Nonlinear Model Predictive Multivariable Control for Trajectory Tracking of a Type of Delta-Parallel Robot

In this paper, an optimal robust nonlinear model predictive controller based on harmony search algorithm is designed for a type of 3-DOF translational parallel robot. Dynamic model of the mechanism is derived using Lagrange method and the model predictive controller augmented by uncertainty estimator is designed and stability is proved by Lyapanov theorem. Performance of the designed controller is evaluated in different conditions such as presence of disturbance and parameter variation. Furthermore, an optimal trajectory consisting four circular obstacles is designed as the reference trajectory of the robot. In order to obtain the optimum control parameters, a cost function combining control signal rate and error is considered and minimized by harmony search algorithm. In order to compare the performance of the designed controller with other nonlinear controllers, two controllers, an optimal sliding mode and a feedback linearization controller are also designed and their results are compared. Simulation results depict the desirable performance of the three controllers in spite of disturbance and model uncertainty, however, error criteria indicate priority of the robust nonlinear model predictive controller over the two other controllers.