A Simple Learning Strategy for Feedback Linearization Control of Aerial Package Delivery Robot

摘要

This paper develops a simple learning (SL) strategy for feedback linearization control (FLC) algorithm for uncertain nonlinear systems. The SL strategy that uses desired closed-loop error dynamics updates the controller coefficients and the disturbance term in the feedback control law, while traditional feedforward control law is designed based on the nominal model by using FLC method. In this strategy, the desired closed-loop error function is minimized by using the gradient-descent method to find the adaptation rules for feedback controller gains and estimated disturbance. In addition, the system stability for an nth order uncertain nonlinear system is proven by using a Lyapunov function candidate. To test the efficiency and efficacy of the SL-FLC framework, the package delivery problem of a tilt-rotor tricopter unmanned aerial vehicle is studied in real-time. The experimental results show that the SL-FLC framework results in a better path tracking performance than the traditional FLC method, while maintaining the nominal control performance in the absence of uncertainties and exhibiting robust control performance in the presence of uncertainties.