Multiobjective flexible job-shop scheduling optimization for manufacturing servitization

Abstract

Purpose With the development trend of China’s service-oriented manufacturing moving toward intelligence and personalization, the deep integration of manufacturing and service has become a synergistic challenge for enterprises. Design/methodology/approach An improved migratory bird optimization (IMBO) algorithm is proposed to solve the multiobjective FJSP model. First, this paper designs an integer encoding method based on job-machine. The algorithm adopts the greedy decoding method to obtain the optimal scheduling solution. Second, this paper combines three initialization rules to enhance the quality of the initial population. Third, three neighborhood search strategies are combined to improve the search capability and convergence of the solution space. Furthermore, the IMBO algorithm introduces the concepts of nondominated ranking and crowding degree to update the population better. Finally, the optimal solution is obtained after multiple iterations. Findings Through the simulation of 15 benchmark studies and a production example of a furniture enterprise, the IMBO algorithm is compared with three other algorithms: the improved particle swarm optimization algorithm, the global and local search with reinitialization-based genetic algorithm and the hybrid grey wolf optimization algorithm. The experiment results show the effectiveness of the IMBO algorithm in solving the multiobjective FJSP. Practical implications The study does not consider the influence of disturbance factors, such as emergency interventions and equipment failures, on scheduling in actual production processing. It is necessary to further study the dynamic FJSP problem. Originality/value The study proposes an IMBO algorithm to solve the multiobjective FJSP problem. It also uses three initialization rules to broaden the range of the solution space. The study applies multiple crossover strategies to avoid the algorithm falling into local optimality.