基于Markov的主城景观空间变异规律动态仿真

Dynamic Simulation of Spatial Variation Law of Main Urban Landscape Based on Markov

为准确预测景观空间变化趋势,提高景观规划的合理性,提出基于Markov的主城景观空间变异规律动态仿真。利用遥感技术采集景观图像,经几何校正、图像融合、信息损失评估等过程完成预处理,增强图像质量,提高信息利用率;结合景观形状、香农多样性指数分析景观空间的自相关性,避免重复采样;将气候、经济、环境等因素作为影响空间变异的驱动因子,计算景观利用变化幅度、利用动态度,建立重心转移模型,确定空间变异核心位置;采用Markov预测模型,构建景观空间变异转移概率矩阵,通过上述矩阵即可模拟变异规律。仿真结果表明,随着空间幅度半径的增加,除建筑景观外,其余景观的数量与面积均表现出缓慢上升、显著上升和趋于平稳的变异规律,可看出空间幅度对景观变异存在一定敏感性。

In order to accurately predict the change trend of landscape space and improve the rationality of landscape planning,a dynamic simulation of landscape spatial variation law of the main city based on Markov is proposed.Landscape images were collected by remote sensing technology,and preprocessed by geometric correction,image fusion and information loss evaluation to enhance image quality and improve information utilization.The autocorrelation of landscape space was analyzed by combining landscape shape and Shannon diversity index to avoid repeated sampling.Taking climate,economy,environment and other factors as the driving factors affecting spatial variation,the landscape utilization change range and utilization dynamics were calculated,the center of gravity transfer model was established,and the core position of spatial variation was determined.The Markov prediction model was used to construct the landscape spatial variation transfer probability matrix,through which the variation law could be simulated.The simulation experiment shows that with the increase of spatial amplitude radius,the number and area of other landscapes,except architectural landscape,show the variation law of slow rise,significant rise and stabilization,which shows that spatial amplitude is sensitive to landscape variation.