【作者】 崔宇； 雷海平； 吴跃民；

【Author】 CUI Yu, LEI Hai-ping, WU Yue-min (Wanyuan Assessment & Consultation Co., LTD. Hunan, Changsha 410011,China)

【机构】 湖南万源评估咨询有限公司；

【摘要】 研究目的:由于部分参数的设置不尽合理和调查资料的偏差,造成了湖南省某些县市农用地分等结果与实际状况不符。通过人工神经网络对全省农用地进行等别判定,可为等别平衡提供依据和参考。研究方法:使用MATLAB语言,构建广义回归神经网络模型(GRNN),使用聚类分析划分农用地等别。研究结果与结论:湖南省农用地划分为10个等别,等别较高的农用地主要分布在洞庭湖区、长株潭地区及部分河谷地区,等别较低的农用地主要分布在湘西及湘南的山地区,这与各县实际情况较为吻合。更多还原

【Abstract】 The gradation of agricultural land doesn’t accord with the fact in some counties of Hunan Province, because some parameters and data are not rational. The paper aims at offering references and bases for gradation-balance through gradating the agricultural land by using artificial neural network. Using MATLAB language, the method of generalized regression neural network (GRNN) was employed, and using cluster analysis the 10 ranks of agricultural land in Hunan Province were marked. We can clearly find that the best agricultural lands distribute over Dongting lake zone, Changsha-Zhuzhou-Xiangtan zone, and riverside, and the worst agricultural lands distribute over the western and southern mountainous area. Which is according with the fact in Hunan Province.更多还原