电厂烟气含氧量智能混合预测方法_湛腾西.pdfVIP

31 8 Vol1 31 No1 8 2010 8 Ch inese Journal of Scientific Instrum ent Aug. 20 10 * 湛腾西, 郭观七 ( 4 14006) : , RBF ; ; , , ; - , , , : ; ; RBF ; ; ; : TP273 : A : 510. 8099 Intelligent hybrid prediction method of the flue gas oxygen content in power plant Zhan Tengx,i Guo Guanqi (School of Inf orm ation and Comm unication Eng ineering, H unan Institute of Science and Technology, Yueyang 414006, China) Abstract: A mi ing at the fact that the pow er p lant flue gas oxygen content is hard to pred ict effecti ely, an intelligent hybrid pred iction method is proposed. F irstly, RBF neural newt ork and principal component analysismethod are ut-i lized to reduce the dmi ension of the input ariab les; secondly, the case-based reasoningmethod is adopted to predict the flue gas oxygen content from the abo e analysis results. Then in order to reflect the tmi e cumulati e effect of the flue gas oxygen conten,t the process neural netw ork m ethod is used to pred ict the current flue gas oxygen con tent a-l ue. F inally the final flue gas oxygen content is obtained using ariance-co ariancemethod ofw eight combination fore- casting. The results of actual operation data analysis and industrial trial operation show that the proposed intelligent hybrid prediction model has high accuracy and robustness, and can effecti ely sol e the prob lem of pred icting flue gas oxygen content. Key words: flue gas oxygen con tent; intelligent hybrid prediction method; RBF neural newt ork; principal compo- nent analysis; case-based reasoning; process neural network , 1 ,