Enhancing Land Use Functions Through the Optimised Allocation of Land Use

Land use functions refer to the capacity of land systems to provide goods and services, either directly or indirectly. They reflect the intensity of land use and the demands of social development at different stages and serve as an important reference for future land use planning. This study establishes a land use allocation optimisation framework that integrates top-down and bottom-up approaches to enhance land use functions. At the bottom-up level, an artificial neural network (ANN) model is employed to evaluate land suitability based on neighbourhood rules and driving factors. At the top-down level, a multi-agent system (MAS) model is used to integrate trade-offs among land use functions with supply–demand relationships, thereby defining land use constraints during the allocation process. Subsequently, the two models were integrated into a cellular automata (CA) model to achieve the spatial allocation of optimised land use patterns. Xintai City, located in Shandong Province, was selected as the study area. The results indicate that, compared with the 2021 baseline, ecological functions exhibited the greatest improvement, increasing by 24.11%, whereas living and production functions decreased slightly by 4.35% and 0.87%, respectively, resulting in an overall comprehensive functional improvement of 9.32%. Regarding trade-off relationships among land use functions, the intensity of trade-offs between production and ecological functions, as well as between living and ecological functions, decreased significantly by 28.72% and 25.56%, respectively, whereas the trade-off intensity between production and living functions increased slightly by 1.68%. Regarding the supply–demand relationships of land use functions in Xintai City, the surplus of production functions decreased by 13.46%, the surplus of living functions increased by 16.67%, the deficit in ecological functions improved by 33.33%, and the comprehensive functional surplus increased by 23.53%. The proposed method demonstrates substantial effectiveness in enhancing land use functions, alleviating trade-offs among land use functions, and optimising their supply–demand relationships.