ILC-MPC CONTROLLER FOR ROBOTIC MANIPULATORS BASED ON THE ULTRA-LOCAL MODEL

This research explores the possibility of simplifying model predictive control strategy for robotic manipulators and improving the control system’s performance with data-driven learning controllers. The main goal is to synthesize a controller that will be feasible for embedded hardware. Simplifying the robot dynamics is done using the ultra-local model method, and then new equations of motion are used to solve a nonlinear optimization problem in model predictive control. An iter- ative learning controller with a serial structure is added for increased performance when the given task is repetitive. Test simulation is carried out in Matlab to verify the feasibility of the proposed control system. Results of the simulation show that the proposed controller indeed manages to at- tenuate external disturbances and improve performance through the learning process.