摘要
研制的精密运动平台同时具有死区与迟滞的特性,普通的辨识算法难以有效处理。本文设计了一种具有特殊激励函数和结构的神经网络,可以同时描述精密运动平台的非光滑死区和迟滞特性,并引入广义梯度的概念来改进L-M训练算法,实现对所构建神经网络的建模训练。实际辨识结果表明所提出的辨识方法取得了令人满意的结果。
Due to the non-smooth nonlinear dead-zone and hysteresis, it is difficult to use the traditional identification method to identify the model of the precision positioning stage. In this paper, a special neural network with non-smooth activation function is adopted to model the non-smooth nonlinear dead zone coupled with hysteresis in the stage. Moreover, the generalized gradient is introduced into the Levenberg- Marquardt algorithm so that the non- smooth neural network can be trained. The experimental results are presented to illustrate the performance of the proposed identification scheme.
作者
谢扬球
曹泽衍
李林东
陆鑫培
于东鑫
Xie Yangqiu;Cao Zeyan;Li Lindong;Lu Xinpei;Yu Dongxin(College of Material Science and Engineering,Guangxi University,Nanning 530004,China;Guangxi Experiment Centre of Science and Technology,Nanning 530004,China)
出处
《科技通报》
北大核心
2017年第2期77-80,共4页
Bulletin of Science and Technology
基金
国家自然科学基金(61563003)
广西理工科学实验中心课题(YXKT2014028)
有色金属及材料加工新技术教育部重点实验室开放基金(GXKFJ11-0213-A-02-05)
广西大学"大学生创新创业训练计划"(201410593153
201510593148
201610593187)
关键词
神经网络
非光滑非线性
精密运动平台
neural network
non-smooth nonlinearity
precision positioning stage