基于机动车音频信号的交通状态判别研究  

Study on Traffic Condition Discrimination based on Vehicle Audio Signals

作　　者：姚雪娇 蔡铭[2] YAO Xuejiao CAI Min(Shenzhen Urban Transport Planning Center, Beijing 100010, China Engineering College, Zhongshan University, Guangzhou 510006, China)

机构地区：[1]深圳市城市交通规划设计研究中心,北京100010 [2]中山大学工学院,广州510006

出　　处：《中国公共安全（学术版）》2017年第3期86-90,共5页China Public Security(Academy Edition)

基　　金：国家自然科学基金(11574407);广东省科技计划项目(2015B010110005);广州市科技计划项目(2017040220142)

摘　　要：道路交通流监测是智能交通发展的重要方向,道路的交通状态是驾驶员在出行中最为关注的信息。应用道路机动车的音频信息,判别交通状态是本文的研究内容。相比于现有的监测技术,音频信号处理技术有成本低、信息覆盖度高、不受光线影响等优点,是对现有监测手段的一种重要补充。在不同道路密度情况下,机动车运行所产生的噪声信息不同。本文以不同的道路交通状态下的音频信息为研究对象,运用车辆运行的音频信号特征进行交通状态的判别研究,通过对音频信号的时域与频域特征分析,对特征属性进行选择、提取,进而利用支持向量机(SVM)分类器对特征数据进行训练与识别分类。本文研究运用道路交通音频信号,判别三种道路状态:畅通、缓行、拥堵,测试样本准确率达到96.2014%。Monitoring road traffic flow is an important development direction of intelligent transportation system. And the road traffic condition is what drivers care most. The main content of this paper is to distinguish traffic condition by the information of road vehicle audio. Compared with current monitoring technology skills, audio signal has many advantages, such as lower cost, bigger coverage area, no light influence, and so on. It is a proper supplement to the existing monitoring methods. On different condition of road density, the noise is different. So this paper takes different road traffic condition＇s audio information as the object of study, and uses the audio signal characteristics to identify the traffic condition. By analyzing the time domain and frequency domain of audio signal, the feature attributes are selected and extracted, and then the support vector machine（SVM） classifier is used to train and classify the feature data. In this paper, based on the road traffic audio signal, to distinguish three kinds of condition： smooth, amble, road congestion, the test sample accuracy rate reached 96.2014%.

关 键 词：机动车音频 交通状态判别 时域频域分析 SVM 

分 类 号：D631.5[政治法律—政治学]