【摘要】：隨著國防工業(yè)的不斷發(fā)展以及科學(xué)技術(shù)水平的不斷提高,大型航天產(chǎn)品的智能裝配技術(shù)得以發(fā)展,移動視覺測量機器人在智能裝配的過程中起到了關(guān)鍵性的作用。視覺測量和路徑規(guī)劃是移動視覺測量機器人任務(wù)規(guī)劃系統(tǒng)中的兩個重要的研究方向。對于待裝配的被測物,我們需要確定機器人的一個最優(yōu)觀測位置,使其在該位置進行測量產(chǎn)生的誤差最小,讓機器人能夠更好地指導(dǎo)裝配過程。本文以單目視覺為測量方案,提出了一種機器人最優(yōu)觀測位置確定的方法。路徑規(guī)劃的目的一般是確定機器人在工作環(huán)境中移動的最短路徑,本文對A*算法進行改進,提出了基于改進A*算法的路徑規(guī)劃方法�；跂鸥穹ǹ紤]環(huán)境中標志點、被測物和觀測位置等因素提出了三種不同的地圖構(gòu)建方法。本文的主要研究工作如下:第一,對移動視覺測量機器人觀測位置規(guī)劃問題進行研究。建立單目視覺測量模型。根據(jù)單目視覺成像原理確定測量方案,以此為基礎(chǔ)根據(jù)幾何關(guān)系和坐標轉(zhuǎn)換分別針對二維和三維情況建立視覺測量模型。針對不同的工況建立最優(yōu)觀測位置測量模型,確定最優(yōu)觀測位置。第二,對移動視覺測量機器人的路徑規(guī)劃方法進行研究。闡述A*算法的實現(xiàn)過程,基于A*算法在柵格地圖中進行路徑規(guī)劃方法的仿真。針對該方法會出現(xiàn)的碰撞問題,改進路徑規(guī)劃方法,使其能夠避免碰撞問題。第三,對移動視覺測量機器人的地圖構(gòu)建方法進行研究。與傳統(tǒng)地圖構(gòu)建不同,本文的地圖構(gòu)建引入標志點和被測物等因素,并以此為基礎(chǔ),分別針對單獨考慮標志點因素,綜合考慮標志點和被測物因素,綜合考慮標志點、被測物和觀測位置因素三種不同的地圖構(gòu)建方法,利用基于改進A*算法的路徑規(guī)劃方法進行仿真,并分析對比不同方法的優(yōu)缺點和應(yīng)用場合。第四,通過實驗驗證最優(yōu)觀察位置測量模型和路徑規(guī)劃方法,并且同時驗證移動測量的優(yōu)勢。針對不同的實驗搭建相應(yīng)的實驗環(huán)境,通過機器人自身CCD相機和天頂全局相機對被測物進行測量,分析測量誤差,根據(jù)得出的結(jié)論去驗證本文提出的理論。
[Abstract]:With the development of national defense industry and the improvement of science and technology, intelligent assembly technology of large-scale aerospace products has been developed. Mobile vision measuring robot plays a key role in the intelligent assembly process. Vision measurement and path planning are two important research directions in task planning system of mobile vision measuring robot. For the object to be assembled, we need to determine an optimal observation position of the robot, so that the measurement error of the robot can be minimized, so that the robot can better guide the assembly process. In this paper, a method for determining the optimal observation position of a robot is proposed, which is based on monocular vision. The purpose of path planning is to determine the shortest path of robot moving in the working environment. In this paper, we improve the A * algorithm and propose a path planning method based on the improved A * algorithm. Based on the grid method, three different map construction methods are proposed based on the factors such as the mark points, the measured objects and the observation position in the environment. The main work of this paper is as follows: first, the position planning of mobile vision measuring robot is studied. A monocular visual measurement model was established. According to the principle of monocular vision imaging, the measurement scheme is determined, and based on the geometric relationship and coordinate transformation, the visual measurement models are established for two-dimensional and three-dimensional situations, respectively. An optimal observation position measurement model is established for different working conditions to determine the optimal observation position. Secondly, the path planning method of mobile vision measuring robot is studied. The realization process of A * algorithm is described, and the simulation of path planning method based on A * algorithm in grid map is carried out. To solve the collision problem, the path planning method is improved to avoid the collision problem. Thirdly, the map construction method of mobile vision measuring robot is studied. Different from the traditional map construction, the map construction in this paper introduces the factors such as mark point and object under test, and on this basis, separately considering the factor of mark point, synthetically considering the factor of mark point and object being tested, synthetically considering the mark point, considering separately the factor of mark point, synthetically considering the factor of mark point. There are three different map construction methods: the measured object and the observed position factor. The path planning method based on the improved A * algorithm is used to simulate, and the advantages and disadvantages of the different methods are analyzed and compared. Fourth, the optimal observation position measurement model and path planning method are verified by experiments, and the advantages of mobile measurement are also verified. According to the different experimental environment, the measured object is measured by the robot's own CCD camera and the zenith global camera, the measurement error is analyzed, and the theory proposed in this paper is verified according to the conclusion.
【學(xué)位授予單位】：哈爾濱工業(yè)大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2017
【分類號】：TP391.41;TP242

【參考文獻】

相關(guān)期刊論文 前10條

1 劉建華;楊建國;劉華平;耿鵬;高蒙;;基于勢場蟻群算法的移動機器人全局路徑規(guī)劃方法[J];農(nóng)業(yè)機械學(xué)報;2015年09期

2 袁立;;現(xiàn)代飛機數(shù)字化柔性裝配生產(chǎn)線[J];航空科學(xué)技術(shù);2011年05期

3 趙娟平;高憲文;劉金剛;符秀輝;;移動機器人路徑規(guī)劃的參數(shù)模糊自適應(yīng)窗口蟻群優(yōu)化算法[J];控制與決策;2011年07期

4 張辰貝西;黃志球;;自動導(dǎo)航車(AGV)發(fā)展綜述[J];中國制造業(yè)信息化;2010年01期

5 齊勇;魏志強;殷波;費云瑞;于忠達;莊曉東;;增強蟻群算法的機器人最優(yōu)路徑規(guī)劃[J];哈爾濱工業(yè)大學(xué)學(xué)報;2009年03期

6 李德仁;;移動測量技術(shù)及其應(yīng)用[J];地理空間信息;2006年04期

7 朱慶保;;復(fù)雜環(huán)境下的機器人路徑規(guī)劃螞蟻算法[J];自動化學(xué)報;2006年04期

8 郭英;盧秀山;馮尊德;;車載測量系統(tǒng)中兩傳感器相對位置關(guān)系的確定[J];工程勘察;2005年02期

9 楊淑瑩,董潔,曹作良;基于視覺導(dǎo)航系統(tǒng)的標志識別方法[J];天津理工學(xué)院學(xué)報;2002年03期

10 何大闊,王福利;基于可進化性的快速遺傳算法[J];東北大學(xué)學(xué)報;2002年07期

相關(guān)碩士學(xué)位論文 前2條

1 高超;導(dǎo)彈柔性裝配系統(tǒng)關(guān)鍵技術(shù)研究[D];哈爾濱工業(yè)大學(xué);2014年

2 吳強;車載導(dǎo)航數(shù)據(jù)獲取與處理技術(shù)研究[D];西安電子科技大學(xué);2009年

，

本文編號：2404722