Gait Analysis and Estimation of GRF Towards Control of Lower-Limb Exoskeleton

Abstract

This thesis presents the development of a comprehensive lower-limb exoskeleton framework aimed at augmenting human mobility and physical performance. A detailed kinematic analysis was carried out to define joint trajectories and limb configurations, employing Denavit-Hartenberg (D-H) approach suited for human-exoskeleton alignment. To validate the model and ensure its biomechanical relevance, experimental gait data was obtained using Gait Lab systems under IOR (Istituto Ortopedico Rizzoli) method for human motion capture protocols. Further, Ground Reaction Force (GRF) estimation was performed as a critical element in kinematic modeling, allowing the system to reflect realistic load-bearing conditions during walking. Unlike purely analytical approaches, a learning-based estimation method was adopted for more accurate and adaptive results. The control techniques help in controlling the system and getting desired motion. To enhance system efficiency and effectiveness, advanced optimization techniques like GA, PSO and GA-PSO were implemented to fine-tune control parameters and ensure minimal tracking error with robust performance.

The integrated approach of combining biomechanics, gait analysis and control optimization culminates in a scalable and adaptive solution for human-exoskeleton synergy, laying the foundation for future development in assistive and rehabilitative robotic systems.