Back to Search Start Over

DEVELOPMENT AND TESTING OF AUTOMATED CONTROL SYSTEM FOR SEA BUCKTHORN BERRY HARVESTING ROBOT “AGROBOT”.

Authors :
Stupans, Andrejs
Maksimkins, Pavels
Senfelds, Armands
Ribickis, Leonids
Source :
Engineering for Rural Development - International Scientific Conference. 2024, p887-893. 7p.
Publication Year :
2024

Abstract

The paper describes the control system of an autonomous agriculture robot and evaluates its operation using a laboratory stand and a digital environment. AgroBot is an autonomous sea buckthorn berry harvesting robot. It automatically finds the bush branches to harvest, cuts them, and stores them in a box. AgroBot consists of a 3 DOF (Degrees of Freedom) Cartesian mobile platform and a Hyundai HH7 industrial robot arm with 6 DOF. The control system is specifically designed for real-time operation, enabling AgroBot to adapt to dynamic environmental conditions (wind, varying light) that complicate target tracking. The control system consists of two modules operating as two separate programs. The first module is the Computer Vision Module (CVM) which has high-level control of AgroBot operation. It uses feed from cameras to find the cutting spot and sends commands to approach, cut, store, and search. The module is written in Python. The second module is the Robot Control Module (RCM) which receives high-level commands from CVM and manages lower-level control of the Hyundai Controller (HC). RCM calculates the robot trajectory to a target position, communicates with both CVM and HC in parallel threads and handles errors. As HC requires time-critical control, RCM is written in C + + . RCM and CVM are running on one computer communicating via sockets. As the mobile platform hardware is in the development stage and is not available at the moment, the digital twin of the robot is created to test the system’s performance in a simulated environment. The input to the digital twin is the same as for the actual robot. It is the x, y, z position and orientation A, B, C using Euler angles. The digital twin visualization is developed in the Unity game engine. Matlab Robotic Toolbox is used with the Levenberg-Marquardt solver algorithm to calculate inverse kinematics of 9 DOF robot. The paper focuses on the Robot Control Module architecture and control system’s testing. [ABSTRACT FROM AUTHOR]

Details

Language :
English
ISSN :
16913043
Database :
Academic Search Index
Journal :
Engineering for Rural Development - International Scientific Conference
Publication Type :
Conference
Accession number :
178284780
Full Text :
https://doi.org/10.22616/ERDev.2024.23.TF175