| 基于PSO-SVM的离心式压缩机转速模型的研究 |
| |
| 引用本文： | 李雪洁,;孙川川,;刘大铭,;翟小宁. 基于PSO-SVM的离心式压缩机转速模型的研究[J]. 宁夏工程技术, 2014, 0(2): 174-177 |
| |
| 作者姓名： | 李雪洁, 孙川川, 刘大铭, 翟小宁 |
| |
| 作者单位： | [1]宁夏大学物理电气信息学院,宁夏银川750021; [2]中国石油宁夏石化公司,宁夏银川750021 |
| |
| 摘 要： | 针对宁夏某石化公司离心式CO2压缩机透平转速预测难以实现问题,引入PSO-SVM回归模型对离心式压缩机透平转速进行预测.分析选取离心式压缩机透平转速作为模型因变量,通过相关分析从采集量中选取高相关度预测因子,运用粒子群算法选择模型最优参数,利用支持向量机的方法建立模型进行预测,与传统的SVM模型进行对比,该模型得到了良好效果,能够有效预测压缩机透平转速.
|
| 关 键 词： | 离心式压缩机 支持向量机 回归算法 粒子群 |
PSO-SVM based model of centrifugal compressor turbine speed |
| |
| Affiliation: | LI Xuejie, SUN Chuanchaun, LIU Darning, ZHAI Xiaonin (1.School of Physics Electronics Information Engineering, Ningxia University, Yinchuan 750021, China; 2.PetroChina Ningxia Petrochemical Company, Yinchuan 750021, China) |
| |
| Abstract: | For problem of centrifugal CO2 compressor turbine speed prediction in Ningxia petrochemical company, PSO-SVM model was introduced to predict centrifugal C02 compressor turbine speed, centrifugal compressor turbine speed was selected as model＇s dependent variable , high relevance predicting factor was selected through correlation analysis from acquisition data, optimal parameters were selected by way of particle swarm optimization （PSO）, forecast model was established on the basis method of support vector machine（SYM）. Compared with normal SVM model, there are good effects for compressor turbine speed prediction. |
| |
| Keywords: | centrifugal compressor SVM recession algorithms particle swarm ontimization |
| 本文献已被 CNKI 维普 等数据库收录！ |
|
