Improved Initialization of State-Space Artificial Neural Networks

被引:0
|
作者
Schoukens, Maarten [1 ]
机构
[1] Eindhoven Univ Technol, Dept Elect Engn, NL-5612 AZ Eindhoven, Netherlands
基金
欧盟地平线“2020”;
关键词
NONLINEAR-SYSTEM IDENTIFICATION; MODELS;
D O I
暂无
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
The identification of black-box nonlinear state-space models requires a flexible representation of the state and output equation. Artificial neural networks have proven to provide such a representation. However, as in many identification problems, a nonlinear optimization problem needs to be solved to obtain the model parameters (layer weights and biases). A well-thought initialization of these model parameters can often avoid that the nonlinear optimization algorithm converges to a poorly performing local minimum of the considered cost function. This paper introduces an improved initialization approach for nonlinear state-space models represented as a recurrent artificial neural network and emphasizes the importance of including an explicit linear term in the model structure. Some of the neural network weights are initialized starting from a linear approximation of the nonlinear system, while others are initialized using random values or zeros. The effectiveness of the proposed initialization approach over previously proposed methods is illustrated on two benchmark examples.
引用
收藏
页码:1913 / 1918
页数:6
相关论文
共 50 条
  • [31] Coupling LSTM neural networks and state-space models through analytically tractable inference
    Vuong, Van-Dai
    Nguyen, Luong-Ha
    Goulet, James-A.
    INTERNATIONAL JOURNAL OF FORECASTING, 2025, 41 (01) : 128 - 140
  • [32] One for all: Universal material model based on minimal state-space neural networks
    Bonatti, Colin
    Mohr, Dirk
    SCIENCE ADVANCES, 2021, 7 (26)
  • [33] Bayesian Training and Committees of State-Space Neural Networks for Online Travel Time Prediction
    van Hinstiergen, C. P. I. j
    van Lint, J. W. C.
    van Zuylen, H. J.
    TRANSPORTATION RESEARCH RECORD, 2009, (2105) : 118 - 126
  • [34] Artificial Neural Networks for Stochastic Control of Nonliner State Space Systems
    Gorji, Ali A.
    Menhaj, Mohammad B.
    2008 IEEE INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS, VOLS 1-8, 2008, : 147 - 154
  • [35] Improved state-space approach to microwave circuit simulation
    Shi, K
    Jastrzebski, AK
    ELECTRONICS LETTERS, 1996, 32 (12) : 1108 - 1109
  • [36] A new look at state-space models for neural data
    Paninski, Liam
    Ahmadian, Yashar
    Ferreira, Daniel Gil
    Koyama, Shinsuke
    Rad, Kamiar Rahnama
    Vidne, Michael
    Vogelstein, Joshua
    Wu, Wei
    JOURNAL OF COMPUTATIONAL NEUROSCIENCE, 2010, 29 (1-2) : 107 - 126
  • [37] A new look at state-space models for neural data
    Liam Paninski
    Yashar Ahmadian
    Daniel Gil Ferreira
    Shinsuke Koyama
    Kamiar Rahnama Rad
    Michael Vidne
    Joshua Vogelstein
    Wei Wu
    Journal of Computational Neuroscience, 2010, 29 : 107 - 126
  • [38] Switching state-space modeling of neural signal dynamics
    He, Mingjian
    Das, Proloy
    Hotan, Gladia
    Purdon, Patrick L.
    PLOS COMPUTATIONAL BIOLOGY, 2023, 19 (08)
  • [39] State-space neural network for modelling, prediction and control
    Zamarreño, JM
    Vega, P
    García, LD
    Francisco, M
    CONTROL ENGINEERING PRACTICE, 2000, 8 (09) : 1063 - 1075
  • [40] AN INITIALIZATION METHOD FOR FEEDFORWARD ARTIFICIAL NEURAL NETWORKS USING POLYNOMIAL BASES
    Varnava, Thanasis M.
    Meade, Andrew J., Jr.
    ADVANCES IN DATA SCIENCE AND ADAPTIVE ANALYSIS, 2011, 3 (03) : 385 - 400