Research On Coating Weight Control System For Hot-Dip Galvanizing Steel Based On MFAC

With the continuous improvement of the national economy and the comprehensive national strength,China's iron and steel smelting technology develop rapidly,but some links need to be improved and optimized.The thickness of galvanized steel strip is the key factor affecting the durability of steel.In this thesis,the thickness of galvanized steel is taken as the research object,the existing control system of zinc coating thickness is analyzed.This thesis analyzes the key equipment of galvanized layer thickness control including details of the structure and mechanism of the air knife.Features of several common brands of air knife are compared.Thickness gauge is galvanized layer thickness testing equipment.Due to the zinc pot temperature is too high,the gauge placed about 120 meters away from the air knife,resulting in a long delay in the control system.There is no universal method to overcome the negative impact.This thesis analyzes the control scheme combining predictor control and Smith feedforward control.This method has some limitations because it relies too much on the exact model.Because of the complexity of zinc plating process,this thesis uses a model-free adaptive control algorithm(MFAC)that does not rely on mathematical models.Model-free self-adaptation is reflected in the fact that the MFAC can calculate control law with the structure,order and parameters of the controlled object unknown.The controller uses the pseudo-gradient parameter which is time-varying to simulate the controlled object locally through a pan-model.The pseudo-gradient is constantly adjusted according to the input and output of the system to improve the estimation accuracy of the model.Galvanization process has a long delay caused by the measuring instrument.The thesis uses the estimated compensation method to improve the basic MFAC algorithm,and uses the approximate differential of the past time to estimate the system output in the future to overcome the impact of the big lag.The controller accepts the deviation of the expected output and the estimated compensation for the control law calculation.This kind of estimated compensation method can effectively solve the controller design problem for the object with large lag.In order to deal with the complicated disturbance existing in galvanization process,the thesis uses the tracking differentiator to filter the output of the system,which can effectively restrain the influence of the noise signal on the control law calculation.The simulation results show that the anti-delay and anti-disturbance MFAC controller is suitable for galvanizing process,and the controlled object can quickly track the change of thickness target value.