| 基于CA的城市土地利用演变人工神经网络模拟 |
| |
| 引用本文： | 赵晶.基于CA的城市土地利用演变人工神经网络模拟[J].兰州大学学报(自然科学版),2006,42(5):27-31. |
| |
| 作者姓名： | 赵晶 |
| |
| 作者单位： | 同济大学,海洋地质国家重点实验室,上海,200092 |
| |
| 基金项目： | 同济大学校科研和教改项目 |
| |
| 摘 要： | 基于CA原理,利用学习矢量量化神经网络从不同时相遥感数据中挖掘土地利用演变的内在规律,自动找到土地利用元胞的转换规则,并以该规则反演和预测土地利用格局.在上海市区典型边缘带的应用显示,挖掘出的元胞转换规则,与同期上海城市发展状况相吻合,表明该模型可以满足土地利用演变模拟预测的要求,大大缩短了建立CA转换规则所需时间.若能增加社会、经济因素的影响,减少网络训练时的信息浪费,将进一步优化模拟效果.
|
| 关 键 词： | 元胞自动机 人工神经网络 土地利用 模拟 转换规则 |
| 文章编号： | 0455-2059（2006）05-0027-05 |
| 收稿时间： | 03 14 2006 12:00AM |
| 修稿时间： | 2006-03-14 |
A simulation of urban land use evolution based on artificial neural network and cellular automata |
| |
| ZHAO Jing.A simulation of urban land use evolution based on artificial neural network and cellular automata[J].Journal of Lanzhou University(Natural Science),2006,42(5):27-31. |
| |
| Authors: | ZHAO Jing |
| |
| Institution: | State Key Laboratory of Marine Geology, Tongji University, Shanghai 200092, China |
| |
| Abstract: | By establishing an integrated model made up of CA and Learning Vector Quantization Network (LVQ Network),evolution rules of land cellular automata were mined,which would be used to invert the land use pattern and predict land cellular status in future.The integrated CA-LVQ model indicates that evolution rules of land cellular automata found in past land use pattern were quite coincident with simultaneous social evolution in the research area in Shanghai;so such an integrated model can be used to simulate land use evolution by distinctly shortening required time.The integrated CA-LVQ model can be further improved by appending to it social and economic factors and reducing LVQ's information waste. |
| |
| Keywords: | cellular automata artificial neural network land use simulation evolution rule |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
