元胞自动机（CA）是研究土地利用变化的过程和趋势的有力工具。传统的土地利用CA模型往往假设元胞是只含单一土地类型的纯质单元，难以表达元胞内混合土地利用结构在时空上的连续变化。基于混合元胞的CA模型在元胞状态、元胞邻域、转化规则等方面都与传统CA有明显差异，如何构建混合元胞CA并有效评价其模拟精度是一个有待解决的问题。因此，本项目拟开展基于混合元胞CA模型的土地利用结构变化模拟与评价研究。研究内容包括：（1）混合元胞内多维土地利用变量连续变化的机理挖掘方法，以及耦合亚元胞尺度土地竞争与定量转化机制的混合元胞CA建模方法；（2）混合元胞CA模型模拟精度与土地利用结构相似度的评价方法；（3）典型区域土地利用结构变化模拟的实验验证，评价模型实际应用的精度并进行情景模拟和分析。本项目研究将推进混合土地利用研究从“静态格局”到“动态模拟”的发展，并可作为地理元胞自动机理论、方法和应用的重要发展和补充。

Cellular Automata (CA) have been widely used for analyzing the pattern, process, and driving forces, and simulating the trends of land use change. However, the cells of traditional CA models only contain one land use type, which ignores the mixed land use pattern and changes occurred within each cell. Therefore, conventional CA models are unable to simulate the continuous change of land use in spatial and temporal terms. Simulating land use change based on mix land use cell is challenging because of the differences in cell state, neighborhoods of a cell, transition rules, space-time continuity, etc. How to establish a Mixed-cell CA model and evaluate the accuracy of the simulation results remains unsolved. Therefore, based on the characteristics of mixed cells, this study aims to develop a modeling approach for mixed land use change simulation with Mixed-cell CA model and provide an evaluation method for validating its simulation results...The main research contents are as follow: 1. The mechanism analysis method of the continuous change of multi-dimensional land use variables, and the Mixed-cell CA model by coupling the mechanism of land use competition and quantitative conversion at the sub-cell scale, 2. The evaluation method for validating the simulation accuracy and similarity of mixed land use structure, 3. An experimental proof of Mixed-cell CA model in typical regions, which can be used to verify the model’s simulation accuracy and future scenario analysis in practice. This study promotes the development of mixed land use research from static pattern to the dynamic simulation and can serve as an important supplement of the theory and application of CA.