An improved PSO algorithm for optimizing land resource allocation

The current focus of land resource allocation optimization is mainly on optimizing land use structure and spatial patterns, with relatively little research combining the two aspects. While some exploration has been undertaken, there is still a lack of practical and effective technical means and methods. This study utilized the differential evolution algorithm to optimize the land use structure in the research area for the year 2028, and the results were analyzed and validated. Subsequently, these optimized land use structure results were integrated as quantitative constraints into the particle swarm optimization algorithm. Multiple constraint conditions were set for particle update optimization, including basic grid, spatial agglomeration, land type conversion, and terrain slope rules, leading to the construction of a DE-PSO model for optimizing land resource allocation. The results indicate that the optimized land use spatial pattern in 2028 aligns with the actual situation in the research area, showing an increased degree of land use spatial agglomeration and a tendency towards a more reasonable overall land use spatial pattern. The water area has also significantly increased. The overall accuracy of the model is 90.21%, with a Kappa coefficient of 0.8563, indicating that the established model performs well.