Abstract:Throughout many fields, practitioners often rely on differential equations to model systems. Yet, for many applications, the theoretical derivation of such equations and/or accurate resolution of their solutions may be intractable. Instead, recently developed methods, including those based on parameter estimation, operator subset selection, and neural networks, allow for the data-driven discovery of both ordinary and partial differential equations (PDEs), on a spectrum of interpretability. The success of these strategies is often contingent upon the correct identification of representative equations from noisy observations of state variables and, as importantly and intertwined with that, the mathematical strategies utilized to enforce those equations. Specifically, the latter has been commonly addressed via unconstrained optimization strategies. Representing the PDE as a neural network, we propose to discover the PDE by solving a constrained optimization problem and using an intermediate state representation similar to a Physics-Informed Neural Network (PINN). The objective function of this constrained optimization problem promotes matching the data, while the constraints require that the PDE is satisfied at several spatial collocation points. We present a penalty method and a widely used trust-region barrier method to solve this constrained optimization problem, and we compare these methods on numerical examples. Our results on the Burgers' and the Korteweg-De Vreis equations demonstrate that the latter constrained method outperforms the penalty method, particularly for higher noise levels or fewer collocation points. For both methods, we solve these discovered neural network PDEs with classical methods, such as finite difference methods, as opposed to PINNs-type methods relying on automatic differentiation. We briefly highlight other small, yet crucial, implementation details.

Physics-informed genetic programming for discovery of partial differential equations from scarce and noisy data

Physics-informed learning of governing equations from scarce data

Physics-Guided Discovery of Highly Nonlinear Parametric Partial Differential Equations

Discovery of partial differential equations from highly noisy and sparse data with physics-informed information criterion

Physics-informed identification of PDEs with LASSO regression, examples of groundwater-related equations

Robust data-driven discovery of governing physical laws with error bars

Identification of Partial Differential Equations from Noisy Data with Integrated Knowledge Discovery and Embedding Using Evolutionary Neural Networks

Physics-constrained robust learning of open-form partial differential equations from limited and noisy data

Deep-learning of parametric partial differential equations from sparse and noisy data

Rapid Bayesian identification of sparse nonlinear dynamics from scarce and noisy data

DL-PDE: Deep-learning based data-driven discovery of partial differential equations from discrete and noisy data

Deep-learning based discovery of partial differential equations in integral form from sparse and noisy data

Physics-informed AI and ML-based sparse system identification algorithm for discovery of PDE's representing nonlinear dynamic systems

Bayesian data-driven discovery of partial differential equations with variable coefficients

Physics-Informed Gaussian Process Regression Generalizes Linear PDE Solvers

PHYSICS DISCOVERY OF ENGINEERING APPLICATIONS WITH CONSTRAINED OPTIMIZATION AND GENETIC PROGRAMMING

Constrained or Unconstrained? Neural-Network-Based Equation Discovery from Data

Deep Generative Modeling for Identification of Noisy, Non-Stationary Dynamical Systems

Dimensional homogeneity constrained gene expression programming for discovering governing equations

Automatically discovering ordinary differential equations from data with sparse regression

A Framework Based on Symbolic Regression Coupled with eXtended Physics-Informed Neural Networks for Gray-Box Learning of Equations of Motion from Data