基于压电薄膜和射频识别的ETC系统研究

发布时间：2018-10-17 15:15

【摘要】：电子不停车收费系统(Electronic Toll Collection System，简称ETC)，是目前世界上最先进的路桥收费方式，无需停车即可顺利扣费。ETC具有通行效率高、快捷方便的优点，有效缓解了收费站拥堵的交通状况。计重收费作为遏制货运车辆超限超载和规范运输市场秩序的有效手段，目前已在全国很多省份普遍实施。传统的计重收费采用整车静态称重的方法，这在一定程度上影响了交通流的顺畅。为了提高计重收费的通行效率，本文对计重收费条件下的不停车收费进行了探讨。动态称重是计重收费条件下不停车收费采用的车辆称重方法，与静态称重相比，动态称重具有节约空间、通行效率高的优点。本文的主要研究工作如下： 1、深入研究了计重设备前置式和后置式两种车道布局模式的系统工作逻辑。 2、对压电传感器信号进行基于小波多分辨分析的去噪处理。压电信号中除包含变化缓慢的轴重信号外，还包含各种噪声信号。与轴重信号相比，噪声信号频率较高，可以利用小波多分辨分析滤除高频干扰信号。具体方法是：首先对压电传感器产生的称重信号进行多尺度小波变换，获得小波变换系数，然后选择合理的阈值，对包含噪声的高频小波系数进行处理，最后对信号进行重构，达到消噪的目的。 3、对压电信号波形进行奇异性分析。针对不规则行车行为，本文采用基于小波变换的奇异性分析和BP神经网络分类判别的方法进行识别。具体方法是：首先对压电信号进行多尺度小波变换，提取高频小波系数矩阵，在此基础上，采用相重构技术构造特征矩阵，，然后对特征矩阵进行矩阵分解，得到矩阵奇异值，最后将压电传感器AD值、称重现场的实测温度值、矩阵奇异值作为BP神经网络的输入，对车辆真实轴重进行有效识别。
[Abstract]:(Electronic Toll Collection System, is the most advanced road and bridge toll collection method in the world at present. ETC has the advantages of high traffic efficiency, fast and convenient, and effectively alleviates the traffic condition of toll station. As an effective means to curb the overloading of freight vehicles and standardize the order of transportation market, weight-calculating charge has been widely implemented in many provinces of China. The traditional method of weighing the whole vehicle by static weighing affects the smooth flow of traffic to a certain extent. In order to improve the efficiency of weighing charge, this paper discusses the non-stop charge under the condition of weighing charge. Dynamic weighing is a vehicle weighing method which is not used for charging under the condition of weighing and charging. Compared with static weighing, dynamic weighing has the advantages of saving space and high traffic efficiency. The main research work of this paper is as follows: 1. The system logic of front and rear lane layout of weighing equipment is studied in depth. 2. The piezoelectric sensor signal is based on wavelet multi-wavelet Denoising processing of resolution analysis. In addition to the slowly changing shafting load signal, the voltage signal also contains various noise signals. Compared with the axle load signal, the noise signal frequency is higher, and the high frequency interference signal can be filtered by wavelet multi-resolution analysis. The specific methods are as follows: first, the weighing signal produced by piezoelectric sensor is transformed by multi-scale wavelet transform, then the wavelet transform coefficient is obtained, then the reasonable threshold is selected to process the high-frequency wavelet coefficient including noise, and finally the signal is reconstructed. To achieve the purpose of noise elimination. 3, the singularity analysis of the waveform of EV signal is carried out. In view of irregular driving behavior, the singularity analysis based on wavelet transform and BP neural network classification and discrimination method are adopted in this paper. The specific methods are as follows: firstly, the piezoelectric signal is transformed by multi-scale wavelet transform, and the high frequency wavelet coefficient matrix is extracted. On this basis, the phase reconstruction technique is used to construct the characteristic matrix, and then the matrix is decomposed to obtain the singular value of the matrix. Finally, the AD value of piezoelectric sensor, the measured temperature value of weighing site and the singular value of matrix are used as the input of BP neural network to effectively identify the real axle load of vehicle.
【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

【相似文献】