An Approach to Hit Point Prediction for Ping-Pong Robot
Wei Wei · Jiangnan daxue xuebao. Ziran kexue ban · 2007
The prediction of the position of hit point of ping-pong and the velocity of the ball when reaching the hit point in the ping-pong robot project has been a research focus.The state function model of the flying phase of ping-pong and the motion model of the bounce process of ping-pong were set up,respectively.Based on these two models,an approach to predict the hit point of ping-pong,mainly by numerical iteration solution of the state function,was developed.An offline learning method to determine the parameters in the models was developed to accomplish prediction tasks in different conditions.Considering the considerable noise contained in the coordinate of ping-pong,special attention was paid to improve the fault-tolerance in the prediction process.According to the experimental data,the prediction accuracy of this approach is high,and the time consumed by the approach itself satisfied the requirement of real time performance.