Design, fabrication, and soft impact modeling and simulation of a collision-resilient foldable micro quadcopter

Abstract

Despite the appreciable advancements in mobile robot navigation and obstacle avoidance algorithms using an abundance of sensors and sensor fusion methods, the navigation of moving robots through confined and cluttered spaces is still a great challenge. The physical interaction and collision between mobile robots and surrounding obstacles in these environments are unavoidable. This becomes more concerning for the case of Vertical Take-Off and Landing (VTOL) Unmanned Aerial Vehicles (UAV) that the system is naturally unstable, and a minor fault or disturbance may result in severe crashes. In this thesis, a Collision-Resilient Quad-rotor Micro Aerial Vehicle (MAV) is designed using the Origami-inspired design fabrication techniques. The quadcopter is lightweight (220g max) and is designed for outdoor inspection and surveillance missions. This compliant drone provides an stable flight for a duration of 5 - 10 min, depends on the flight condition. A dynamic model is derived for the quadcopter to represent the realistic features designed for this specific UAV. In addition to that, the impact of the compliant body to surrounding obstacles is modeled as visco-elastic contact force and is added to the quadcopter’s dynamic model. The contact dynamic friction force between the protective soft bumpers and the surface is also modeled. The developed dynamic model is then used to simulate the impact of the collisionresilient quadcopter in two different simulation environments; MATLAB Simulink and ROS Gazebo. A cascaded PID control scheme is suggested for low-level (attitude) and high-level (global position) control of the drone in experiments and simulation. The result of these soft impact simulations closely imitate the collision-resilient properties of the actual quadcopter in experiments. Coefficient of Restitution (CoR) for the compliant drone impact, both in simulations and experiments, is in the interval [0.5 0.6]. This shows a great capacity for the drone to dampen the collisions.