Distance-based Autonomous Docking for UAV-MP in GPS-denied Environment

In this paper, we study the docking problem between a UAV and a mobile platform (MP) by estimating the relative position through distance information and its own displacement, and design the approaching trajectory for eliminating the need of any global positioning systems or other infrastructures. The main challenges of this problem are to estimate the relative position between the UAV and the MP and design controllers for the navigation trajectory control. Since the MP is not equipped with odometer and its motion changes from time to time, the UAV is iteratively navigated using the estimated relative positions to make it infinitely close to the MP, so that the MP finally enters the field of view of the camera and accomplishes the autonomous docking. A simulation experiment is carried out and the theoretical results are verified.

1 Introduction

In the past few years, multi-rotor unmanned air vehicle (UAV) technology has been developed to a large extent and has led to a large number of applications, in areas such as agriculture, transport and exploration [1]–[4]. How to navigate a UAV to land on a MP is always a popular research topic. The first step in completing the landing task is, of course, navigation, i.e., controlling the UAV to fly over the MP (the MP comes within the UAV’s field-of-view (FOV) of camera). Depending on the positioning methods, there are two main types of autonomous docking: absolute positioning (GPS, Beidou, MCS) and self-search navigation (V-SLAM) [5]–[8]. The former is based on the signal strength of the base setup which determines its accuracy in absolute coordinate systems and will lead to deviations or failures in navigation when the signal is limited. On the other hand, the latter requires a large amount of search until the MP appears within the FOV of the UAV, often requiring significant computing power and time. Therefore, it is necessary to propose autonomous navigation algorithms for UAV the navigation algorithm is only for UAV, which do not depend on external infrastructures, and do not need a mass of computing resources and too much search time. In turn, the method of using relative positioning in navigation has become a research direction.