This paper presents a robust control approach for robot manipulators in the presence of uncertain dynamics due to the unknown load. The optimal method of θ-D is used to obtain the law of optimal feedback control. The optimal method θ-D is based on the approximate solution of the Hamilton–Jacobi–Bellman equation (H-J-B) through the perturbation process. A part of disturbances is added to the system cost function to ensure optimality and achieve global stability. Adjustable parameters in the system disturbance components allow for flexible adjustment of the system performance. The synthesized problem of nonlinear optimal control was solved by the θ-D method, which provides an approximate analytical feedback solution. The structural controller of the neural network was analyzed on the basis of the H-J-B equation. Bayesian regression method of training for the learning process of the neural network was estimated. The results of the simulation showed that the intelligent controller built on the basis of nonlinear optimal control leads the robot manipulator to the desired position, fully providing a criterion for system quality at large load variations

Authors: M. P. Belov, D. K. Tran

Direction: Electrical Engineering

Keywords: Optimal method θ-D, intelligent controller, artificial neural network

View full article