基于改进支持向量机的作物叶水势软测量建模

Soft Sensor Modeling of Leaf Water Potential Based on Improved Support Vector Machine

在标准最小二乘支持向量机(least square supportvector machine,LS-SVM)的基础上,利用改进的粒子群算法(i mproved particle swarmopti mization,IPSO)来优化LS-SVM模型参数,提出了基于IPSO-LS-SVM的软测量建模方法,建立了作物叶水势软测量模型.仿真结果表明,该方法比基本LS-SVM和PSO-LS-SVM模型具有更高的精度,能够很好地预测作物叶水势信息.

Based on study on least square support vector machine（LS-SVM）,the paper presents an improved particle swarm optimization（IPSO） algorithm to select the parameters of LS-SVM.The soft sensor modeling of the leaf water potential is established based on IPSO-LS-SVM.Simulation results indicate that the method based on IPSO-LS-SVM is of a higher accuracy than the basic LS-SVM and LS-SVM based on PSO,which can well predict leaf water potential.