Research On Optimization Control Of Smelting Process Of Fused Magnesium Furnace

Fused magnesia,a kind of widely-used refractory material,is mainly produced by the electrical fused magnesia furnace.The smelting process of electrical fused magnesia furnace is characterized by the nonlinearity,strong coupling,frequent changes in production boundary conditions and severe random disturbance.It is hard to accurately adjust the position of three-phase electrode to control the melting current value at the optimum value only using the experience of workers,leading to low product output of fused magnesia and high energy consumption per ton.In this thesis,the optimal control method for the smelting process of electrical fused magnesia furnace was designed,and the appropriate current value was given,so as to effectively reduce the energy consumption per ton and increase the product output of fused magnesia.The smelting process of electrical fused magnesia furnace can be considered as a typical batch production process,in which the iterative learning control method can fully explore the iterative character of the process and update the signal in the time dimension and batch dimension.On the basis of in-depth analysis and research on smelting process of electrical fused magnesia furnace,this thesis proposed an iterative learning control-based optimal control framework for the smelting process of electrical fused magnesia furnace using the iterative learning control idea.First,combined with the operating experience of smelting process,the in-batch optimization model based on case-based reasoning was designed for the frequent changes in working conditions and raw material boundary conditions during the smelting process of electrical fused magnesia furnace in the batch.To optimize the iterative learning control timer shaft,the unit consumption deviation predicted at the previous time point was introduced into the case description,which was used to guide the setting of current value at this moment.In view of the characteristics of smelting process of electrical fused magnesia furnace,the case retrieval,case reuse,case correction and case storage strategies were designed.The case attribute weight given using the traditional manual trial-and-error method may produce the inappropriate case solution,so the case attribute weight was solved using the similar rough set theory in case retrieval.Finally,the simulation experiment was performed to verify the effectiveness of model.Second,the inter-batch optimization model based on PI-type iterative learning control algorithm was designed in view of the problems of uncertain disturbance and uncontrollable random disturbance among different batches of electrical fused magnesia furnace.To optimize the iterative learning control timer shaft,the PI-type iterative learning control algorithm was adopted in this thesis,and the actual unit consumption deviation of the previous batch was used to guide the setting of current value of this batch.Finally,the simulation experiment was performed combined with the in-batch optimization module and inter-batch optimization module,thus verifying the effectiveness of optimal control model of the smelting process of electrical fused magnesia furnace based on iterative learning control.The experimental results demonstrate that the model established in this thesis can effectively reduce the energy consumption per ton and improve the product output of fused magnesia.