Design, modelling and control of a rotorcraft landing gear for uneven ground conditions

Abstract

The ability to perform vertical take-off and landing, hovering and lateral flight provides rotorcrafts crucial advantages over other aircrafts and land vehicles for operations in remote areas. However, a major limitation of rotorcrafts is the requirement of a flat surface to land, increasing the difficulty and risk of landing operations on rough terrain or unstable surfaces. This limitation is mainly due to the use of conventional landing gear like skids or wheels. The growing use of Unmanned Aerial Vehicles (UAVs) also increases the necessity for more landing autonomy of these systems.
This thesis presents the investigation into the development of an adaptive robotic landing gear for a small UAV that enhances the landing capabilities of current rotorcrafts. This landing gear consists in a legged system that is able to sense and adapt the position of its legs to the terrain conditions.
This research covers the development of effective tools for the design and testing of the control system using software and hardware platforms. Mathematical models using multibody system dynamics are developed and implemented in software simulations. A hardware robot is designed and built to validate the simulation results.
The system proposed in this thesis consists in a landing gear with four robotic legs that uses an Inertial Measurement Unit (IMU) to sense the body attitude, Force Sensing Resistors (FSR) to measure feet pressure and a distance sensor to detect ground approach. The actuators used are position-controlled servo motors that also provide angular position feedback. The control strategy provides position commands to coordinate the motion of all joints based on attitude and foot pressure information. It offers the advantage of being position-controlled, so it is easier to be implemented in hardware systems using low-cost components, and at the same time, the feet forcecontrol and leg design add compliance to the system.
Through software simulations and laboratory experiments the system successfully landed on a 20° slope surface, substantially increasing the current slope landing limit.