Nonlinear model reduction by deep autoencoder of noise response data

In this paper a novel model order reduction method for nonlinear systems is proposed. Differently from existing ones, the proposed method provides a suitable non-linear projection, which we refer to as control-oriented deep autoencoder (CoDA), in an easily implementable manner. This is done by combining noise response data based model reduction, whose control theoretic optimality was recently proven by the author, with stacked autoencoder design via deep learning.

[1]  Stefan Volkwein,et al.  Galerkin Proper Orthogonal Decomposition Methods for a General Equation in Fluid Dynamics , 2002, SIAM J. Numer. Anal..

[2]  Dong Yu,et al.  Deep Learning: Methods and Applications , 2014, Found. Trends Signal Process..

[3]  J. Peraire,et al.  Balanced Model Reduction via the Proper Orthogonal Decomposition , 2002 .

[4]  Kenji Kashima,et al.  Optimality of simulation-based nonlinear model reduction: Stochastic controllability perspective , 2016, 2016 American Control Conference (ACC).

[5]  Yee Whye Teh,et al.  A Fast Learning Algorithm for Deep Belief Nets , 2006, Neural Computation.

[6]  Takayuki Ishizaki,et al.  Model Reduction and Clusterization of Large-Scale Bidirectional Networks , 2014, IEEE Transactions on Automatic Control.

[7]  P. Holmes,et al.  Turbulence, Coherent Structures, Dynamical Systems and Symmetry , 1996 .

[8]  Yoshua. Bengio,et al.  Learning Deep Architectures for AI , 2007, Found. Trends Mach. Learn..

[9]  Alessandro Astolfi,et al.  Model Reduction by Moment Matching for Linear and Nonlinear Systems , 2010, IEEE Transactions on Automatic Control.

[10]  Nathan van de Wouw,et al.  Model Reduction for Nonlinear Systems by Incremental Balanced Truncation , 2014, IEEE Transactions on Automatic Control.

[11]  Ioannis Karatzas,et al.  Brownian Motion and Stochastic Calculus , 1987 .

[12]  Marc'Aurelio Ranzato,et al.  Building high-level features using large scale unsupervised learning , 2011, 2013 IEEE International Conference on Acoustics, Speech and Signal Processing.

[13]  Antoine Girard,et al.  Clustered model reduction of positive directed networks , 2015, Autom..

[14]  Kenji Kashima,et al.  Noise Response Data Reveal Novel Controllability Gramian for Nonlinear Network Dynamics , 2016, Scientific Reports.

[15]  J. Marsden,et al.  A subspace approach to balanced truncation for model reduction of nonlinear control systems , 2002 .

[16]  J. M. A. Scherpen,et al.  Balancing for nonlinear systems , 1993 .

[17]  Albert-László Barabási,et al.  Controllability of complex networks , 2011, Nature.