一种基于改进型深度学习的非线性建模方法  

A Nonlinear Modeling Method Based on Improved Deep Learning

作　　者：王盈旭 韩红桂 郭民 WANG Yingxu;HAN Honggui;GUO Min(College of Automation,Faculty of Information Technology,Beijing University of Technology,Beijing 100124,China;Beijing Key Laboratory of Computational Intelligence and Intelligent System,Beifing 100124,China;Engineering Research Center of Digital Community,Ministry of Education,Beijing 100124,China)

机构地区：[1]北京工业大学信息学部自动化学院,北京100124 [2]计算智能与智能系统北京市重点实验室,北京100124 [3]数字社区教育部工程研究中心,北京100124

出　　处：《信息与控制》2018年第6期680-686,共7页Information and Control

基　　金：国家自然科学基金资助项目(61622301);北京市自然科学基金资助项目(4172005);科技部水专项资助项目(2017ZX07104)

摘　　要：围绕非线性系统的建模问题,提出了一种基于改进型深度学习的非线性建模方法.首先,设计了基于高斯径向基函数的深度信念网络训练模型;其次,利用对比分歧算法对径向基函数的权值、中心和宽度进行调整,并利用反向传播对网络连接权值进行优化;最后,将获得的改进型深度学习方法应用于非线性系统建模.实验结果验证了该算法的有效性和可行性.To model nonlinear systems,we propose a nonlinear modeling method based on an improved deep learning algorithm. First,we design a learning model of a deep belief network based on the Gaussian radial basis function. Second,we adjust the weight,center,and width of the radial basis functions by a contrastive divergence algorithm and optimize the weights of the deep belief network using a back-propagation algorithm. Finally,we apply the improved deep learning algorithm to model nonlinear systems. The experimental results verify the effectiveness and feasibility of the algorithm.

关 键 词：深度学习 非线性建模 径向基函数 深度信念网络 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]