Research On Temperature Control Method Of Continuous Annealing Furnace

As the improtant equipment of the continuous hot-dip galvanizing process,continuous annealing furnace can directly affect the quality and galvanizing effect of strip steel.Temperature control is an important part of the annealing furnace control system.Effective temperature control of the furnace can significantly improve product quality and production efficiency,reduce energy consumption and achieve the goal of energy saving and environmental protection.Because of the characteristics of non-linear temperature control,large time delay and parameter time-varying,the traditional PID control strategy is difficult to meet the control requirements.Therefore,designing a good control method to improve the control effect of furnace temperature is of great importance to production.Based on the actual situation,taking the high strength steel modification project of the third cold rolling plant of Benxi Iron and Steel as the background,taking the furnace of the 1870 continuous hot-dip galvanizing line as the research object,the process characteristics of the furnace were studied in detail.In this thesis,the furnace temperature control loop,including the furnace temperature control system and gas combustion control system,were in-depth analysised.The controller design and control effect verification need to be based on the object model.In the process of modeling the furnace temperature control system,taking into account the furnace temperature control system of large inertia,pure lag,strong coupling and parameters change and so on,the system is identified by RBF neural network optimized by orthogonal least square method.This thesis analyzed the factors that affect the furnace temperature,choose the gas and combustion air flow as the main input of the system.OLS_RBF is used to train and verify the network model through a large amount of data collected in the field.The generalization results showed that the furnace temperature model established by OLS_RBF neural network had a good identification effect.Based on the mathematic model of the furnace temperature,the advantages and disadvantages of the differential prior PID control method in the original control system were analyzed according to the process characteristics and control difficulties of the furnace temperature.A Dahlin algorithm controller based on the CMAC neural network is designed.This controller combines the non-overshooting stability control of the Dahlin algorithm with the learning ability of the CMAC neural network,making up for the drawbacks of the long settling time in Dahlin controller simulation.Through the simulation experiment,the thesis compared the control effects of differential prior PID control,Dahlin control and CMAC_Dahlin control respectively.The results show that the CMAC_Dahlin controller has the advantages of small overshoot,stable and short adjustment time.This method also avoided the cumbersome parameter setting process of the original differential PID control method.