【摘要】：隨著信息化技術(shù)在道路交通領(lǐng)域的廣泛應(yīng)用,以數(shù)據(jù)挖掘技術(shù)為核心的道路車輛監(jiān)控管理和輔助決策已經(jīng)成為現(xiàn)代智能交通的重要發(fā)展方向之一。數(shù)據(jù)挖掘可以很好地解決道路交通領(lǐng)域中信息海量但缺乏深入研究的問題,然而,我國目前對于數(shù)據(jù)挖掘技術(shù)及其在車輛監(jiān)控系統(tǒng)中的應(yīng)用研究還處于起步階段。 基于以上背景,本文討論了數(shù)據(jù)挖掘技術(shù)和其中的聚類分析技術(shù),包括五種常用聚類分析技術(shù)的定義、基本思想和主要算法步驟；根據(jù)數(shù)據(jù)挖掘技術(shù)的應(yīng)用場景,提出了車輛監(jiān)控系統(tǒng)的總體設(shè)計方案,包括系統(tǒng)的技術(shù)指標(biāo)、總體架構(gòu)和技術(shù)路線；分析了車輛監(jiān)控系統(tǒng)數(shù)據(jù)的特點,完成了車輛數(shù)據(jù)的預(yù)處理過程；將數(shù)據(jù)挖掘技術(shù)與車輛應(yīng)用系統(tǒng)的場景相結(jié)合,提出了一種具有噪聲的基于密度聚類算法的優(yōu)化方法,該方法具有聚類速度快、抗噪性能良好以及可以發(fā)現(xiàn)任意形狀的空間聚類的優(yōu)點,并將改進后的聚類算法應(yīng)用于車輛監(jiān)控系統(tǒng)之中,實現(xiàn)了車輛監(jiān)控系統(tǒng)中超速多發(fā)路段的發(fā)現(xiàn)和重點車輛信息監(jiān)控的功能；本文最后展示了車輛監(jiān)控系統(tǒng)的相關(guān)界面,分析了超速多發(fā)路段的聚類分析和重點車輛信息監(jiān)控的結(jié)果。實驗結(jié)果表明,改進后的聚類算法應(yīng)用于車輛監(jiān)控系統(tǒng)之中,能夠有效地實現(xiàn)車輛監(jiān)控系統(tǒng)中發(fā)現(xiàn)超速多發(fā)路段和監(jiān)控重點車輛信息的功能,并顯著提升了發(fā)現(xiàn)超速多發(fā)路段的準(zhǔn)確性,從而為道路交通監(jiān)管機構(gòu)的決策提供了有力的支持。
[Abstract]:With the wide application of information technology in the field of road traffic, data mining technology as the core of the monitoring and management of road vehicles and auxiliary decision-making has become one of the important development directions of modern intelligent transportation. Data mining can solve the problem of huge amount of information but lack of in-depth research in the field of road traffic. However, the research on data mining technology and its application in vehicle monitoring system is still in its infancy in our country. Based on the above background, this paper discusses the data mining technology and clustering analysis technology, including five commonly used clustering analysis technology definition, basic ideas and main algorithm steps. The overall design scheme of the vehicle monitoring system is put forward, including the technical index, the overall structure and the technical route of the system, the characteristics of the vehicle monitoring system data are analyzed, and the preprocessing process of the vehicle data is completed. Combining the data mining technology with the scene of vehicle application system, an optimization method based on density clustering algorithm with noise is proposed, which has fast clustering speed. The improved clustering algorithm is applied to vehicle monitoring system, which has good anti-noise performance and can find the advantages of arbitrary shape spatial clustering. Realized the detection of overspeed section in the vehicle monitoring system and the function of key vehicle information monitoring. Finally, the paper showed the related interface of the vehicle monitoring system. The results of clustering analysis and monitoring of key vehicles are analyzed. The experimental results show that the improved clustering algorithm is applied to the vehicle monitoring system, which can effectively realize the function of finding overspeed sections and monitoring key vehicle information in the vehicle monitoring system. It also improves the accuracy of finding overspeed section, and provides strong support for the decision of road traffic supervision organization.
【學(xué)位授予單位】：北京郵電大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2015
【分類號】：U495;TP311.13

【參考文獻】

相關(guān)期刊論文 前7條

1 周水庚,周傲英,曹晶;基于數(shù)據(jù)分區(qū)的DBSCAN算法[J];計算機研究與發(fā)展;2000年10期

2 榮秋生,顏君彪,郭國強;基于DBSCAN聚類算法的研究與實現(xiàn)[J];計算機應(yīng)用;2004年04期

3 蔣暉;陳允鋒;;數(shù)據(jù)挖掘及其一種關(guān)聯(lián)規(guī)則算法[J];計算機與數(shù)字工程;2011年06期

4 宋明,劉宗田;基于數(shù)據(jù)交疊分區(qū)的并行DBSCAN算法[J];計算機應(yīng)用研究;2004年07期

5 陳睿,韓春梅,朱健,李強;高速公路應(yīng)急救援指揮軟件系統(tǒng)的設(shè)計[J];交通與計算機;2001年05期

6 吳章龍;;堅持人本管理 重在關(guān)鍵環(huán)節(jié)——關(guān)于道路運輸企業(yè)安全管理的探討[J];交通企業(yè)管理;2006年06期

7 錢衛(wèi)寧 ,魏藜 ,王焱 ,錢海蕾 ,周傲英;一個面向大規(guī)模數(shù)據(jù)庫的數(shù)據(jù)挖掘系統(tǒng)[J];軟件學(xué)報;2002年08期

，

本文編號：2218722