Autonomous Flight and Landing for Multi-Rotor using Integrated Navigation and Vision Detection

In this paper, we design an INS/GPS integrated navigation system based on federal Kalman filter for autonomous flight of multi-rotor which contains several parallel subfilter and guarantee the stability of system. At the same time, we present a discrete path generation algorithm for the multi-rotor trajectory tracking. During the autonomous landing stage, on-board computer runs a real-time land-marker detection algorithm and fuses IMU data to improve the landing precision. We also use a compensator to compensate the position controller and present a landing strategy to avoid the system's instability brought from frequent switch of controllers while landing.