Dual-Channel Adaptive NMPC for Quadrotor under Instantaneous Impact and Payload Disturbances

TL;DR

This paper introduces a dual-channel adaptive NMPC control architecture for quadrotors, enhancing their ability to handle impact forces and payload uncertainties during object capture tasks.

Contribution

It proposes a novel DCA-NMPC framework combining nonlinear MPC with adaptive controllers to improve robustness against impact and payload variations.

Findings

Successfully resists impact forces during simulations

Adapts to uncertain payload inertial parameters

Enhances quadrotor stability during object capture

Abstract

Capturing target objects using the quadrotor has gained increasing popularity in recent years, but most studies focus on capturing lightweight objects. The instantaneous contact force generated when capturing objects of a certain mass, along with the payload uncertainty after attachment, will pose significant challenges to the quadrotor control. This paper proposes a novel control architecture, namely Dual-Channel Adaptive Nonlinear Model Predictive Control (DCA-NMPC), which cascades a nonlinear model predictive control with two lower-level model reference adaptive controllers and can resist drastic impact and adapt to uncertain inertial parameters. Numerical simulation experiments are performed for validation.