基于RBF神经网络模型的臭氧浓度软测量研究  

Research on software sensor of ozone concentration based on RBF neural networks model

作　　者：张海传[1] 刘钟阳[1] 许东卫[1] 王宁会[1] 

机构地区：[1]大连理工大学电气工程学院,辽宁大连116024

出　　处：《大连理工大学学报》2010年第6期1020-1023,共4页Journal of Dalian University of Technology

基　　金：辽宁省自然科学基金资助项目(20031079);教育部科学技术研究资助项目(重点项目104062)

摘　　要：目前的臭氧浓度在线分析测试仪器在臭氧发生器中的应用受到很大限制.通过监测影响臭氧产生浓度的6个参量,基于RBF神经网络模型实现了臭氧浓度软测量.该模型采用梯度下降法确定RBF基函数的中心及输出层权值,可离线和在线校正所建立的神经网络模型.实验证明,软测量模型输出结果与臭氧浓度分析仪测量结果的绝对误差小于5g/m3的达93%以上,绝对误差小于1g/m3的达33%以上,且响应时间小于1s.The ozone concentration on-line analyzer apparatus is limited in the application of ozone generator greatly.The software sensor of the ozone concentration based on RBF neural networks model is presented with measuring six parameters which influence the ozone concentration.A gradient descent algorithm is introduced to decide the center of the RBF function and the weights of output layer in the model.It can be calibrated off-line or on-line.The experimental results prove that the absolute error of more than 93% is less than 5g/m3 and more than 33% is less than 1g/m3 between the output of model and the result measuring with the ozone concentration on-line analyzer,and the response time is less than 1s.

关 键 词：臭氧浓度 径向基函数(RBF) 神经网络 软测量 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置] TQ123[自动化与计算机技术—控制科学与工程]