Simultaneous Identification and Denoising of Dynamical Systems

DOI10.1137/22M1486303MaRDI QIDQ6113941zbMATH OpenOpenAlexFDO

Authors Jeffrey M. Hokanson, Gianluca Iaccarino, Alireza Doostan

Publication date 11 July 2023

Published in SIAM Journal on Scientific Computing (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/2203.13837

zbMATH Keywords

dynamical systemsinverse problemsmodel discoveryparameter estimationsparse recovery

Mathematics Subject Classification ID

Inverse problems involving ordinary differential equations (34A55) Numerical solution of inverse problems involving ordinary differential equations (65L09) Methods of successive quadratic programming type (90C55) System identification (93B30)

Abstract: In recent years there has been a push to discover the governing equations dynamical systems directly from measurements of the state, often motivated by systems that are too complex to directly model. Although there has been substantial work put into such a discovery, doing so in the case of large noise has proved challenging. Here we develop an algorithm for Simultaneous Identification and Denoising of a Dynamical System (SIDDS). We infer the noise in the state measurements by requiring that the denoised data satisfies the dynamical system with an equality constraint. This is unlike existing work where the mismatch in the dynamical system is treated as a penalty in the objective. We assume the dynamics is represented in a pre-defined basis and develop a sequential quadratic programming approach to solve the SIDDS problem featuring a direct solution of KKT system with a specialized preconditioner. In addition, we show how we can include sparsity promoting regularization using an iteratively reweighted least squares approach. The resulting algorithm leads to estimates of the dynamical system that approximately achieve the Cram'er-Rao lower bound and, with sparsity promotion, can correctly identify the sparsity structure for higher levels of noise than existing techniques. Moreover, because SIDDS decouples the data from the evolution of the dynamical system, we show how to modify the problem to accurately identify systems from low sample rate measurements. The inverse problem approach and solution framework used by SIDDS has the potential to be expanded to related problems identifying governing equations from noisy data.

Recommendations

• scientific article; zbMATH DE number 4147938
• Nonlinear dynamical system identification with dynamic noise and observational noise
• scientific article; zbMATH DE number 782641
• Identification of dynamic systems from noisy data: Single factor case
• scientific article; zbMATH DE number 65815
• scientific article; zbMATH DE number 4114530
• Approximate identification of dynamical systems
• scientific article; zbMATH DE number 4121252
• Deterministic identification of dynamical systems
• Non-parametric identification of homogeneous dynamical systems

Cites work

• A data scalable augmented Lagrangian KKT preconditioner for large-scale inverse problems
• A Sequential Quadratic Programming Method Without A Penalty Function or a Filter for Nonlinear Equality Constrained Optimization
• A Supernodal Approach to Sparse Partial Pivoting
• An Algorithm for Degenerate Nonlinear Programming with Rapid Local Convergence
• Data-driven operator inference for nonintrusive projection-based model reduction
• Deep learning of dynamics and signal-noise decomposition with time-stepping constraints
• Discovering governing equations from data by sparse identification of nonlinear dynamical systems
• Enhancing sparsity by reweighted \(\ell _{1}\) minimization
• scientific article; zbMATH DE number 46726 (Why is no real title available?)
• scientific article; zbMATH DE number 852536 (Why is no real title available?)
• scientific article; zbMATH DE number 6159604 (Why is no real title available?)
• Improved iteratively reweighted least squares for unconstrained smoothed _q minimization
• Iterative hard thresholding for compressed sensing
• Iteratively reweighted least squares minimization for sparse recovery
• Numerical aspects for approximating governing equations using data
• On optimization techniques for solving nonlinear inverse problems
• Preconditioned all-at-once methods for large, sparse parameter estimation problems
• Sparse identification of nonlinear dynamical systems via reweighted \(\ell_1\)-regularized least squares
• Sparse identification of nonlinear dynamics for model predictive control in the low-data limit
• Tensor Decompositions and Applications
• The Differentiation of Pseudo-Inverses and Nonlinear Least Squares Problems Whose Variables Separate

Cited in

(11)

• IDENT: identifying differential equations with numerical time evolution
• Sparse identification of nonlinear dynamical systems via reweighted \(\ell_1\)-regularized least squares
• scientific article; zbMATH DE number 4147938 (Why is no real title available?)
• scientific article; zbMATH DE number 65815 (Why is no real title available?)
• On the convergence of the SINDy algorithm
• How entropic regression beats the outliers problem in nonlinear system identification
• Identification of MultiObject Dynamical Systems: Consistency and Fisher Information
• Derivative-based SINDy (DSINDy): addressing the challenge of discovering governing equations from noisy data
• Dynamic deconvolution and identification of independent autoregressive sources
• Quadrature rule based discovery of dynamics by data-driven denoising
• Asymptotic consistency of the WSINDy algorithm in the limit of continuum data

This page was built for publication: Simultaneous Identification and Denoising of Dynamical Systems

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6113941)