Autoregressive Noise

GitHub
GazeboC++RoboticsSimulationControl Systems

Gazebo simulator plugin for realistic noise injection in differential drive robots.

This Gazebo simulator plugin introduces realistic noise characteristics into differential drive robots. By injecting autoregressive noise into wheel encoder measurements, it simulates the drift and lateral slippage that can occur on inclined or uneven surfaces.

The goal is to achieve more robust and reliable control algorithms by exposing them to data that closely mirrors real-world uncertainties. Integrating this plugin into Gazebo helps developers refine navigation and localization strategies by anticipating and compensating for natural sensor drift in robotics.