Research On Automatic Registration System Of Gravure Press Based On BP Neural Network PID Control

The automatic registration control system is the key device of gravure printing electronic equipment,which plays an important role in improving the accuracy of printing overprinting and reducing the rejection rate.With the gradual development of digital,intelligent and green printing equipment in China,the calibration device based on traditional PID controller has been unable to meet the needs of printing electronics for high-speed and high-precision production.Based on the above background,the automatic alignment model of gravure press and the control strategy based on BP neural network PID are deeply studied in this paper.The main contents are as follows.(1)The cause of alignment error of gravure press was studied,and the two-color alignment error model of gravure press was established.The mathematical relationship between alignment error and belt tension and speed in alignment system was revealed.(2)Aiming at the problem of low convergence efficiency and easy to fall into local minimum of conventional BP neural network,an adaptive learning rate algorithm is proposed according to the convergence speed of error function of neural network.Experiments show that the algorithm has good convergence characteristics.According to the descent of error function of neural network,a dynamic adjustable small batch gradient descent algorithm is proposed.Experiments show that the algorithm has high convergence speed and accuracy.(3)According to the characteristics of BP neural network PID controller and gravure press alignment error model,the alignment controller based on BP neural network is designed and the simulation model of the controller is established.Under different external disturbances,the BP neural network PID controller and the conventional PID controller are compared and simulated.The results show that the BP neural network PID controller has better control performance.(4)With STM32F407 as the control core,through hardware selection and the design of registration image detection algorithm based on Hough transform,combined with BP neural network PID controller model,an automatic registration control physical system was built on a two-color gravure press.The experimental results show that the PID controller based on BP neural network designed in this paper can be applied to the automatic registration system of gravure printing equipment at low speed.