基于Multi-Agent系统的土地利用情景模拟

Scenario simulation of land use based on Multi-Agent system

以常州市新北区为研究区,运用多智能体空间建模方法,利用Eclipse和Swarm平台,对研究区不同情景下2006—2020年的土地利用情况进行了模拟,为区域规划决策和土地利用调控提供辅助技术支持。研究结果表明:经济社会快速发展情景中城镇化速度较快,城乡用地转换明显,农村居民点整治力度较大;经济社会中速发展情景中城镇用地空间扩张和人口增长都减缓速度,耕地资源的流失得到一定程度的控制;经济社会慢速发展情景中城乡用地变化速度明显放缓,耕地资源流失得到有效控制。多智能体模型突出考虑了土地利用过程中相关主体行为对土地利用变化的影响,并在建模过程中融合了土地利用情景分析,模型构建适用性强,可以较好地模拟不同情景下的土地利用变化情况。

A Multi-Agent model of a new technology based on the platforms of Eclipse and Swarm was built to simulate the land-use changes of Xinbei District of Changzhou in different scenarios from 2006 to 2020. The results indicated that： in scenario with high economic and social development seed, the urbanization went fast, the urban-rural land conversion was obvious, and the renovation and the reconstruction of the ru- ral settlement were enhanced. In scenario with medium economic and social development seed, the urban expansion and population growth slowed down, and the cultivated land loss decreased in a certain extent. In scenario with low economic and social development seed, the urban-rural land conversion decreased ob- viously, and the cultivated land loss was effectively prevented. The land-use changes in various scenarios could be effectively simulated by the Muhi-Agent model which absorbed different social and economic fac- tors relevant to the land use.