基于径向基函数神经网络的植烟土地适宜性评价
|
摘要
在Matlab7.0环境下,根据人工神经网络的理论和方法,以重庆市彭水县植烟土地的实测数据及评价标准构建径向基函数神经网络模型,并进行模型训练及样本评价;在ArcGIS技术支持下,进行不同尺度土地适宜性评价及精度检验.结果表明:采用最近邻聚类学习算法选取聚类中心,模型具有较强非线性处理能力和逼近能力,并具有学习时间短,网络运算速度快,性能稳定等优点;通过模型评价结果和检验值的验证,发现用径向基函数神经网络模型评价土地适宜性,具有评价精度高,使用方便,适应性强等优点,因此可望将其用于区域土地资源生态环境分类评价研究.
Taking into account the non-linear relationship between assessment indicators and assessment grades, the experimental results showed a high accuracy, thus demonstrating that this method could be widely extended. In this paper, a Radial Basis Function neural network model (RBF) was established for evaluating land suitability in Pengshui county of Chongqing based on the theory and methodology of neural network with Matlab 7.0 and the data obtained from the tobacco-planted land and criteria of suitability evaluation. Land suitability evaluation and its accuracy tests were performed at different scales based on ArcGIS. The results were as follows. The model proved its capability to approach function and treat non-linear problem by using nearest neighbor cluster algorithm to select the clustering center. The main advantage of the RBF-based model was its accuracy, time saving, fast running and stability behavior. The good agreement between the assessed and the tested data was observed. Because of these benefits, it is believed that the model could find application in classifying and assessing regional land resources.
Taking into account the non-linear relationship between assessment indicators and assessment grades, the experimental results showed a high accuracy, thus demonstrating that this method could be widely extended. In this paper, a Radial Basis Function neural network model (RBF) was established for evaluating land suitability in Pengshui county of Chongqing based on the theory and methodology of neural network with Matlab 7.0 and the data obtained from the tobacco-planted land and criteria of suitability evaluation. Land suitability evaluation and its accuracy tests were performed at different scales based on ArcGIS. The results were as follows. The model proved its capability to approach function and treat non-linear problem by using nearest neighbor cluster algorithm to select the clustering center. The main advantage of the RBF-based model was its accuracy, time saving, fast running and stability behavior. The good agreement between the assessed and the tested data was observed. Because of these benefits, it is believed that the model could find application in classifying and assessing regional land resources.
引文
[1]卢学强,梁雪慧,卢学军.神经网络方法及其在非线性时间序列预测中的应用[J].系统工程理论与实践,1997,(6):97-99.
[2]张静怡,陆桂华,徐小明.自组织特征映射神经网络方法在水文分区中的应用[J].水利学报,2005,36(2):163-165.
[3]葛晓光,吴潇,钱凯.自组织人工神经网络与聚类法在矿区沉积物分类中实用性对比[J].煤炭学报,2006,31(2):169-173.
[4]Mehmet E,Serdar D,Erkan Z E,et al.The Classification of Human Tremor Signals Using Artificial Neural Network[J].Expert Systems with Applications,2007,(33):754-761.
[5]Moisen G G,Frescino T S Ccomparing Five Modeling Techniques for Predicting Forest Characteristics[J].Ecol Model,2002,(157):209-225.
[6]曹先珍.基于径向基函数网络的经济时序预测模型[J].武汉工业学院学报,2005,24(1):88-90.
[7]Olajide J O,Igbeka J C,Afolabi TJ,et al.Prediction of Oil Yield from Groundnut Kernels in a Hydraulic Press Using Artificial Neural Network(ANN)[J].Journal of Food Engineering,2007,(81):643-646.
[8]Eddy E T,Laurent L,IsamS,et al.Use of Artificial Neural Network Si mulation Metamodelling to Assess Groundwater Contaminationin a Road Projuct[J].Mathematical and Computer Modelling,2007,(4):766-776.
[9]徐胜祥,贺立源,黄魏,等.人工神经网络在柑桔生产专家系统中的应用[J].计算机应用研究,2006,(2):138-141.
[10]王炜,吴耿锋,张博锋,等.径向基函数(RBF)神经网络及其应用[J].地震,2005,25(2):1-25.
[11]张运陶,杨晓丽.自适应径向基神经网络及其应用[J].计算机与应用化学,2006,23(1):59-63.
[12]杨彦军,杨宇,康志宏.径向基函数神经网络及其在插值计算中的应用[J].新疆石油地质,2005,26(2):209-211.
[13]雷铁安,吴作伟,杨周妮.径向基网络在结构设计与分析中的应用[J].钢结构,2005,20(2):57-59.
[14]郭宗楼.径向基函数网络模型在水质评价中的应用[J].浙江大学学报,2001,27(3):335-338.
[15]鲁如坤.土壤农业化学分析方法[M].北京:中国农业出版社,1999,205-206.
[16]刘国顺.烟草栽培学[M].北京:中国农业出版社,2003;苏德成.烟草栽培[M].北京:中国财政经济出版社,2000.
[17]黄昌勇.土壤学[M].北京:中国农业出版社,2000.
[18]石月珍.径向基函数网络模型在区域生态承载能力综合评价中的应用[J].四川环境,2005,24(5):48-51.
[2]张静怡,陆桂华,徐小明.自组织特征映射神经网络方法在水文分区中的应用[J].水利学报,2005,36(2):163-165.
[3]葛晓光,吴潇,钱凯.自组织人工神经网络与聚类法在矿区沉积物分类中实用性对比[J].煤炭学报,2006,31(2):169-173.
[4]Mehmet E,Serdar D,Erkan Z E,et al.The Classification of Human Tremor Signals Using Artificial Neural Network[J].Expert Systems with Applications,2007,(33):754-761.
[5]Moisen G G,Frescino T S Ccomparing Five Modeling Techniques for Predicting Forest Characteristics[J].Ecol Model,2002,(157):209-225.
[6]曹先珍.基于径向基函数网络的经济时序预测模型[J].武汉工业学院学报,2005,24(1):88-90.
[7]Olajide J O,Igbeka J C,Afolabi TJ,et al.Prediction of Oil Yield from Groundnut Kernels in a Hydraulic Press Using Artificial Neural Network(ANN)[J].Journal of Food Engineering,2007,(81):643-646.
[8]Eddy E T,Laurent L,IsamS,et al.Use of Artificial Neural Network Si mulation Metamodelling to Assess Groundwater Contaminationin a Road Projuct[J].Mathematical and Computer Modelling,2007,(4):766-776.
[9]徐胜祥,贺立源,黄魏,等.人工神经网络在柑桔生产专家系统中的应用[J].计算机应用研究,2006,(2):138-141.
[10]王炜,吴耿锋,张博锋,等.径向基函数(RBF)神经网络及其应用[J].地震,2005,25(2):1-25.
[11]张运陶,杨晓丽.自适应径向基神经网络及其应用[J].计算机与应用化学,2006,23(1):59-63.
[12]杨彦军,杨宇,康志宏.径向基函数神经网络及其在插值计算中的应用[J].新疆石油地质,2005,26(2):209-211.
[13]雷铁安,吴作伟,杨周妮.径向基网络在结构设计与分析中的应用[J].钢结构,2005,20(2):57-59.
[14]郭宗楼.径向基函数网络模型在水质评价中的应用[J].浙江大学学报,2001,27(3):335-338.
[15]鲁如坤.土壤农业化学分析方法[M].北京:中国农业出版社,1999,205-206.
[16]刘国顺.烟草栽培学[M].北京:中国农业出版社,2003;苏德成.烟草栽培[M].北京:中国财政经济出版社,2000.
[17]黄昌勇.土壤学[M].北京:中国农业出版社,2000.
[18]石月珍.径向基函数网络模型在区域生态承载能力综合评价中的应用[J].四川环境,2005,24(5):48-51.
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心