摘要
针对下水道可燃气体传感器非线性、选择性差和交叉敏感的特点,建立了一种基于粒子群算法(PSO)支持向量回归机(SVR)的下水道可燃气体分析预测模型。该模型通过引入粒子群算法对支持向量回归机的重要参数进行优化,从而实现了支持向量回归机的参数自动判定,用于下水道可燃气体的定量分析。仿真结果表明:基于粒子群的支持向量回归机下水道可燃气体分析预测模型优于SVR模型,具有较好的泛化性能和较高的预测精度。
According to the characteristics of non-linearity,poor selectivity and cross-sensitivity in sewer combustible gas sensor,an analysis prediction model of sewer combustible gas based on the PSO-SVR machine is established.By particle swarm optimization(PSO) algorithm,the important parameters of SVR are optimized to realize the automatic determination of parameters of the SVR machine for quantitative analysis on combustible gas in the sewer.The simulation results show that this model is superior to SVR model,with better generalization performance and higher prediction accuracy.
出处
《测控技术》
CSCD
2015年第2期20-23,共4页
Measurement & Control Technology
基金
自贡市科技局科技服务民生专项(2011S079)
自贡市重点科技计划项目(2013D06)
关键词
支持向量回归机
粒子群算法
可燃气体
预测模型
support vector regression
particle swarm optimization(PSO) algorithm
combustible gas
prediction model