基于多智能體MAS的智能交通控制系統(tǒng)的研究
本文關(guān)鍵詞: 多智能體 智能交通控制 智能體通訊 模型 仿真 出處:《長沙理工大學(xué)》2007年碩士論文 論文類型:學(xué)位論文
【摘要】: 交通問題是全球普遍關(guān)注的重要問題,傳統(tǒng)的基于精確數(shù)學(xué)模型的交通控制方法不僅復(fù)雜而且在實際應(yīng)用中往往達(dá)不到預(yù)期的控制效果。而多智能體(Multi-Agent)技術(shù)是人工智能領(lǐng)域中的一個研究熱點,單個Agent能夠自主地、主動地對對象進(jìn)行控制,并且Agent之間可以進(jìn)行協(xié)作從而使整個系統(tǒng)達(dá)到更大的智能性。 本文在論述Agent理論和多智能主體系統(tǒng)(Multi-Agent System,MAS)技術(shù)的基礎(chǔ)上,針對目前智能交通控制系統(tǒng)的現(xiàn)狀及交通系統(tǒng)存在的非線性、不確定性、時變性和不完全性等問題,將Agent技術(shù)應(yīng)用于交通控制系統(tǒng),提出了一種基于多智能體的智能交通控制系統(tǒng)結(jié)構(gòu)。該系統(tǒng)在每個交叉口均設(shè)置一個交叉口Agent,,由中央Agent監(jiān)督與管理,并探討了各個交通元素Agent的功能以及它們的協(xié)調(diào)合作關(guān)系?刂葡到y(tǒng)采用部分全局的方法規(guī)劃整個控制過程。交叉口Agent內(nèi)部控制策略采用廣義的知識模型實現(xiàn),采用預(yù)測、協(xié)調(diào)、感應(yīng)調(diào)整三段式的控制流程,用動態(tài)規(guī)劃的學(xué)習(xí)方法制定初始控制方案,采用博弈論實現(xiàn)與相鄰交叉口Agent及與中央Agent的協(xié)調(diào),采用點對點直接通訊方式完成與相鄰交叉口Agent及與中央Agent的通訊。同時運用Delph7.0對基于多智能主體的智能交通控制系統(tǒng)進(jìn)行了初步仿真。仿真結(jié)果表明:基于多智能主體的智能交通控制系統(tǒng)擁有一種全新的智能交通控制結(jié)構(gòu),它能夠適應(yīng)交通系統(tǒng)的復(fù)雜性和隨機(jī)性,較好地克服傳統(tǒng)交通控制系統(tǒng)所存在的缺陷。
[Abstract]:The traffic problem is an important problem all over the world, the traditional traffic control methods based on precise mathematics models is not only complex but also in practical applications are often not up to the expected control effect. The multi-agent system (Multi-Agent) technology is a hot research topic in the field of artificial intelligence, a single Agent can independently control active objects, and can carry out collaboration to make the whole system to achieve greater intelligence Agent.
This paper discusses the Agent theory and multi agent system (Multi-Agent, System, MAS) on the basis of technology, aiming at the nonlinear, existing traffic system and intelligent traffic control system uncertainty, time-varying and uncompleteness, applies Agent technology to traffic control system, put forward a kind of intelligent traffic control the system structure based on Multi-Agent system. The system sets an intersection Agent at each intersection, by the central Agent supervision and management, all traffic elements Agent functions and their coordination and cooperation is discussed. Method of control system using partial global planning to control the whole process. The intersection of the Agent internal control strategy is adopted the realization of the generalized knowledge model, prediction, coordination, induction adjustment control three stage process, making the initial control scheme by learning dynamic programming method, using game On the realization and adjacent intersection of Agent and coordination with the central Agent, using point-to-point direct communication mode and adjacent intersection with central Agent and Agent communication. At the same time Delph7.0 a preliminary simulation of the intelligent traffic control system based on multi-agent system application. The simulation results show that the intelligent traffic control system of multi agent have a new intelligent traffic control based on the structure, it can adapt to the complexity of traffic system and randomness, overcome the drawbacks of traditional traffic control system.
【學(xué)位授予單位】:長沙理工大學(xué)
【學(xué)位級別】:碩士
【學(xué)位授予年份】:2007
【分類號】:U495
【參考文獻(xiàn)】
相關(guān)期刊論文 前8條
1 歐海濤,張衛(wèi)東,張文淵,許曉鳴;基于多智能體技術(shù)的城市智能交通控制系統(tǒng)[J];電子學(xué)報;2000年12期
2 郭紅霞,吳捷,張端金,王春茹;多Agent技術(shù)的研究進(jìn)展[J];河南科學(xué);2004年02期
3 王俊松,崔世鋼;Multi-Agent技術(shù)及應(yīng)用[J];計算機(jī)工程與應(yīng)用;2003年18期
4 馮大可,王云莉,段廣洪;實現(xiàn)多智能體系統(tǒng)的關(guān)鍵技術(shù)與應(yīng)用[J];計算機(jī)集成制造系統(tǒng)-CIMS;2001年07期
5 宋厚冰,蔡遠(yuǎn)利;交通系統(tǒng)計算機(jī)仿真研究綜述[J];計算機(jī)仿真;2004年03期
6 劉志剛,潘和平,李德仁;智能主體主要理論的綜述與分析[J];計算機(jī)應(yīng)用研究;2002年07期
7 陸化普,朱茵;21世紀(jì)的智能交通系統(tǒng)——現(xiàn)代科技的融合[J];交通運輸工程與信息學(xué)報;2003年02期
8 鹿玲杰;申徐洲;;AGENT技術(shù)在智能交通控制中的應(yīng)用[J];交通與計算機(jī);2006年01期
本文編號:1522666
本文鏈接:http://sikaile.net/kejilunwen/rengongzhinen/1522666.html