摘要
针对印刷机墨量控制的非线性特性,提出了一种基于改进BP神经网络的PID控制算法。运用神经元的自学习、自适应特点,对检测的印刷品色差缺陷进行实时墨量控制。仿真实验表明:该控制方法具有良好的动、静态性能和较强的鲁棒性与自适应性。
Aimed at the non-linear characteristics of printer ink control,a PID control algorithm based on improved BP neural network was presented,and real-time ink control for the color defect of printed matter is realized by applying the self-learning and adaptive features of neuron.The result of simulation test shows that the control method has good steady and dynamic performance as well as strong robustness and adaptability.
出处
《包装工程》
CAS
CSCD
北大核心
2010年第19期28-31,42,共5页
Packaging Engineering
关键词
墨量控制
神经网络
自学习
ink control
neural network
self-learning
