Abstract: Aiming at the hybrid flow shop scheduling problem, an improved multi-objective salp swarm optimization algorithm based on Q-learning is proposed, simultaneously considering the minimization of makespan and equipment processing energy consumption. To enhance the convergence speed of the algorithm, a strategy combining chaotic mapping and heuristic rules is employed to generate diversified initial populations. To balance the global search capability and local exploitation capability of the algorithm, a Q-learning adaptive selection strategy is introduced in the selection of leaders proportion. To improve the optimization accuracy of the algorithm, an effective variable neighborhood search strategy is proposed to strengthen the local exploitation capability. Experimental validations conducted on public datasets demonstrate that the proposed algorithm can effectively solve the multi-objective optimization problem in hybrid flow shop scheduling.

Key words: multi-objective hybrid flow shop scheduling, salp swarm algorithm, Q-learning, chaotic mapping, variable neighborhood search