【作者】 朱强；

【导师】 徐建闽； 林卫民；

【作者基本信息】 华南理工大学 ， 控制工程， 2013， 硕士

【摘要】 随着智能交通的发展，车辆自动检测技术作为智能交通系统的关键组成部分得到了越来越多的关注。为了给交通优化管理、基础交通建设决策等提供基础交通数据，车辆自动检测技术的深入研究对智能交通的发展具有重要意义。考虑到目前国内外在实际工程应用中的车辆检测识别的研究应用现状，本文基于低成本、低功耗、易实现、多用性的原则，在检测设备和检测方法上进行了较深入的研究分析及应用验证。本文分析了国内外车辆自动识别技术，尤其是车型自动识别技术的发展现状，提出基于磁阻效应的车辆检测技术。采用三轴数字型各向异性磁阻效应传感器作为检测器的传感器件，利用平均滑动波形处理以及自适应阈值滤波解决波形噪声问题，归一化提取标准车型特征向量和聚类分析，最后采用加权欧氏距离的方法来识别车型。按照车型自动识别算法的实现过程，本文设计了基于无线传感网络的车辆检测系统，提供大范围实时的车辆数据，为智能交通控制系统进行最优控制提供服务，以此达到公路交通的优化管理。最后，车辆检测系统在实际路况进行实地测试，并进行数据分析处理，查看统计报表，整体性能稳定可靠，能达到预期的设计效果。总的来说，本文设计的车型自动识别系统成本低，对硬件要求较低，同时车型检测算法易于在各类单片机实现，深入完善该系统将具有广阔的应用前景。更多还原

【Abstract】 With the intelligent transportation development, the vehicle automatic detectiontechnology as the intelligent transportation system is more and more attention. In order toprovide base traffic data for optimize traffic management and the decision-making, in-depthstudy on automatic vehicle detection technology development of intelligent transportation issignificance. Considering the vehicle identification application status on the current domesticand international engineering applications, this paper tests methods for a more in-depthresearch and analysis detection equipment and application authentication based on low cost,low power consumption, easy to implement, the principle usage.This paper analyzes the domestic and foreign automatic vehicle identification technology,especially the development status, and uses the vehicle detection technology base on magnetoresistive effect. To solve the noise problem waveform, this paper uses three-axis digitalanisotropic magneto resistive sensor as a detector of the sensors, and use the average slipwaveform for processing and adaptive threshold filtering, and finally using the weightedEuclidean distance method to identify vehicle models. In order to achieve optimalmanagement of road traffic, according to the implementation process of the automatic vehiclerecognition algorithm, this paper designs a vehicle detection system based on wireless sensornetworks which provides a wide range of real-time vehicle data, and services the intelligenttraffic control system for the optimal. Finally, the vehicle detection system is applied to theactual road conditions testing, and process data analysis in the vehicle detection system, andthe overall performance is stable and reliable which is able to achieve the desired designeffect. Overall, automatic vehicle identification system designed by the paper is low cost, lowhardware requirements, and is easy to use the vehicle detection algorithm for MCU, and itwill be broad application prospects after in-depth improvement.更多还原