An adaptive agent-based optimization model for spatial planning: A case study of Anyue County, China

Abstract

Spatial planning is a complex land use allocation process involving multiple land use stakeholders, and resolving potential conflicts among stakeholders presents a challenging issue. This study proposed an innovative combination of an agent-based model and heuristic methods (machine learning and artificial intelligence approaches) to address spatial planning issues. The agent-based model simulates the decision-making processes of stakeholders in land use allocation, the machine learning approach obtains the nonlinear behavioral rules of land use agents, and the artificial intelligence approach provides a flexible optimization framework that can incorporate agents’ preferences into land use allocation. The integration of the agent-based model and heuristic methods enables us to adaptively explore nonlinear relationships between agent behaviors and decision-making environments and efficiently identify solutions to land use allocation in a spatially explicit way. The results show that the optimal allocation solutions obtained by the agent-based model are more applicable based on the support of the factual evidence than those obtained by the non-agent-based model. The proposed model can integrate the simulated local decision of stakeholders and global optimization of the specified objectives in land use planning, and thus provide a flexible theoretical framework to support the reform of China’s spatial planning system.