A Method of Cooperative Pursuing and Shooting of Multi Robots
Wende Ke, Bingrong Hong, Cui Gang, Zesu Cai, Zhiping Peng · International Conference on Electric Information and Control Engineering · 2012
The Q learning algorithm with ¦Ð calculus was proposed to solve the problems of big learning space, slow learning velocity and lacking the ability of reasoning and decision in the traditional reinforcement used in cooperative pursuing and shooting of multi robots. The model of pursuing and shooting problem was constructed, the definitions of ¦Ð cooperating calculus were shown, and the reward functions of Q learning in sensing and moving to the ball were introduced. The validity was proved through the experiments.