Multi-Objective Optimization and Allocation of Water Resources in Hancheng City Based on NSGA Algorithm and TOPSIS-CCDM Decision-Making Model

Intelligent algorithms and decision models are key tools for improving the efficiency and adaptability of multi-objective optimization and allocation, and for achieving sustainable utilization of water resources. This study takes Hancheng City as a case study to develop a water resource optimization allocation model based on economic, social, and ecological benefits, analyzing and predicting the supply and demand of conventional and unconventional water resources in the study area. The model is solved using the NSGA algorithm, and solutions are screened from the Pareto front using the TOPSIS-CCDM two-level decision model, with the RSR method used for comparative verification. The results show that the schemes II-2022-21 (water shortage of 17,802.35 m3/d, economic benefits of 21,019,556.17 yuan, pollutant emissions of 745.92 tons), II-2027-ACS (shortage of 14,098.76 m3/d, economic benefits of 29,401,252.75 yuan, emissions of 712.07 tons), and II-2032-ACS (shortage of 12,709.33 m3/d, economic benefits of 36,660,367.83 yuan, emissions of 700.96 tons) are in line with the water resource allocation planning for Hancheng City before 2035. These schemes not only meet the regional planning requirements but also maximize economic benefits while minimizing water shortages and pollutant emissions. The study finds that NSGA-II has an advantage in selecting more coordinated schemes, while NSGA-III focuses more on the selectivity of specific targets. Although the TOPSIS-CCDM model performs well in comprehensive evaluation, it also exposes limitations such as sensitivity to data fluctuations and high computational complexity. By developing and applying advanced optimization and decision models, this study provides a scientific water resource allocation scheme for Hancheng City, supporting the sustainable management of regional water resources, and offering a reference for future research in addressing data uncertainties and improving computational efficiency.