本文關(guān)鍵詞：車聯(lián)網(wǎng)條件下的混合動(dòng)力客車車載傳感器實(shí)時(shí)數(shù)據(jù)預(yù)處理研究　出處：《重慶大學(xué)》2014年碩士論文　論文類型：學(xué)位論文

  更多相關(guān)文章： 物聯(lián)網(wǎng) 車聯(lián)網(wǎng) 情報(bào)定量分析 數(shù)據(jù)預(yù)處理 實(shí)時(shí)數(shù)據(jù)清洗

【摘要】：在計(jì)算機(jī)、互聯(lián)網(wǎng)以及移動(dòng)通信出現(xiàn)以后，物聯(lián)網(wǎng)引領(lǐng)了新一代的信息技術(shù)革命。2009年美國在首次提出“智慧地球”這一新穎概念的同時(shí)，將物聯(lián)網(wǎng)作為振興經(jīng)濟(jì)、提高國家實(shí)力的重點(diǎn)領(lǐng)域之一。同一年，我國溫總理提出了“感知中國”，并把物聯(lián)網(wǎng)正式列入國家五大新興戰(zhàn)略性支柱產(chǎn)業(yè)之一，同時(shí)寫入政府工作報(bào)告。在信息社會(huì)，物聯(lián)網(wǎng)主要應(yīng)用在醫(yī)療、電子政務(wù)、電網(wǎng)、教育、交通、城市管理等領(lǐng)域。物聯(lián)網(wǎng)應(yīng)用的同時(shí)產(chǎn)生了數(shù)以萬計(jì)的實(shí)時(shí)數(shù)據(jù)，隨著數(shù)據(jù)量的急劇增長和數(shù)據(jù)種類的不斷增加，催生了大數(shù)據(jù)時(shí)代的到來。 車聯(lián)網(wǎng)作為物聯(lián)網(wǎng)在智能公交中的典型應(yīng)用，車聯(lián)網(wǎng)條件下的車載數(shù)據(jù)除了具有物聯(lián)網(wǎng)大數(shù)據(jù)一般性的特征，而且具有自身的結(jié)構(gòu)性質(zhì)。本文在總結(jié)數(shù)據(jù)預(yù)處理、數(shù)據(jù)清洗和數(shù)據(jù)質(zhì)量等情報(bào)定量分析的相關(guān)文獻(xiàn)下，探討了實(shí)時(shí)數(shù)據(jù)清洗處理定量分析的研究現(xiàn)狀，分析了車聯(lián)網(wǎng)實(shí)時(shí)數(shù)據(jù)的主要質(zhì)量問題和特征。在研究一般傳感數(shù)據(jù)和傳感信息的處理方法下，基于滑動(dòng)窗口借鑒使用各種情報(bào)定量分析方法對(duì)車載傳感實(shí)時(shí)數(shù)據(jù)進(jìn)行清洗等預(yù)處理。 針對(duì)所研究的問題，本文介紹了車聯(lián)網(wǎng)網(wǎng)絡(luò)架構(gòu)和車載數(shù)據(jù)框架，，結(jié)合車載數(shù)據(jù)的海量、異構(gòu)、實(shí)時(shí)等主要特征，提出了車聯(lián)網(wǎng)條件下的車載數(shù)據(jù)實(shí)時(shí)清洗的預(yù)處理框架，然后以連續(xù)型和開關(guān)量信號(hào)兩種主要的車載類型實(shí)時(shí)數(shù)據(jù)的清洗處理算法。最后以城市混合動(dòng)力客車的天然氣氣瓶壓力為代表的連續(xù)型變量和剎車制動(dòng)為代表的車載開關(guān)量信號(hào)實(shí)時(shí)數(shù)據(jù)為基礎(chǔ)，分別研究分析了相關(guān)的實(shí)時(shí)清洗預(yù)處理算法。針對(duì)各自的數(shù)據(jù)特征，本文以滑動(dòng)窗口來保證數(shù)據(jù)清洗的實(shí)時(shí)性，運(yùn)用萊茵達(dá)準(zhǔn)則、滑動(dòng)平均濾波、小波變換、貝葉斯決策理論等多種定量分析方法進(jìn)行混合清洗處理使用。運(yùn)用這些情報(bào)定量分析方法對(duì)數(shù)據(jù)進(jìn)行預(yù)處理，能夠有效地處理掉錯(cuò)誤、噪聲、缺失等異常數(shù)據(jù)，得到準(zhǔn)確的數(shù)據(jù)信息，為駕乘人員提供實(shí)時(shí)、準(zhǔn)確、可靠的公共交通信息服務(wù)，從而改善公共交通的智能化運(yùn)營管理。 由于車載數(shù)據(jù)的特征和數(shù)據(jù)實(shí)時(shí)清洗算法的復(fù)雜性，加之寫作時(shí)間比較倉促和自身的水平能力有限，文中有錯(cuò)誤之處還請(qǐng)批評(píng)指正。
[Abstract]:After the advent of computers, the Internet of things and mobile communications, the Internet of things led a new generation of information technology revolutions. In 2009, the United States first proposed the new concept of "intelligent earth" at the same time. In the same year, Premier Wen proposed "perceiving China" and formally listed the Internet of things as one of the five emerging strategic pillar industries. In the information society, the Internet of things is mainly used in medical, e-government, power grid, education, transportation. The application of the Internet of things has produced tens of thousands of real-time data in the field of urban management. With the rapid growth of data volume and the constant increase of data types, the advent of the era of big data has been accelerated. As a typical application of the Internet of things in the intelligent public transportation, the vehicle-borne data under the condition of the vehicle networking has the general characteristics of the big data of the Internet of things. And has its own structural properties. This paper summarizes the data preprocessing, data cleaning and data quality and other information quantitative analysis literature, discusses the research status of real-time data cleaning quantitative analysis. This paper analyzes the main quality problems and characteristics of the real-time data of the vehicle network, and studies the processing methods of the general sensing data and sensing information. Based on sliding window, various intelligence quantitative analysis methods are used to preprocess the real time data of vehicle sensor. Aiming at the problems, this paper introduces the network architecture and data frame of vehicle network, combining with the main characteristics of vehicle data, such as mass, heterogeneity, real-time and so on. In this paper, a pre-processing framework for real-time cleaning of on-board data under the condition of vehicle networking is proposed. Finally, the continuous variable and brake brake are used as the representative of the gas cylinder pressure of the city hybrid electric bus. The on-board switch signal is based on real-time data. According to their data characteristics, this paper uses sliding window to ensure the real-time of data cleaning, using Rhinda criterion, moving average filter, wavelet transform. Bayesian decision theory and other quantitative analysis methods are used for mixed cleaning. Using these information quantitative analysis methods to preprocess the data can effectively deal with the abnormal data such as error noise and missing. Get accurate data information, provide real time, accurate and reliable public transportation information service for drivers, and improve the intelligent operation management of public transportation. Due to the characteristics of on-board data and the complexity of data real-time cleaning algorithm, coupled with the short writing time and limited ability of its own level, there are some errors in this paper, which should be criticized and corrected.
【學(xué)位授予單位】：重慶大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2014
【分類號(hào)】：TP391.44;TN929.5;TP311.13

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 黃國兵;馮龍喜;王軍平;馮樂雯;;基于ARM的開關(guān)量信號(hào)采集模板的設(shè)計(jì)[J];電子技術(shù);2009年03期

2 賈曉豐;;基于物聯(lián)網(wǎng)的大數(shù)據(jù)量實(shí)時(shí)信息交換策略研究[J];電子政務(wù);2011年04期

3 諸彤宇;王家川;陳智宏;;車聯(lián)網(wǎng)技術(shù)初探[J];公路交通科技(應(yīng)用技術(shù)版);2011年05期

4 黃曉斌;鐘輝新;;大數(shù)據(jù)時(shí)代企業(yè)競爭情報(bào)研究的創(chuàng)新與發(fā)展[J];圖書與情報(bào);2012年06期

5 裴益軒;郭民;;滑動(dòng)平均法的基本原理及應(yīng)用[J];火炮發(fā)射與控制學(xué)報(bào);2001年01期

6 王自強(qiáng);段愛玲;張德賢;;基于支持向量數(shù)據(jù)描述的高效異常數(shù)據(jù)檢測算法[J];吉林大學(xué)學(xué)報(bào)(工學(xué)版);2009年02期

7 韓京宇;徐立臻;董逸生;;數(shù)據(jù)質(zhì)量研究綜述[J];計(jì)算機(jī)科學(xué);2008年02期

8 張德喜;黃浩;;一種適合于大數(shù)據(jù)集處理的混合EM算法[J];計(jì)算機(jī)應(yīng)用;2006年08期

9 亓開元;趙卓峰;房俊;馬強(qiáng);;針對(duì)高速數(shù)據(jù)流的大規(guī)模數(shù)據(jù)實(shí)時(shí)處理方法[J];計(jì)算機(jī)學(xué)報(bào);2012年03期

10 畢強(qiáng),史海燕;網(wǎng)絡(luò)信息集成服務(wù)研究綜述[J];情報(bào)理論與實(shí)踐;2004年01期

本文編號(hào)：1421124