Distributed Flow Shop Scheduling for Heterogeneous Serial Lines with Dueling Double DQN Improved Discrete Particle Swarm Optimization

With the development of technology and the changing market demands, the influence of distributed small-batch flexible production modes in the manufacturing industry has been expanding. Meanwhile, production lines with random failures and finite buffer capacities are receiving academic attention due to their closer resemblance to real-world scenarios compared to fault-free production lines. Therefore, a distributed flow shop scheduling problem (DFSP) for the heterogeneous serial lines system model with Bernoulli machines is formulated in this paper. Considering the due time constraint of different jobs in real production, this scheduling problem involves optimizing both the total completion time and the total tardiness of the system. A novel velocity-free discrete particle swarm optimization (DPSO) Algorithm, enhanced by Dueling Double Deep Q Network (D3QN), is proposed in this paper to solve the aforementioned multi-objective problem (MOP). Specifically, during the discrete particle swarm search process, D3QN is employed to intelligently select exploration directions and local search strategies based on the current state information of the population. This approach enables the attainment of a high-quality Pareto front within a limited optimization process.