土地利用配置的混沌蚁群优化算法研究

土地利用优化配置是促进土地可持续发展的重要举措,然而现有研究缺乏有效求解土地利用优化配置模型的新型混合式智能优化算法。本文结合蚁群算法和混沌模型,形成混沌蚁群优化（Chaos Ant Colony Optimization,CACO）算法,并以广州市增城区为研究区,对土地利用现状进行优化配置;然后在数量结构、目标函数值、空间布局等方面将优化结果与土地现状及标准蚁群算法优化结果进行对比分析。结果表明：① CACO算法能在满足多种约束条件下,有效解决多目标土地利用优化配置问题;② 与标准蚁群算法相比,CACO算法能增加土地利用的经济效益7.18亿元、生态效益0.33亿元、社会效益1.13%,同时降低地类转换成本1.15%;③ CACO算法能使土地利用现状空间分布多样性和均匀性的下降控制在1.30%以内,同时缩减地块数量8.86%,并使平均斑块大小增加9.77%,从而提升土地集约利用水平,更合理地配置各现状地类的空间分布,为研究区土地利用的科学规划与决策提供支持。

Land Use Optimization Allocation Based on Chaos Ant Colony Algorithm

The optimal allocation of land use is an important and effective measures of promoting the sustainable development of the land. However, existing research was lack of efficient methods in the optimization allocation for the quantitative structure and spatial layout of land use by using original mixed algorithm. Therefore, this paper combined the ant colony optimization algorithm (ACO) with the chaos model, and proposed a hybrid and self-adapt chaos ant colony optimization algorithm (CACO). After that, in order to verify the feasibility and efficiency of the CACO, the Zengcheng district of Guangzhou was selected as the study case. The CACO was utilized to solve the model of land use optimization allocation based on the evaluation of actual land sustainable use. Finally, this study made some comparative analysis of the results of the CACO and the actual land use and the results of the ACO respectively in three main aspects: the quantitative structure of land use, the spatial layout of land use and the multiple objective functions. The results showed that: firstly, the CACO can effectively solve the complex problems of multi-objective land use optimization allocation under multiple constraint conditions; secondly, compared with the ACO, the coordination between economic benefit and ecological effectiveness in the CACO was weakened slightly. The CACO rose all others objective functions’ values. For example, economic benefits increased by 7.18 billion yuan, ecological effectiveness increased by 0.33 billion yuan, social benefits increased by 1.13%, while the land conversion costs shrank by 1.15%. Thirdly, compared with the ACO, the CACO decreased the diversity and evenness index of actual land use spatial distribution within 1.30%, made the number of total land patches reduced about 8.86%, and the average patches size increased 9.77%. The level of the land intensive use was improved. Therefore, the CACO could reasonably optimize actual various land use types to appropriate spatial layout, and supply useful technical support for scientific land use planning and decisions making.