Physics Informed Neural Networks(PINNs) for Burgers' equation

PINNs are neural networks that undergo training to execute supervised learning tasks while adhering to certain laws of physics, which are typically defined by basic nonlinear partial differential equations. This paper concentrates on the advancements in addressing burger equation problems using data-driven approaches for partial differential equations. Here the algorithms are separated into continuous time designs. The neural networks that emerge from this process are part of a new class of data-efficient approximators for universal functions, which automatically include all fundamental physical principles as prior information. U sing these networks, this research demonstrates how to deduce solutions to partial differential equations.