Research On Longitudinal Register Control In Roll-To-roll Process

Register accuracy is the most important index in the electronic axis gravure printing systems,and also is a key factor in determining product quality and reducing production costs.In addition,most domestic high-end electronic gravure printing machines are imported,and the control system is also a foreign high-precision system.At present,most of the industrial applications of electronic axis gravure printing systems are still dominated by PID control.In order to improve the register accuracy to satisfy the production needs of higher-precision products,such as the printing of solar thin film batteries and flexible circuit boards,this paper studies the longitudinal register control method of electronic axis gravure printing systems in the printing production stage at a stable speed.This paper mainly starts from two mathematical models of the electronic axis gravure printing systems.Then,two validated reliable mathematical models are analyzed respectively,and different control strategies are adopted to finish the register control of the electronic axis gravure printing machine system according to the two mathematical models.One of the control methods is to design a reduced-order observer based on the mathematical model,the closedloop analysis and algorithm comparison analysis prove that the observer estimate of this method can converge,and the full state feedback control is completed according to the state estimated by the observer.Based on the state observer,the linear quadratic regulator(LQR)algorithm is used to complete the final state feedback control,and the particle swarm optimization algorithm is used to optimize the parameters of the LQR control algorithm.The simulation results show that the optimized LQR control effect has greatly improved at the adjustment time and register accuracy of the register control.The another of the control methods is to correct the mathematical model based on the idea of model error modeling to obtain an accurate mathematical model which is closer to the actual system.And then we design an adaptive model reference algorithm based on this accurate mathematical model.Compared with the existing algorithms,the adjustment time and control accuracy of the system are both improved.For the two control algorithms designed in this paper,the first control method requires high computing power,but its performance also improves with the improvement of computing power,so it is likely to be widely used in the rapid iteration of future science and technology;the second control method can be quickly applied in the system with a large number of open-loop data,and can improve the control effect in a short time.