【摘要】：交通仿真技術(shù)是智能交通技術(shù)的重要組成部分。通過交通仿真,可以對智能交通控制算法的效果進(jìn)行驗(yàn)證。傳統(tǒng)的交通仿真無法滿足當(dāng)前城市快速路對仿真方面的需求,使得快速路的算法仿真實(shí)現(xiàn)極為困難。 本文首先分析了城市快速路的仿真需求,在此基礎(chǔ)上設(shè)計(jì)了快速路仿真平臺的總體框架,并對框架內(nèi)的各個(gè)功能進(jìn)行逐一實(shí)現(xiàn),最后對仿真平臺效果進(jìn)行驗(yàn)證。全文的主要研究內(nèi)容和成果總結(jié)如下： 1)通過設(shè)計(jì)TransModeler的二次開發(fā)框架,打通TransModeler軟件內(nèi)部與外部算法之間的通道,將外部算法通過GISDK導(dǎo)入TransModeler中,解決了傳統(tǒng)仿真平臺無法導(dǎo)入外部算法的問題。 2)針對快速路控制算法在仿真中的實(shí)時(shí)性與閉環(huán)在線方面的要求,本文設(shè)計(jì)了實(shí)時(shí)交通流數(shù)據(jù)庫與仿真流程控制器。通過TransModeler的TsmAPI接口實(shí)時(shí)獲取檢測器采集交通流數(shù)據(jù),并將采集的瞬時(shí)值轉(zhuǎn)化為平均值,存入數(shù)據(jù)庫中,供外部算法驗(yàn)證時(shí)使用,解決了仿真平臺的實(shí)時(shí)性要求。同時(shí),通過運(yùn)用系統(tǒng)底層的回調(diào)函數(shù),控制整個(gè)仿真流程,根據(jù)控制方案的要求對控制算法循環(huán)調(diào)用,實(shí)現(xiàn)了對算法的閉環(huán)在線控制。 3)部分控制算法不僅需要實(shí)時(shí)數(shù)據(jù),同時(shí)需要未來的預(yù)測數(shù)據(jù),本文利用卡爾曼濾波算法,對快速路的交通流進(jìn)行短時(shí)交通流預(yù)測,通過仿真驗(yàn)證,預(yù)測效果精度可以滿足要求,是一種有效的快速路短時(shí)交通流預(yù)測算法,能夠直接為控制算法提供預(yù)測數(shù)據(jù)支持。 4)在上述工作的基礎(chǔ)上,選取北京四通橋區(qū)域作為控制背景,搭建快速路仿真平臺,測試仿真平臺對基于實(shí)時(shí)區(qū)域交通狀態(tài)的入口匝道與輔路口的協(xié)調(diào)控制策略的支持情況,結(jié)果表明仿真平臺能夠較好支持協(xié)調(diào)控制策略的仿真驗(yàn)證要求。
[Abstract]:Traffic simulation technology is an important part of intelligent transportation technology. Through traffic simulation, the effect of intelligent traffic control algorithm can be verified. The traditional traffic simulation can not meet the needs of the current urban expressway simulation, so it is very difficult to realize the algorithm simulation of the expressway. In this paper, the simulation requirements of urban expressway are analyzed firstly, and the overall framework of the simulation platform is designed, and the functions of the simulation platform are realized one by one. Finally, the simulation platform effect is verified. The main research contents and achievements are summarized as follows: 1) through the design of the secondary development framework of TransModeler, the channels between the internal and external algorithms of TransModeler software are opened, and the external algorithms are imported into TransModeler through GISDK. The problem that the traditional simulation platform can not import the external algorithm is solved. 2) according to the requirement of real-time and closed-loop control algorithm in simulation, the real-time traffic flow database and simulation flow controller are designed in this paper. Through the TsmAPI interface of TransModeler, the real-time acquisition data of traffic flow is acquired by the detector, and the collected instantaneous value is transformed into the average value, which is stored in the database for use in the verification of the external algorithm. The real-time requirement of the simulation platform is solved. At the same time, the closed-loop on-line control of the algorithm is realized by using the callback function of the bottom layer of the system to control the whole simulation process and calling the control algorithm cycle according to the requirements of the control scheme. 3) some control algorithms need not only real-time data, but also forecast data in the future. In this paper, Kalman filter algorithm is used to predict the short-term traffic flow of expressway, which is verified by simulation. The prediction accuracy can meet the requirements and is an effective short-term traffic flow prediction algorithm for expressway. It can directly support the prediction data for the control algorithm. 4) on the basis of the above-mentioned work, the Beijing Sitong Bridge area is selected as the control background, and the expressway simulation platform is built to test the support of the simulation platform for the coordinated control strategy of the on-ramp and auxiliary junctions based on the real-time regional traffic condition. The simulation results show that the simulation platform can better support the simulation verification requirements of the coordinated control strategy.
【學(xué)位授予單位】：浙江大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2014
【分類號】：U495;U491

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 耿彥斌;李海峰;柴大勝;;基于Trans Modeler的城市快速路交通設(shè)計(jì)方案的微觀仿真評價(jià)[J];道路交通與安全;2009年01期

2 唐敏;宋杰;;嵌入式數(shù)據(jù)庫SQLite的原理與應(yīng)用[J];電腦知識與技術(shù);2008年04期

3 楊宏鵬;王阿川;王妍瑋;;GIS二次開發(fā)方法與實(shí)現(xiàn)[J];信息技術(shù);2008年08期

4 逄玉俊;柳明;李元;;k均值聚類分析在過程改進(jìn)中的應(yīng)用[J];華中科技大學(xué)學(xué)報(bào)(自然科學(xué)版);2009年S1期

5 陳學(xué)文;王殿海;金盛;王彩霞;;城市快速路人口匝道控制方法[J];吉林大學(xué)學(xué)報(bào)(工學(xué)版);2008年S1期

6 熊勝輝;李星毅;施化吉;;基于元胞自動(dòng)機(jī)的快速路交織區(qū)交通流仿真建模[J];計(jì)算機(jī)應(yīng)用;2010年02期

7 馬殨;張玉梅;;高速公路入口匝道控制算法綜述[J];計(jì)算機(jī)仿真;2009年01期

8 姜舟;;快速路型式的選擇探析[J];交通標(biāo)準(zhǔn)化;2010年13期

9 曲昭偉;王立明;金盛;;快速路出口匝道銜接道路控制仿真研究[J];交通信息與安全;2009年02期

10 余俊昌;干宏程;楊珍珍;;快速路控制策略仿真研究[J];計(jì)算機(jī)與現(xiàn)代化;2011年09期

相關(guān)博士學(xué)位論文 前1條

1 孫偉力;基于VMS發(fā)布誘導(dǎo)信息的快速路網(wǎng)動(dòng)態(tài)最優(yōu)控制[D];浙江大學(xué);2013年

，

本文編號：2456358