基于PSO-LSSVM的浓缩池溢流水浓度预测

Prediction for overflow water concentration of concentrated tank based on PSO-LSSVM

目前针对溢流水浊度测量时存在的设备昂贵、可靠性差以及寿命短等问题,对浓缩池煤泥水处理过程带来的不利影响,提出了一种基于PSO-LSSVM的溢流水快速可靠的预测方法。根据现场获得的数据组建溢流水浊度数据库,并将其分为训练集与测试集,构建预测模型,并以粒子群算法(PSO)优化最小二乘支持向量机(LS-SVM)模型中的相关参数。经仿真验证,预测值精度可以达到92.38%,表明基于PSO-LSSVM的浓缩池溢流水浓度预测模型可以较好地实现溢流浊度的预测。

Aiming at the adverse effects of high cost, relatively poor reliability and short service life of overflow water turbidity measurement equipment on concentrated tank slime water treatment process, a quick and reliable prediction approach of over- flow water based on PSO-LSSVM was put forward. The overflow water turbidity data base on the basis of the actual data was built and divided into training set and test set to build forecasting model, and the particle swarm optimization （PSO） was used to optimize the relevant parameters of the least squares support vector machine （LS-SVM） model. Verified by simulation, the prediction accuracy could reach 92.38%, which showed that the overflow concentration prediction model of concentrated tank based on POS-LSSVM was able to preferably realize the overflow turbidity prediction.