电熔镁炉熔炼过程的建模

东北大学硕士学位论文AbstractModeling of Electrical Smelting Furnace for Magnesia SmeltingProcessAbstractEnergy conservation is one of the trends of modern industrial development.Improvingenergy efficiency accords with the long-term interests of human development.However,limited to the current technical conditions,a number of industrial processes are essential tohigh pollution and high energy consumption,forcing people to study and improve theindustrial processes.Electrical-smelted magnesia is a sort of sophisticated fireproof material which has theproperties of high temperature resistibility and dense structure.It is an important material.Theobjective of this research is the production of electrical-smelted magnesia.It is widely used inmetallurgy and aerospace industry and other fields.The three-phase AC electrical smeltingfurnace for magnesia(ESFM)is an important equipment for producing the electrical-smeltedmagnesia.It belongs to a kind of mine hot electric arc furnace.The magnesite powder andlight-burning magnesium powder are heated by the furnace charge resistance and theelectrodes arc,this is a high energy consumption process.In addition,the process has featuressuch as time-varying,multivariable,strong coupling.Facing the problems,researches include the following related parts in this thesis:(1)The physical and chemical changes of furnace in normal operating conditions areresearched.And the principle of heat transfer between material batches in ESFM is analyzeddeeply in this thesis;(2)According to thermal principle and the actual site conditions,a current set approach isproposed through heat transfer calculating,which can conserve the energy.And the value ofthe consuming energy can be gottern;(3)Combining PLS and iterative learning control (ILC),adding constraints,and then usingparticle swarm optimization (PSO)to tracke the set-point value.And the simulation resultsshow that the algorithm is effective.Key words:Energy Conservation;Electrical Smelting Furnace for Magnesia (ESFM);PartialLeast Squares;Iterative Learning control-Ⅱ