Control approaches for the most modern exoskeleton devices are based on the use of the potentiometric proportional sensors. This allows setting the velocity of the movement of the exoskeleton links, but has significant peculiarities, which are concluded in a large time delay for processing the control signal and increased sensitivity of such sensors, which leads to increased injury risk during control. The use of muscle biopotentials for control of an exoskeleton device also makes it possible to take into account the physiological characteristics of the operator for using the exoskeleton in various areas of human activity. The development of control algorithms of the exoskeleton, along with the use of the activity of human muscle groups’ data, is essential for expanding the functionality of a human-machine system such as the “operator-exoskeleton”. The paper considers the interaction of a human and an exoskeleton drive based on mathematical models of a DC motor with a current feedback loop and a muscle duplex. A calibration algorithm is proposed to determine the parameters of the muscle duplex model in order to form a database that corresponds to an individual operator and reflects its individual characteristics. The technique for setting the parameters of the control system in the exoskeleton calibration mode is given. Paper presents the results of experiments with the developed algorithm on full-scale stand, simulating the arm exoskeleton with the electric drive, located in the elbow joint and controlling algorithms based on the electromyogram of the biceps brachii and triceps brachii of the operator. The structure and features of the stand developed in the laboratory of robotics and mechatronics of IPMech RAS are shown. A comparative characteristic of the control quality of the electric drive, which is part of the exoskeleton, with the proposed algorithm in relation to one operator when changing by another one was worked out. At the same time, the following control quality indicators were evaluated – over-regulation, time to set the specified position, and accuracy of positioning the control point of the exoskeleton link. The present work was supported by the Ministry of Science and Higher Education within the framework of the Russian State Assignment under contract No. AAAA-A20-120011690138-6.

M. M. Knyazkov

Papers

Control System Calibration Algorithm for Exoskeleton Under the Individual Specificities of the Operator