Data driven system ID

This repository contains several examples of the data driven system identification techique called SINDy. SINDy (Sparse Identification of Nonlinear Dynamics) is a powerful data-driven method for discovering the governing equations of complex dynamical systems. It uses sparse regression techniques to identify a set of functions that best describe the system's behavior from time-series data.

Steps:

1. Data Collection: Time-series measurements of the system state are gathered
2. Library Construction: A library of candidate nonlinear functions is created
3. Sparse Regression: The algorithm identifies the most relevant terms from the library to describe the system dynamics
4. Model Formation: The selected terms are combined to form the governing equations

List of examples

• Example 1: Exponential & trigonometric
• Example 2: Linear system
• Example 3: Cubic
• Example 4: Lorenz system

References

An introduction to Sparse Identification of Nonlinear Dynamical systems (SINDy)

PYSINDy

Decoding Dynamics: A Quick Guide to SINDy