Control and Analysis of an Omnidirectional Autonomous Robot Based on Software Approach and Multimedia Database

Abstract

Robocup competition is an international event for research on fully autonomous robot control and related subject like: Artificial intelligence, Image processing, machine learning, robot path planning, and obstacle avoidance. Robot soccer is more than controlling a single autonomous robot. It also deals with interacting with other objects, i.e. interacting with the ball and interacting with team mates and opponent robots. Hence, the ability to model other objects and their dynamic behaviour is inevitable to be successful. In this paper new practical software based methods for control, analysis, decision making and trajectory of an autonomous robot in Middle Size Soccer Robot league (MSL) are presented. In a robots soccer game, the environment is highly competitive and dynamic. In order to work in the dynamically changing environment, the software of a soccer robot system should have the features of flexibility, real-time control and adaptation. Omni-directional mobile robots have many applications especially in robotic soccer competitions. Omni directional navigation system, omni-vision system and kicking mechanism in such mobile robots have not ever been combined. This situation brings the idea of a robot with no head direction into existence, a comprehensive omni directional mobile robot. Such a robot can respond more quickly and it would be capable for more sophisticated behaviours with multi-sensor data fusion algorithm for global localization. Despite recent advances, effective control and self-localization methods of omni-directional mobile robots remain as important and challenging issues. For this purpose, we utilize the sensor data fusion method in the control system parameters, self localization and world modelling. A vision-based self-localization and the conventional odometry systems are fused for robust self-localization. The methods have been tested in the many Robocup competition field middle size robots. The localization algorithm includes filtering, sharing and integration of the data for different types of objects recognized in the environment. This paper has tried to focus on description of areas such as omni directional mechanisms, omni-vision sensor for object detection, robot path planning, multimedia database management system for sophisticated offline analysis and other subjects related to mobile robot's software. The results are satisfactory which has already been successfully implemented in ADRO RoboCup team. This project is still in progress and some new interesting methods are described in the current report.