Physics Informed Neural Network Based Time-Independent Schrödinger Equation Solver

This paper presents a novel approach for solving the time-independent Schrödinger equation for any arbitrary potential using Physics-Informed Neural Networks (PINNs). PINNs seamlessly embed physical principles into neural networks, allowing precise quantum wavefunction predictions. Extensive experimentation highlights its superior performance over conventional solvers. This innovative framework advances quantum mechanics simulations and underscores machine learning's potential in tackling intricate physical phenomena. Our model's exceptional efficiency and accuracy extend to non-ground state energy levels, with a maximum relative error below 1%.