本文關鍵詞： 拆除機器人 SLAM 多傳感器信息融合定位 機器人操作系ROS　出處：《安徽工業(yè)大學》2017年碩士論文　論文類型：學位論文

【摘要】：隨著城市的改造及工業(yè)的發(fā)展,拆除機器人的使用日益廣泛。目前拆除機器人不具備自主移動能力,以人工遙控操作方式為主,一定程度上限制了作業(yè)效率與精度。拆除機器人的自主移動,需要借助傳感器信息同時進行空間自定位與環(huán)境感知,這個過程被稱為即時定位與地圖構建(SLAM)。該技術涉及傳感器信息處理以及多個數(shù)學模型,是拆除機器人自主移動的前提。本文首先從SLAM的一般性問題出發(fā),闡述SLAM中定位與建圖的關系;建立SLAM的概率學模型,以擴展卡爾曼濾波和粒子濾波算法為基礎,對SLAM算法的具體實現(xiàn)進行探討;在MATLAB中對兩種SLAM算法進行仿真對比實驗,以性能較好的算法作為拆除機器即時定位與地圖構建的理論基礎。其次,針對拆除機器人SLAM過程中的定位問題展開分具體析:探討拆除機器人定位過程中存在的問題與難點,對拆除機器人SLAM過程中涉及的模型及兩種定位方式進行了分析與建模,并在MATLAB中針對拆除機器人掃描匹配定位進行了算法驗證;在此基礎上,基于多傳感器信息融合技術,融合里程計定位與掃描匹配定位信息,解決履帶滑移帶來的定位問題;以上述理論為基礎,通過融合定位算法對Fast SLAM算法進行改進,作為拆除機器人即時定位與地圖構建算法。最后,以上述改進的SLAM算法為基礎,基于開源機器人操作系統(tǒng)ROS的開源包進行編碼改進,在機器人仿真器Gazebo中建立仿真環(huán)境,對比里程計模型定位與融合算法定位的建圖效果。在此基礎上,基于拆除機器人進行硬件平臺和軟件平臺的搭建,在實驗室對改進后的算法進行實驗,構建了精度較高的實驗室地圖,完成了對拆除機器人即時定位與地圖構建算法研究。
[Abstract]:With the improvement of city and the development of industry, the demolition robot is used more and more widely. At present, the demolition robot does not have the ability to move independently, and it is mainly operated by manual remote control. To a certain extent, the efficiency and precision of the operation are limited. In order to move the robot independently, it is necessary to use the sensor information to simultaneously carry out the space self-localization and the environment perception. This process is called instant location and map building. The technology involves sensor information processing and multiple mathematical models, which is the premise of removing robot's autonomous movement. This paper starts with the general problem of SLAM. This paper expounds the relationship between location and map building in SLAM, establishes the probabilistic model of SLAM, discusses the realization of SLAM algorithm based on extended Kalman filter and particle filter algorithm, and makes a simulation and comparison experiment on two SLAM algorithms in MATLAB. The algorithm with good performance is used as the theoretical basis of the real time location and map construction of the demolition machine. Secondly, the localization problems in the SLAM process of the demolition robot are analyzed in detail: the problems and difficulties in the localization process of the demolition robot are discussed. This paper analyzes and models the models and two kinds of localization methods involved in the process of removing robot SLAM, and verifies the algorithm of scanning matching localization in MATLAB, and based on this, based on multi-sensor information fusion technology, The location problem caused by crawler slip is solved by combining the location information of mileometer and scanning, and based on the above theory, the Fast SLAM algorithm is improved by the fusion localization algorithm. Finally, based on the improved SLAM algorithm, the coding of the open source package based on the open source robot operating system ROS is improved, and the simulation environment is established in the robot simulator Gazebo. On the basis of this, the hardware platform and software platform are built based on the demolition robot, and the improved algorithm is experimented in the laboratory. The laboratory map with high precision is constructed, and the algorithms of real time localization and map construction of the dismantled robot are studied.
【學位授予單位】：安徽工業(yè)大學
【學位級別】：碩士
【學位授予年份】：2017
【分類號】：TP242

【參考文獻】

相關期刊論文 前8條

1 芮強;王紅巖;王欽龍;萬麗;蓋江濤;周廣明;;履帶車輛轉向性能參數(shù)分析與試驗研究[J];機械工程學報;2015年12期

2 王法勝;魯明羽;趙清杰;袁澤劍;;粒子濾波算法[J];計算機學報;2014年08期

3 朱磊;樊繼壯;趙杰;吳曉光;;未知環(huán)境下的移動機器人SLAM方法[J];華中科技大學學報(自然科學版);2011年07期

4 李世飛;王平;沈振康;;迭代最近點算法研究進展[J];信號處理;2009年10期

5 石杏喜;趙春霞;郭劍輝;;基于PF/CUKF/EKF的移動機器人SLAM框架算法[J];電子學報;2009年08期

6 賈銀亮;張煥春;經(jīng)亞枝;;Bresenham直線生成算法的改進[J];中國圖象圖形學報;2008年01期

7 鞠純純;何波;劉保龍;王永清;;基于粒子濾波器的SLAM的仿真研究[J];系統(tǒng)仿真學報;2007年16期

8 羅榮華,洪炳昒;移動機器人同時定位與地圖創(chuàng)建研究進展[J];機器人;2004年02期

相關碩士學位論文 前2條

1 王帆;基于卡爾曼濾波和粒子濾波的移動機器人同時定位與地圖創(chuàng)建研究[D];西安工程大學;2012年

2 王培勛;基于子地圖連接的機器人同時定位與地圖構建研究[D];中國海洋大學;2010年

，

本文編號：1501758