(1. 中南大學(xué) 信息科學(xué)與工程學(xué)院，長沙 410083；
2. 湖南第一師范學(xué)院 信息科學(xué)與工程系，長沙 410205；
3. 內(nèi)蒙古鑫旺再生資源有限公司，鄂爾多斯 014300)
摘 要: 分解溫度是氧化鋁晶種分解工序中的關(guān)鍵工藝參數(shù)。為精確控制分解溫度，運用機理分析與參數(shù)辨識相結(jié)合的方法建立帶板式換熱器種分槽系統(tǒng)的非線性動態(tài)模型，并利用實際生產(chǎn)過程數(shù)據(jù)驗證模型的正確性。提出一種基于不可測干擾預(yù)測的非線性模型預(yù)測控制(DP-NMPC)方法，利用時間序列分析方法建立系統(tǒng)中不可測擾動的自適應(yīng)預(yù)測模型，并以此模型對分解溫度預(yù)測模型進行校正。基于實際生產(chǎn)過程數(shù)據(jù)的仿真研究表明，相比常規(guī)NMPC，該方法提高了預(yù)測模型的精度，使控制系統(tǒng)能快速跟蹤系統(tǒng)設(shè)定值，更好地抑制超調(diào)，因而其抗干擾能力更強，能對晶種分解溫度進行有效控制。由于該方法適用于具有不可測非白噪聲強干擾過程的模型預(yù)測控制，具有顯著的實用價值。
關(guān)鍵字: 氧化鋁；晶種分解；擾動預(yù)測；非線性預(yù)測控制
（1. School of Information Science and Engineering, Central South University, Changsha 410083, China;
2. Department of Information Science and Engineering, Hunan First Normal University, Changsha 410205, China;
3. Inner Mongolia Xinwang Renewable Resources Co., Ltd., Erdos 014300, China）
Abstract:The decomposition temperature is the key technological parameter in alumina seed precipitation process. In order to control the decomposition temperature precisely, a nonlinear dynamic model of the precipitator equipped with a plate heat exchanger in alumina tri-hydrate precipitation was built by mechanism analysis and parameter estimation, and the accuracy of the model was proved by the simulation with actual process data. A nonlinear model predictive control (DP-NMPC) method based on the unmeasured disturbances prediction was proposed, which applies the analysis of time series to build an adaptive predictive model of unmeasured disturbances in the precipitator system, and then revises the decomposition temperature predictive model. Comparing with the common NMPC, the proposed method is more effective in controlling decomposition temperature, which improves the accuracy of the predictive model, performs a quick following of set point changes, and has a better reduction of overshoot and a stronger rejection of disturbances. That method can be applied to the process with strong unmeasured nonwhite disturbances, and has remarkable practical value.
Key words: alumina; seed precipitation; disturbances prediction; nonlinear model predictive control
