arXiv subjects

Prakash Vedula

Publications and source records attributed to Prakash Vedula.

25 records · Page 2Linked to original sources

arXiv2019

A priori analysis on deep learning of subgrid-scale parameterizations for Kraichnan turbulence

In the present study, we investigate different data-driven parameterizations for large eddy simulation of two-dimensional turbulence in the \emph{a priori} settings. These models utilize resolved flow field variables on the coarser grid to estimate the subgrid-scale stresses. We use data-driven closure models based on localized learning that employs multilayer feedforward artificial neural network (ANN) with point-to-point mapping and neighboring stencil data mapping, and convolutional neural network (CNN) fed by data snapshots of the whole domain. The performance of these data-driven closure models is measured through a probability density function and is compared with the dynamic Smagorinksy model (DSM). The quantitative performance is evaluated using the cross-correlation coefficient between the true and predicted stresses. We analyze different frameworks in terms of the amount of training data, selection of input and output features, their characteristics in modeling with accuracy, and training and deployment computational time. We also demonstrate computational gain that can be achieved using the intelligent eddy viscosity model that learns eddy viscosity computed by the DSM instead of subgrid-scale stresses. We detail the hyperparameters optimization of these models using the grid search algorithm.

physics.comp-ph↗

arXiv2019

A novel quantum grid search algorithm and its application

In this paper we present a novel quantum algorithm, namely the quantum grid search algorithm, to solve a special search problem. Suppose $ k $ non-empty buckets are given, such that each bucket contains some marked and some unmarked items. In one trial an item is selected from each of the $ k $ buckets. If every selected item is a marked item, then the search is considered successful. This search problem can also be formulated as the problem of finding a "marked path" associated with specified bounds on a discrete grid. Our algorithm essentially uses several Grover search operators in parallel to efficiently solve such problems. We also present an extension of our algorithm combined with a binary search algorithm in order to efficiently solve global trajectory optimization problems. Estimates of the expected run times of the algorithms are also presented, and it is proved that our proposed algorithms offer exponential improvement over pure classical search algorithms, while a traditional Grover's search algorithm offers only a quadratic speedup. We note that this gain comes at the cost of increased complexity of the quantum circuitry. The implication of such exponential gains in performance is that many high dimensional optimization problems, which are intractable for classical computers, can be efficiently solved by our proposed quantum grid search algorithm.

quant-ph↗

arXiv2019

Trajectory optimization using quantum computing

We present a framework wherein the trajectory optimization problem (or a problem involving calculus of variations) is formulated as a search problem in a discrete space. A distinctive feature of our work is the treatment of discretization of the optimization problem wherein we discretize not only independent variables (such as time) but also dependent variables. Our discretization scheme enables a reduction in computational cost through selection of coarse-grained states. It further facilitates the solution of the trajectory optimization problem via classical discrete search algorithms including deterministic and stochastic methods for obtaining a global optimum. This framework also allows us to efficiently use quantum computational algorithms for global trajectory optimization. We demonstrate that the discrete search problem can be solved by a variety of techniques including a deterministic exhaustive search in the physical space or the coefficient space, a randomized search algorithm, a quantum search algorithm or by employing a combination of randomized and quantum search algorithms depending on the nature of the problem. We illustrate our methods by solving some canonical problems in trajectory optimization. We also present a comparative study of the performances of different methods in solving our example problems. Finally, we make a case for using quantum search algorithms as they offer a quadratic speed-up in comparison to the traditional non-quantum algorithms.

math.OC↗

arXiv2019

Invariant compact finite difference schemes

In this paper, we propose a method, that is based on equivariant moving frames, for development of high order accurate invariant compact finite difference schemes that preserve Lie symmetries of underlying partial differential equations. In this method, variable transformations that are obtained from the extended symmetry groups of PDEs are used to transform independent and dependent variables, and derivative terms of compact finite difference schemes (constructed for these PDEs) such that the resulting schemes are invariant under the chosen symmetry groups. The unknown symmetry parameters that arise from the application of these transformations are determined through selection of convenient moving frames. In some cases, owing to selection of convenient moving frames, numerical representation of invariant (or symmetry preserving) compact numerical schemes, that are developed through the proposed method, are found to be notably simpler than that of standard, noninvariant compact numerical schemes. Further, the order of accuracy of these invariant compact schemes can be arbitrarily set to a desired order by considering suitable compact finite difference algorithms. Application of the proposed method is demonstrated through construction of invariant compact finite difference schemes for some common linear and nonlinear PDEs (including the linear advection-diffusion equation in 1D/2D, the inviscid and viscous Burgers' equations in 1D). Results obtained from our numerical simulations indicate that invariant compact finite difference schemes not only inherit selected symmetry properties of underlying PDEs, but are also comparably more accurate than the standard, non-invariant base numerical schemes considered here.

math-ph↗

arXiv2018

Data-driven deconvolution for large eddy simulations of Kraichnan turbulence

In this article, we demonstrate the use of artificial neural networks as optimal maps which are utilized for convolution and deconvolution of coarse-grained fields to account for sub-grid scale turbulence effects. We demonstrate that an effective eddy-viscosity is predicted by our purely data-driven large eddy simulation framework without explicit utilization of phenomenological arguments. In addition, our data-driven framework precludes the knowledge of true sub-grid stress information during the training phase due to its focus on estimating an effective filter and its inverse so that grid-resolved variables may be related to direct numerical simulation data statistically. The proposed predictive framework is also combined with a statistical truncation mechanism for ensuring numerical realizability in an explicit formulation. Through this we seek to unite structural and functional modeling strategies for modeling non-linear partial differential equations using reduced degrees of freedom. Both a priori and a posteriori results are shown for a two-dimensional decaying turbulence case in addition to a detailed description of validation and testing. A hyperparameter sensitivity study also shows that the proposed dual network framework simplifies learning complexity and is viable with exceedingly simple network architectures. Our findings indicate that the proposed framework approximates a robust and stable sub-grid closure which compares favorably to the Smagorinsky and Leith hypotheses for capturing the theoretical $k^{-3}$ scaling in Kraichnan turbulence.

physics.flu-dyn↗

arXiv2018

Sub-grid modelling for two-dimensional turbulence using neural networks

In this investigation, a data-driven turbulence closure framework is introduced and deployed for the sub-grid modelling of Kraichnan turbulence. The novelty of the proposed method lies in the fact that snapshots from high-fidelity numerical data are used to inform artificial neural networks for predicting the turbulence source term through localized grid-resolved information. In particular, our proposed methodology successfully establishes a map between inputs given by stencils of the vorticity and the streamfunction along with information from two well-known eddy-viscosity kernels. Through this we predict the sub-grid vorticity forcing in a temporally and spatially dynamic fashion. Our study is both a-priori and a-posteriori in nature. In the former, we present an extensive hyper-parameter optimization analysis in addition to learning quantification through probability density function based validation of sub-grid predictions. In the latter, we analyse the performance of our framework for flow evolution in a classical decaying two-dimensional turbulence test case in the presence of errors related to temporal and spatial discretization. Statistical assessments in the form of angle-averaged kinetic energy spectra demonstrate the promise of the proposed methodology for sub-grid quantity inference. In addition, it is also observed that some measure of a-posteriori error must be considered during optimal model selection for greater accuracy. The results in this article thus represent a promising development in the formalization of a framework for generation of heuristic-free turbulence closures from data.

physics.flu-dyn↗

arXiv2017

Generalized deconvolution procedure for structural modeling of turbulence

Approximate deconvolution forms a mathematical framework for the structural modeling of turbulence. The sub-filter scale flow quantities are typically recovered by using the Van Cittert iterative procedure. In this paper, however, we put forth a generalized approach for the iterative deconvolution process of sub-filter scale recovery of turbulent flows by introducing Krylov space iterative methods. Their accuracy and efficiency are demonstrated through a systematic a-priori analysis of solving the Kraichnan and Kolmogorov homogeneous isotropic turbulence problems in two- and three-dimensional domains, respectively. Our numerical assessments show that the conjugate gradient based iterative techniques lead to significantly improved performance over the Van Cittert procedure and offer great promise for approximate deconvolution turbulence models. In fact, our energy spectra analysis illustrates that a substantially longer inertial range can be recovered by using the proposed procedure equipped with the BiCGSTAB iterative scheme. This trend is also confirmed by capturing tails of the probability density function of turbulent flow quantities.

physics.flu-dyn↗