Higher-order Energy-conserving Methods for Schrodinger-Poisson Equations
We combined Implicit-Explicit (ImEx) Runge Kutta methods with projection based relaxation to conserve mass and maintain energy balance( or conservation). arXiv link
Hi there, I’m Manvendra (Manu).
I am a researcher working at the intersection of Physics, Numerical Methods and Scientific Machine Learning.
About
Background
I obtained my Ph.D. in Cosmology from IISER Mohali under the guidance of Prof. Jasjeet S. Bagla. After that I worked as a postdoctoral researcher at Ahmedabad University. Currently, I am a postdoc at KAUST in David Ketcheson's Numerical Maths group.
Research Interests
- Numerical Methods for PDEs
- Cosmology
- Scientific Machine Learning
- Physics-informed ML
Recent Projects
Integral based PINNs
We use conservation-based loss instead of usual strong form of hyperbolic PDEs to enhance learning of solutions with shocks. The work was presented at ICLR 2024 Workshop on AI4DifferentialEquations In Science You can read the article here.