【作者】 姜平；

【导师】 陈无畏；

【作者基本信息】 合肥工业大学 ， 车辆工程， 2011， 博士

【摘要】 随着国民经济的飞速发展和人们生活水平的提高,汽车在我们国家应用越来越广泛,但由汽车引起的一系列问题也随之产生。比如在用车使用年限较长,新车排放达标率低,交通管理、道路设施等方面还不完善,汽车排放污染状况日益严重。为了有效地控制城市汽车排放污染物,必须掌握汽车在实际路况上的排放量。研究行驶工况是汽车排放和燃油消耗的基础工作,也是确定汽车排放总量及其环境影响的重要依据。研究我国城市道路实际的行驶工况,不能硬套国外的行驶工况标准。目前,国内外对汽车行驶工况的研究多处于理论和实验中,在具体城市上的应用并不多见,特别是我们国家虽然对某些城市的汽车行驶工况进行了研究,但应用于实际的很少,只是局限在某些理论分析上,且理论分析往往采用传统的多元统计分析方法。主要是因为在汽车实际行驶工况研究中存在着大量不确定因素,每个城市每种车型的行驶工况都不一样。自20世纪90年代末以来,研究机构和高校都从理论方面对汽车行驶工况进行了大量的研究。本文在借鉴国内外汽车行驶工况构建方法的同时,采用短行程方法、马尔科夫方法、小波变换方法和聚类方法对汽车行驶工况进行了整体构建和分类构建,为我国混合交通行驶工况的研究提供了新的思路,提高了行驶工况的构建精度。本文主要包括以下内容:(1)以合肥市为例,对典型道路行驶工况进行了实验数据的整理。(2)建立了短行程特征参数的数学模型,分析了合肥市行驶工况的特点。采用传统的短行程方法对城市道路的行驶工况进行了构建,提取了反映汽车行驶工况的特征参数,把构建好的代表性行驶工况与实验数据进行了对比分析。(3)考虑我国混合交通的特点,采用小波变换方法对汽车行驶工况实验数据进行了压缩和重构,对重构后的行驶工况数据进行了工况构建。(4)汽车行驶工况可以看作一个随机过程,利用极大似然估计法和马尔科夫方法对汽车行驶工况进行了构建,构建的代表性行驶工况与传统方法、小波变换方法进行了比较,并同国外标准行驶工况进行了对比分析。(5)采用主成分分析和聚类分析方法把汽车行驶工况分成拥挤行驶工况和畅通行驶工况,并分别构建了两类代表性行驶工况。(6)采用因子分析法得到与燃油消耗相关的两个因子,对整个行驶工况进行聚类分析,得到与燃油消耗相关的低速、中速和高速行驶工况。每类分别按照马尔科夫方法和新的分类准则把行驶工况分成不同片段,根据转移概率和比功率构建候选工况,根据综合加权测试法选取代表性行驶工况。(7)总结了全文的研究内容,对下一步的研究提出一些建议。更多还原

【Abstract】 With the rapid development of the national economy and the improvement of people’s living standards, the cars are more and more widely used in our country, but a series of problems that the automobiles caused have appeared. For example, the cars in use have longer lives, the emission standard-reaching rate of new cars is low, traffic management and road infrastructure, etc, are not perfect, automobile emission pollutions are getting worse.In order to effectively control emission pollutants of urban vehicles, the vehicle emissions on real-road driving conditions must be mastered. The research of the driving cycle is basic work of vehicle emissions and fuel consumption and is also an important basis to determine vehicle emissions and environmental impacts of vehicle emissions.Standards of foreign driving cycles can not be mechanically transplanted for the real driving cycle research of cities of our country. The present research on the vehicle driving cycle is only theoretical and experimental at home and abroad, practical application in specific cities is rare, especially in our country although driving cycles of vehicles of some cities are studied, very few are used in practice, only limited to some theoretical analyses, and theoretical analyses often use traditional methods of multivariate statistical theory. Mainly because there are a lot of uncertainty during the research of actual vehicle driving cycle, driving cycles of each city and every vehicle type are different. Since the late 90s during the 20th century a number of theoretical research of the vehicle driving cycle are made by research institutions and universities. In this paper, learned from the design methods of the vehicle driving cycle at home and abroad, driving cycles are wholly studied and classified by the short trip method, the Markov method, the wavelet transform method and the cluster method, which provide the new thinking for the driving cycle research of mixed traffic in our country and improve the accuracy of the driving cycle construction. The research contents of this dissertation are listed as follow:(1) The experimental data are treated in typical road driving cycles by an example of Hefei.(2) The mathematic models of parameters of short trip are set up and the characteristics of driving cycle in Hefei are analyzed. The driving cycle is constructed by the traditional method of short trip, the parameters of the vehicle driving cycle are reflected, the constructed representative driving cycle is compared with the experimental data.(3) According to the characteristics of mixed traffic in China the experimental data of the vehicle driving cycle are compressed and reconstructed by the wavelet transform method, the reconstructed driving cycle data are constructed.(4) The vehicle driving cycle can be viewed as a random process, which is constructed by using the maximum likelihood estimation method and the Markov method. The constructed representative driving cycle is compared with the traditional method, wavelet transform method and foreign standard driving cycles.(5) The vehicle driving cycle is divided into crowded cycle and unblocked cycle by principle component analysis and cluster analysis, two groups of representative driving cycles are separately constructed.(6) Two factors related to fuel consumption are obtained using factor analysis, the total driving cycle is clustered and the low-speed, medium-speed and high-speed driving cycles related with fuel consumption are obtained. The driving cycle of each class is divided into different segments according to the Markov method and new classification criteria, the candidate driving cycles are constructed according to the transition probability and the specific power, the representative driving cycle is selected by the Composite Performance Measure.(7) The study work of the whole dissertation is concluded, and then the advice is given for the next step work.更多还原