Process planning issues based on the integration of NSGA-II and reinforcement learning

This paper focuses on the multi-objective flexible process planning and scheduling (IPPS) problem in the enterprise production process. Given its NP-hard characteristics and the difficulties faced by traditional methods in solving it, a solution algorithm integrating the non-dominated sorting genetic algorithm-II (NSGA-II) and reinforcement learning is proposed. Through innovative coding methods, population initialization, genetic operations, and dynamic parameter adjustment strategies, the performance of the algorithm is effectively improved. Experiments show that the algorithm is accurate and efficient in finding multi-objective optimal solutions, significantly reducing the number of crossovers and mutations and the computation time. It provides an effective approach for solving such problems in intelligent production and enhances the adaptability of decision-making and the operability of scheduling.