In this notebook we investigate different PINN implementations for harmonic oscillators, covering all three different settings: (1) the underdamped case with δ < ω_0, (2) the critically damped case with δ = ω_0 and (3) the overdamped case with δ > ω_0.
Check out the notebook here!
The following blog posts/articles were particularly useful:
- Lagaris, I. E., et al., "Artificial neural networks for solving ordinary and partial differential equations", 1998.
- Raissi, M., et al., "Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations", 2019.
- Moseley B., "So what is a Physics-Informed Neural Network?", 2021.
- beltoforion.de, "Damped Harmonic Oscillator".