基于PID神经元网络的印刷品质量控制研究

Research on the printing quality control based on PID neural network

针对印刷品质量控制系统这种多输入多输出的复杂动态系统,选用具有较强解耦功能的多控制量PID神经元网络对金属薄板印刷机的墨键开度进行调节。从标准模板上获取4种颜色油墨的网点面积率作为目标值,在线检测到的网点面积率作为反馈值,一同作为控制系统的输入,系统输出为墨键开度的调整值,从而构成闭环控制,保证印刷品质量的稳定性。仿真结果表明,该控制方法具有较高的控制精度和较快的响应速度,能够满足印刷品质量控制的要求。

For the printing quality control system with multiple input and output variables,it proposes the method of multivariable PID neural network with strong decoupling function to adjust the ink key opening degree of metal sheet printing machine,obtains the dot area of four kinds of color ink from the standard template. Taking the dot area from online detect as the input to the control system,it realizes that the output of the system is the opening value of the ink key. This forms a closed loop control to ensure the stability of the quality of printing. The simulation results demonstrate that the method has high control precision and fast response speed,and meets the requirements of quality control of printing.