發(fā)布時(shí)間：2018-04-10 17:02

  本文選題：交通流預(yù)測(cè) + 多模型融合�。� 參考：《華南理工大學(xué)》2014年碩士論文

【摘要】：隨著我國經(jīng)濟(jì)的持續(xù)穩(wěn)步快速發(fā)展，城市化進(jìn)程的步伐日益加快，城市道路交通基礎(chǔ)設(shè)施供給速度的緩慢與城市的交通需求的飛速發(fā)展之間的矛盾日益突出，交通擁堵問題已經(jīng)嚴(yán)重阻礙了城市的可持續(xù)發(fā)展，對(duì)人們的日常生活與工作造成了嚴(yán)重影響。國內(nèi)外發(fā)展實(shí)踐表明，在現(xiàn)有交通設(shè)施的基礎(chǔ)上通過更加信息化、智能化的管理系統(tǒng)來提高交通管理水平，是從根本上解決城市道路交通問題的有效手段。南沙區(qū)智能交通管控平臺(tái)項(xiàng)目以“提高管理水平、保障城市暢通、改善交通秩序、提升服務(wù)水平”為目標(biāo)。城市道路交通流的精確預(yù)測(cè)以及交通狀態(tài)判別是該系統(tǒng)中的重要組成部分。本文選題來自南沙區(qū)智能交通管控平臺(tái)項(xiàng)目。 本文研究的重點(diǎn)是短時(shí)交通流的預(yù)測(cè)以及道路交通狀態(tài)的判別這兩方面，，本文研究的目的是為智能交通管控平臺(tái)提供有力的技術(shù)支持，使得管理更具有科學(xué)性。主要包括以下幾方面的研究： （1）根據(jù)交通流預(yù)測(cè)以及狀態(tài)判別的需要，提出了交通數(shù)據(jù)預(yù)處理的方法。文章簡(jiǎn)要介紹了故障數(shù)據(jù)的識(shí)別以及處理技術(shù)，通過預(yù)處理保證了數(shù)據(jù)的質(zhì)量，使得交通流預(yù)測(cè)以及狀態(tài)判別結(jié)果的準(zhǔn)確性有所保證。 （2）前期的大量研究表明，僅僅選用單一預(yù)測(cè)模型很難滿足預(yù)期的精度要求。本文在以BP神經(jīng)網(wǎng)絡(luò)進(jìn)行短時(shí)交通流預(yù)測(cè)研究的基礎(chǔ)上，詳述了其建模過程，并對(duì)其中的不足給出了改進(jìn)的方法。然后選用小波神經(jīng)網(wǎng)絡(luò)建立短時(shí)交通流預(yù)測(cè)模型，最后提出了基于數(shù)據(jù)融合的多模型融合預(yù)測(cè)算法。通過運(yùn)用MATLAB進(jìn)行仿真對(duì)比，從而驗(yàn)證了多模型融合預(yù)測(cè)算法的有效性。 （3）根據(jù)道路交通狀態(tài)的模糊不確定性，本文通過模糊綜合評(píng)價(jià)的方法來對(duì)道路交通狀態(tài)進(jìn)行判別，并選用模糊層次分析法對(duì)其中的權(quán)重系數(shù)進(jìn)行確定，以此降低人為因素的影響,最后通過實(shí)例分析驗(yàn)證了算法的可行性。 （4）以南沙區(qū)智能交通管控平臺(tái)項(xiàng)目為基礎(chǔ)，介紹了其中交通流預(yù)測(cè)與狀態(tài)判別的實(shí)現(xiàn)過程。為大中城市解決相關(guān)問題提供了參考。
[Abstract]:With the steady and rapid development of our economy, the pace of urbanization is accelerating day by day, and the contradiction between the slow supply of urban road traffic infrastructure and the rapid development of urban traffic demand is becoming increasingly prominent.The problem of traffic congestion has seriously hindered the sustainable development of the city and has a serious impact on people's daily life and work.The development practice at home and abroad shows that improving the level of traffic management through more information and intelligent management system on the basis of existing traffic facilities is an effective means to fundamentally solve urban road traffic problems.Nansha intelligent traffic control platform project aims to improve the management level, ensure the smooth flow of the city, improve the traffic order and improve the service level.The accurate prediction of urban road traffic flow and the identification of traffic state are important parts of the system.This paper selected topics from the Nansha District Intelligent Transportation Control platform project.This paper focuses on the prediction of short-term traffic flow and the identification of road traffic status. The purpose of this study is to provide strong technical support for the intelligent traffic control platform and make the management more scientific.Mainly includes the following aspects of research:1) according to the need of traffic flow prediction and state discrimination, a method of traffic data preprocessing is proposed.This paper briefly introduces the identification and processing technology of the fault data. The quality of the data is guaranteed by preprocessing, which ensures the accuracy of the traffic flow prediction and the result of state discrimination.2) A large number of previous studies have shown that it is difficult to meet the expected accuracy requirements by using a single prediction model.Based on the research of short time traffic flow prediction based on BP neural network, the modeling process of BP neural network is described in detail, and an improved method is given for its shortcomings.Then wavelet neural network is used to build a short-term traffic flow prediction model. Finally, a multi-model fusion prediction algorithm based on data fusion is proposed.The effectiveness of the multi-model fusion prediction algorithm is verified by simulation and comparison with MATLAB.3) according to the fuzzy uncertainty of the road traffic state, this paper uses the fuzzy comprehensive evaluation method to distinguish the road traffic state, and selects the fuzzy analytic hierarchy process (FAHP) to determine the weight coefficient of the road traffic state.In order to reduce the influence of human factors, the feasibility of the algorithm is verified by an example.4) based on the intelligent traffic control platform project in Nansha District, the realization process of traffic flow prediction and state discrimination is introduced.It provides a reference for large and medium cities to solve related problems.
【學(xué)位授予單位】：華南理工大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2014
【分類號(hào)】：U491.1

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 吳殿廷,李東方;層次分析法的不足及其改進(jìn)的途徑[J];北京師范大學(xué)學(xué)報(bào)(自然科學(xué)版);2004年02期

2 楊兆升;張茂雷;;基于模糊綜合評(píng)判的道路交通狀態(tài)分析模型[J];公路交通科技;2010年09期

3 馮相昭;鄒驥;郭光明;;城市交通擁堵的外部成本估算[J];環(huán)境與可持續(xù)發(fā)展;2009年03期

4 項(xiàng)文強(qiáng);張華;王Y

本文編號(hào)：1732108