發(fā)布時(shí)間：2018-05-30 00:47

  本文選題：換道行為 + 支持向量機(jī)��； 參考：《昆明理工大學(xué)》2017年碩士論文

【摘要】：換道行為是車(chē)輛在行駛過(guò)程中的隨機(jī)動(dòng)作,該行為有可能產(chǎn)生交通沖突點(diǎn),甚至導(dǎo)致不同程度的交通事故發(fā)生。一般根據(jù)換道過(guò)程平順或猛烈,大致將其分為安全性換道行為、風(fēng)險(xiǎn)性換道行為兩類(lèi),換道行為過(guò)于猛烈將增加交通事故風(fēng)險(xiǎn)。伴隨車(chē)輛換道預(yù)警系統(tǒng)的不斷發(fā)展,車(chē)輛換道預(yù)警系統(tǒng)的出現(xiàn),有望在車(chē)輛進(jìn)行換道的起始階段準(zhǔn)確識(shí)別出兩類(lèi)不同的換道行為,并自動(dòng)對(duì)存在風(fēng)險(xiǎn)的換道操作給出相應(yīng)預(yù)警或干預(yù)。本文運(yùn)用模式識(shí)別技術(shù)中的支持向量機(jī)方法,嘗試建立一種能有效識(shí)別區(qū)分安全換道和風(fēng)險(xiǎn)換道兩種駕駛行為的識(shí)別模型。借助KMRTDS駕駛模擬仿真平臺(tái)開(kāi)展安全換道和風(fēng)險(xiǎn)換道模擬實(shí)驗(yàn),用采集到的駕駛操縱數(shù)據(jù)和車(chē)輛運(yùn)行數(shù)據(jù)提取訓(xùn)練及測(cè)試識(shí)別模型的樣本。本文研究意義在于,當(dāng)識(shí)別模型判斷出某一換道行為,因操作過(guò)于猛烈而超出一般安全換道行為閾值時(shí),換道預(yù)警系統(tǒng)第一時(shí)間報(bào)警提示駕駛員,以避免事故發(fā)生。分類(lèi)模型的識(shí)別效果主要受識(shí)別時(shí)間窗大小、特征參數(shù)、模型自身參數(shù)等方面的影響,本文運(yùn)用ROC.運(yùn)行結(jié)果,綜合AUC值和模型的分類(lèi)準(zhǔn)確率對(duì)識(shí)別效果進(jìn)行分析,通過(guò)“最優(yōu)時(shí)間窗確定、最優(yōu)特征參數(shù)提取、最優(yōu)算法尋找模型參數(shù)”三個(gè)步驟,建立最優(yōu)識(shí)別模型。首先,將定義的換道行為起始時(shí)刻作為時(shí)間窗中點(diǎn),向前向后分別取相同時(shí)間段(0.5s、1s、1.5s)形成3個(gè)時(shí)間窗(1s、2s、3s),并經(jīng)過(guò)對(duì)比分析三個(gè)時(shí)間窗下的模型識(shí)別效果,確定了最優(yōu)時(shí)間窗為2s。其次,運(yùn)用逐步回歸分析、因子分析、多維偏好分析三種方法對(duì)原有的特征參數(shù)進(jìn)行降維處理。其中,經(jīng)逐步回歸分析法提取的特征參數(shù)訓(xùn)練后的分類(lèi)器性能最好,故將此方法提取的參數(shù)作為最優(yōu)特征參數(shù)。最后,運(yùn)用了枚舉算法、粒子群算法、遺傳算法等三種算法進(jìn)行參數(shù)尋優(yōu)。其中,遺傳算法的分類(lèi)準(zhǔn)確率最低,枚舉算法和粒子群算法分類(lèi)準(zhǔn)確率較為接近,但后者的AUC值為0.992,接近原始數(shù)據(jù)下的0.996,較好的彌補(bǔ)了因數(shù)據(jù)降維處理而損失的信息,所以選取粒子群算法為最優(yōu)算法。在最優(yōu)時(shí)間窗、最優(yōu)特征參數(shù)、最優(yōu)模型參數(shù)確定后,借助LIBSVM算法,在MATLAB中訓(xùn)練、驗(yàn)證識(shí)別模型,得到模型最終總體識(shí)別率為92.55%,基本能準(zhǔn)確識(shí)別出車(chē)道保持、安全換道、風(fēng)險(xiǎn)換道三種行為,達(dá)到了運(yùn)用較小樣本量建立較高識(shí)別率模型的預(yù)期,取得較好的識(shí)別效果。
[Abstract]:The behavior of changing the road is the random action of the vehicle in the course of driving, which may produce traffic conflict points and even lead to traffic accidents of different degrees. Generally according to the smooth or violent course of changing the road, it can be roughly divided into two types of safe changing, and the risk of changing the road is two kinds, and the risk of traffic accident will be increased if the change of course is too violent. With the continuous development of the vehicle early warning system, the emergence of the vehicle early warning system, it is expected to accurately identify two different types of road change behavior in the initial phase of the vehicle change. And give the corresponding warning or intervention to the change operation of the risk automatically. In this paper, the support vector machine (SVM) method in pattern recognition technology is used to establish a recognition model which can effectively identify and distinguish two driving behaviors: safe change and risk change. With the help of the KMRTDS driving simulation platform, the simulation experiments of safe and risk change are carried out, and the samples of training and testing identification model are extracted from the collected driving control data and vehicle running data. The research significance of this paper is that when the identification model judges a certain changing behavior and exceeds the threshold of the general safe changing behavior because of the heavy operation, the early warning system of changing the channel will alert the driver in the first time to avoid the accident. The recognition effect of the classification model is mainly affected by the size of the time window, the characteristic parameters and the model's own parameters. The result of the operation is based on the analysis of the AUC value and the classification accuracy of the model. The optimal recognition model is established through the three steps of "determining the optimal time window, extracting the optimal feature parameters and finding the parameters of the model by the optimal algorithm". Firstly, the starting time of the change behavior is taken as the midpoint of the time window, and the same time interval is taken forward and backward, respectively.) three time windows are formed, which are 1 sm ~ 2 s ~ 3 s ~ (3), and the optimal time window is determined to be 2 s by comparing and analyzing the model recognition effect under the three time windows. Secondly, stepwise regression analysis, factor analysis and multidimensional preference analysis are used to reduce the dimension of the original characteristic parameters. Among them, the performance of the classifier trained by stepwise regression analysis is the best, so the parameters extracted by this method are regarded as the optimal feature parameters. Finally, three kinds of algorithms, enumeration algorithm, particle swarm optimization algorithm and genetic algorithm, are used to optimize the parameters. The classification accuracy of genetic algorithm is the lowest, the classification accuracy of enumeration algorithm and particle swarm optimization algorithm is close, but the AUC value of the latter is 0.992, which is close to 0.996 under the original data, which makes up for the loss of information due to the dimensionality reduction processing. So the particle swarm optimization algorithm is chosen as the optimal algorithm. After the optimal time window, the optimal feature parameter and the optimal model parameter are determined, the recognition model is trained in MATLAB with the help of LIBSVM algorithm, and the overall recognition rate of the model is 92.55. The three behaviors of risk changing reach the expectation of establishing a higher recognition rate model by using smaller sample size and obtain better recognition effect.
【學(xué)位授予單位】：昆明理工大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類(lèi)號(hào)】：U491

【參考文獻(xiàn)】

相關(guān)期刊論文 前9條

1 丁潔云;黨睿娜;王建強(qiáng);李克強(qiáng);;駕駛?cè)藫Q道決策分析及意圖識(shí)別算法設(shè)計(jì)[J];清華大學(xué)學(xué)報(bào)(自然科學(xué)版);2015年07期

2 袁偉;張亞岐;王暢;;支持向量機(jī)在換道行為識(shí)別中的應(yīng)用研究[J];計(jì)算機(jī)工程與設(shè)計(jì);2013年02期

3 呂岸;胡振程;陳慧;;基于高斯混合隱馬爾科夫模型的高速公路超車(chē)行為辨識(shí)與分析[J];汽車(chē)工程;2010年07期

4 郭鳳香;熊堅(jiān);秦雅琴;萬(wàn)華森;;基于駕駛模擬實(shí)驗(yàn)的85%位車(chē)速預(yù)測(cè)模型[J];交通科技與經(jīng)濟(jì);2010年03期

5 張曉龍;江川;駱名劍;;ROC分析技術(shù)在機(jī)器學(xué)習(xí)中的應(yīng)用[J];計(jì)算機(jī)工程與應(yīng)用;2007年04期

6 鄭勇濤,劉玉樹(shù);支持向量機(jī)解決多分類(lèi)問(wèn)題研究[J];計(jì)算機(jī)工程與應(yīng)用;2005年23期

7 唐發(fā)明,王仲東,陳綿云;支持向量機(jī)多類(lèi)分類(lèi)算法研究[J];控制與決策;2005年07期

8 游家興;如何正確運(yùn)用因子分析法進(jìn)行綜合評(píng)價(jià)[J];統(tǒng)計(jì)教育;2003年05期

9 王冬梅,沈頌東;逐步回歸分析法[J];工業(yè)技術(shù)經(jīng)濟(jì);1997年03期

相關(guān)博士學(xué)位論文 前4條

1 王暢;車(chē)輛換道預(yù)警的若干關(guān)鍵問(wèn)題研究[D];長(zhǎng)安大學(xué);2012年

2 彭金栓;基于視覺(jué)特性與車(chē)輛相對(duì)運(yùn)動(dòng)的駕駛?cè)藫Q道意圖識(shí)別方法[D];長(zhǎng)安大學(xué);2012年

3 賈立山;體現(xiàn)駕駛員特性的車(chē)道偏離預(yù)警系統(tǒng)關(guān)鍵技術(shù)研究[D];華中科技大學(xué);2011年

4 顧柏園;基于單目視覺(jué)的安全車(chē)距預(yù)警系統(tǒng)研究[D];吉林大學(xué);2006年

相關(guān)碩士學(xué)位論文 前4條

1 張利丹;面向車(chē)道偏離預(yù)警系統(tǒng)的駕駛員無(wú)意識(shí)車(chē)道偏離識(shí)別方法研究[D];吉林大學(xué);2015年

2 孫純;基于駕駛?cè)艘曈X(jué)特性的換道意圖識(shí)別[D];長(zhǎng)安大學(xué);2012年

3 彭璐;支持向量機(jī)分類(lèi)算法研究與應(yīng)用[D];湖南大學(xué);2007年

4 駱名劍;基于ROC的分類(lèi)算法評(píng)價(jià)方法[D];武漢科技大學(xué);2005年

，

本文編號(hào)：1953192