Method of data forward generation with partial differential equations for machine learning modeling in fluid mechanics
This addresses the data scarcity issue for researchers and engineers in fluid mechanics, but it is incremental as it builds on existing PDE-based methods with specific neural network integrations.
The study tackled the problem of expensive or inaccessible high-fidelity data in AI for fluid mechanics by proposing a data forward generation method from partial differential equations (PDEs), which successfully trained models like Poisson-NN and WTCNN with excellent generalization and accuracy without any DNS data, and data following physical laws improved convergence rate, generalization, and accuracy compared to Gaussian random field data.
Artificial intelligence (AI) for fluid mechanics has become attractive topic. High-fidelity data is one of most critical issues for the successful applications of AI in fluid mechanics, however, it is expensively obtained or even inaccessible. This study proposes a high-efficient data forward generation method from the partial differential equations (PDEs). Specifically, the solutions of the PDEs are first generated either following a random field (e.g. Gaussian random field, GRF, computational complexity O(NlogN), N is the number of spatial points) or physical laws (e.g. a kind of spectra, computational complexity O(NM), M is the number of modes), then the source terms, boundary conditions and initial conditions are computed to satisfy PDEs. Thus, the data pairs of source terms, boundary conditions and initial conditions with corresponding solutions of PDEs can be constructed. A Poisson neural network (Poisson-NN) embedded in projection method and a wavelet transform convolutional neuro network (WTCNN) embedded in multigrid numerical simulation for solving incompressible Navier-Stokes equations is respectively proposed. The feasibility of generated data for training Poisson-NN and WTCNN is validated. The results indicate that even without any DNS data, the generated data can train these two models with excellent generalization and accuracy. The data following physical laws can significantly improve the convergence rate, generalization and accuracy than that generated following GRF.