bahamiriyan 11 ارسال شده در 2 مهر، 2012 In this study, a SCARArobot manipulator is simulated under PD and learning based controllers. The trajectory following performance of the robot is studied against these controllers. The adaptive/learning hybrid controller scheme and learning controller method are utilized as learning based controllers. The results of simulations show that, learning algorithm based controllers reduce the position tracking error effectively. The hybrid adaptive/learning controller has similar performance as the learning controller although it uses partial state information and compensates both mechanical and electrical dynamics, whereas the learning controller needs both position and velocity measurements neglecting electrical dynamics. [h=2][/h]
ارسال های توصیه شده