Application of an Inertial Sensor Unit for Position Estimation and Motion Control of the Lower-Extremity Powered Exoskeleton

Abstract

Introduction. The problem of controlling the lower-extremity powered exoskeleton motion was investigated. To solve it, it was proposed to use a program control and feedback control. The formation of control in the form of feedback required an assessment of the state of the exoskeleton (rotation angles, angular velocities, and accelerations of the links). The possibility of using an inertial measuring unit to estimate angular velocities and accelerations of exoskeleton links was considered. The work objective was to develop laws for the formation of the exoskeleton motion control, which could provide the stability of the program motion and use the measurements of encoders, micromechanical gyroscopes and accelerometers.Materials and Methods. Previously performed mathematical modeling of the exoskeleton dynamics was used to form a program control. It was proposed to equip the exoskeleton with inertial sensor units. This solution made it possible to evaluate the state vector of the exoskeleton and to use these estimates in a feedback loop. A mathematical model of measurements of these sensors was described. The proposed version is suitable for control systems of three-link exoskeletons of the lower extremities and can be expanded to the case of multi-link exoskeleton designs.Results. New laws of exoskeleton motion control based on a mathematical model of the system dynamics and using measurement information from encoders and inertial information sensors were proposed. Numerical simulation of exoskeleton motion was performed in the Wolfram Mathematica mathematical package. Its results confirmed the operability of the proposed control and the possibility of using an inertial sensor unit to assess the exoskeleton state. The numerical simulation results for the following program movements were presented: lifting the exoskeleton from a sitting position to a vertical position, and stabilization of the vertical equilibrium position.Discussion and Conclusions. The proposed control can be applied in exoskeletons for medical purposes, e.g., in the task of verticalization of patients with dysfunctions of the musculoskeletal system. The possibility of using measurement information obtained from inertial measurements units in the problem of estimating the state of exoskeleton links was demonstrated. The use of inertial sensors will make it possible to determine the angular acceleration of the exoskeleton links, avoiding numerical differentiation of the measurement information received from the encoders. The obtained estimates of angular acceleration allow us to introduce feedback on angular accelerations into the control system, which opens up the possibility of improving transients in controlling the exoskeleton motion.

Keywords

• exoskeleton
• mathematical model
• inertial sensors
• control
• state estimation