Quadcopter Control Based on Experimental Dynamic Identification

The development of reliable and robust flight control systems is an exigent concern in the field of unmanned aerial vehicles (UAVs), particularly in light of the inherent open-loop instability associated with quadrotors, which renders the direct application of controllers impracticable without prior design. In this investigation, we advance a method for regulating quadcopter flight based on experimental dynamic identification. Our approach involves utilizing a black-box direct methodology to construct a model of the quadrotor's attitude behaviors through data collection and the application of various linear and nonlinear representations. The veracity of the models is subsequently ascertained via stringent validation protocols. Consequently, predicated on the resulting dynamic model, a control system is developed to optimize the quadcopter's stability and maneuverability. This approach to controller design engenders a dependable and efficient solution for regulating quadcopter flight.