A Study on the Dynamic Weight-Driven AHP-NSGA-II Framework for High-Dimensional Multi-Objective Optimization

Abstract

This paper investigates an efficient optimization framework for complex systems through the synergistic integration of the Analytic Hierarchy Process (AHP) and Non-dominated Sorting Genetic Algorithm (NSGA-II). The proposed methodology addresses multi-criteria decision-making challenges by systematically resolving competing priorities while maintaining solution quality. AHP is leveraged to derive objective weights through structured pairwise comparisons, while NSGA-II identifies Pareto-optimal solutions that optimize system-wide performance metrics. Computational experiments reveal that the hybrid approach achieves superior optimization outcomes with enhanced operational efficiency compared to conventional optimization techniques, particularly in scenarios requiring trade-off analysis between divergent objectives. The framework demonstrates versatility across diverse application contexts including industrial process design, logistics network configuration, and infrastructure development planning. Key innovations include a dynamic weight adjustment mechanism and an adaptive fitness evaluation process that collectively improve solution convergence characteristics and decision space exploration capabilities.