污水处理控制系统设计与出水参数的软测量*

作者：徐方舟，潘 丰* 日期：2010-11-15/span> 浏览：3212 查看PDF文档

污水处理控制系统设计与出水参数的软测量*

徐方舟，潘 丰*

(江南大学 通信与控制工程学院，江苏 无锡 214122)

摘要：随着工业技术的发展，各种污染已经严重影响了人类的生存环境。近年来人们由于对水环境保护意识不断增强，更多的技术被要求应用于污水处理过程之中。针对某焦化厂污水处理系统，设计了一套基于WINCC、PLC的污水处理控制系统，从而实现了污水处理系统自动化控制。利用已测得进水数据，基于最小二乘支持向量机（LSSVM）建立了软测量模型，并且引入小波分析这一数学分析方法，选用径向基小波核函数。针对支持向量机参数，采用自适应粒子群优化（APSO）算法对其进行了调整。实验结果表明，此方法能够实现对污水处理系统出水BOD5、COD浓度的软测量预测。
关键词：控制系统；最小二乘支持向量机；核函数；自适应粒子群优化
中图分类号：TH185；TP273
文献标识码：A文章编号：1001-4551(2010)09-0042-04

Design for the sewage treatment control system and soft sensing of the water parameters
XV Fangzhou, PAN Feng
(School of Network Engineering Structures, Jiangnan University, Wuxi 214122, China)

Abstract: With the quick development of the industrial technology, all kinds of pollution severely influenced the environment of human. Recently, nations pay attention to water protecting project, more technology is demanded in waste water treatment control system. Aiming at the waste water treatment system of coking plant, a suit control system was designed which based on WINCC, PLC and intelligent instruments, the automation of control schedule and business management was achieved. Through the input water data, the softsensing model was built with least squares support vector machine(LSSVM). With the method of wavelet analysis, the WRBF kernel function was adopted. For the parameters selection in LSSVM, the adaptive particle swarm optimization(APSO) algorithm was used. The test results indicate that the method can realize the soft sensing of the BOD5、COD concentrations in the whole waste water treatment control system.
Key words: control system; least squares support vector machine(LSSVM); kernel function; adaptive particle swarm optimization(APSO)