基于ROS的移动机器人建图导航技术研究

发布时间：2018-03-24 12:40

  本文选题：地图创建　切入点：ROS　出处：《哈尔滨工业大学》2017年硕士论文

【摘要】：近年来,移动机器人在各方面的应用日益广泛,与之相关的技术已在国内外机器人领域掀起研究热潮。对环境的认识和定位从而实现自主导航是移动机器人智能化的重要标志和特征。未知环境下的即时定位与地图构建(SLAM)一直是移动机器人技术领域的研究热点。本项目面向室内服务机器人系统,开展基于ROS的移动机器人建图导航相关技术研究,具体包括基于ROS的移动机器人操作系统开发、室内环境自主探索建图、以及局部路径规划等方面,具体研究内容如下。首先,设计了移动机器人的硬件结构,完成机器人的传感器等硬件系统搭建。在软件控制系统方面,基于模块化、分布式计算的设计思想,开发了基于ROS的移动机器人系统平台,对移动机器人的传感器信息采集、移动控制、建图导航及环境观测等节点进行了设计,最终搭建了一套完整的建图导航机器人系统。该机器人能够进行实时的地图创建与导航,同时具备较好的人机交互能力。其次,针对目前SLAM系统中状态估计无迹卡尔曼滤波算法(UKF)计算量大的问题,提出了基于局部采样的UKF状态估计算法。根据算法中状态向量的特性,在UKF算法的UT采样过程中采用仅对部分与当前状态估计相关数据进行采样的策略,降低了状态估计的计算复杂度。对新的采样算法进行公式推导分析,并通过仿真分析,验证了提出的算法在保证滤波定位精度的同时降低计算量,提高机器人建图的实时性。提出了基于全局路径的改进动态窗口(DWA)局部路径规划算法。首先通过对传统的DWA算法进行公式分析及仿真实验研究,针对其评价函数中存在非必需项和特殊环境导航效果差的问题,提出了基于全局路径规划的DWA算法,并通过MATLAB实验仿真分析证明改进方法的有效性与鲁棒性。最后,利用搭建的移动机器人实验平台对机器人的自主探索SLAM进行了实验,验证了基于局部采样的UKF定位算法和改进的DWA局部路径规划的算法的可行性。
[Abstract]:In recent years, mobile robots have been widely used in various fields. The related technologies have aroused a research boom in the field of robotics at home and abroad. Understanding and locating the environment to realize autonomous navigation is an important sign and feature of the intelligent mobile robot. Map building (slam) has always been a hot topic in the field of mobile robot technology. Research on mobile robot mapping and navigation technology based on ROS is carried out, including the development of mobile robot operating system based on ROS, autonomous exploration and mapping of indoor environment, and local path planning and so on. The specific research contents are as follows. The hardware structure of the mobile robot is designed, and the hardware system such as the sensor of the robot is built. In the aspect of software control system, the platform of mobile robot system based on ROS is developed based on the design idea of modularization and distributed computing. The sensor information collection, mobile control, map building and navigation and environment observation of mobile robot are designed. Finally, a complete mapping navigation robot system is built. The robot can create and navigate maps in real time. At the same time, it has good human-computer interaction ability. Secondly, aiming at the problem that the state estimation unscented Kalman filter algorithm in SLAM system is computationally large, A state estimation algorithm for UKF based on local sampling is proposed. According to the characteristics of the state vector in the algorithm, only part of the data related to the current state estimation is sampled in the UT sampling process of the UKF algorithm. The computational complexity of state estimation is reduced. The formula derivation and analysis of the new sampling algorithm are carried out, and the simulation results show that the proposed algorithm not only ensures the accuracy of filtering location, but also reduces the computational complexity. An improved dynamic window DWA-based local path planning algorithm based on global path is proposed to improve the real-time performance of robot mapping. Firstly, the traditional DWA algorithm is studied by formula analysis and simulation experiments. In order to solve the problem of non-essential item and poor navigation effect in special environment, the DWA algorithm based on global path planning is proposed, and the effectiveness and robustness of the improved method are proved by MATLAB simulation. Based on the experimental platform of mobile robot, the autonomous exploration SLAM of the robot is tested, and the feasibility of the UKF location algorithm based on local sampling and the improved DWA local path planning algorithm is verified.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP242
，

本文编号：1658333