کنفرانس مهندسی برق ایران

صفحه اصلی / سی و سومین کنفرانس بین المللی مهندسی برق

Experimental Study and Implementation of a Generalized Predictive Controller on Delta Parallel Robot Based on Actuator Identification

نویسندگان :

Hasan Jalali¹ Behnam Moradkhani² Hossein Damavandi³ Mehdi Tale Masouleh⁴ Ahmad Kalhor⁵

1- University of Tehran 2- University of Tehran 3- K.N.Toosi University of Technology 4- University of Tehran 5- University of Tehran

کلمات کلیدی :

Delta Parallel Robot،GPC،CARIMA model،Kinematics

چکیده :

Model Predictive Controllers are among the most robust conventional controllers and they are being used in a wide range of applications. Model Predictive Controllers are known for their high computational load compared to other conventional control schemes. Applying Model Predictive Controllers on different systems in different fields of study has increased in the area of robotics. In this study, three Generalized Predictive Controllers have been designed and implemented on the Delta Parallel Robot actuators. These controllers have been designed based on system identification of actuators results obtained in a previous study. High velocities of the end-effector and the input disturbance power are the factors found out to be important and effective on the controllers' tracking. The results of the Generalized Predictive Controllers have been compared to the ones with PIDs both in MATLAB Simulink simulations and in practice, taking into consideration the factors mentioned earlier. As a result, the designed Generalized Predictive Controller demonstrates the best performance, achieving the lowest Mean Squared Error with an average of 3.90 and Root Mean Squared Error with an average of 1.97, significantly outperforming the PID-based methods.