Garcia, Gonzalo (University of Kansas), Kim, A Ram (University of Kansas), Jackson, Ethan (Microsoft Research), Keshmiri, Shawn (University of Kansas), Shukla, Daksh (University of Kansas)

Modeling and Flight Control of a Commercial Nano Quadrotor

Scheduled for presentation during the "Control Architecture - III" (WeC3), Wednesday, June 14, 2017, 16:40−17:00, Salon CD

2017 International Conference on Unmanned Aircraft Systems, June 13-16, 2017, Miami Marriott Biscayne Bay, Miami, FL,

This information is tentative and subject to change. Compiled on April 19, 2024

Abstract

This work presents the flight test validation of an autonomous nano quadrotor called Crazyflie 2.0 in arbitrary trajectory tracking. The complexity of the control tasks were due to the low rigidity and high flexibility of propellers and airframe components, high sensitivity of vehicle inertial dynamic model to the battery location, uncalibrated motors, and unsteady battery depletion rate. Keeping the Crazyflie embedded inner control loop system, different outer loop guidance logics were designed for trajectory following. The guidance logics were running on an external computer and communicated via a radio link with the quadrotor in real-time. Attitude estimations were done using OptiTrack position information. An accurate dynamic model of aircraft was developed using principal moments of inertia and motor performance parameters. The Crazyflie dynamic model was used for the 6-DOF simulations and LQR and MPC implementations. All three guidance techniques, PID, LQR, and MPC, were validated through actual flight tests and fulfilled tracking tasks with an acceptable performance.