Parameter Estimation Based on Linear Regression for Commercial Quadrotors

Over the last decade, the rapid evolution in communications technology, mechanics and electronics has propelled access to Unmanned Aerial Vehicles (UAVs). Among this class of robots, quadrotors stand out due to maneuverability and Vertical Take Off and Landing (VTOL) capability. This work proposes a methodology to solve the model parameter estimation problem of a commercial quadrotor. Mathematical models are derived and implementation details are given. Experimental results, obtained in an open source environment associated with an accurate motion capture system, verify that the estimated dynamic model approximates to the real dynamics.