[1]张超,严洪森.基于最优结构多维泰勒网的含噪声非线性时变系统辨识[J].东南大学学报(自然科学版),2017,47(6):1086-1093.[doi:10.3969/j.issn.1001-0505.2017.06.002]
 Zhang Chao,Yan Hongsen.Identification of nonlinear time-varying system with noise based on multi-dimensional Taylor network with optimal structure[J].Journal of Southeast University (Natural Science Edition),2017,47(6):1086-1093.[doi:10.3969/j.issn.1001-0505.2017.06.002]
点击复制

基于最优结构多维泰勒网的含噪声非线性时变系统辨识()
分享到:

《东南大学学报(自然科学版)》[ISSN:1001-0505/CN:32-1178/N]

卷:
47
期数:
2017年第6期
页码:
1086-1093
栏目:
材料科学与工程
出版日期:
2017-11-20

文章信息/Info

Title:
Identification of nonlinear time-varying system with noise based on multi-dimensional Taylor network with optimal structure
作者:
张超123严洪森13
1东南大学自动化学院, 南京 210096; 2河南工学院计算机科学与技术系, 新乡 453003; 3东南大学复杂工程系统测量与控制教育部重点实验室, 南京 210096
Author(s):
Zhang Chao123 Yan Hongsen13
1School of Automation, Southeast University, Nanjing 210096, China
2Department of Computer Science and Technology, Henan Institute of Technology, Xinxiang 453003, China
3Key Laboratory of Measurement and Control of Complex Systems of Engineering, Ministry of Education, Southeast University, Nanjing 210096, China
关键词:
辨识 非线性时变系统 多维泰勒网 噪声干扰 剪枝算法
Keywords:
identification nonlinear time-varying system multi-dimensional Taylor network noise disturbance pruning algorithm
分类号:
TB391
DOI:
10.3969/j.issn.1001-0505.2017.06.002
摘要:
针对具有噪声干扰的非线性时变系统建模时存在的困难,建立了一种具有最优结构和最佳泛化能力的多维泰勒网模型,以实现对该系统的辨识.首先,为了能够快速反映系统输入/输出的变化,以多维泰勒网的连接权系数作为时变参数,并由带可变遗忘因子的递推最小二乘算法对其进行训练,进而讨论了辨识方案的稳定性.然后,为了避免维数灾难并满足实时性要求,给出了选择多维泰勒网有效回归项的改进权衰减法,以使多维泰勒网同时具有最小结构和最佳的泛化能力.最后,通过算例说明基于最优结构的多维泰勒网在含噪声非线性时变系统辨识问题中应用的方法,算例结果验证了该方法的有效性.
Abstract:
Aiming at the modeling difficulties of the nonlinear time-varying system with noise disturbance, a multi-dimensional Taylor network(MTN)model with optimal structure and optimum generalization ability is established to implement the identification of the system. Firstly, to rapidly reflect the input-output changes of the system, the link weight coefficients of MTN are taken as the time-varying parameters, and then the recursive least-squares algorithm with a variable forgetting factor is adopted to train the system and the stability of the identification scheme is addressed. Secondly, to avoid the dimension curse and meet the real-time requirements, an improved pruning algorithm is developed to choose the effective regression items of MTN, which provides the network with the simplest structure and optimum generalization ability. Finally, an example is given to illustrate the application of the MTN with minimum structure in the identification of a nonlinear time-varying system with noise disturbance, and the experimental results demonstrate the effectiveness of the proposed method.

参考文献/References:

[1] Hunt K J, Sbarbaro D, Z·bikowski R, et al. Neural networks for control systems—A survey[J]. Automatica, 1992, 28(6): 1083-1112. DOI:10.1016/0005-1098(92)90053-i.
[2] Peng J, Dubay R. Identification and adaptive neural network control of a DC motor system with dead-zone characteristics[J]. ISA Trans, 2011, 50(4): 588-598. DOI:10.1016/j.isatra.2011.06.005.
[3] 林屹, 严洪森. 多尺度正负反馈交替论模型及其应用 [J]. 控制理论与应用, 2016, 33(7): 879-888. DOI:10.7641/CTA.2016.50742.
Lin Yi, Yan Hongsen. The model of multi-scale alternate positive negative feedbackics and its applications [J]. Control Theory & Applications, 2016, 33(7): 879-888. DOI:10.7641/CTA.2016.50742. (in Chinese)
[4] 周博, 严洪森. 一类具有羊群效应的多重多维泰勒网动力学模型[J]. 控制理论与应用, 2015, 32(7): 963-969. DOI:10.7641/CTA.2015.40628.
Zhou Bo, Yan Hongsen. A class of multiple multi-dimensional Taylor networks dynamics model with herd behavior[J]. Control Theory & Applications, 2015, 32(7): 963-969. DOI:10.7641/CTA.2015.40628. (in Chinese)
[5] Li J, Zheng Y, Lin Z. Recursive identification of time-varying systems: Self-tuning and matrix RLS algorithms[J]. Systems & Control Letters, 2014, 66: 104-110. DOI:10.1016/j.sysconle.2014.01.004.
[6] Qin B, Cai Y, Champagne B, et al. A low-complexity variable forgetting factor constant modulus RLS algorithm for blind adaptive beamforming[J]. Signal Processing, 2014, 105: 277-282. DOI:10.1016/j.sigpro.2014.06.003.
[7] Kuo R J, Wu P, Wang C P. An intelligent sales forecasting system through integration of artificial neural networks and fuzzy neural networks with fuzzy weight elimination[J]. Neural Networks, 2002, 15(7): 909-925. DOI:10.1016/s0893-6080(02)00064-3.
[8] Xie L, Wei H, Zhao J, et al. Automatic feature extraction based structure decomposition method for multi-classification[J]. Neurocomputing, 2016, 173: 744-750. DOI:10.1016/j.neucom.2015.08.025.
[9] 朱红路. 热工信号自适应滤波及其在建模与控制中的应用 [D]. 北京: 华北电力大学控制与计算机工程学院, 2010.
[10] Plett G L. Adaptive inverse control of linear and nonlinear systems using dynamic neural networks [J]. IEEE Transactions on Neural Networks, 2003, 14(2): 360-376. DOI:10.1109/TNN.2003.809412.
[11] 林屹, 严洪森, 周博. 基于多维泰勒网的自适应混沌时间序列多步预测[J]. 东南大学学报(自然科学版), 2015, 45(2): 281-288. DOI:10.3969/j.issn.1001-0505.2015.02.016.
Lin Yi, Yan Hongsen, Zhou Bo. Adaptive multi-step prediction of chaotic time series based on multi-dimensional Taylor network[J]. Journal of Southeast University(Natural Science Edition), 2015, 45(2): 281-288. DOI:10.3969/j.issn.1001-0505.2015.02.016. (in Chinese)
[12] Klambauer G. Mathematical analysis [M]. New York: Marcel Dekker Inc, 1975: 236-237.
[13] 林屹, 严洪森, 周博. 基于多维泰勒网的非线性时间序列预测方法及其应用[J]. 控制与决策, 2014,29(5): 795-801. DOI:10.13195/j.kzyjc.2013.0198.
Lin Yi, Yan Hongsen, Zhou Bo. Nonlinear time series prediction method based on multi-dimensional Taylor network and its applications[J]. Control and Decision, 2014, 29(5): 795-801. DOI:10.13195/j.kzyjc.2013.0198. (in Chinese)
[14] Kreiberg D, Söderström T, Yang-Wallentin F. Errors-in-variables system identification using structural equation modeling[J]. Automatica, 2016, 66: 218-230. DOI:10.1016/j.automatica.2015.12.007.
[15] Wawrzynski P, Tanwani A K. Autonomous reinforcement learning with experience replay [J]. Neural Networks, 2013, 41: 156-167. DOI:10.1016/j.neunet.2012.11.007.

相似文献/References:

[1]尹家洪.一种用Fourier级数辨识延时线性系统的算法[J].东南大学学报(自然科学版),1993,23(6):24.[doi:10.3969/j.issn.1001-0505.1993.06.004]
 Yin Jiahong.Algorithm to the Parameter Estimation of Linear Delayed Systems Via Fourier Series[J].Journal of Southeast University (Natural Science Edition),1993,23(6):24.[doi:10.3969/j.issn.1001-0505.1993.06.004]
[2]黄东,宋文忠.一类递推辨识算法的评价和比较[J].东南大学学报(自然科学版),1989,19(4):90.[doi:10.3969/j.issn.1001-0505.1989.04.013]
 Huang Dong Song Wenzhong Research Institute of Automation.Evaluation and Comparison ot A Class ot Recursive Identification Algorithms[J].Journal of Southeast University (Natural Science Edition),1989,19(6):90.[doi:10.3969/j.issn.1001-0505.1989.04.013]

备注/Memo

备注/Memo:
收稿日期: 2017-06-22.
作者简介: 张超(1983—),男,博士生,讲师;严洪森(联系人),男,博士,教授,博士生导师, hsyan@seu.edu.cn.
基金项目: 国家自然科学基金资助项目(61673112, 60934008)、中央高校基本科研业务费专项资金资助项目(2242017K10003, 2242014K10031)、江苏高校优势学科建设工程资助项目.
引用本文: 张超,严洪森.基于最优结构多维泰勒网的含噪声非线性时变系统辨识[J].东南大学学报(自然科学版),2017,47(6):1086-1093. DOI:10.3969/j.issn.1001-0505.2017.06.002.
更新日期/Last Update: 2017-11-20