基于归一化最小均方算法的自适应核RBFNN

doi:10.13190/j.jbupt.2021-132

北京邮电大学学报 ›› 2022, Vol. 45 ›› Issue (2): 29-35.doi: 10.13190/j.jbupt.2021-132

基于归一化最小均方算法的自适应核RBFNN

火元莲¹, 巩琪¹, 齐永锋², 安娅琦¹

1. 西北师范大学物理与电子工程学院, 兰州 730070;
2. 西北师范大学计算机科学与工程学院, 兰州 730070

收稿日期:2021-06-24 发布日期:2021-12-16
作者简介:火元莲(1973—),女,副教授,硕士生导师,邮箱:hylqqq@nwnu.edu.cn。
基金资助:
国家自然科学基金项目(61561044)

Adaptive Kernel RBFNN Based on Normalized Least Mean Square Algorithm

HUO Yuanlian¹, GONG Qi¹, QI Yongfeng², AN Yaqi¹

1. College of Physics and Electronic Engineering, Northwest Normal University, Lanzhou 730070, China;
2. College of Computer Science and Engineering, Northwest Normal University, Lanzhou 730070, China

Received:2021-06-24 Published:2021-12-16

摘要/Abstract

摘要： 为了使自适应核径向基函数神经网络(RBFNN)有更好的收敛速度和稳态误差,提出了以归一化最小均方为学习算法对自适应核RBFNN进行优化的方法。在梯度下降算法的基础上,通过一个可变的步长因子,对归一化最小均方(NLMS)算法进行推导,并将其作为学习算法对自适应核RBFNN的权系数及偏差进行更新训练。在非线性系统辨识及模式分类中的仿真实验结果表明,使用NLMS学习算法训练自适应核RBFNN相较于其他学习算法下的自适应核RBFNN,具有更快的收敛速度及相对较小的稳态误差。

关键词: 自适应滤波, RBF神经网络, 归一化最小均方算法, 非线性系统辨识

Abstract: To make the adaptive kernel radial basis function neural network (RBFNN) exhibit the characteristics of fast convergence and steady-state error, a method that optimizes the adaptive kernel RBFNN by using the normalized least mean square as the learning algorithm is proposed. Based on the gradient descent algorithm, we derive the normalized least mean square (NLMS) algorithm with a variable step factor, and use it as a learning algorithm to update the weights and the biases of the adaptive kernel RBFNN. The simulation results in nonlinear system identification and pattern classification show that using NLMS learning algorithm to train adaptive kernel RBFNN has faster convergence speed and relatively less steady-state error compared with other learning algorithms.

Key words: adaptive filtering, radial basis function neural network, normalized least mean square algorithm, nonlinear system identification

中图分类号:

TN713

火元莲, 巩琪, 齐永锋, 安娅琦. 基于归一化最小均方算法的自适应核RBFNN[J]. 北京邮电大学学报, 2022, 45(2): 29-35.

HUO Yuanlian, GONG Qi, QI Yongfeng, AN Yaqi. Adaptive Kernel RBFNN Based on Normalized Least Mean Square Algorithm[J]. Journal of Beijing University of Posts and Telecommunications, 2022, 45(2): 29-35.

参考文献

[1] AFTAB W, MOINUDDIN M, SHAIKH M S. A novel kernel for RBF based neural networks[J/OL]. Abstract and Applied Analysis, 2014[2021-05-20]. https://www.hindawi.com/journals/aaa/2014/176253.
[2] YIN X, ZHANG Q C, WANG H, et al. RBFNN-based minimum entropy filtering for a class of stochastic nonli-near systems[J]. IEEE Transactions on Automatic Control, 2020, 65(1):376-381.
[3] 翟莹莹,左丽,张恩德.基于参数优化的RBF神经网络结构设计算法[J].东北大学学报(自然科学版), 2020, 41(2):176-181, 187. ZHAI Y Y, ZUO L, ZHANG E D. Algorithm for structure design of RBF neural network based on parameter optimization[J]. Journal of Northeastern University (Natural Science), 2020, 41(2):176-181, 187.
[4] HAN H G, MA M L, QIAO J F. Accelerated gradient algorithm for RBF neural network[J]. Neurocomputing, 2021, 441:237-247.
[5] ZENG X P, ZHAO H Q, JIN W D, et al. Identification of nonlinear dynamic systems using convex combinations of multiple adaptive radius basis function networks[J]. Measurement, 2013, 46(1):628-638.
[6] SUN Y, CAO M H, SUN Y Q, et al. Uncertain data stream algorithm based on clustering RBF neural network[J/OL]. Microprocessors and Microsystems, 2021, 81[2021-05-20]. https://www.sciencedirect.com/science/article/pii/S0141933120308760.
[7] KHAN S, NASEEM I, TOGNERI R, et al. A novel adaptive kernel for the RBF neural networks[J]. Circuits, Systems, and Signal Processing, 2017, 36(4):1639-1653.
[8] ATIF S M, KHAN S, NASEEM I, et al. Multi-kernel fusion for RBF neural networks[EB/OL].2020(2020-07-26)[2021-06-10]. https://arxiv.org/abs/2007.02592.
[9] KHAN S, AHMAD J, SADIQ A, et al. Spatio-temporal RBF neural networks[C]//20183rd International Conference on Emerging Trends in Engineering, Sciences and Technology (ICEEST). Piscataway, NJ:IEEE Press, 2018:238-246.
[10] KHAN S, NASEEM I, MALIK M A, et al. A fractional gradient descent-based RBF neural network[J]. Circuits, Systems, and Signal Processing, 2018, 37(12):5311-5332.
[11] 张兴兰,曹长修,梅彬.一种新型径向基函数神经网络学习算法:递归正交最小二乘法(ROLS)[J].重庆大学学报(自然科学版), 2002, 25(10):56-60. ZHANG X L, CAO C X, MEI B. A new learning algorithm based on radial basis function neural network:recursive orthogonal least squares (ROLS) algorithm[J]. Journal of Chongqing University (Natural Science Edition), 2002, 25(10):56-60.
[12] HAN H G, QIAO J F. Adaptive computation algorithm for RBF neural network[J]. IEEE Transactions on Neural Networks and Learning Systems, 2012, 23(2):342-347.
[13] 迪尼兹.自适应滤波算法与实现[M].北京:电子工业出版社, 2014:1-67. DINIZ. Adaptive filtering algorithms and practical implementation[M]. Beijing:Publishing House of Electronics Industry, 2014:1-67.
[14] KHAN S, AHMED N, MALIK M A, et al. FLMF:fractional least mean fourth algorithm for channel estimation in non-Gaussian environment[C]//2017 International Conference on Information and Communication Technology Convergence (ICTC). Piscataway, NJ:IEEE Press, 2017:466-470.

基于归一化最小均方算法的自适应核RBFNN

Adaptive Kernel RBFNN Based on Normalized Least Mean Square Algorithm

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 2

编辑推荐

Metrics

本文评价

[1]	火元莲, 连培君, 齐永锋, 王丹凤. 一种改进的凸组合仿射投影符号算法[J]. 北京邮电大学学报, 2022, 45(1): 102-107.
[2]	曾乐雅, 许华, 王天睿. 改进的凸组合最小均方算法[J]. 北京邮电大学学报, 2016, 39(4): 114-117.