Combining Genetic Algorithms and Supervised Learning for Enhanced Efficiency

This chapter presents a hybrid optimization methodology combining Genetic Algorithms (GAs) and supervised learning to enhance production flow in reconfigurable manufacturing systems (RMSs). The proposed approach leverages the adaptive capabilities of neural networks to guide the evolutionary process of GAs, improving solution feasibility and quality. The method addresses multi-objective optimization challenges, focusing on minimizing production costs and time while maximizing energy efficiency. Experimental results demonstrate that the hybrid GA achieves superior performance compared to traditional GA methods, with a 25% reduction in computational time and a 30% improvement in solution quality. The Pareto front analysis highlights better trade-offs between conflicting objectives, offering decision-makers a broader spectrum of optimal solutions. Furthermore, task scheduling results show a significant decrease in idle machine time, optimizing resource utilization by 35%. A detailed case study in a flexible workshop illustrates the practical application of the methodology, showcasing its scalability and adaptability to dynamic production environments. These results underscore the hybrid GA’s robustness and efficiency in addressing the complexities of RMS scheduling. Compared to existing methods, this approach delivers faster convergence and higher-quality solutions, establishing its potential as a novel contribution to intelligent optimization in modern manufacturing systems.