Covariance Intersection Fusion Approach for Gait Estimation of Lower Limb Rehabilitation Exoskeleton Robot

Abstract

This paper aims to study the gait estimation problem of lower limb rehabilitation exoskeleton robot, and proposes a gait estimation strategy based on covariance inter-section (CI) fusion. First, the discrete mathematical model of exoskeleton rehabilitation robot is obtained by Lagrange system discretization principle. Then, two groups of angle sensors with unavoidable measurement errors are employed to obtain the joint angles of the exoskeleton robot, and the gait of the robot is estimated by CI fusion filtering algorithm. Finally, simulation experiments show that CI fusion filtering algorithm can effectively overcome the interference of process noise, and greatly improve the accuracy of gait estimation during gait tracking.