Model predictive control of a serial link robot manipulator

Abstract

This paper presents a novel approach for model predictive control dynamics applied to a two-link manipulator robot. The technique involves the initial step of linearizing the inherently nonlinear dynamic model of the robot through the application of feedback linearization control. Once the linear model is derived, a predictive control strategy is developed to enhance the robot's overall performance. To achieve this, we introduce a quadratic criterion, and the associated parameters are meticulously computed to induce specific behaviors within the closed-loop system. The primary objective of this study is to effectively control the arm robot, guiding it from an initial configuration to a desired final configuration by employing the predictive control approach, all while minimizing the quadratic criterion. The chosen criterion serves as a crucial metric in evaluating the system's response and optimizing its behavior during the control process. To validate the effectiveness of the proposed method, a series of simulation results are provided, showcasing the system's performance under various conditions. These simulations serve as empirical evidence supporting the feasibility and efficiency of the introduced model predictive control approach for the dynamics of a twolink manipulator robot. The outcomes highlight the method's potential for real-world applications and contribute to the ongoing advancement of control strategies in robotic systems.