Implementation of multi agent robot systems
Studies on artificial intelligence for multi-agent system
Figure 1. Robot agent
In this project, an autonomous multi-robot system was developed. Each robot or agent should be totally autonomous in that it has basic motion capabilities such as move or obstacle avoidance. For this each robot is equipped with servo-motors, a camera and a Pentium-II computer. The robots have also wireless eithernet cards so that they can communicate with each other. A human user can interact with them through a host computer which monitors the group status and provides information to the human user. The project objectives comprise of several sub-goals:
1. Omni-directional robot mechanism: Omni-directional mobility is advantageous for autonomous agents since it makes the control strategy simple while giving better motion capability.
2. Stand-alone vision board: A camera is the most powerful sensor for autonomous mobile robots. Since the processing time for camera signal is usually long compared to control algorithms, a fast processing vision board is essential for real time vision processing. A stand-alone vision board will be developed using DSPs and EPLDs.
3. Positioning: Accurate mobile robot positioning is very essential for good performance. A positioning algorithm using onboard sensors will be developed for accurate positioning of each robot.
4. Man-machine interface: To make the system user friendly, graphical or haptic interface is considered. The user can participate in the task performing process or can interact with the robots.
5. Multi-robot control strategy: The control strategy for each robot's behavior and overall behavior of the group will be considered.
