|Full text PDF:||http://www.theseus.fi/handle/10024/72212|
This thesis introduces the development of team strategy for Botnia RoboCup Small Size League (SSL) Robot Team. It aims to optimize the strategy software and apply STP architecture for team strategy development to obtain satisfactory performance. Meanwhile, introduction to several important algorithms, such as ERRT, safe navigation and STP architecture is another target. The main contribution of this thesis is that it optimizes the strategy software, thereby providing more stable and convenient platform for testing and research. Also, it applies STP architecture in the team strategy, for the control of both overall team and single robots behaviors, and thus obtains good performance as expected and our system can make appropriate strategy according to different circumstances. Moreover, it successfully utilizes probability theory to scientifically evaluate the actual field situation for decision making based on mathematical model. The thesis mainly develops on Qt at Linux platform and simulates by Matlab at Windows platform. The implementation methods in this thesis are typical soft-ware engineering approaches, which is planning, developing, debugging and test-ing. For sake of debugging convenience, the algorithm was firstly simulated on Matlab and then implemented in the software. It can be concluded that STP architecture is very useful for team behavior control and mathematical theory support is very important and necessary for a system.