基于压电薄膜的车型分类研究

发布时间：2018-09-08 19:08

【摘要】：中国社会主义经济持续快速的增长,以及人民的物质生活水平不断提高,带动了汽车产业的发展,汽车的购买量不断增加,拥有私家车已经成为了一个非常普遍的现象。这一方面的确给人民的日常生活带来了极大地便利,同时推动了社会经济的进一步发展。但另一方面也产生了一系列的社会问题,例如交通环境的日益恶化、重大交通事故率不断升高、城市道路交通日益堵塞。同时,受利益地驱动以及运输竞争关系的影响,道路上行驶的超载和超限的运输车辆逐年增多,其对道路造成的破坏以及其他经济社会损失令人触目惊心。对现有的车型分类方法进行了全面系统的研究,分析了各种方法的优缺点的基础上,提出了本文的方案。本文的主要工作有以下几个方面:(1)本文设计了基于压电薄膜的车辆识别系统,通过高灵敏度的压电薄膜轴的合理铺设,尽可能准确地测量数据。经过信号增强和滤波之后,整理出车辆的轴距、轴载、轴数、轮数、车重等参数。(2)分析整理车辆参数对于车型分类的作用,选取对分类有用的特征。本文重点探索了不同车型,不同轴数车辆的轴荷分配情况。通过轴载特征的引入有效提高车型识别率。(3)为了尽可能的加强车辆特征的使用,本文设计了两层分类器,第一层分类器利用明显特征对车型进行完全准确的分类,将可能的车型限定在一定的范围之内;二层分类器是在分析支持向量机各个核函数和参数选取基础上设计的精细分类器,利用该分类器在第一层的准确判别基础上进行进一步的细分。实验表明,本文设计的两层智能分类器,可以有效利用压电薄膜测得的数据,同时,文中重点考察的轴载数据在分类器中起到了较好的分类作用。
[Abstract]:The sustained rapid growth of China's socialist economy and the continuous improvement of the people's material standard of living have led to the development of the automobile industry and the increasing purchase of cars. Owning private cars has become a very common phenomenon. On the one hand, it brings great convenience to people's daily life and promotes the further development of social economy. But on the other hand, there are a series of social problems, such as the worsening of traffic environment, the increasing traffic accident rate and the increasing congestion of urban road traffic. At the same time, under the influence of benefit driving and transportation competition, the number of overloaded and overloaded transport vehicles on the road increases year by year, and the damage to the road and other economic and social losses are shocking. On the basis of analyzing the advantages and disadvantages of the existing vehicle classification methods, the paper puts forward the scheme of this paper. The main work of this paper is as follows: (1) the vehicle recognition system based on piezoelectric film is designed in this paper. Through the reasonable laying of high sensitivity piezoelectric film shaft, the data can be measured as accurately as possible. After signal enhancement and filtering, the vehicle wheelbase, axle load, axle number, wheel number, vehicle weight and other parameters are sorted out. (2) the effect of vehicle parameters on vehicle classification is analyzed, and the useful features are selected. This paper focuses on the distribution of axle load of different vehicle models and different axle numbers. In order to strengthen the use of vehicle features as much as possible, a two-layer classifier is designed in this paper. The first layer classifier uses obvious features to classify vehicle models completely and accurately. The two-layer classifier is a fine classifier designed on the basis of analyzing the kernel function and parameter selection of support vector machine. The classifier is used for further subdivision on the basis of the accuracy of the first layer. The experiments show that the two-layer intelligent classifier designed in this paper can effectively utilize the measured data from piezoelectric film. Meanwhile, the axial load data studied in this paper play a better role in classifying the classifier.
【学位授予单位】：天津大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：U495;TP391.41

【参考文献】