A Fuzzy-Expert enhanced NSGA-II approach for sustainable agricultural systems

This study proposes a Fuzzy-Expert-NSGA-II algorithm, an enhanced NSGA-II approach incorporating fuzzy expert systems, for multi-objective optimization of agricultural planting strategies. The framework integrates expert rules, fuzzy mathematics, and evolutionary computation to establish a comprehensive optimization system considering economic benefits, ecological sustainability, and management efficiency. Using Chehe Village in Shanxi Province, China as a case study, we developed a multi-period crop planning model encompassing 41 crop types and 54 cultivated plots, incorporating constraints related to land types, crop rotation rules, and market uncertainties. The algorithm innovatively introduces a Hybrid Adaptive Local Search (HALS) mechanism and an Expert Rule-based Repair (ERB) module, significantly improving solution quality and feasibility. Experimental results demonstrate superior performance compared to standard NSGA-II, MOPSO, and MOEA/D, achieving better hypervolume indicator (HV = 0.892) and constraint satisfaction rate (1.2%). Simulation projections for 2024-2030 indicate that the optimized solution could increase average profits by 23% while maintaining high biodiversity levels (Simpson index 0.72-0.83). This research provides a data-driven decision support tool for agricultural land use planning, effectively balancing economic and ecological objectives.