Using AI libraries for Incompressible Computational Fluid Dynamics

Recently, there has been a huge effort focused on developing highly efficient open source libraries to perform Artificial Intelligence (AI) related computations on different computer architectures (for example, CPUs, GPUs and new AI processors). This has not only made the algorithms based on these libraries highly efficient and portable between different architectures, but also has substantially simplified the entry barrier to develop methods using AI. Here, we present a novel methodology to bring the power of both AI software and hardware into the field of numerical modelling by repurposing AI methods, such as Convolutional Neural Networks (CNNs), for the standard operations required in the field of the numerical solution of Partial Differential Equations (PDEs). The aim of this work is to bring the high performance, architecture agnosticism and ease of use into the field of the numerical solution of PDEs. We use the proposed methodology to solve the advection-diffusion equation, the non-linear Burgers equation and incompressible flow past a bluff body. For the latter, a convolutional neural network is used as a multigrid solver in order to enforce the incompressibility constraint. We show that the presented methodology can solve all these problems using repurposed AI libraries in an efficient way, and presents a new avenue to explore in the development of methods to solve PDEs and Computational Fluid Dynamics problems with implicit methods.

Solving the Discretised Multiphase Flow Equations with Interface Capturing on Structured Grids Using Machine Learning Libraries

This paper solves the multiphase flow equations with interface capturing using the AI4PDEs approach (Artificial Intelligence for Partial Differential Equations). The solver within AI4PDEs uses tools from machine learning (ML) libraries to solve (exactly) partial differential equations (PDEs) that have been discretised using numerical methods. Convolutional layers can be used to express the discretisations as a neural network, whose weights are determined by the numerical method, rather than by training. To solve the system, a multigrid solver is implemented through a neural network with a U-Net architecture. Immiscible two-phase flow is modelled by the 3D incompressible Navier-Stokes equations with surface tension and advection of a volume fraction field, which describes the interface between the fluids. A new compressive algebraic volume-of-fluids method is introduced, based on a residual formulation using Petrov-Galerkin for accuracy and designed with AI4PDEs in mind. High-order finite-element based schemes are chosen to model a collapsing water column and a rising bubble. Results compare well with experimental data and other numerical results from the literature, demonstrating that, for the first time, finite element discretisations of multiphase flows can be solved using the neural network solver from the AI4PDEs approach. A benefit of expressing numerical discretisations as neural networks is that the code can run, without modification, on CPUs, GPUs or the latest accelerators designed especially to run AI codes.

Analyzing Convergence in Quantum Neural Networks: Deviations from Neural Tangent Kernels

A quantum neural network (QNN) is a parameterized mapping efficiently implementable on near-term Noisy Intermediate-Scale Quantum (NISQ) computers. It can be used for supervised learning when combined with classical gradient-based optimizers. Despite the existing empirical and theoretical investigations, the convergence of QNN training is not fully understood. Inspired by the success of the neural tangent kernels (NTKs) in probing into the dynamics of classical neural networks, a recent line of works proposes to study over-parameterized QNNs by examining a quantum version of tangent kernels. In this work, we study the dynamics of QNNs and show that contrary to popular belief it is qualitatively different from that of any kernel regression: due to the unitarity of quantum operations, there is a non-negligible deviation from the tangent kernel regression derived at the random initialization. As a result of the deviation, we prove the at-most sublinear convergence for QNNs with Pauli measurements, which is beyond the explanatory power of any kernel regression dynamics. We then present the actual dynamics of QNNs in the limit of over-parameterization. The new dynamics capture the change of convergence rate during training and implies that the range of measurements is crucial to the fast QNN convergence.