Omnidirectional robot modeling and simulation

TL;DR

This paper presents a modeling and simulation approach for omnidirectional robots, focusing on efficient and accurate simulation to aid development and testing of robotic algorithms safely and cost-effectively.

Contribution

It introduces a two-step modeling process for omnidirectional robots, addressing motor non-linearities and overall behavior for improved simulation accuracy.

Findings

Successful modeling of motor non-linear features.

Accurate simulation of robot velocity response.

Efficient simulation approach for robot fleets.

Abstract

A robot simulation system is a basic need for any robotics application. With it, developers' teams of robots can test their algorithms and make initial calibrations without risk of damage to the real robots, assuring safety. However, building these simulation environments is usually time-consuming work, and when considering robot fleets, the simulation reveals to be computing expensive. With it, developers building teams of robots can test their algorithms and make initial calibrations without risk of damage to the real robots, assuring safety. An omnidirectional robot from the 5DPO robotics soccer team served to test this approach. The modeling issue was divided into two steps: modeling the motor's non-linear features and modeling the general behavior of the robot. A proper fitting of the robot was reached, considering the velocity robot's response.