摘要
硫容量是衡量合成渣脱硫能力的关键参数,是如何选择渣系的重要依据。目前,国内外预测硫容量的模型主要以机理模型为主,但是由于许多参数无法测量,经常使得预测无法进行,或者预测值与实际值有较大误差。应用改进的LS-SVM对精炼过程的渣系终点硫容量进行预测,它克服了冶炼现场数据量小,参数难以准确测量等问题,仿真结果表明该方法可以大幅度提高预测的准确度。
The sulfur capacity is one of the key parameters which can measure the desulphurization ability of the slag system.It is an important basis for choosing the slag system.At present,the prediction models of the sulfur capacity mostly are mechanism models at home and aboard.But because some of the parameters of the mechanism models can not be measured,the prediction can not proceed,or the errors between the predicted values and the real values are too big.The improved LS-SVM is used to predict the final sulfur capacity of the slag system,it overcomes the problems of less data and the parameters can not be measured exactly.The result of the simulation shows that the improved LS-SVM can significantly improve the prediction accuracy.
出处
《控制工程》
CSCD
北大核心
2010年第S1期49-51,共3页
Control Engineering of China
基金
国家高新技术研究发展计划基金资助项目(2007AA041401
2007AA04Z194)
