Design And Implementation Of Communication And Decision System For Robocup Small-size Soccer Robot

With the continuous development of science and technology, robots can be widelyapplied, the soccer robot has also been rapid development. Robot soccer game is a high-techcombat activity which blends with multidisciplinary and involves a number of research areas,including automatic control, artificial intelligence, wireless communications, computerscience and pattern recognition and so on. In this paper, we focus on RoboCup soccer robot asthe research object, and study wireless communication subsystem and decision-makingsubsystem in the robot game system. The main research work and contributions of thedissertation are summarized as follows:(1) The wireless communication mode and protocol of robot soccer game is designed.Moreover, a set of wireless communication subsystem based on NRF2401A monolithic radiofrequency transceiver chip is exploited, and the hardware circuit diagram and software flowchart of design of wireless transceiver is given. The test results show that the subsystem cannot only satisfy the requirement of the robot soccer communication, but also have highersystem stability and stronger anti-jamming capability.(2) By analyzing the coordinated control strategy of soccer robot, a reasonable divisionmethod of field area and the condition of role assignment and change are given. In addition,the specific offensive and defensive strategies of soccer robot game were designed and adetailed program flow chart is also given.(3) For time delay problem exists in robot soccer game system, a mathematical model ofsmall size soccer robot system is built. By using extended Kalman filter algorithm in thestrategy design, we can predict the football movement ahead of a few periods. The soccerrobots can also move in advance according to the forecast data, and the time delay problem ofthe system can be solved effectively.(4) The simulation and experiment results for offensive and defensive game in thesimulation platform show the feasibility and practicality of the strategy. The MATLABsimulation results show that it can accurately predict football position with the extendedKalman filter algorithm, and the accurate predictive value can provide a guarantee for ourrobot targeted movement.