based Model in terms of makespan and CPU time.