摘要
利用神经元网络的逼近能力,采用一般训练与特别训练相结合的方法对神经元网络进行训练,求得了机床伺服系统的前馈补偿控制器,弥补了系统的动态延迟.计算机仿真实验表明,采用这种前馈加反馈的监督控制方法,不仅可确保控制系统的稳定性和鲁棒性,而且可有效地提高系统的精度和自适应能力.
A neural network based servomechanism for CNC machine tool is presented in the paper.The method incorporating the generalized learning with the specialized learning is adopted to train the network.The simulation results indicate that by adopting such a control and monitor method as combining feedforward with feedback,the stability and robustness of the system are guaranteed,and the precision and adaptability of the system are effectively improved.
出处
《沈阳工业大学学报》
EI
CAS
2004年第1期64-67,共4页
Journal of Shenyang University of Technology
关键词
神经元网络
特别训练
伺服系统
数控机床
neural network
specialized training
servomechanism
CNC machine tool
