Research Of Hybrid Intelligent Optimal Control Method And Its Application Of Aluminum Electrolytic Processes

In Aluminum electrolyte pot control system, the main target is to get high current efficiency and low energy consumption. The intelligent control by computer is one of the most effective methods to fulfill these targets. However, technology indicators (such as current efficiency, the DC consumption, etc.) of Aluminum electrolytic industrial processes is often difficult to on-line measure, correlate closely to settings (such as setting NB, setting voltage) of bottom control loop .Dynamic identity among them often has strong non-linearity, coupling and is hard to describe by precise model. Integrated complexity varying with working condition and cell condition's operating conditions makes it difficult to achieve optimal control by primary control method. The control system of Aluminum electrolytic industrial processes must use the combination of a variety of control methods to realize optimal control in the semi-closed-loop process of industrial production. Substance of control problem of optimal control in the process of Aluminum electrolytic production is to control possible the controllable parameters in the process of production into technological desired ranges .In this paper, A hybrid intelligent control method for process optimal operation is proposed, which controls the technique indices into the desired ranges by on-line adjusting the set-points of the control loops according to the operation working condition and cell condition of electrolytic cells.The main contents and achievements of this paper include:①An idea is presented by paper that we regard electrolyte pot's cell condition and operating working condition as impact factors of intelligent control and decision: The definition of concepts of cell condition and working condition is to reflect its own production performance. Slot-machines change the value of control variables according to cell condition,working condition changed so as to improve control accuracy to realize energy saving and consumption.②It is presented that a hybrid intelligent optimization control method: Case-based reasoning model for the control loop preset, Rule-based reasoning model for working conditions diagnosed, Rule-based reasoning model for cell conditions diagnosed, Classification and identification of data stream model based on neural network are set up by using artificial intelligence; A suitable expert system mode is set up based on these models so that series electrolysis's working condition and cells conditions are real-time diagnosed. And then, amendment of control circuit's setting values is reasoned according to improved intelligent control strategies and is sent to lower machines control through communicational interfaces to realize optimal operation of Aluminum electrolytic process.The proposed hybrid intelligent optimal control method has been effectively used in the Guizhou Branch of Aluminum. System has been running in the six electrolytic cells, which belong to Electrolytic Cell Line 3 of Electrolytic Aluminum Plant 2 of China Aluminum Guizhou branch, from January in 2009 year to now. Although more than one year's on-line test, voltage curve basically shows sinusoidal (or cosine) waveform, and control system of cells is running smoothly. The main economic and technical indicators are good. For example, anode effect coefficient of No.3325 electrolytic cell is less than 0.08, the average temperature of it drops by 10.1℃and current efficiency increases by 4.05%.