Search

In this paper, a new posture prediction method is proposed and evaluated on human leg, as being a physiologically constrained three-link arm. The main posture prediction solution is focused on optimizing the manipulability and the human performance of the leg. The forward kinematics is used to define the feasible workspace of the human leg in sagittal plane. Using an effective optimization-based human performance measure that incorporates a new objective function of musculoskeletal discomfort, the optimal posture is obtained.

This paper proposes a non-singular terminal sliding mode combined with time delay controller for fast tracking trajectory of uncertain second-order nonlinear systems in presence of external disturbances. The motivation for using Non-Singular Terminal Sliding Mode (NSTSM) mainly relies on its appreciable features, such as simplicity of design and implementation, high precision and fast convergence. However, the NSTSM problem is the well-known chattering, which degrades the controlled system. Then, using the time delay to estimate uncertainties and disturbances allows chattering reduction. The stability and the robustness of the proposed controller are verified by using the classical Lyapunov criterion. The proposed controller is implemented in real time on the seven-degrees-of-freedom (7-DOF) ANAT robot arm and compared with the Sliding Mode with Time Delay Control (SMTDC) with classical sliding surface to prove its superiority.