The rising prevalence of mobility impairments due to accidents, strokes, and aging has significantly increased the demand for physiotherapy services. Conventional rehabilitation is repetitive and physically demanding for both patients and therapists. While exoskeleton robots offer a promising alternative, their high costs limit access for researchers and students seeking practical experience. This paper presents a cost-effective laboratory simulator for knee exoskeleton robots using three DC motors to represent the knee joint, resistance forces, and assistive mechanisms. The simulator\\\\\\\\\\\\\\\'s realism stems from detailed modeling of physical parameters including inertia, mechanical stiffness, and damping coefficients of the couplings, which together create a tunable and realistic experimental platform. The mechanical design is developed in SolidWorks and the dynamic simulation is performed in MATLAB. To achieve robust performance, a multi-agent leader-follower control architecture is implemented where each motor operates as an independent agent. This architecture is combined with a hybrid controller that integrates PID and sliding mode control along with a disturbance observer to compensate for uncertainties. Physiologically realistic walking gait patterns are generated using musculoskeletal modeling in SCONE, and applied to the system. Simulation results demonstrate that the proposed control strategy successfully tracks the desired velocity and torque profiles with high accuracy and minimal error. The simulator\\\\\\\\\\\\\\\'s detailed physical modeling, combined with its simplicity, safety, and low cost, makes it a practical tool for educational and research applications in rehabilitation robotics.

Keywords