Optimal Control of a Wheeled-Exoskeleton Using Linear Quadratic Regulator and Kalman Filter

Exoskeletons are increasingly gaining attention due to their role in medical rehabilitation and restoring mobility in individuals with movement impairments. However, traditional exoskeleton designs face limitations due to the use of a rigid structure and reliance on complex control algorithms. This work proposes integrating exoskeletons with a wheeled base to enhance mobility, alongside developing an optimal control strategy using a Linear Quadratic Regulator (LQR) combined with a Kalman filter to stabilize the system's vertical equilibrium. The dynamic model of the system is derived using Euler-Lagrange equations, yielding nonlinear differential equations that are linearized and written in state-space form. A control strategy based on minimizing a quadratic cost function is implemented to optimize energy efficiency. To address disturbances from sensor noise, the Kalman filter is employed to improve algorithmic robustness. Simulation results demonstrate the transient response of the exoskeleton's joint angles to initial conditions under feedback control and noise filtering. The controller successfully tracks predefined trajectories while maintaining equilibrium, validating the algorithm's ability to stabilize the system. Additionally, the proposed model's energyefficient performance highlights its potential for advancing the design of wheeled exoskeletons in medical and assistive applications.