| 基于GARBF神经网络的土壤有效锌空间插值方法研究 |
| |
| 引用本文： | 董敏,王昌全,李冰,唐敦义,杨娟,宋薇平.基于GARBF神经网络的土壤有效锌空间插值方法研究[J].土壤学报,2010,47(1):42-50. |
| |
| 作者姓名： | 董敏 王昌全 李冰 唐敦义 杨娟 宋薇平 |
| |
| 作者单位： | 1. 四川农业大学资源环境学院,四川雅安,625014
2. 成都市龙泉驿区农村发展局,成都,610100 |
| |
| 基金项目： | 四川省教育厅重点项目（2006ZD003，2006A013，07ZS002）、四川省科技厅项目（2008JY0095）资助 |
| |
| 摘 要： | 以土壤有效锌为研究对象,构建遗传径向基函数(GARBF)神经网络对该元素属性值进行空间插值,以训练样本集的测定值与预测值之间的决定系数、逼近误差及检验样本的插值误差为评判标准,比较GARBF神经网络、径向基函数(RBF)神经网络、普通克里格(Ordinary Kriging)的拟合能力和空间插值能力。结果表明:同一区域两种抽样方案(a、b)下三种插值方法对训练样本的拟合能力为GARBFRBFOr-dinary Kriging。以平均绝对误差和误差均方根作为插值精度的评价指标,GARBF与RBF神经网络相比,训练样本的逼近误差分别降低0.22～0.25(a方案)和0.10～0.11(b方案),检验样本的插值误差分别降低0.13～0.11(a方案)和0.02～0.13(b方案);GARBF神经网络与Ordinary Kriging相比,训练样本的逼近误差分别降低1.12～1.40(a方案)和1.45～1.88(b方案),检验样本的插值误差分别降低0.20～0.24(a方案)和0.14～0.32(b方案),GARBF神经网络的误差最小,插值精度最高。从GARBF神经网络的插值图可以看出,遗传算法避免了神经网络容易陷入局部最优点,扩大了对土壤中相关空间信息的搜索范围,在一定程度上避免了类似克里格插值的"平滑效应"。
|
| 关 键 词： | GARBF神经网络 RBF神经网络 普通克里格 空间插值 |
Study on soil available zinc with ga-rbf-neural-network-based spatial interpolation method |
| |
| Dong Min,Wang Changquan,Li Bing,Tang Dunyi,Yang Juan and Song Weiping.Study on soil available zinc with ga-rbf-neural-network-based spatial interpolation method[J].Acta Pedologica Sinica,2010,47(1):42-50. |
| |
| Authors: | Dong Min Wang Changquan Li Bing Tang Dunyi Yang Juan and Song Weiping |
| |
| Institution: | Colloge of Resources and Environment, Sichuan Agricultural University;Colloge of Resources and Environment, Sichuan Agricultural University;Colloge of Resources and Environment, Sichuan Agricultural University;Rural Development Bureau, Long Quanyi District;Colloge of Resources and Environment, Sichuan Agricultural University;Colloge of Resources and Environment, Sichuan Agricultural University |
| |
| Abstract: | A spatial interpolation method based on GARBF neural network was used to study available zinc in soil.Comparison between GARBF neural network,RBF neural network and Ordinary Kriging interpolation method in fitting capacity and spatial interpolation capacity,with determination coefficient between measured values and predicted values of the training sample set,approximate error and interpolation error of test samples cited as criteria for evaluation.Results show that in terms of the fitting capacity,the three... |
| |
| Keywords: | GARBF Neural Network RBF Neural Network Ordinary Kriging method Spatial interpolation |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
| 点击此处可从《土壤学报》浏览原始摘要信息 |
| 点击此处可从《土壤学报》下载免费的PDF全文 |
|
