An Aerial Robot Path Follower Based on the ‘Carrot Chasing’ Algorithm

Abstract

This paper presents a three-dimensional path follower implementation for an aerial robot based on the carrot-chasing algorithm. The main objective was to improve the performance of the position controller of the PX4 autopilot when following a list of waypoints. This autopilot is widely used in the aerial robotics community, but we needed to improve its performance for navigation in cluttered environments. Different simulations have been carried out under the ROS (Robotic Operating System) environment for the comparison between the position controller of the PX4 and the proposed path follower. In addition, we have implemented different modes to generate the path from the input list of waypoints that are also analyzed in our simulation environment.

Autonomous 3D Exploration of Large Structures Using an UAV Equipped with a 2D LIDAR

“…The software architecture is based upon the Robot Operating System (ROS) software framework by Quigley et al [46], which allows a hardware-agnostic design through the use of its interprocess communication interfaces. The different components are described in the following: Sensor : The sensor is a Hokuyo 30LX 2D laser sensor with a range of 30 m; Software for basic commands execution : UAL (UAV Abstraction Layer): A software-interface for hardware abstraction [47] which handles the standard commands to control the vehicle such as velocity control, taking-off, and landing; Path Follower: Software to follow a waypoint sequence [48], while also adjusting vehicle yaw, so that in every segment the Hokuyo sensor is aligned with the movement. Octomap : Occupancy octree for world representation using the octomap framework [42], as previously detailed in Section 3.1. The world representation is shared among all the components; Flyby manoeuvre : The manoeuvre executed to collect data around the target to gather 3D information with the 2D laser, as described in Section 3.2; Path Planning : The Lazy Theta* any-angle deterministic planner proposed by [43] and adapted in previous work of the authors [44] due to the advantages mentioned in Section 3.3; Exploration Strategy : Frontier Algorithm : The classical and widely-used frontier exploration algorithm presented in [28]; The implementation used is an extension of [49], that generates neighbours taking the sensor range into account as presented in Section 3.4; Frontier Management : Combines and orders the operational requirements, such as safety distance, observation manoeuvre visibility, or obstacle detection, with exploration optimisation.…”

“…Software for basic commands execution : UAL (UAV Abstraction Layer): A software-interface for hardware abstraction [47] which handles the standard commands to control the vehicle such as velocity control, taking-off, and landing; Path Follower: Software to follow a waypoint sequence [48], while also adjusting vehicle yaw, so that in every segment the Hokuyo sensor is aligned with the movement.…”

“…Path Follower: Software to follow a waypoint sequence [48], while also adjusting vehicle yaw, so that in every segment the Hokuyo sensor is aligned with the movement.…”

A Stable Path Following Algorithm Based on Adaptive Target Points for Unmanned Surface Vehicles

“…One of the methods used in the path following algorithm is the carrot chasing algorithm in the literature. It is preferred by many researchers (Perez et al, 2019;Chung et al, 2020;Jin et al, 2020) due to its simplicity and easy applicability. Moreover, it is an effective technique to catch the desired path for USVs.…”