An Embedded Platform for Positioning and Obstacle Detection for Small Unmanned Aerial Vehicles
Salvatore Ponte,Gennaro Ariante,Umberto Papa,Giuseppe Del Core
DOI: https://doi.org/10.3390/electronics9071175
IF: 2.9
2020-07-19
Electronics
Abstract:Unmanned Aerial Vehicles (UAV) with on-board augmentation systems (UAS, Unmanned Aircraft System) have penetrated into civil and general-purpose applications, due to advances in battery technology, control components, avionics and rapidly falling prices. This paper describes the conceptual design and the validation campaigns performed for an embedded precision Positioning, field mapping, Obstacle Detection and Avoiding (PODA) platform, which uses commercial-off-the-shelf sensors, i.e., a 10-Degrees-of-Freedom Inertial Measurement Unit (10-DoF IMU) and a Light Detection and Ranging (LiDAR), managed by an Arduino Mega 2560 microcontroller with Wi-Fi capabilities. The PODA system, designed and tested for a commercial small quadcopter (Parrot Drones SAS Ar.Drone 2.0, Paris, France), estimates position, attitude and distance of the rotorcraft from an obstacle or a landing area, sending data to a PC-based ground station. The main design issues are presented, such as the necessary corrections of the IMU data (i.e., biases and measurement noise), and Kalman filtering techniques for attitude estimation, data fusion and position estimation from accelerometer data. The real-time multiple-sensor optimal state estimation algorithm, developed for the PODA platform and implemented on the Arduino, has been tested in typical aerospace application scenarios, such as General Visual Inspection (GVI), automatic landing and obstacle detection. Experimental results and simulations of various missions show the effectiveness of the approach.
engineering, electrical & electronic,computer science, information systems,physics, applied