numerical-pde

Runge–Kutta Physics-Informed Neural Networks with time-discrete losses (Gauss/Radau/Lobatto) in PyTorch.

Risk-sensitive stochastic control framework implementing nonlinear HJB solvers, policy iteration, and Monte Carlo validation for linear-quadratic diffusion models.

Finite difference experiments for PDEs: Poisson equation solvers and convection schemes · 偏微分方程有限差分实验：泊松方程求解与对流格式

Verified FV and physics-informed (PINN) solvers for semiconductor device equations, built toward adjoint doping optimization of 4H-SiC power devices.

JEPA-style self-supervised model for elliptic PDEs: predicts masked latents instead of minimising PDE residuals.

Fourier and Chebyshev pseudospectral methods in MATLAB, compared with finite differences and P1 finite elements (Warwick MA395 essay)

本仓库是论文 "High-Order Accurate Entropy Stable Schemes for Relativistic Hydrodynamics with General Synge-Type Equation of State" (Journal of Scientific Computing, 2024, 98:43) 的官方完整数值实验代码实现。