Sélectionner une page

Robot Club Toulon 2019 robots have participated in the RoboCup in Sydney. 

They are based on a 3 wheels holonomic autonomous plateform wighting 30kg. Its size is 50cm*50cm and it height is 80cm.

Positionning 

Global positionning is done using an omnidirectional camera at the top of the robot. This one uses a special custom mirror designed in order that an horizontal plane is not distorted in the image.

This global positionning in the merged with odometric data and inertial sensor data using a Kalman filter in order to compute a precise position of the robot all the time and to detect sliding.

Kicking system

Kicking system relies on a coil gun striking balls using a 350g iron rod propelled by a 4700uF / 450V capacitor coupled with a 3kg copper coil. Plunger acceleration in the kicking system is approximately 100G and power during the impulse is approximately 70KW.

Environment perception

Each robot have the ability of knowing its environment using the omnidirectionnal camera, a second front camera and two SICK TIM361 Lidars. All the data from these sensors are gathered, processed and merged using an embedded computer (LattePanda Alpha) which will be replaced in the 2020 version in order to be more reliable and to have more computational power.

Team strategy

Team strategy is computed autonomously by a « base station ». This is a laptop computer located near the field and communicating with each robot, getting their different environment perceptions in order to build a « map of the world » describing the soccer field and woth teams as precisely as possible. The ball position and the robots positions and move are very important in order to build the team strategy, computed using an AI based on rules for the moments.

Ball catcher

Ball catcher is a very important part of the robot. It allows to cath the ball from the opponent or for received a pass. Based on two wheels mounted on a vertical sliding system, our ball catcher was not satisfying in the 2019 robot and will be fully redesigned in the 2020 version.