ABSTRACT Territorial spatial optimization is a key strategy to promote long‐term regional sustainable development. Although ant colony optimization (ACO) is widely used in territorial spatial optimization, it is prone to local optima and offers limited integration of spatial suitability and regional heterogeneity. To address these problems, an adaptive ACO for territorial spatial optimization (TS‐AACO) is proposed in this study. Specifically, the multi‐scale geographically weighted regression (MGWR) is integrated into the suitability evaluation mechanism to construct ACO's objective function, enhancing its global search capability. In addition, an adaptive pheromone update strategy is designed and incorporated with the k ‐opt local search algorithm to improve both the convergence speed and solution quality. Experimental results demonstrate that compared to current mainstream optimization algorithms, TS‐AACO significantly improves convergence speed by 21%. The optimized territorial spatial layout shows a more orderly landscape pattern, with greater patch compactness and a significant reduction in conflict areas, providing robust support for rational territorial resource utilization.
Li et al. (Tue,) studied this question.