Flexibility in robotic manipulators offers several benefits in terms of weight, maneuverability, operational speed and power consumption. However, these merits are obtained at the cost of control complexity. This paper addresses the control problem of flexible joint manipulator by proposing two laws; Linear Quadratic Regulator (LQR) and nonlinear backstepping. The objective is to accurately track the Link’s position by minimizing the vibrations due to joint’s flexibility. The novelty of the derived Lagrangian-based dynamic model lies in consideration of both critical effects; viscous damping and gravity. The system is then linearized using input-output feedback linearization. The designed control laws are then subjected to various test inputs to characterize and compare the tracking performance. Moreover, robustness of the control laws is evaluated based on applying bounded matched disturbance. Comparative results dictate that backstepping outperforms than LQR. Result of this research can be used in applications deploying flexible robotic manipulators including medical, space and industrial automation.

Keywords: Robotic manipulator, Flexible joint, Linear control, Robust control, Feedback linearization, Robustness