Neural PID admittance control of a robot

Abstract

In order to achieve the goal of the operator feeling as if a wearable robot is natural extension of the body, in this paper we develop a novel admittance control for a wearable robot. This controller uses three force sensors to generate reference signals in task space. It has PID form and can be easily realized. In the task space, in order to minimize steady-state error with respect to uncertainties in robot control, we extend the popular neural PD control into neural PID control. This novel control is a natural combination of industrial linear PID control and neural compensation. Novel sufficient conditions of semiglobal asymptotic stability are proposed via stability analysis in task space.