ANFIS在土地利用变化模拟中的应用  

The Application of Adaptive Neuron-Fuzzy Inference System in Simulating Land Use Change

作　　者：李心玉[1] 葛莹[2] 张遵忠[1] 

机构地区：[1]江苏省有色金属华东地质勘查局,南京210007 [2]河海大学土木工程学院,南京210098

出　　处：《地理科学进展》2009年第2期187-192,共6页Progress in Geography

基　　金：国家自然科学基金项目(40672040)

摘　　要：土地利用变化在空间维和时间维上是一个渐进的、不确定的复杂过程,而模糊理论正是解决不确定性现象的一种合适的方法,所以尝试运用模糊推理理论对土地利用变化进行深入的探讨。以江苏省南通市崇川区为研究区,建立了基于自适应神经模糊推理系统(ANFIS)的土地利用变化模糊推理模型,通过利用ANFIS训练获得模型的隶属函数及参数,并运用该模型对研究区进行土地利用变化的模拟和预测。研究结果表明,通过利用ANFIS建立的模型,基本上可以模拟研究区复杂而不确定的土地利用变化过程,同时ANFIS可以有效地简化模糊推理模型结构,使得模型更具灵活性。因此,为土地利用变化模拟提供了另一种可行的解决思路。The land use and land cover change （LUCC） is one of the most important projects on global environmental changes. The process of land use change resembles a progressive and non- deterministic process both spatially and temporally. Fuzzy set theory offers a way to represent and handle uncertainty present in the continuous real world. So, the fuzzy inference theory has important significance for researching on land use change that was discussed deeply in the study. Using Chongchuan district in Nantong City of Jiangsu Province as a case, the paper mainly studied the fuzzy inference model of land use change based on Adaptive Neuron-Fuzzy Inference System（ANFIS for short） and simulated and forecasted the transition progress of the study area by using the model. The results confirm the potential of fuzzy inference to produce realistic simulations of the land use change progress. The land use model based on ANFIS could reflect the complexity and uncertainty of land use change. It is convenience to obtain the parameters and fuzzy rules of the model by the ANFIS from the history data and predigest the model structure. So the study supplied another possible idea for simulating the land use change.

关 键 词：土地利用变化 模糊推理 ANFIS 南通市崇川区 

分 类 号：F301[经济管理—产业经济] F224