A predictive algorithm with fine interpolation for vision-guided robots

摘要

A predictive algorithm with fine interpolation for the optimization of path tracking for vision guided robotic manipulators is presented. The look-ahead trajectory information extracted from the image sensor is processed in a way that it can be used in the sampling rate of the joint control loop. The predictive control law is obtained from the short-term future trajectory and the minimization of differences between the state-variables of each linearized robot joint and the corresponding state-variables of an internal path generator model. Simulations and experimental results for a vision-guided hydraulic robot are presented.