Bilinear residual Neural Network for the identification and forecasting of dynamical systems

Due to the increasing availability of large-scale observation and simulation datasets, data-driven representations arise as efficient and relevant computation representations of dynamical systems for a wide range of applications, where model-driven models based on ordinary differential equation remain the state-of-the-art approaches. In this work, we investigate neural networks (NN) as physically-sound data-driven representations of such systems. Reinterpreting Runge-Kutta methods as graphical models, we consider a residual NN architecture and introduce bilinear layers to embed non-linearities which are intrinsic features of dynamical systems. From numerical experiments for classic dynamical systems, we demonstrate the relevance of the proposed NN-based architecture both in terms of forecasting performance and model identification.

[1]  Redouane Lguensat,et al.  The Analog Data Assimilation , 2017 .

[2]  Zhizhen Zhao,et al.  Analog forecasting with dynamics-adapted kernels , 2014, 1412.3831.

[3]  R. Vaillancourt,et al.  Behind and beyond the Matlab ODE suite , 2000 .

[4]  Geir Evensen,et al.  Sequential data assimilation , 2009 .

[5]  Richard J. Field,et al.  Oscillating chemical reactions and nonlinear dynamics , 1989 .

[6]  E. Lorenz Deterministic nonperiodic flow , 1963 .

[7]  J N Kutz,et al.  Model selection for dynamical systems via sparse regression and information criteria , 2017, Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences.

[8]  S. Brunton,et al.  Discovering governing equations from data by sparse identification of nonlinear dynamical systems , 2015, Proceedings of the National Academy of Sciences.

[9]  Subhransu Maji,et al.  Bilinear CNN Models for Fine-Grained Visual Recognition , 2015, 2015 IEEE International Conference on Computer Vision (ICCV).

[10]  Jan Swevers,et al.  Identification of nonlinear systems using Polynomial Nonlinear State Space models , 2010, Autom..

[11]  Charalampos Tsitouras,et al.  Neural networks with multidimensional transfer functions , 2002, IEEE Trans. Neural Networks.

[12]  Dong-Chul Park,et al.  Complex-bilinear recurrent neural network for equalization of a digital satellite channel , 2002, IEEE Trans. Neural Networks.

[13]  Guigang Zhang,et al.  Deep Learning , 2016, Int. J. Semantic Comput..

[14]  Angelos A. Anastassi Constructing Runge–Kutta methods with the use of artificial neural networks , 2013, Neural Computing and Applications.