基于RS-SVM的TN软测量建模研究

采用粗糙集理论(RS)约简属性,在保留重要信息的前提下消除冗余信息,简化了模型结构。而支持向量机(SVM)是一种基于统计学习理论的新型学习机,本文根据TN(总氮)难于在线测量的情况,采用RS-SVM方法,用某城市污水处理厂的实际水质参数数据,建立了出水TN基于粗糙集-支持向量机的软测量模型。和未经粗糙集预处理的支持向量机模型及粗糙集-BP神经网络(RS-BPNN)模型进行了比较,选择RS-SVM模型作为最终的软测量模型。结果表明,有粗糙集预处理后,不仅测量值的误差值更小,而且大大降低了输人数据的维数,减小了模型的规模,更有利于软测量模型的实用化。同时也表明支持向量机作为建立软测量模型的工具,具有良好的性能,比神经网络更加具有优势。

Studing About TN Soft Measure Model BASED ON RS-SVM

It results in simplification of the model structure to reduce the attributions of model, and eliminate superfluous data by rough set. And support vector machine is a kind of new machine learning method based on statistic. By those methods,the TN soft measure model based on rough set using support vector machinels established, using practical data of water quality parameters in some municipal wastewater treatment plant. After contract with SVM(RS isn' t used)model and RS-BPNN model, choice RS-SVM model. The result indicates that the error is smaller when the rough set is used than isn' t used, and the dimensions of the input data are decreased greatly. Meanwhile it indicates that support vector machine is advantaged as method of soft measure model than artificial neural network.