Towards trajectory planning from a given path for multirotor aerial robots trajectory tracking
| Autor: | Manuel Castillo-Lopez, Miguel A. Olivares-Mendez, Holger Voos, Jose Luis Sanchez-Lopez |
|---|---|
| Rok vydání: | 2018 |
| Předmět: |
Computer science [C05] [Engineering
computing & technology] 0301 basic medicine 0209 industrial biotechnology Computer science 02 engineering and technology Sciences informatiques [C05] [Ingénierie informatique & technologie] Computer Science::Robotics 03 medical and health sciences Jerk Acceleration 030104 developmental biology 020901 industrial engineering & automation Control theory Trajectory Robot Quaternion Focus (optics) Multirotor |
| Zdroj: | 2018 International Conference on Unmanned Aircraft Systems (ICUAS), Dallas 12-15 June 2018. Dallas, USA (2018). BASE-Bielefeld Academic Search Engine |
| Popis: | Planning feasible trajectories given desired collision-free paths is an essential capability of multirotor aerial robots that enables the trajectory tracking task, in contrast to path following. This paper presents a trajectory planner for multirotor aerial robots carefully designed considering the requirements of real applications such as aerial inspection or package delivery, unlike other research works that focus on aggressive maneuvering. Our planned trajectory is formed by a set of polynomials of two kinds, acceleration/deceleration and constant velocity. The trajectory planning is carried out by means of an optimization that minimizes the trajectory tracking time, applying some typical constraints as m-continuity or limits on velocity, acceleration and jerk, but also the maximum distance between the trajectory and the given path. Our trajectory planner has been tested in real flights with a big and heavy aerial platform such the one that would be used in a real operation. Our experiments demonstrate that the proposed trajectory planner is suitable for real applications and it is positively influencing the controller for the trajectory tracking task. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|