Research On Automatic Ink-adjusting System Based On Artificial Neural Networks

With increasing of the requirement of printing's quality and printing speed,traditional ways to adjust ink can not meet the real-time controlling and the needs for high-quality printed materials,so automatic printing ink-adjusting system has become an important part for the online quality detection and control system.High-performance automatic printing ink adjusted system is an effective way which can reduce scraps and improve print quality.The automatic printing ink-adjusting system based on artificial neural network was designed.With the data of printing's quality inspection,the calculation between color and ink can be completed.Besides these,a new control strategy based on neural network is presented for the motor of ink-adjusting.In this paper,overall scheme of automatic ink-adjusting system based on artificial neural networks is designed firstly.Each module's function is introduced in detail,and development and advantage of neural networks is discussed.Conversion module between color and ink and control module based on neural network are highlighted.Secondly,a model calculation software is established by Visual C++6.0 to the calculation of RGB-XYZ color space conversion.By analyzing traditional methods of color space conversion,a new method of color space conversion based on improved BP neural network is presented to realize the non-linear conversion between three-dimension color space and four-dimension color space of CMYK.The experiment's results show that the method can meet actual needs through neural network's training and testing.Finally,traditional control algorithm can not meet the needs of control because of the non-linear characteristic of the printing's ink control.So a new control algorithm based on neural network is presented to complete ink control in the paper.Brushless DC motors are control body of neural network PID controller,and performance of the neural network PID controller is analyzed and tested through experiments.The result of simulation shows that the system has not only good steady and dynamic performance but also high robustness and adaptability.