(中南大學(xué) 信息科學(xué)與工程學(xué)院，長沙 410083)
摘 要: 針對鉛鋅燒結(jié)工藝過程，提出一種集成的燒結(jié)終點(diǎn)預(yù)測模型。首先利用具有自適應(yīng)能力和逼近任意非線性函數(shù)的神經(jīng)網(wǎng)絡(luò)技術(shù)建立燒結(jié)終點(diǎn)預(yù)測模型；其次利用總結(jié)專家知識和操作者經(jīng)驗(yàn)的模糊邏輯規(guī)則建立燒結(jié)終點(diǎn)預(yù)測模型；最后采用最優(yōu)組合算法綜合集成兩種模型作為燒結(jié)終點(diǎn)的預(yù)測模型。實(shí)際預(yù)測結(jié)果表明：利用集成方法建立的預(yù)測模型能 夠較為準(zhǔn)確地反映實(shí)際生產(chǎn)過程，具有較高的預(yù)測精度。
關(guān)鍵字: 鉛鋅燒結(jié)過程；燒結(jié)終點(diǎn)；集成建模；最優(yōu)組合算法
burning through point of sintering
process based on optimal
combination algorithm
（College of Information Science and Engineering,
Central South University, Changsha 410083,China）
Abstract:For the complicated Pb-Zn sintering process, an app roach that integrates two models to predicting the burning through point(BTP) was presented.The first method uses BP neural network and Levenberg-Marquardt algorithm to predict BTP; the second method uses metallurgical and control expert knowledge and skilled operator experiences of the sintering process.Two results are combined by using optimal combination algorithm as general modeling results of the burning through point. Through the analysis of the industrial site modeling experiment,the results show that the integration of two model can show the real-time p roduction exactly.
Key words: Pb-Zn sintering process； burning through point; integration modeling; optimal combination algorithm
