基于GIS和ANN的农户生计脆弱性的空间模拟分析  

Spatial Simulation Using GIS and Artificial Neural Network for Household Livelihood Vulnerability

作　　者：刘进[1] 甘淑[1] 吕杰[1] 闫海青[1] 袁欣婷[1] 刘晨晨[1] 谢长波[1] 

机构地区：[1]昆明理工大学国土资源工程学院,云南昆明650031

出　　处：《山地学报》2012年第5期622-627,共6页Mountain Research

基　　金：国家自然科学基金项目(71163023)~~

摘　　要：基于农户生计脆弱性测定方式的不完善和云贵高原缺乏农户生计脆弱性研究,选取云南省宜良县为案例区,构建了农户生计脆弱性评价指标体系,包括农户面临的风险、农户生计资本和面对风险的应急能力,并运用GIS与BP神经网络模拟区域的风险度指数、农户生计资本和应急能力指数的空间分布格局。在此基础上,得到农户生计脆弱性指数,结果表明坝区的生计脆弱性指数为山区>半山区>坝区,且各自原因不同。可见,运用BP神经网络模拟生计脆弱性简便实用,是一种可行的实践方法。Aiming at the status that evaluation method of farmers＇ livelihood vulnerability is not perfect and Yungui Plateau is lacking in the study of farmers＇ livelihood vulnerability, Yiliang Country of Yunnan Province are selected as an example to construct the index system of farmers＇ livelihood vulnerability including the risks farmer faced, household livelihood and the emergency capability. The spatial distribution pattern of the Risk Index, the Farmers＇ Livelihood Index and the Emergency Capability Index are simulated with tools of GIS and BP Natural Network. Then,the livelihood vulnerability index is discussed. The results show that the villages＇ vulnerability index in Mountain District is bigger than in Middle-mountain and the Middle-mountain District is bigger than Dam District, and the vulnerably reasons are variable. The results also indicate that application of BP Neural Network to simula- ting livelihood vulnerability index is convenient, precise and feasible,which can be an alternative approach to simu- lating livelihood vulnerability.

关 键 词：生计脆弱性 BP神经网络 GIS 云贵高原 

分 类 号：X22[环境科学与工程—环境科学]