SYNTHESIS FOR EXOSKELETON ELECTRIC DRIVE CONTROL SYSTEM WITH AN OBSERVER AND COMPENSATION OF DISTURBANCES WITH THE NEURAL NETWORK

The article proposes a method of observer and compensation of indefinite disturbance to the exoskeleton electric drive. A mathematical model of the exoskeleton electric drive control system has been developed, taking into account the nonlinear components and the interaction force of the exoskeleton with the lower limb. The proportional differential (PD) control law combines an adaptive neural network RBF (Radial Basis Function) with a linear-quadratic regulator as the main foundation for the observer and compensation for uncertain disturbances in the control system. The PD control is used for the stability of the main part of the model. The adaptive part of the regulator is used to compensate for the deviation of the system characteristics from the dominant linear model to improve performance. The advantage of the supply regulator is resistance with strong nonlinearities and external disturbances of the system without abandoning the PD regulator, which is well known to many engineers. The proposed control scheme provides limited system states and parameter estimation. The simulation results in Matlab/Simulink show that the proposed PD-linear quadratic controller with adaptive compensation using a neural network is more efficient than a conventional PD-linear quadratic controller.

Authors: M. P. Belov, D. D. Truong

Direction: Electrical Engineering

Keywords: Exoskeleton, electric drive, exoskeleton electric drive, control system, disturbance observer, disturbance compensation, neural network

---

View full article