State Estimation of the Self-Balancing Lower Limb Exoskeleton Using Extended Kalman Filter

Abstract

The safe operation of self-balancing lower limb exoskeletons heavily depends on real-time and precise monitoring of essential state variables, including position, velocity, and posture, necessitating the development of more refined and efficient measurement update techniques. This paper delves into the realm of state estimation strategies, leveraging the Extended Kalman Filter (EKF) algorithm. By integrating diverse information sources like inertial measurement units (IMUs), joint encoders, and planned bipedal trajectory, it successfully accomplishes precise estimation of the exoskeleton’s torso and foot states within the global coordinate system.The EKF-based state estimation algorithm introduced in this paper notably enhances the overall state estimation’s accuracy and robustness by effectively mitigating the accumulation and propagation of IMU measurement errors at the exoskeleton’s torso. This advancement establishes a robust foundation for stable and precise control of self-balancing exoskeleton robots. Simulation experiments have further substantiated the high-precision capabilities of the proposed exoskeleton state estimation method.