Soft-Landing of Multi-Rotor Drones using a Robust Nonlinear Control and Wind Modeling

Grasping, manipulation, and inspection by multirotor systems require soft landing without any bumps; hence, the one-shot landing subject is critical due to aerodynamics effects under a multirotor unmanned aerial vehicle (UAV). One of the tasks in the HYFLIERS project is landing on a rack of pipes for inspection, mainly measurement of the pipe thickness and corrosion. The rack of pipes generates an unknown disturbance caused by the induced airflow by the propellers during the landing phase. The modeling of this problem is developed for two cases, landing on the ground and rack of pipes. The ground effect modeling is straightforward; however, the rack of pipes imposes more uncertainty on the system modeling. The source of aerodynamics disturbance also could be either external wind or the one caused by the UAV’s propellers near the pipes or ground. This work proposes a solution for the one-shot landing of a quadrotor considering the ground effect. First, the induced wind by the rotors near the ground is computed and then the reflection model of that near the ground is defined. Modeling of the quadrotor considering the wind in the environment is done. Next, the reflected wind by the ground is set in the wind model of the system. The uncertainty in the modeling exists due to interference of airflow under the UAV and behavior of that, so a robust nonlinear control is selected to control the system. The correction gain of the sliding mode controller was defined based on the steady-state thrust that plays the role of an upper bound of uncertainty. A simulation has been successfully done to present the advantages of the soft landing method considering the ground effect. The resultant input thrust decreased smoothly near the pipes that compensated the ground effect thrust.Research partially supported by the HYFLIERS (H2020-ICT-25-2017-779411) and PILOTING (H2020-ICT-2019-871542) projects funded by the European Union's H2020 programme, and the ARTIC project (RTI2018-102224-B-I00), funded by the Spanish Ministerio de Economia, Industria, y Competitividad