摘要
利用C#语言结合ArcEngineGIS平台编程对广州市花都区土地利用演变CA模拟进行研究,利用逻辑回归方法确定特征变量概率,并结合约束条件、领域作用和随机作用项共同确定元胞转换概率,通过与转换概率阈值的比较确定元胞在下一时刻的转化状态。结果表明,模拟的总体精度为79.3%,MoranI指数与实际情况较接近,说明该模型具有较强的实用性。
The land use change of Huadu in Guangzhou was simulated based on CA model with C# language and ArcEngine GIS platform. The probability of characteristic variable was determined with logistic regression method,and the cellular transformation probability was determined combining with constraint condition,field action and randomly action. The transforming state of cellular in the next time was determined through comparison of transition probability threshold. The results showed that the stimulated general accuracy was 79.3%,and the Moran I index was close to the actual situation,which indicated that the model had stronger practicality.
出处
《安徽农业科学》
CAS
北大核心
2010年第11期5791-5793,共3页
Journal of Anhui Agricultural Sciences
基金
国家自然科学基金资助项目(30970482)
关键词
土地利用变化
元胞自动机
逻辑回归方法
Land use change
Cellular automaton
Logistic regression method
