论文部分内容阅读
为合理利用多智能体算法解决城市扩张动态模拟问题,基于地理学理论和社会学规律对粒子群算法进行有针对性的改进,提出分段式粒子群算法(SPSO),并结合元胞自动机模拟复杂时空过程的能力,构建出适用于城市扩张模拟的地理元胞自动机SPSO-CA。在SPSO-CA中我们利用多时像的土地利用数据、交通路网数据和地形数据,挖掘出1995~2000年南京城市扩张的土地转换规则。再由此规则实现1995~2008年的南京市城市扩张过程的动态模拟。最后对比SPSO-CA、PSOCA及NULL模型结果得:SPSO-CA总精度86.3%,Kappa系数为0.792,Moran’s I为0.078,PSO-CA总精度83.6%,Kappa系数为0.755,Moran’s I为0.054,NULL模型总精度81.9%,Kappa系数为0.741,真实的Moran’s I为0.072。这表明无论是总精度还是空间一致性,SPSO-CA都优于PSO-CA和NULL模型,即用SPSO-CA模拟城市扩张是可行的。
In order to make rational use of the multi-agent algorithm to solve the problem of dynamic urban expansion, the PSO algorithm is improved based on the theory of geography and the law of sociology. A Particle Swarm Optimization (SPSO) algorithm is proposed. Combined with cellular automaton Simulate the ability of complex space-time process, and construct the geographic cellular automaton SPSO-CA which is suitable for urban expansion simulation. In SPSO-CA, we use land-use data, road network data and topographic data for a long time to figure out the land conversion rules for Nanjing city expansion from 1995 to 2000. Then this rule realizes the dynamic simulation of urban expansion in Nanjing City from 1995 to 2008. Finally, the results of SPSO-CA, PSOCA and NULL model showed that the overall accuracy of SPSO-CA was 86.3%, Kappa coefficient was 0.792, Moran’s I was 0.078, PSO-CA overall accuracy was 83.6%, Kappa coefficient was 0.755, Moran’s I was 0.054, NULL The overall accuracy of the model was 81.9% with a Kappa coefficient of 0.741 and a true Moran’s I of 0.072. This indicates that SPSO-CA is superior to PSO-CA and NULL models in terms of overall accuracy and spatial consistency, that is, using SPSO-CA to simulate urban expansion is feasible.