Agile Flight Control Using Incremental Nonlinear Dynamic Inversion with Highly Realistic Model of Actuator

Aggressive flight is the central prerequisite to expanding the application of quadrotors into complex and agile tasks. However, it will introduce severe model uncertainty and significant interference, which poses significant challenges to flight control. This paper proposes a novel incremental nonlinear dynamic inversion (INDI)- based controller to address the urgent demand for response speed in aggressive flights without compromising the disturbances rejection and uncertainty tolerance ability. First, through mechanistic modeling theory and wind tunnel tests, a high-fidelity nonlinear dynamic representation of quadrotor thrust units was established, which provided a more precise depiction of UAV dynamics, thus markedly reducing the model uncertainty. Second, we theoretically analyzed and demonstrated that introducing a high-fidelity thrust unit model into the INDI algorithm could notably improve the system's dynamic response while maintaining excellent robustness simultaneously, and a control scheme was designed. Finally, we evaluated the performances of the proposed method in simulation and the results clearly showed the superiority of our method in terms of response and convergence speed, trajectory tracking accuracy.