MSc Thesis Defense: Araz Haghnazari Esfahlan, DESIGN AND CONTROL OF A SOFT ROBOTIC EXOSKELETON FOR PHYSICAL ASSISTANCE,
DESIGN AND CONTROL OF A SOFT ROBOTIC EXOSKELETON FOR PHYSICAL ASSISTANCE
Araz Haghnazari Esfahlan
Mechatronics Engineering, MSc. Thesis, 2025
Thesis Jury
Asst. Prof. Melih Turkseven (Thesis Supervisor)
Prof Volkan Patoglu
Assoc. Prof. Nazila Nikdel
Date & Time: December 19th, 2025 – 10:30 AM
Place: FASS G018
Keywords : Soft exoskeleton, Pneumatic soft actuator, Lagrangian dynamics, Control of soft robots, Piece-wise Constant Curvature
Abstract
Soft exoskeletons offer promising solutions for human assistance and rehabilitation due to their inherent compliance, safety, and adaptability. However, the development of dynamic models and effective control strategies for these systems remains challenging because of nonlinear material behavior, distributed actuation, and complex deformation characteristics. This thesis presents the design, modeling, simulation, and control of a pneumatically driven soft elbow exoskeleton. A reduced-order dynamic model is developed using Lagrangian dynamics and piecewise constant curvature kinematics. The model, which is pneumatically driven, incorporates elastic and damping effects act as a spring and damper. Simulation results are compared with experimental measurements to assess the model's ability to accurately emulate the dynamics of the physical system. The results demonstrate that the developed model captures the steady-state behavior of the soft actuator with acceptable accuracy, while discrepancies in transient response are undeniable. In addition, a model-based control strategy is implemented and evaluated in simulation to move the tip position toward desired points. The controller is tested using multiple randomly selected reference positions. The results show consistent convergence and stable behavior for all selected desired positions, indicating robustness with respect to variations in the desired tip position.Overall, this work demonstrates that the reduced-order Lagrangian model, combined with piecewise constant curvature kinematics, can provide a good model for soft actuators, yielding acceptable results. The findings establish useful information for future development toward assistive and rehabilitative applications.