An Improved NSGAII Algorithm for Flexible Job-shop Scheduling Problem Based on A New Decoding Mechanism

The flexible job-shop scheduling problem (FJSP) is a famous combinatorial optimization problem. FJSP is widely used in process manufacturing industry, so it is of great significance to study FJSP to improve production efficiency. In this paper, a non-dominated sorting genetic algorithm II based on a new decoding mechanism (NSGAII_NDM) is proposed. Firstly, the opposite-based learning strategy is introduced on the basis of NSGAII, and the sufficiency of global search is improved by generating the opposite value. Secondly, an improved adaptive crossover and mutation probability strategy is added, which can dynamically adjust the inherent crossover and mutation probability with the number of iterations and improve the convergence speed of the algorithm. Furthermore, based on fully active scheduling strategy, a new decoding mechanism is proposed, which can get a better scheduling scheme by arranging the machines and starting time of each process more reasonably. Finally, the improved algorithm is tested in a standard test case, which verifies the superiority of NSGAII_NDM in solving the flexible job-shop scheduling problem.