Lagrangian and Hamiltonian Neural Networks With a Dissipative System

The paper studies the applicability of Lagrangian and Hamiltonian neural network models to a dissipative mechanical system whose Lagrangian, Hamiltonian, and total energy explicitly depend on time. Its experiments use simulated one-dimensional, one-component harmonic oscillators with and without damping to compare the two neural-network approaches. The resource targets neural modeling of dissipative dynamical systems.