Meta-Learning Augmented MPC for Disturbance-Aware Motion Planning and Control of Quadrotors

A major challenge in autonomous flights is unknown disturbances, which can jeopardize safety and cause collisions, especially in obstacle-rich environments. This letter presents a disturbance-aware motion planning and control framework for autonomous aerial flights. The framework is composed of two key components: a disturbance-aware motion planner and a tracking controller. The motion planner consists of a predictive control scheme and an online-adapted learned disturbance model. The tracking controller, developed using contraction control methods, ensures safety bounds on the quadrotor’s behavior near obstacles with respect to the motion plan. The algorithm is tested in simulations with a quadrotor facing strong crosswind and ground-induced disturbances.

I. Introduction

Enhancing the autonomy of unmanned aerial vehicles (UAVs) has made safe autonomous landing in harsh environments a key research challenge. This capability is relevant in various domains and applications, including air mobility, search and rescue, and drone delivery [1], [2], [3]. Developing robust quadrotor landing algorithms is challenging due to disturbances and safety-critical constraints near obstacles. Therefore, planning and control algorithms must account for these disturbances and their effects on UAV performance.