信息与控制2011，Vol.40Issue(4)：532-536,5.DOI:10.3724/SP.J.1219.2011.00532
基于PSO优选参数的SVR水质参数遥感反演模型
A Model for Water Quality Remote Retrieval Based on Support Vector Regression with Parameters Optimized by Particle Swarm Optimization
摘要
Abstract
In order to improve water quality retrieval accuracy of multi-spectral image, a model is put forward for water quality remote retrieval based on support vector regression (S VR) with parameters optimized by particle swarm optimization (PSO). Based on high-resolution multi-spectral remote SPOT-5 data and the water quality field data, The model uses CV (cross validation ) to estimate the generalization error and adopts PSO to optimize parameters of S VR model. Thus, automatic global optimization of model parameters is achieved, and the water quality is retrieved by the trained SVR. The proposed model is applied to the water quality retrievals of Weihe River in Shaanxi province. The experiment result shows that the developed model is more accurate than the routine linear regression model. It provides a new approach for remote sensing and monitoring of inland river environments.关键词
高分辨遥感影像/粒子群优化算法/支持向量回归/参数优选/水质反演Key words
high-resolution remote sensing image/ particle swarm optimization algorithm/ support vector regression/ parameter optimization/ water quality retrieval分类
资源环境引用本文复制引用
何同弟,李见为,黄鸿..基于PSO优选参数的SVR水质参数遥感反演模型[J].信息与控制,2011,40(4):532-536,5.基金项目
国家自然科学基金资助项目(40671133) （40671133）
重庆市科技攻关重点资助项目(CSTC2009AB2231). （CSTC2009AB2231）
