2023年全國碩士研究生考試考研英語一試題真題(含答案詳解+作文范文)_第1頁
已閱讀1頁,還剩11頁未讀 繼續(xù)免費閱讀

下載本文檔

版權(quán)說明:本文檔由用戶提供并上傳,收益歸屬內(nèi)容提供方,若內(nèi)容存在侵權(quán),請進(jìn)行舉報或認(rèn)領(lǐng)

文檔簡介

1、<p>  本科生畢業(yè)設(shè)計(論文)</p><p><b>  外文文獻(xiàn)翻譯</b></p><p>  畢業(yè)設(shè)計題目: 交通燈智能控制系統(tǒng)</p><p>  學(xué) 院: 信息科學(xué)與工程學(xué)院 </p><p>  專業(yè)班級: 測控技術(shù)與儀器0703班 </p><p> 

2、 學(xué)生姓名: 王欣 </p><p>  指導(dǎo)教師: 桑海峰 </p><p>  2011年 3月 19日</p><p><b>  外文原文</b></p><p>  Intelligent Traffic Light Control</

3、p><p>  Marco Wiering, Jelle van Veenen, Jilles Vreeken, and Arne Koopman Intelligent Systems Group</p><p>  Institute of Information and Computing Sciences Utrecht University</p><p> 

4、 Padualaan 14, 3508TB Utrecht, The Netherlands</p><p>  email: marco@cs.uu.nl</p><p>  July 9, 2004</p><p><b>  Abstract</b></p><p>  Vehicular travel is in

5、creasing throughout the world, particularly in large urban areas.Therefore the need arises for simulating and optimizing traffic control algorithms to better accommodate this increasing demand. In this paper we study the

6、 simulation and optimization of traffic light controllers in a city and present an adaptive optimization algorithm based on reinforcement learning. We have implemented a traffic light simulator, Green Light District, tha

7、t allows us to experiment with differe</p><p>  Keywords: Intelligent Traffic Light Control, Reinforcement Learning, Multi-Agent Systems (MAS), Smart Infrastructures, Transportation Research</p><p

8、>  1 Introduction</p><p>  Transportation research has the goal to optimize transportation flow of people and goods.As the number of road users constantly increases, and resources provided by current infr

9、astructures are limited, intelligent control of traffic will become a very important issue in the future. However, some limitations to the usage of intelligent traffic control exist. Avoiding traffic jams for example is

10、thought to be beneficial to both environment and economy, but improved traffic-flow may also lead to an</p><p>  There are several models for traffic simulation. In our research we focus on microscopic model

11、s that model the behavior of individual vehicles, and thereby can simulate dynamics of groups of vehicles. Research has shown that such models yield realistic behavior [Nagel and Schreckenberg, 1992, Wahle and Schreckenb

12、erg, 2001].</p><p>  Cars in urban traffic can experience long travel times due to inefficient traffic light control. Optimal control of traffic lights using sophisticated sensors and intelligent optimizatio

13、n algorithms might therefore be very beneficial. Optimization of traffic light switching increases road capacity and traffic flow, and can prevent traffic congestions. Traffic light control is a complex optimization prob

14、lem and several intelligent algorithms, such as fuzzy logic, evolutionary algorithms, and rein</p><p>  In our approach, reinforcement learning [Sutton and Barto, 1998, Kaelbling et al., 1996] with road-user

15、-based value functions [Wiering, 2000] is used to determine optimal decisions for each traffic light. The decision is based on a cumulative vote of all road users standing for a traffic junction, where each car votes usi

16、ng its estimated advantage (or gain) of setting its light to green. The gain-value is the difference between the total time it expects to wait during the rest of its trip if the</p><p>  We compare the perfo

17、rmance of our model-based RL method to that of other controllers using the Green Light District simulator (GLD). GLD is a traffic simulator that allows us to design arbitrary infrastructures and traffic patterns, monitor

18、 traffic flow statistics such as average waiting times, and test different traffic light controllers. The experimental results show that in crowded traffic, the RL controllers outperform all other tested non-adaptive con

19、trollers. We also test the use of the le</p><p>  This paper is organized as follows. Section 2 describes how traffic can be modelled, predicted, and controlled. In section 3 reinforcement learning is expla

20、ined and some of its applications are shown. Section 4 surveys several previous approaches to traffic light control, and introduces our new algorithm. Section 5 describes the simulator we used for our experiments, and in

21、 section 6 our experiments and their results are given. We conclude in section 7.</p><p>  2 Modelling and Controlling Traffic</p><p>  In this section, we focus on the use of information techno

22、logy in transportation. A lot of ground can be gained in this area, and Intelligent Transportation Systems (ITS) gained interest of several governments and commercial companies [Ten-T expert group on ITS, 2002, White Pap

23、er, 2001, EPA98, 1998].</p><p>  ITS research includes in-car safety systems, simulating effects of infrastructural changes, route planning, optimization of transport, and smart infrastructures. Its main goa

24、ls are: improving safety, minimizing travel time, and increasing the capacity of infrastructures. Such improvements are beneficial to health, economy, and the environment, and this shows in the allocated budget for ITS.&

25、lt;/p><p>  In this paper we are mainly interested in the optimization of traffic flow, thus effectively minimizing average traveling (or waiting) times for cars. A common tool for analyzing traffic is the traf

26、fic simulator. In this section we will first describe two techniques commonly used to model traffic. We will then describe how models can be used to obtain real-time traffic information or predict traffic conditions. Aft

27、erwards we describe how information can be communicated as a means of controlling </p><p>  2.1 Modelling Traffic.</p><p>  Traffic dynamics bare resemblance with, for example, the dynamics of f

28、luids and those of sand in a pipe. Different approaches to modelling traffic flow can be used to explain phenomena specific to traffic, like the spontaneous formation of traffic jams. There are two common approaches for

29、modelling traffic; macroscopic and microscopic models.</p><p>  2.1.1 Macroscopic models.</p><p>  Macroscopic traffic models are based on gas-kinetic models and use equations relating traffic d

30、ensity to velocity [Lighthill and Whitham, 1955, Helbing et al., 2002]. These equations can be extended with terms for build-up and relaxation of pressure to account for phenomena like stop-and-go traffic and spontaneous

31、 congestions [Helbing et al., 2002, Jin and Zhang, 2003, Broucke and Varaiya, 1996]. Although macroscopic models can be tuned to simulate certain driver behaviors, they do not offer a d</p><p>  2.1.2 Micros

32、copic models.</p><p>  In contrast to macroscopic models, microscopic traffic models offer a way of simulating various driver behaviors. A microscopic model consists of an infrastructure that is occupied by

33、a set of vehicles. Each vehicle interacts with its environment according to its own rules. Depending on these rules, different kinds of behavior emerge when groups of vehicles interact.</p><p>  Cellular Aut

34、omata. One specific way of designing and simulating (simple) driving rules of cars on an infrastructure, is by using cellular automata (CA). CA use discrete partially connected cells that can be in a specific state. For

35、example, a road-cell can contain a car or is empty. Local transition rules determine the dynamics of the system and even simple rules can lead to chaotic dynamics. Nagel and Schreckenberg (1992) describe a CA model for t

36、raffic simulation. At each discrete time-step, v</p><p>  Cognitive Multi-Agent Systems. A more advanced approach to traffic simulation and optimization is the Cognitive Multi-Agent System approach (CMAS), i

37、n which agents interact and communicate with each other and the infrastructure. A cognitive agent is an entity that autonomously tries to reach some goal state using minimal effort. It receives information from the envir

38、onment using its sensors, believes certain things about its environment, and uses these beliefs and inputs to select an action. Bec</p><p>  Dia (2002) used a CMAS based on a study of real drivers to model t

39、he drivers’ response to travel information. In a survey taken at a congested corridor, factors influencing the choice of route and departure time were studied. The results were used to model a driver population, where dr

40、ivers respond to presented travel information differently. Using this population, the effect of different information systems on the area where the survey was taken could be simulated. The research seems promising,</p

41、><p>  A traffic prediction model that has been applied to a real-life situation, is described in [Wahle and Schreckenberg, 2001]. The model is a multi-agent system (MAS) where driving agents occupy a simulated

42、 infrastructure similar to a real one. Each agent has two layers of control; one for the (simple) driving decision, and one for tactical decisions like route choice. The real world situation was modelled by using detecti

43、on devices already installed. From these devices, information about the numbe</p><p><b>  中文譯文</b></p><p><b>  智能交通燈控制</b></p><p>  馬克 威寧,簡麗 范 威,吉爾 威瑞肯,安瑞 庫普

44、曼</p><p><b>  智能系統(tǒng)小組</b></p><p>  烏得勒支大學(xué)信息與計算科學(xué)研究所</p><p>  荷蘭烏得勒支Padualaan14號</p><p>  郵箱:marco@cs.uu.nl</p><p><b>  2004年7月9日</b>

45、</p><p><b>  摘要</b></p><p>  世界各地的車輛運行逐漸增多,尤其是在一個大的本地區(qū)域。因此就需要有關(guān)交通控制的模擬與優(yōu)化算法,來更好的地適應(yīng)日益增長的需求。在文中,我們學(xué)習(xí)了在城市中的模擬與優(yōu)化的交通燈控制器,以及目前基于強化學(xué)習(xí)的自適應(yīng)優(yōu)化算法。我們已經(jīng)實行了一個交通等模擬器,綠燈區(qū),這允許我們用不同的基礎(chǔ)設(shè)施和不同的交通控制器去實驗

46、。實驗結(jié)果表明,在所有基礎(chǔ)設(shè)施的研究領(lǐng)域內(nèi),我們的自適應(yīng)交通燈控制器優(yōu)于其他固定的控制器。</p><p>  關(guān)鍵字:智能交通燈控制,強化學(xué)習(xí),多代理系統(tǒng)(MAS),智能基礎(chǔ)設(shè)施,運輸研究</p><p><b>  1 介紹</b></p><p>  運輸研究的目的是優(yōu)化人流和物流。隨著道路使用者的數(shù)量不斷上漲,當(dāng)前基礎(chǔ)設(shè)施所提供的資源受

47、到限制,在未來,交通智能控制將會成為一個非常重要的問題。然而,一些交通智能控制使用受限問題的存在。避免交通堵塞,例如,被認(rèn)為是對環(huán)境和經(jīng)濟有益的,但是增加的交通流也可能導(dǎo)致資源需求的增加。[萊文森,2003]。</p><p>  這有幾個交通仿真模型。在我們的研究中,我們專注于那些具有個體車輛行為的微觀模型,從而更好的模擬群體車輛的動力學(xué)。研究表明,這種模型的出現(xiàn)具有現(xiàn)實意義[Nagel and Schreck

48、enberg,1992,Wahle and Schreckenberg,2001]。</p><p>  汽車在城市交通中經(jīng)歷了漫長的運行時間,要歸因于低效的交通燈控制。因此,使用成熟傳感器和智能優(yōu)化算法的交通燈優(yōu)化控制可能是有益的。優(yōu)化的交通燈切換增加了道路的容量和人流,能阻止交通堵塞。交通燈控制是一個復(fù)雜的優(yōu)化問題和幾個智能算法,例如模糊邏輯、遺傳算法和強化學(xué)習(xí)(RL)已被應(yīng)用去試圖解決問題。在本文中,我們描

49、述了一種對交通燈控制,基于模型的、多代理的強化學(xué)習(xí)算法。</p><p>  我們的方法,強化學(xué)習(xí)[Sutton and Barto,1998,Kaelbling,1996]和基于道路使用者的價值功能[威寧,2000]被用來決定每個交通燈的優(yōu)化選擇。這個決定是基于道路使用者站了一個交叉路口的累積投票,在那里每輛汽車使用其估計選票的優(yōu)勢(或增益)設(shè)置它的光的綠色。在其余路程,它的所有等待時間里,如果信號燈現(xiàn)在是紅色

50、的或者綠色的,那么增益的值是不同的。汽車直到到達(dá)目的地后的等待時間,是通過監(jiān)測汽車流過基礎(chǔ)設(shè)施和應(yīng)用強化學(xué)習(xí)(RL)算法而估算出來的。</p><p>  本文寫作安排如下。第二部分描述了交通是如何被建立、預(yù)測和控制的。在第三部分解釋了什么是強化學(xué)習(xí)和一些它的應(yīng)用。第四部分調(diào)查了幾個以前交通控制的方法,介紹了我們的新算法。第五部分描述了我們實驗中所使用的仿真器,以及第六部分給出我們的實驗和實驗結(jié)果。在第七部分我們

51、得出結(jié)論。</p><p><b>  2 建立和控制交通</b></p><p>  在這一部分,我們專注于在交通運輸方面所使用的信息技術(shù)。在這個區(qū)域增加了大量的土地,并且一些政府和商業(yè)公司在交通智能系統(tǒng)(ITS)方面獲得了利潤。[Ten-T expert group on ITS,2002,白皮書,2001,EPA98,1998]。</p><

52、p>  交通智能系統(tǒng)(ITS)研究包括車內(nèi)安全系統(tǒng),基礎(chǔ)設(shè)施改變所引起的仿真效果,路途規(guī)劃,優(yōu)化運輸和智能的基礎(chǔ)設(shè)施。其主要目標(biāo)是:提高安全性、減少運行時間、增加基礎(chǔ)設(shè)施的能力。這種改進(jìn)有益健康、經(jīng)濟、環(huán)境,這表現(xiàn)在交通智能系統(tǒng)的分配預(yù)算方面。</p><p>  在本文中,我們主要對車流的優(yōu)化感興趣,從而有效減少平均運行(或者等待)的車輛次數(shù)。一種常見的分析交通的工具就是交通仿真器。在這部分中,我們將首先

53、描述兩種常用于交通模型的技術(shù)。然后我們將描述模型是如何用來獲取實時交通信息或者預(yù)測交通情況的。后來,我們描述信息是如何作為一種控制交通的手段來進(jìn)行溝通的,在這樣的交通條件下,溝通產(chǎn)生了什么樣的影響。最后,我們描述了所有的汽車都使用計算機進(jìn)行控制的研究。</p><p><b>  2.1 建立交通</b></p><p>  與交通動力學(xué)僅有的相似之處是,例如,流體力

54、學(xué)和管內(nèi)的沙子。建立車流模型的不同方法是用來解釋交通的特殊現(xiàn)象的,就像自發(fā)形成的交通堵塞狀況。有兩種普遍的方法去建立交通:宏觀和微觀模型。</p><p>  2.1.1 宏觀模型</p><p>  宏觀交通模型是基于gas-kinetic模型的,利用了關(guān)于交通密度和速度的方程式[Lighthill and Whitham,1955,Helbing et al.,2002]。這些方程式可

55、以延長積累和放松壓力,歸因于類似的停停走走的交通和自發(fā)的擁堵的現(xiàn)象。[Helbing et al.,2002,Jin and Zhang, 2003,Broucke and Varaiya,1996]。盡管宏觀模型可以來模擬一些特定的可調(diào)驅(qū)動行為,但是他們不能提供一個直接的、靈活的建立和優(yōu)化交通的方法,這使他們不太適合我們的研究。</p><p>  2.1.2 微觀模型</p><p>

56、  與宏觀模型相對比的,微觀交通模型提供了一種仿真各種各樣司機行為的方法。一個微觀模型由一組車輛占據(jù)的基礎(chǔ)設(shè)施組成。每輛車都根據(jù)自己的規(guī)則,和周圍的環(huán)境產(chǎn)生作用。根據(jù)這些規(guī)則,當(dāng)很多車輛互相作用時,不同種類的行為就會出現(xiàn)。</p><p>  元胞自動機。一個在基礎(chǔ)設(shè)施上的具體設(shè)計和仿真(簡單的)汽車駕駛規(guī)則,利用了元胞自動機(CA)。元胞自動機運用離散的部分連接細(xì)胞,那些細(xì)胞就能處于一種特殊的狀態(tài)下。例如,一個

57、道路細(xì)胞可以包含一輛汽車或者也可以是空的。當(dāng)?shù)氐霓D(zhuǎn)換規(guī)則決定了系統(tǒng)的動力學(xué),甚至簡單的規(guī)則可以導(dǎo)致混沌動力學(xué)。Nagel and Schreckenberg (1992)描述了這種用于交通仿真的元胞自動機模型。在每個離散的時間--步長內(nèi),車輛在一定數(shù)值上增加自身的速度,直到他們的最大速度。萬一如果車速較慢的車行駛在前面,那么車輛的速度將會降低,避免沖撞。一些無規(guī)則性是通過增加每輛車的小幾率減速而被介紹的。實驗表明,在單一路段上,當(dāng)交通密

58、度增加時,元胞自動機模型的現(xiàn)實行為會以起始波浪那樣的形式出現(xiàn)。</p><p>  認(rèn)知的多代理系統(tǒng)。一個更先進(jìn)的交通仿真和優(yōu)化方法是認(rèn)知多代理系統(tǒng)方式(CMAS),這些代理互相作用并且與其他代理和基礎(chǔ)設(shè)施相溝通。一個認(rèn)知代理是一個整體,這個整體利用最小的努力,試著去達(dá)成一些目標(biāo)狀態(tài)。他利用自己的傳感器接收來自環(huán)境的信息,對這些來自環(huán)境的信息產(chǎn)生信任,利用這些信任,輸入信號來選擇一個行動。因為每個代理都是一個單獨

59、的整體,他可以優(yōu)化他的選擇行為(例如使用學(xué)習(xí)能力)。此外,不同的代理有不同的傳感器、目標(biāo)、行為和學(xué)習(xí)能力,利用異構(gòu)多代理系統(tǒng),從而使我們利用一個非常廣泛的(微觀)交通模型去實驗。</p><p>  Dia(2002)使用了一個基于真正司機模型的、認(rèn)知多代理系統(tǒng)的運行信息回應(yīng)。在調(diào)查中,采取了一個擁擠的走廊,選擇了有影響因素的路線和起飛時間來研究。這個結(jié)果以前被用于建立一個司機數(shù)量模型,司機對現(xiàn)有的運行信息回應(yīng)不

60、同。利用這一數(shù)量,在研究的區(qū)域內(nèi),不同信息系統(tǒng)的影響能被仿真出來。盡管目前沒有結(jié)論,但是這樣的研究似乎是有希望的。</p><p><b>  2.2 預(yù)測交通</b></p><p>  對于優(yōu)化控制來說,預(yù)測交通條件的能力是重要的。例如,如果我們在現(xiàn)有的條件下,能夠知道哪些道路將會在未來堵塞,這些信息能夠傳輸給道路使用者,那么他們就能規(guī)避這條道路,從而緩解整個系統(tǒng)

61、的擁堵現(xiàn)象。另外,如果我們能準(zhǔn)確的預(yù)測不同駕駛策略的后果,一個優(yōu)化決定(或者至少對這個區(qū)間路段的優(yōu)化預(yù)測)就能通過預(yù)測結(jié)果進(jìn)行比較。</p><p>  在一個交叉路口,最簡單形式的交通預(yù)測就是通過在某段時間內(nèi)測量交通,或者假設(shè)下一時段的交通和現(xiàn)在相同[Ledoux,1996],一個交通燈下,神經(jīng)網(wǎng)絡(luò)被用于表示對一列排隊的長期預(yù)測。一個多感知層[Rumelhart et al.,1986]被訓(xùn)練來預(yù)測下一個時間-

62、-步長的排隊長度,長期預(yù)測可以由迭代法一步預(yù)測制成。當(dāng)預(yù)測以十個步長進(jìn)行時,生成的網(wǎng)絡(luò)將十分精確,但是它還沒有被納入到控制器里。</p><p>  應(yīng)用于真實生活情況的交通預(yù)測模型被描述在[Wahle and Schreckenberg, 2001]。該模型是一種占用一個仿真基礎(chǔ)設(shè)施的多代理系統(tǒng)(MAS)。每個代理都有兩層控制:一個是(簡單的)駕駛決定,另一個是類似于路徑選擇的決定戰(zhàn)術(shù)策略。利用已經(jīng)安裝好的探測

溫馨提示

  • 1. 本站所有資源如無特殊說明,都需要本地電腦安裝OFFICE2007和PDF閱讀器。圖紙軟件為CAD,CAXA,PROE,UG,SolidWorks等.壓縮文件請下載最新的WinRAR軟件解壓。
  • 2. 本站的文檔不包含任何第三方提供的附件圖紙等,如果需要附件,請聯(lián)系上傳者。文件的所有權(quán)益歸上傳用戶所有。
  • 3. 本站RAR壓縮包中若帶圖紙,網(wǎng)頁內(nèi)容里面會有圖紙預(yù)覽,若沒有圖紙預(yù)覽就沒有圖紙。
  • 4. 未經(jīng)權(quán)益所有人同意不得將文件中的內(nèi)容挪作商業(yè)或盈利用途。
  • 5. 眾賞文庫僅提供信息存儲空間,僅對用戶上傳內(nèi)容的表現(xiàn)方式做保護(hù)處理,對用戶上傳分享的文檔內(nèi)容本身不做任何修改或編輯,并不能對任何下載內(nèi)容負(fù)責(zé)。
  • 6. 下載文件中如有侵權(quán)或不適當(dāng)內(nèi)容,請與我們聯(lián)系,我們立即糾正。
  • 7. 本站不保證下載資源的準(zhǔn)確性、安全性和完整性, 同時也不承擔(dān)用戶因使用這些下載資源對自己和他人造成任何形式的傷害或損失。

評論

0/150

提交評論