Genetic Programming based Optimisation of Soccer Agents
Authors: Christopher Lazarus, Huosheng Hu
Abstract
This paper describes our approach on evolving a robot controller to generate a set of goalkeeper behaviours in the RoboCup domain based on the genetic programming (GP) technique. The goalkeeper agent is a software construct that operates in a simulated environment provided by the Soccer Server. Our research is to determine the conditions that need to be met for an evolved goalkeeper agent to perform adequately in a defensive situation. A framework is developed to test the agent’s performance in the simulator. The impact of the sensors information that operates as the input to our GP implementation is addressed. The framework by incorporating it within two established soccer agents is evaluated, and then the offline performance of each agent is tested. The experimental results show that our approach is able to produce a robust robot controller and the framework is flexible enough to be readily incorporated into other existing soccer agent frameworks.
| Attachment | Size |
|---|---|
| License+Agreement+2008_Essex.pdf | 71.56 KB |
| IPROMS2008_slides.ppt | 1.08 MB |
Hi, authors,
Thanks for the very interesting work. What is your future expectation of this work? Do you think this or other similar techniques will one day be used to totally help evolve a robot, at least for a specific scenario, like this soccer RoboCup.
Thanks.
Hi Ze Ji,
Thank you for your comments. We are currently working on improving this work. Similar techniques are already being being employed in the RoboCup simulator league but due to this being a hard problem, most evolutionary work using GP makes use of human-coded high-level behaviours as part of the GP's functions and terminals. In our work, we are proposing to evolve behaviours using lower-level primitives. We hope our techniques will help to further work in this area and in the future we may be able to fully evolve soccer agents controllers without making use of any human-coded controller layers.
Dear authors and all conference participants,
Welcome to the robotics session of this IPROMS 2008 conference.
My name is Ze Ji, and I am the co-chair of this session.
If you need any assist regarding this session, please feel free to contact me via email: JiZ1 (at) cf.ac.uk.
First, I would thank the authors for contributing to IPROMS 2008.
For this paper, I have realised that the video has not been uploaded. Could the authors please prepare a video presentation and upload it? This will be helpful for others understanding your paper better. Here is the instructions on how to create a video presentation
In order to access the full text papers and video presentations and other functionalities, all conference participants are encouraged to register on the IRPOMS website online for free.
You are very welcome to post questions after reading the papers, and I believe the authors will be there very happy to answer you.
I hope you will enjoy this session.
Kind regards,
Ze Ji
Robotics Session Co-Chair
IPROMS 2008