【摘要】：近年來,移動機(jī)器人技術(shù)水平有著飛躍式的發(fā)展,不僅在生產(chǎn)生活中被廣泛應(yīng)用,而且對許多國家工業(yè)、國防以及國民經(jīng)濟(jì)產(chǎn)生了巨大的影響。它在實(shí)際應(yīng)用中可代替人工在危險(xiǎn)(有毒、輻射等)、復(fù)雜環(huán)境中作業(yè),可從事在人所不及(太空、深海等)的空間環(huán)境中。其中兩輪自平衡機(jī)器人是移動機(jī)器人中應(yīng)用較為廣泛的一種,它是集環(huán)境感知、動態(tài)決策與規(guī)劃等多功能于一體的智能化機(jī)器系統(tǒng),它融合了多級倒立擺的不穩(wěn)定性、多變量、非線性和強(qiáng)耦合等特點(diǎn)。由于兩輪機(jī)器人可以實(shí)現(xiàn)多輪機(jī)器人無法實(shí)現(xiàn)的復(fù)雜動作,因此對它們的深入研究具有重要的理論及現(xiàn)實(shí)意義。本文針對自平衡機(jī)器人在實(shí)際應(yīng)用中存在穩(wěn)定性不佳、人工經(jīng)驗(yàn)選擇控制器權(quán)值結(jié)果較差等問題,展開了系統(tǒng)研究,在基于線性二次性最優(yōu)控制器和自平衡機(jī)器人模型的基礎(chǔ)上,提出了一個(gè)基于蟻群算法優(yōu)化機(jī)器人穩(wěn)定性的模型。之后,在保證了機(jī)器人安全性、平穩(wěn)性的基礎(chǔ)上,針對機(jī)器人路徑規(guī)劃中前期環(huán)境地圖信息采用劃分方式不理想和后期單一算法優(yōu)化路徑不佳等問題,提出一種改進(jìn)蜂巢柵格法來優(yōu)化前期環(huán)境信息并采用一種基于遺傳-蟻群的混合算法來解決后期最優(yōu)路徑的問題,使系統(tǒng)平穩(wěn)且無碰撞的完成路徑尋優(yōu)。本文研究工作主要包括:(1)本文提出一個(gè)基于蟻群算法優(yōu)化機(jī)器人穩(wěn)定性的模型,用以選擇適合控制器的參數(shù)權(quán)值,克服了經(jīng)驗(yàn)選擇參數(shù)的不確定性和耗時(shí)性。此模型對最優(yōu)控制器的性能指標(biāo)函數(shù)對于狀態(tài)量的權(quán)陣參數(shù)進(jìn)行優(yōu)化,通過最優(yōu)的權(quán)陣結(jié)果得到控制器解值,最終決定移動機(jī)器人的穩(wěn)定指標(biāo)結(jié)果。該優(yōu)化方式更優(yōu)于傳統(tǒng)人工選擇參數(shù)的方法,不僅優(yōu)化了機(jī)器人的穩(wěn)定性,還降低了實(shí)踐中相關(guān)系統(tǒng)應(yīng)用的危險(xiǎn)系數(shù)。(2)本文提出一種改進(jìn)蜂巢柵格法來處理路徑優(yōu)化前期環(huán)境信息,它結(jié)合了自然生物現(xiàn)象和幾何學(xué)理論,克服了傳統(tǒng)柵格法在處理地圖信息時(shí)占用過多可行空間、增加機(jī)器人碰撞率的缺點(diǎn),并且改善了蜂巢柵格法編碼時(shí)不連續(xù)、不能根據(jù)起始點(diǎn)和終點(diǎn)的位置控制編碼規(guī)則的問題。該方法采用自適應(yīng)編碼方式,靈活操作地圖信息的設(shè)定,使柵格具有連續(xù)性,更利于最優(yōu)路徑的搜索,有效的減少機(jī)器人行進(jìn)的路徑長度和路徑搜索時(shí)間,提高機(jī)器人的安全性。(3)本文在處理路徑規(guī)劃后期搜索最優(yōu)路徑的問題中,選定蟻群和遺傳算法作為混合方法的研究基礎(chǔ),通過實(shí)驗(yàn)對比分析,結(jié)合二者的優(yōu)缺點(diǎn),采用了一種基于遺傳-蟻群的混合算法對路徑進(jìn)行尋優(yōu),克服了傳統(tǒng)路徑規(guī)劃采用單一優(yōu)化算法搜索路徑時(shí)穩(wěn)態(tài)迭代次數(shù)和穩(wěn)態(tài)時(shí)路徑長度不理想的問題,并通過設(shè)置30次、80次和150次的迭代次數(shù),可以得到多混合算法均優(yōu)于單一算法的處理結(jié)果。
[Abstract]:In recent years, the level of mobile robot technology has been developed by leaps and bounds, which is not only widely used in production and life, but also has a great impact on industry, national defense and national economy in many countries. It can be used instead of artificial in dangerous (toxic, radiation, etc.), complex environment, can be engaged in the space environment (space, deep sea, etc.). Two-wheeled self-balancing robot is widely used in mobile robots. It is an intelligent machine system which integrates environmental perception, dynamic decision-making and planning. It combines the instability and multivariable of multi-stage inverted pendulum. Nonlinear and strong coupling and other characteristics. Because the two-wheeled robot can realize the complex action that the multi-wheeled robot can not realize, it is of great theoretical and practical significance to study them deeply. In this paper, a systematic study is carried out to solve the problems of poor stability and poor weight value of artificial experience selection controller in the practical application of self-balancing robot. Based on the linear quadratic optimal controller and the self-balancing robot model, a model based on ant colony algorithm is proposed to optimize the stability of the robot. After that, on the basis of ensuring the safety and stability of the robot, the environmental map information in the early stage of robot path planning is not well divided and the path optimization by a single algorithm in the later stage is not good. An improved beehive grid method is proposed to optimize the environmental information in the early stage and a hybrid algorithm based on genetic ant colony is used to solve the problem of optimal path in the later stage, which makes the system stable and collision-free. The main research work of this paper is as follows: (1) in this paper, a model based on ant colony algorithm to optimize the stability of robot is proposed, which can select the parameter weights suitable for the controller and overcome the uncertainty and consumption of empirical selection parameters. In this model, the performance index function of the optimal controller is optimized for the weight matrix parameters of the state quantity, and the solution value of the controller is obtained by the optimal weight matrix result, and finally the stability index result of the mobile robot is determined. This optimization method is better than the traditional manual parameter selection method, and not only optimizes the stability of the robot. It also reduces the risk coefficient of the application of related systems in practice. (2) in this paper, an improved hive grid method is proposed to deal with the environmental information in the early stage of path optimization, which combines natural biological phenomena and geometry theory. The shortcomings of the traditional grid method in dealing with map information are overcome, such as taking up too much feasible space and increasing the collision rate of the robot, and the problem of discontinuity in the coding of the hive grid method is improved, and the coding rules can not be controlled according to the position of the starting point and the end point. In this method, the adaptive coding method is adopted to flexibly operate the setting of map information, so that the grid has continuity, which is more conducive to the search of the optimal path, and effectively reduces the path length and path search time of the robot. Improve the safety of the robot. (3) in dealing with the problem of searching the optimal path in the later stage of path planning, ant colony and genetic algorithm are selected as the research basis of the hybrid method, and the advantages and disadvantages of the two methods are compared and analyzed by experiments. A hybrid algorithm based on genetic ant colony is used to optimize the path, which overcomes the problem that the number of steady-state iterations and the path length in steady state are not ideal when the traditional path planning uses a single optimization algorithm to search the path, and by setting 30 times, With 80 iterations and 150 iterations, the processing results of multi-hybrid algorithm are better than those of single algorithm.
【學(xué)位授予單位】：昆明理工大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2017
【分類號】：TP242

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 章俠;王祥榮;;人工智能背景下的機(jī)器人[J];科技展望;2016年33期

2 曾辰;許瑛;;一種蜂巢柵格下機(jī)器人路徑規(guī)劃的蟻群算法[J];機(jī)械科學(xué)與技術(shù);2016年08期

3 李波濤;;人工智能技術(shù)在試題庫建設(shè)中的應(yīng)用[J];軟件工程師;2015年06期

4 符川;屈鐵軍;孫世國;;主動調(diào)頻液柱阻尼器基于遺傳算法的LQR控制優(yōu)化設(shè)計(jì)[J];振動與沖擊;2015年02期

5 史恩秀;陳敏敏;李俊;黃玉美;;基于蟻群算法的移動機(jī)器人全局路徑規(guī)劃方法研究[J];農(nóng)業(yè)機(jī)械學(xué)報(bào);2014年06期

6 沈鳳梅;吳隆;;基于改進(jìn)人工勢場法的移動機(jī)器人自主導(dǎo)航和避障研究[J];制造業(yè)自動化;2013年23期

7 陳超;唐堅(jiān);;基于可視圖法的水面無人艇路徑規(guī)劃設(shè)計(jì)[J];中國造船;2013年01期

8 武俊峰;張繼段;;兩輪自平衡機(jī)器人的LQR改進(jìn)控制[J];哈爾濱理工大學(xué)學(xué)報(bào);2012年06期

9 徐德明;;改進(jìn)的遺傳混合蟻群算法在TSP問題中的應(yīng)用[J];計(jì)算機(jī)時(shí)代;2012年11期

10 劉浩梅;張昌凡;;基于LQR的環(huán)形單級倒立擺穩(wěn)定控制及實(shí)現(xiàn)[J];中南大學(xué)學(xué)報(bào)(自然科學(xué)版);2012年09期

相關(guān)博士學(xué)位論文 前2條

1 劉傳領(lǐng);基于勢場法和遺傳算法的機(jī)器人路徑規(guī)劃技術(shù)研究[D];南京理工大學(xué);2012年

2 王超學(xué);遺傳算法和蟻群算法及其在TSP問題和配電網(wǎng)重構(gòu)問題中的應(yīng)用研究[D];西安理工大學(xué);2007年

相關(guān)碩士學(xué)位論文 前8條

1 汪敏;基于環(huán)境地圖的多移動機(jī)器人協(xié)同機(jī)制研究[D];江蘇科技大學(xué);2016年

2 謝朦;基于多傳感器信息融合的移動機(jī)器人定位研究[D];太原科技大學(xué);2015年

3 袁杰;基于蟻群遺傳混合智能算法求解TSP問題[D];長春工業(yè)大學(xué);2014年

4 彭麗;基于遺傳算法的移動機(jī)器人路徑規(guī)劃[D];長沙理工大學(xué);2013年

5 丁寅;基于遺傳和蟻群的自適應(yīng)路徑規(guī)劃算法研究[D];華中科技大學(xué);2012年

6 孟憲強(qiáng);基于量子遺傳算法的足球機(jī)器人路徑規(guī)劃研究[D];中國海洋大學(xué);2009年

7 帥知春;混合智能算法在移動機(jī)器人導(dǎo)航中的研究及應(yīng)用[D];廣東工業(yè)大學(xué);2007年

8 宋西蒙;倒立擺系統(tǒng)LQR—模糊控制算法研究[D];西安電子科技大學(xué);2006年

，

本文編號：2497101