利用PSO算法在空间上能快速搜索和寻优的特点,在MATLAB 7.0环境下利用改进的粒子群算法从N×M个栅格空间中确定K个目标栅格的位置,使得目标点距离南河、南山及瓷土矿的平均距离是最小的K个,然后结合这K个目标栅格自身土地适宜性条件及邻域土地利用情况来判定该目标栅格是否能演变为窑址.通过此模型模拟了五代至北宋这段历史时期景德镇南河流域窑址景观的演变过程,并对比分析了不同K值和概率阈值下的模拟情况,得出合适的取值.在此取值下,模拟结果得到的Moran's Ⅰ指数与实际情况下的Moran's Ⅰ指数比较接近,说明该模型在古代缺失GIS空间数据的情况下模拟窑址景观的演变同样有效；同时,为研究GIS空间数据缺失下的景观演变提供了新的解决方法.%By taking advantage of PSO algorithm' s characteristics of quick searching and optimizing in space, this paper used improved particle swarm algorithm in MATLAB 7. 0 environment to determine the location of K target grids from N × M grid space, made the average distance the smallest from the target grid to Nanhe, Nanshan and the clay mine. Then, by combining the K target grids' land suitability with the neighboring land utilization, this paper concluded whether target grid could evolve into a kiln site. This model simulated the evolvement of kiln landscape along Nanhe river of Jingdezhen, from the Five Dynasties to Northern Song Dynasty. Through a comparative analysis of different evolutions with different K value and probability threshold, it got the right value. With the right value, the simulated Moran' s I index was close to the actual Moran' s I index. It is proved that the model is effective in simulating kiln landscape evolvement of ancient times. At the same time, this paper gave a new solution to study landscape evolution without GIS spatial data.
展开▼
