Searcharxiv⌕ Search

arXiv subjects

Pratibhamoy Das

Publications and source records attributed to Pratibhamoy Das.

2 recordsLinked to original sources

A singularity detecting uniformly convergent quantum wavelet framework for weakly singular nonlocal Emden--Fowler systems with quantum Mittag--Leffler sources

We develop a quantum wavelet-based computational framework for coupled systems of Caputo-type quantum ($q$-) fractional Emden--Fowler problems with $q$-Mittag--Leffler source terms, where the solutions exhibit singular behavior in time. Here $q\in(0,1)$ denotes the parameter of $q$-calculus. The highest-order Caputo $q$-fractional derivatives are expanded through the Haar wavelet basis, and the lower-order fractional derivatives are approximated through Jackson $q$-fractional integrals of Haar wavelets. The continuous problem reduces to a nonlinear algebraic system on the midpoint collocation points. Under suitable structural assumptions, the Banach fixed-point theorem yields the existence and uniqueness of the continuous solutions and wellposedness of the discrete solutions on collocation points. The nonlinearity of the given problem is dealt with by the damped Newton's iteration. In addition, We prove stability of the scheme and derive rigorous convergence estimates in a weighted norm that remains uniformly controlled despite the appearance of singular weight at $t=0$ in the continuous model. When the highest-order fractional derivatives are Hölder continuous with exponent $σ$, the mathematical error of weakly singular solution and its lower-order fractional derivatives become $\mathcal{O}(τ_L^σ)$, where $τ_L=2^{-(L+1)}$ denotes the width of the finest subintervals for maximum level of resolution $L$. First-order convergence is guaranteed under stronger regularity with $σ=1$. Several test cases are considered to confirm the accuracy and robustness of the present approach, which strongly agrees with the theoretical estimates. In particular, first-order convergence is observed even where the stronger regularity assumption is not satisfied.

math.NA↗

ENPINN: Energy-Norm-Guided Gradient-Enhanced PINNs for Generalized Transport Problems with Sharp Gradients

Physics-informed neural networks (PINNs) have emerged as a meshless alternative to conventional numerical methods for solving partial differential equations (PDEs). However, their limited ability to capture sharp gradients can lead to substantial errors when resolving boundary and interior layers. Here, we introduce an energy-norm-enhanced PINN (ENPINN) that incorporates gradient information and variational structure into the loss function to improve the resolution of layer-dominated solutions. We first examine two related formulations: weak-loss PINNs (WLPINNs), which incorporate test functions into the conventional PINN residual, and gradient-enhanced PINNs (gPINNs), which augment the loss with spatial derivatives of the PDE residual. By analyzing these formulations, we identify their limitations in resolving steep solution gradients and motivate the systematic construction of ENPINN. We establish theoretically how the energy-norm error depends on the ENPINN loss and show that a suitably modified residual-derivative term is essential for accurately capturing boundary layers. We further establish the existence of neural-network approximations with arbitrarily small energy error and derive corresponding derivative bounds, providing a theoretical foundation for the proposed framework. The performance of ENPINN is assessed through systematic comparisons with existing PINN variants for convection-diffusion-reaction problems exhibiting steep gradients. Numerical experiments include a combustion model, a coupled multi-scale system, a two-dimensional Burgers equation with an interior layer, and a three-dimensional time-dependent problem.

math.NA↗