Experimental Validation and Robustness Analysis of Ergodic UAV Control

This study presents an experimental validation of coupled ergodic control for horizontal movement with model predictive control (MPC) for velocity and altitude of multi-rotor Unmanned Aerial Vehicles (UAVs) during search operations in hilly terrain. For ergodic control, we employ a two-dimensional potential-based algorithm called the Heat Equation Driven Area Coverage (HEDAC). This paper aims to compare search performance between simulation-based environments and real-world experiments, validating the most important flight parameters. The motivation stems from the critical role UAVs play in time-sensitive Search and Rescue (SAR) operations, where robustness to different uncertainties and errors is beneficial. Some of the uncertainties impacting flight accuracy are sensor inaccuracies from GPS and compass sensors, wind disturbance, and variability in communication lag duration, the difference between simulated and real flights is being investigated. By conducting two experiments, each consisting of one simulation and two field tests, we evaluate the motion control uncertainty under these unpredictable factors. The results show promising efficiency and robustness of the motion control system.