【作者】 辛忻；

【导师】 罗霞；

【作者基本信息】 西南交通大学 ， 安全技术及工程， 2005， 硕士

【摘要】 近年来,我国经济在快速发展的同时,伴随着机动车数量的激增和城市人口的膨胀,交通拥堵、交通事故和交通引起的环境污染问题也越来越明显地影响着人民的生活和制约着经济的发展。在我国资源、环境矛盾日益突出的今天,城市道路设施的增长空间极为有限,这一现状决定了解决城市交通问题必须要在合理规划建设道路的同时大力发展智能交通运输系统(ITS),智能交通运输系统可以看作是一个互动的平台,将运行的车辆进行实时监控、合理分流以此提高交通运输效率。目前应用的主要监控手段是在城市主要道路上设置足够数量的车辆检测器,以便向监控中心实时发送各路段的车流量信息,从而实现高效的交通管理,已成为解决交通问题的必然趋势。然而,现有的车辆检测器,由于种种原因,尚难以满足沿道路大量设置的要求。 本文论证了车辆声音识别技术的可行性,分析了车辆声音识别的数学模型,制作了以车辆声音识别技术为基础的车辆检测器雏形机并进行了一系列的实地实验。主要内容包括以下几个部分。 (1) 对车辆声音识别技术的研究在语音识别技术的基础上起步。通过对语音识别的数学模型的分析,将隐马尔可夫模型修正后用于车辆声音的识别。 (2) 制作了以车辆运动声音识别技术为基础的车辆检测器雏形机并进行了一系列的实地实验。在文中介绍了该雏形机的结构与实验过程,并分析了实验结果,解决了雏形机存在的部分问题, (3) 该车辆检测器雏形机可全天候工作、便于大量设置且有利于环保,而且具有体积小、重量轻、造价低和可靠性好的特点,具有一定的可行性。更多还原

【Abstract】 Nowadays, accompanied with the quick increase of economy in our country, the automobile amount and population increase rapidly and the environmental contamination problem caused by the traffic jamming and traffic accident significantly influenced our life and economy development. For the prominent contradiction of resources and environment and the limitation of space for increasing road facilities, city’ s traffic problems can only be solved according to two principles: rational planning of road construction and active development of intellect traffic system (ITS). ITS is an interplay platform, it improves the traffic effect through in-situ monitoring and rational branching of vehicles. At the present time, the traffic condition is monitored mainly by setting adequate vehicle monitors. By this means, traffic information can be sent to the monitoring center so that the management of traffic condition can be optimized. However, existing vehicle monitors are not enough to meet the need.In this thesis, the feasibility of the acoustics recognition technique of vehicle’ motion was investigated; its mathematical model was analyzed, a base machine was made based on this technique and series experiments had been performed upon it. The thesis including the following sections:(1) Acoustics recognition technique of vehicle’ motion can be built upon the base of voice recognition technique. It was found that implicit Markovian model could be well applied to acoustics recognition technique of vehicle’ motion.(2) A base machine of vehicle monitor was built based on the acoustics recognition technique of vehicle’ motion. Series experiments were carried out on it. The structure of base machine and experimental process were introduced, and experiment results were analyzed carefully. Some problems of this base machine were solved.(3) The lab-constructed base machine of vehicle monitor can work更多还原