【作者】 张顺广；

【导师】 朱金陵；

【作者基本信息】 西南交通大学 ， 载运工具运用工程， 2003， 硕士

【摘要】 本文是以铁道部科技发展项目“提速列车优化操纵与机车仿真器研究开发”为依托，对机车仿真器的声音系统进行研究：第一，对机车仿真器中声音仿真系统的音效进行改进，增加3D音效，如声音的距离感、方位感、多普勒效应等，使得训练环境音响更加逼真；第二，实现机车仿真器中教员与学员间的语音网络通讯。 声音仿真系统以声音渲染的数学模型、人耳的听觉模型、计算机音频知识为基础。本文分析了实际机车运行过程中的声音情况，并结合机车声音仿真的控制方法与具体要求，比较Windows环境中各种声音媒体控制的方法和特点后，提出了采用DirectSound技术开发声音仿真系统的方法。探讨了利用DirectSound技术实现声音的混音、回放和3D效果的原理及处理过程，实现了机车声音仿真的3D音效和多普勒效应。 以语音压缩编解码技术、网络通信技术和多线程技术为主要理论基础，给出了教员与学员间语音网络通讯的具体实现方案：采用对等模式及UDP通信技术和简捷的数据压缩技术，对全双工声卡编程，在局域网内实现点到点的全双工的、实时的语音通讯。文章详细地讨论了语音网络通讯的采集模块、回放模块、通信模块和压缩解压模块的具体实现，并提出了一种自适应的静音检测算法和数据包乱序调整的策略。最后，分析了语音网络通讯的时延。更多还原

【Abstract】 The paper, which is based on the "Research and Development of the Speed-up Train’s Manipulation of Optimization and Train Simulator" project, concerns the sound system in the Train Simulator. In the paper, the sound system includes two parts contents: firstly, improving the sound effect of the sound simulation system in the Train Simulator and enhancing 3D sound effect, such as sense of distance, sense of orientation, Doppler Effect etc, to make the sound of training environment more verisimilar; secondly, realization of the speech communication in network between teacher and student in the Train Simulator.The sound simulation system is based on the mathematical model of sound rendering technology, the model of human hearing and the audio knowledge in the computer. The paper analyses the sound of train in fact, and combines with the requirement to the sound simulation of the Train Simulator. After evaluating all methods and traits of the controls to sound media in Windows system, the paper brings forward to use the DirectSound technology to develop the sound simulation system. At last, the paper discusses the principle and the process of disposal of using DirectSound technology to realize real time sound synthesization, playback and 3D sound effect, and realize 3D effect and Doppler Effect of the sound simulation.The paper gives a project to realize speech communication between teacher and student in network, which is based on speech’s compression and decompression technology, network communication technology and multithreading technology. The point-to-point full duplex and real time speech communication in network is implemented on LAN by using peer to peer model, UDP communication technology and simple data-compression method, and programming to the full duplex sound card. Then, the paper discusses the realization of the speech’s sampling module, playback module, communication module, compression and decompression module in detail, and gives a self-adaptive silence detection algorithm and a policy to make disorder data package in order. At last, the paper analyses the delay time of thespeech communication in network.更多还原