基于多目标微粒群优化算法的土地利用分区模型

Land use zoning model based on multi-objective particle swami optimization algorithm

土地利用分区实质上是一个多目标的土地利用空间优化问题,将传统分区方法用于解决多目标优化问题明显不足。该文提出了一种基于多目标微粒群优化的土地利用分区模型,构建了土地利用分区的属性约束指标,提取4个能够体现土地利用分区目的及意义的目标函数,即分区间差异性最大区内相似性最强、空间分区集中连片、土地利用效益最大化、土地适宜性评价指数最高,同时考虑了土地利用分区图上最小图斑面积、用途区面积、用途转换规则3类约束条件,并详细阐述了算法的核心思想、编码策略、状态更新机制及其算子等内容,最后以湖北省宜城市为例,对模型的可行性和有效性进行验证,结果表明通过对目标的权重调控可以得到不同目标偏好的土地利用分区方案,该文所构建模型在土地利用规划实践中具有可操作性。

Land use zoning was discussed in this paper based on multi-objective optimization. Constraint indices system of land use zoning was constructed, and four objective functions were comprised of the largest difference among zones and the strongest similarity in zone, the higher connectivity among zones, the maximization of land use efficiency and the best land suitability. Three types of constraint conditions that include the minimum area on the map ,the area of land use zone and the regulation of conversion of land use are also considered. Model of land use zoning based on the multi-objective particle swami optimization algorithm was proposed to solve the multi-objective optimization problem. Detailed descriptions of the algorithm, coding scheme, the mechanism and operator of state updating are given in this paper. At last, take the city of YiCheng in Hubei province as a case study, the results showed that the model presented was effective and feasible.