Synthesis of PID Controller with Neural Network for Nonlinear Electric Drive Exoskeleton

In this paper, a PID controller with a neural network is proposed to compensate for uncertain changes in gravity and friction of the lower limbs human exoskeleton (LLHE) in nolinear electric drive control system to ensure the desired trajectory of movement. The neural network is used to approximate the nonlinear elements of the object model based on the dynamic error of the angles of the joints of the lower extremities of the exoskeleton. The mathematical model of LLHE is built on the sagittal plane, including legs with 2 links, taking into account nonlinear elements and external disturbances. The results of modeling the controlled movement of the exoskeleton in the sagittal plane are presented, showing the high efficiency of the functioning of the PID controller with a neural network in comparison with the proportional-integral differential controller. © 2021 IEEE.