An Efficient Self-Organizing Deep Fuzzy Neural Network for Nonlinear System Modeling

被引:43
作者
Wang, Gongming [1 ]
Qiao, Junfei [1 ]
机构
[1] Beijing Univ Technol, Fac Informat Technol, Beijing 100124, Peoples R China
基金
中国国家自然科学基金;
关键词
Fuzzy neural networks; Feature extraction; Training; Data models; Adaptation models; Neurons; Fuzzy control; Fuzzy neural network; incremental deep pretraining (IDPT); nonlinear system modeling; self-organizing learning; BELIEF NETWORK;
D O I
10.1109/TFUZZ.2021.3077396
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
A fuzzy neural network (FNN) is an effective learning system that combines neural network and fuzzy logic, which has achieved great success in nonlinear system modeling. However, when the input is practical complex data with external disturbance, the existing FNN cannot extract effective input features sufficiently, leading to unsatisfactory performances in learning speed and accuracy. It also fails to achieve a better generalization capability because of its fixed structure size (the number of rule neurons). In this article, an efficient self-organizing FNN (SOFNN) with incremental deep pretraining (IDPT), called IDPT-SOFNN, is developed to overcome these shortcomings. First, IDPT is designed to extract effective features and consider them as the input of the SOFNN. Different from the existing pretraining, the self-growing structure of IDPT improves pretraining efficiency with a more compact structure. Second, the SOFNN can dynamically add and delete neurons according to the current error and error-reduction rate. In this case, it can obtain better modeling performance with a more compact structure as well. Third, as a novel hybrid model with the cascade dual-self-organizing algorithm, the IDPT-SOFNN combines the advantage of IDPT and SOFNN. Moreover, the convergence and stability are analyzed. Finally, simulation studies and comparisons demonstrate that the proposed IDPT-SOFNN has better performances than its peers in learning speed, accuracy, and generalization capability.
引用
收藏
页码:2170 / 2182
页数:13
相关论文
共 36 条
[1]   Air pollutants concentrations forecasting using back propagation neural network based on wavelet decomposition with meteorological conditions [J].
Bai, Yun ;
Li, Yong ;
Wang, Xiaoxue ;
Xie, Jingjing ;
Li, Chuan .
ATMOSPHERIC POLLUTION RESEARCH, 2016, 7 (03) :557-566
[2]   Data preprocessing for river flow forecasting using neural networks: Wavelet transforms and data partitioning [J].
Cannas, Barbara ;
Fanni, Alessandra ;
See, Linda ;
Sias, Giuliana .
PHYSICS AND CHEMISTRY OF THE EARTH, 2006, 31 (18) :1164-1171
[3]   A self-organizing neuro-fuzzy network based on first order effect sensitivity analysis [J].
Chen, Cheng ;
Wang, Fei-Yue .
NEUROCOMPUTING, 2013, 118 :21-32
[4]  
Chen T., 2015, COMPUT SCI
[5]   SOFMLS: Online Self-Organizing Fuzzy Modified Least-Squares Network [J].
de Jesus Rubio, Jose .
IEEE TRANSACTIONS ON FUZZY SYSTEMS, 2009, 17 (06) :1296-1309
[6]   A Hierarchical Fused Fuzzy Deep Neural Network for Data Classification [J].
Deng, Yue ;
Ren, Zhiquan ;
Kong, Youyong ;
Bao, Feng ;
Dai, Qionghai .
IEEE TRANSACTIONS ON FUZZY SYSTEMS, 2017, 25 (04) :1006-1012
[7]   A data ensemble approach for real-time air quality forecasting using extremely randomized trees and deep neural networks [J].
Eslami, Ebrahim ;
Salman, Ahmed Khan ;
Choi, Yunsoo ;
Sayeed, Alqamah ;
Lops, Yannic .
NEURAL COMPUTING & APPLICATIONS, 2020, 32 (11) :7563-7579
[8]   Fuzzy Integrated Cell Formation and Production Scheduling Considering Automated Guided Vehicles and Human Factors [J].
Goli, Alireza ;
Tirkolaee, Erfan Babaee ;
Aydin, Nadi Serhan .
IEEE TRANSACTIONS ON FUZZY SYSTEMS, 2021, 29 (12) :3686-3695
[9]   Efficient self-organizing multilayer neural network for nonlinear system modeling [J].
Han, Hong-Gui ;
Wang, Li-Dan ;
Qiao, Jun-Fei .
NEURAL NETWORKS, 2013, 43 :22-32
[10]   Nonlinear Systems Modeling Based on Self-Organizing Fuzzy-Neural-Network With Adaptive Computation Algorithm [J].
Han, Honggui ;
Wu, Xiao-Long ;
Qiao, Jun-Fei .
IEEE TRANSACTIONS ON CYBERNETICS, 2014, 44 (04) :554-564