Affordable Access

Trajectory planning for robotic maintenance of pasture based on approximation algorithms

Authors
  • Cariou, C.
  • Gobor, Zoltan
Publication Date
Jan 01, 2018
Source
HAL-UPMC
Keywords
Language
English
License
Unknown
External links

Abstract

This paper addresses the problem of trajectory planning of a mobile robot for pasture maintenance comprising mulching weeds, reseeding patches without vegetation and spreading cowpats. Based on the sensor-based acquired data (points of interest), the proposed approach is to first use an approximation algorithm for data clustering in the form of non-convex and convex hulls. These hulls are then delimited by stair-shaped limits with respect to the working width of the robot, and their centres of gravity calculated. To minimise the travelled distance between the centres of gravity of the defined areas, the Travelling Salesman Problem is addressed via an evolutionary algorithm. Finally, kinematic and dynamic properties of the robot are considered in order to generate the final trajectory. The capabilities of the proposed approaches are highlighted through the processing of several datasets.

Report this publication

Statistics

Seen <100 times