A trajectory prediction algorithm based on fuzzy rectification for spinning ball

Trajectory prediction plays a very important role in the process of playing table tennis for robot. Its accuracy determines whether the striking action will succeed or not. This paper first analyzes how the spinning influences the flight model and the rebound model for the table tennis robot. Two models are designed for spinning balls. Meantime, it indicates that the offset on y axis is related to the flight distance on x axis and rotation degree. Then a fuzzy controller is designed for online rectification of the trajectory. And a trajectory prediction algorithm is presented for the spinning ball combined with the algorithms based on the parameter model and experience learning. The experiment results validate the effectiveness of the proposed method.