摘要
将小波神经网络引入到FMS的优化配置工作中,结合实验设计、仿真实验和神经网络训练,提出对高度非线性柔性制造系统建模的一种新方法。文中介绍了该方法的实现原理,通过实例验证用于非线性函数逼近的多维小波神经网络算法的正确性,并针对一个工程实例,说明这种建模方法的可行性。
Wavelet neural networks is introduced into the optimizing configuration of FMS.A new ap proach is put forward to deal with the high nonlinear modeling problem by using processes design, simulating and neural networks. The principle is presented and the ability of wave- let neural networks on nonlinear function approximation is proved by an example.
出处
《制造业自动化》
2004年第12期16-19,70,共5页
Manufacturing Automation
基金
国家自然科学基金资助(50175001)