Dynamical system prediction from sparse observations using deep neural networks with Voronoi tessellation and physics constraint

Abstract

Despite the success of various methods in addressing the issue of spatial reconstruction of dynamical systems with sparse observations, spatio-temporal prediction for sparse fields remains a challenge. Existing Kriging-based frameworks for spatio-temporal sparse field prediction fail to meet the accuracy and inference time required for nonlinear dynamic prediction problems. In this paper, we introduce the Dynamical System Prediction from Sparse Observations using Voronoi Tessellation (DSOVT) framework, an innovative methodology based on Voronoi tessellation which combines convolutional encoder-decoder (CED) and long short-term memory (LSTM) and utilizing Convolutional Long Short-Term Memory (ConvLSTM). By integrating Voronoi tessellations with spatio-temporal deep learning models, DSOVT is adept at predicting dynamical systems with unstructured, sparse, and time-varying observations. CED-LSTM maps Voronoi tessellations into a low-dimensional representation for time series prediction, while ConvLSTM directly uses these tessellations in an end-to-end predictive model. Furthermore, we incorporate physics constraints during the training process for dynamical systems with explicit formulas. Compared to purely data-driven models, our physics-based approach enables the model to learn physical laws within explicitly formulated dynamics, thereby enhancing the robustness and accuracy of rolling forecasts. Numerical experiments on real sea surface data and shallow water systems clearly demonstrate our framework's accuracy and computational efficiency with sparse and time-varying observations.

References in corpus (18)

• Convolutional LSTM Network: A Machine Learning Approach for Precipitation Nowcasting
• Discovering governing equations from data: Sparse identification of nonlinear dynamical systems
• Machine Learning for Fluid Mechanics
• Super-resolution reconstruction of turbulent flows with machine learning
• Reduced-order modeling of advection-dominated systems with recurrent neural networks and convolutional autoencoders
• Machine-learning-based reduced order modeling for unsteady flows around bluff bodies of various shapes
• Sparse identification of nonlinear dynamics with low-dimensionalized flow representations
• Grasping Extreme Aerodynamics on a Low-Dimensional Manifold
• A non-intrusive reduced order modeling framework for quasi-geostrophic turbulence
• Fully Convolutional Mesh Autoencoder using Efficient Spatially Varying Kernels
• Deep-learning assisted reduced order model for high-dimensional flow prediction from sparse data
• Latent-space time evolution of non-intrusive reduced-order models using Gaussian process emulation
• Data-driven transient lift attenuation for extreme vortex gust-airfoil interactions
• Multi-fidelity physics constrained neural networks for dynamical systems
• Bounded nonlinear forecasts of partially observed geophysical systems with physics-constrained deep learning
• DeepKriging: Spatially Dependent Deep Neural Networks for Spatial Prediction
• HOPE: Hierarchical Spatial-temporal Network for Occupancy Flow Prediction
• Dependent relevance determination for smooth and structured sparse regression

Cited by in corpus (1)

• Improving Long-term Autoregressive Spatiotemporal Predictions: A Proof of Concept with Fluid Dynamics