| 最小二乘支持向量机应用于西安霸河口水质预测 |
| |
| 引用本文： | 房平,邵瑞华,司全印,任娟.最小二乘支持向量机应用于西安霸河口水质预测[J].系统工程,2011(6). |
| |
| 作者姓名： | 房平 邵瑞华 司全印 任娟 |
| |
| 作者单位： | 西安建筑科技大学环境与市政学院;西安工程大学环境与化学工程学院;陕西省环保厅;西安工程大学纺织与材料学院; |
| |
| 基金项目： | 渭河水污染防治专项技术研究与示范(2009ZX07212-002); 西安工程大学校管科研项目(2005XG07;2006XG16) |
| |
| 摘 要： | 利用西安霸河10年的水质平均数据作为数据集,建立了基于最小二乘支持向量机的水质预测模型。通过适当的参数选择,其平均相对误差只有4.95%,预测的准确率达到95%。通过实例计算且与误差逆传播(BP)神经网络、RBF神经网络等预测方法进行了对比分析,表明该方法的平均预测精度较传统的神经网络方法提高约4%,且具有收敛速度快、泛化能力强等优点,可有效用于水质预测。
|
| 关 键 词： | 最小二乘支持向量机(LS-SVM) 径向基核函数(RBF) 核函数 结构风险最小化(SRM) 水质预测模型 |
The Application of Least Squares Support Vector Machine Regression in Water Quality Forecast of Xi'an Ba River |
| |
| FANG Ping,SHAO Rui-hua,SI Quan-yin,REN Juan.The Application of Least Squares Support Vector Machine Regression in Water Quality Forecast of Xi'an Ba River[J].Systems Engineering,2011(6). |
| |
| Authors: | FANG Ping SHAO Rui-hua SI Quan-yin REN Juan |
| |
| Institution: | FANG Ping1,2,SHAO Rui-hua1,SI Quan-yin3,REN Juan4(1.Xi'an University of Architecture & Technology,Xi'an Shaanxi 710055,China,2.School of Environmental & Chemical Engineering,Xi'an Polytechnic University,Xi'an 710048,3.Shaanxi Environment Protection Agency,Xi'an 710004,4.School of Textile and Materials,China) |
| |
| Abstract: | Using the data of the water quality data of Xi'an Ba river of nearly ten years recorded as the data set,this paper sets up a new Water Quality forecast model based on Least Squares Support Vector Machine Regression(LS-SVMR).By choosing the model parameter properly,the mean relative percentage error is only about 4.95% and the rate of accuracy comes to 95%.The experimental results indicate that the average predication precision increase by 4 percent,compared to the BP and RBF neural network method,and that t... |
| |
| Keywords: | Least Squares Support Vector Machine Radial Basis Kernel Function Kernels Structural Risk Minimization Water Quality Forecast Model |
| 本文献已被 CNKI 等数据库收录！ |
|
