基于BP神经网络的贵州喀斯特农村地区人居环境评价——以清镇红枫区、毕节鸭池区和关岭-贞丰花江区为例

Rural Human Settlement Environment Assessment at Karst Area in Guizhou Province Based on BP Artificial Neural Network——take Hongfeng of Qinzhen city, Yachi of Bijie city and Huajiang of junction of Zhenfeng country and Guanling country as examples

  • 摘要: 选取具有代表性的清镇红枫区、毕节鸭池区以及关岭贞丰花江区3个典型区域作为研究对象,在总结喀斯特人居环境研究现状以及人居环境评价理论和实践的基础上,结合喀斯特地区的实际环境条件构建宜居指数,对喀斯特农村的村落人居环境进行评价.再基于BP神经网络的特点以及3个研究区29个行政村社会经济调查统计数据特点,构建并训练得到合理的神经网络,最后进行评价.评价结果显示,居住环境最好和较好的村落以红枫区最多,鸭池和花江区的村落大部分处于中下水平,且花江区的部分村落评价结果最差.可见喀斯特人居环境与区域综合地理环境特别是自然条件密切相关,自然环境仍然是喀斯特地区人居环境的主导因素.

     

    Abstract: To evaluate the rural human settlement in Karst landform region of Guizhou Province, three typical areas, Hongfeng community at Qinzhen city, Yachi community at Bijie city and Huajiang community at the junction of Zhenfeng county and Guanling county are selected as the examples. Firstly, assessment index system for actual environmental conditions in Karst is established on the basis of the existing theory and practice in human settlement evaluation. Then the statistic characters about socio-economic survey data of 29 administrative villages and the structure of ANN is combined to build and train a reasonable neural network. Lastly, the BP ANN is used to assess them. Results show that Hongfeng had better human settlement communities, most of Yachi and Huajiang resident level is in the mid or lower grade, and the worst assessment is in Huanjiang. Therefore, rural living environment in Karst area closely related to regional geographic environment, especially natural conditions, and it is still the key for Karst human settlement.

     

/

返回文章
返回