Research On Modeling And Optimal Control Of The Casting Billet Induction Heating Process

At present, continuous casting billet Direct Hot Charge Rolling is an energy-savingtechnology that widely used in steel enterprises, and the hot delivery process generally usesthe gas furnace or the induction heating furnace to make temperature compensation. In recentyears, electromagnetic induction heating technology has the advantages of less energy, fastheating speed, less pollution and easier control, so its application has been widely promoted.However, the temperature control on the continuous casting billet induction heating process ismainly relied on experience to adjust in the actual production, and applied research on thisaspect are still few.This thesis mainly studies the induction heating process of the continuous casting billetin Direct Hot Charge Rolling, analyzes the heat transfer mechanism of steel billet, andattempts to establish a finite difference model and a system identification prediction model ofthe continuous casting billet temperature with the actual production conditions. On the basisof the system identification model, a random search algorithm and a genetic algorithm areused to adjust the control voltage so as to make the casting billet temperature reach the targettemperature after induction heating, and then realize the temperature optimal control ofcontinuous casting billet induction heating process. The conclusion of this master’s thesis is asfollows:(1) Using ANSYS finite element model to simulate the casting billet induction heatingprocess is a common method, with high precision but long time-consuming, so it’s unable tomeet the needs of online application. Based on the finite difference method, the predictionmodel established in this thesis predicts the temperature distribution of the billet cross sectionin the induction heating process, and the final prediction data accords with the ANSYSsimulation data basically shows the prediction model has high precision. And it alsoovercomes the low calculation speed of ANSYS software, so it can provides the reference forrealizing the temperature prediction and control of the continuous casting billet inductionheating process.(2) Using system identification method establishes the billet induction heating processARX model. Based on the induction heating principle, this model utilizes actual data toimplement model identification, and validates model by new observation data. Withforecasting the surface temperature of the billet, the results show that this model has highprediction accuracy and it lays the model foundation for the temperature optimal control. (3) In the actual production, the induction heating control parameters is determined byoperators’ experience commonly, the heated billet is easily to cause uneven temperaturedistribution. On the basis of the ARX model built by system identification, a random searchalgorithm and a genetic algorithm are used to adjust the control voltage, and the resultingcontrol voltage curve can effectively optimize the billet discharging temperature distributionso as to realize the temperature optimal control of the continuous casting billet inductionheating.