A PSO-based hybrid metaheuristic for permutation flowshop scheduling problems.

Area of Science:

• Operations Research
• Computational Intelligence
• Combinatorial Optimization

Background:

• The permutation flowshop scheduling problem (PFSP) is a complex combinatorial optimization challenge.
• Minimizing makespan and total flowtime are critical objectives in manufacturing and production scheduling.
• Existing metaheuristics often face challenges with exploration and diversification in complex search spaces.

Purpose of the Study:

• To develop an efficient hybrid metaheuristic for solving the PFSP.
• To address the dual objectives of minimizing makespan and total flowtime.
• To improve the performance of particle swarm optimization (PSO) for scheduling problems.

Main Methods:

• A hybrid metaheuristic combining Particle Swarm Optimization (PSO) with Simulated Annealing (SA) and Stochastic Variable Neighborhood Search (SVNS).
• Incorporation of SA and SVNS to enhance the exploration capabilities of the PSO algorithm.
• Implementation of a path relinking strategy for solution diversification and escaping local optima.

Main Results:

• The proposed PSO-based hybrid metaheuristic demonstrates competitive performance on benchmark PFSP instances.
• The integration of SA and SVNS effectively improved the exploration of the search space.
• The path relinking strategy enhanced solution diversification, preventing premature convergence.

Conclusions:

• The developed hybrid metaheuristic offers a promising approach for the permutation flowshop scheduling problem.
• The synergistic combination of PSO, SA, SVNS, and path relinking yields superior results.
• The method is effective in minimizing both makespan and total flowtime in complex scheduling scenarios.