A Multi-Sensorial Simultaneous Localization and Mapping (SLAM) System for Low-Cost Micro Aerial Vehicles in GPS-Denied Environments

A Multi-Sensorial Simultaneous Localization and Mapping (SLAM) System for Low-Cost Micro Aerial Vehicles in GPS-Denied Environments

一种在无GPS环境中设计的面向低价微型飞行器的多传感器同步定位成图系统

学术编辑：Gonzalo Pajares Martinsanz
收到：2017年1月25日；接受：2017年4月5日；发布时间：4月8日201

Abstract: One of the main challenges of aerial robots navigation in indoor or GPS-denied environments is position estimation using only the available onboard sensors. This paper presents a Simultaneous Localization and Mapping (SLAM) system that remotely calculates the pose and environment map of different low-cost commercial aerial platforms, whose onboard computing capacity is usually limited. The proposed system adapts to the sensory configuration of the aerial robot, by integrating different state-of-the art SLAM methods based on vision, laser and/or inertial measurements using an Extended Kalman Filter (EKF). To do this, a minimum onboard sensory configuration is supposed, consisting of a monocular camera, an Inertial Measurement Unit (IMU) and an altimeter. It allows to improve the results of well-known monocular visual SLAM methods (LSD-SLAM and ORB-SLAM are tested and compared in this work) by solving scale ambiguity and providing additional information to the EKF.When payload and computational capabilities permit, a 2D laser sensor can be easily incorporated to the SLAM system, obtaining a local 2.5D map and a footprint estimation of the robot position that improves the 6D pose estimation through the EKF. We present some experimental results with two different commercial platforms, and validate the system by applying it to their position control.

简介：空中机器人在无GPS信号的环境中的一个主要挑战是只使用可用的机载传感器的位置估计。本文提出了一种同时建图和定位（SLAM）系统，它可以远程计算不同低价格商用航空平台的位姿和环境地图，远程计算的原因是机载计算能力通常是受限的。该系统适合于空中机器人的传感器配置，通过融合不同的先进的SLAM方法，包括视觉SLAM，激光雷达SLAM和/或者使用EKF的惯性测量。要做到这一点，一种最小的机载传感配置是可以做到的，包括一个单目相机、一个惯性测量单位（IMU）和一个高度计。它可以提高著名的单目视觉SLAM方法的结果（本文测试并对比了LSD-SLAM和ORB-SLAM），因为它解决了尺度不确定性并且提供了额外的EKF信息。当装载能力和计算能力允许的情况下，2D的雷达传感器可以轻松地融入进SLAM系统，从而获得了2.5维地图和机器人位置的指纹估计，然后通过EKF可以提高6D的位姿估计。我们展示了两种不同的商业平台的一些实验结果，并且通过添加进它们的位置控制验证了该系统。

Keywords: aerial robots; SLAM; sensor fusion

关键字：航空机器人；SLAM；传感器融合

1. Introduction

1. 简介

Research on autonomous aerial robots has advanced considerably in the last decades, especially in outdoor applications. Enabled by MEMS inertial sensors and GPS, Unmanned Aerial Vehicles (UAVs) that show an awesome set of flying capabilities in outdoor environments have been developed, ranging from typical flight manoeuvres [1], to collaborative construction tasks [2] or swarm coordination [3], among many other applications.

自主飞行机器人的研究在过去的几十年里已经相当先进，特别是在室外应用。由MEMS惯性传感器和GPS推动，展示了精彩的户外飞行能力的无人机（UAV）已经开发出来了，从典型的飞行演习到协同建设任务，或者群协调，以及其他的很多应用。

Although technological progress has made possible the development of small Micro Aerial Vehicles (MAVs) capable of operating in confined spaces, indoor navigation is still an important challenge for a number of reasons: (i) most indoor environments remain without access to external positioning systems such as GPS; (ii) the onboard computational power is restricted and (iii) the low payload capacity limits the type and number of sensors that can be used. However, there is a growing interest in indoor applications such as surveillance, disaster relief or rescue in GPS-denied environments (such as demolished or semi-collapsed buildings). In these stages is often preferable to use low-cost MAVs that can be easily replaced in case of breakage, damage or total loss.

虽然科技进步使得开发小型的微飞行器(MAV)成为可能，它可以在有限的空间运行，但是室内导航仍然是一个重要的挑战，主要是以下几方面原因：(i)大多数室内环境仍然无法使用外部定位系统，如GPS；(ii)机载能力受限制；(iii)低的荷载能力限制了可以使用的传感器的种类和数量。然而，人们对室内应用的需求越来越强烈，比如监视、赈灾或者营救无GPS的环境（例如被摧毁的或者半倒塌的建筑物）。在这些阶段一般通向于使用低价格的飞行器，因为一旦破坏、损毁或者完全丧失的话可以轻松替代。

From the navigation point of view, state estimation of the six degrees of freedom (6-DoF) of the MAV (attitude and position) is the main challenge that must be tackled to achieve autonomy. The inaccuracy and high drift of MEMS inertial sensors, the limited payload for computation and sensing, and the unstable and fast dynamics of air vehicles are the major difficulties for position estimation. So far, the most robust solutions are based on external sensors, as in [4], where an external trajectometry system directly yields the position and orientation of the robot, or in [5], where an external CCD camera provides the measurements. However, these solutions require a previous preparation of the environment and are not applicable to unknown spaces, in which the MAV must rely on its own onboard sensors to navigate

从导航的角度来看，飞行器的六个自由角度的状态估计(位置和姿态)是解决自治的主要挑战。