【作者】 许冰；

【导师】 秦贵和；

【作者基本信息】 吉林大学 ， 计算机应用技术， 2008， 硕士

【摘要】 随着车载信息设备的不断增多,面向媒体的系统传输协议MOST越来越受到汽车生产商的重视,如何方便的访问MOST网络信息设备更是当前汽车电子研究的一个热点。本文通过运用MOST网络技术和语音识别技术,实现了基于语音接口的MOST网络设备控制。本文对MOST网络音频数据和控制数据的传输原理进行分析,利用MOST网络收发器OS8104和微控制器89S51设计了MOST网络节点,搭建了一个典型的MOST音频网络,并对语音识别芯片RSC300进行了学习和研究,在MOST音频网络上添加了语音接口节点。文中详细介绍了网络节点的软硬件设计,提供了微控制器89S51、MOST网络收发器OS8104、语音识别芯片RSC300和光驱、麦克风等元件的连接电路,以及OS8104路由表配置、光驱控制器、语音训练及识别等程序流程图。本文成功的实现了通过语音接口控制MOST网络设备的功能,验证了MOST网络在保证传输音频数据的高速、可靠和实时性的同时,可以互不干扰的传输控制数据的特性。更多还原

【Abstract】 With the development of vehicle electronic technology, the vehicleinformation process impel deeply. More and more electronic devices andinformation devices come into our life. The appearance of network makesthe in-vehicle information devices get rid of traditional wiring harness, andit will develop to distribution and intelligent. How to visit the vehicleinformation equipment conveniently is becoming key point in automotiveelectronics field. By using speech recognition technology, we have realizeda MOST audio network with speech interface which is based on MOST datatransferring principle. Based on the current experiment condition, we choosesome chips, decide how to connect those chips and embed the softwareinside, and debug the modules. At last, we realize the function that controlMOST network equipment by speech commands.BMW, DaimlerChrysler, Harman/Becker and OASIS Silicon Systemsfounded the MOST Cooperation in 1998. Since then the MOST Cooperationestablished the MOST Specification. They have the target that makes theMOST Specification develop to vehicle multimedia network standard withhigh performance and low cost. MOST is used to multimedia informationsystems. It can transfer synchronous data, asynchronous data and controldata at the same time. MOST bus has only one master. Synchronous datatransferring uses way of circuit-switch. In contrast with CAN bus,asynchronous data and control data transferring use the address to send orreceive data. The bandwidth of the MOST bus’s first generation is 25Mbps.This version of MOST technology can support up to 15 uncompressed,stereo CD-quality audio channels, up to 15 MPEG1 audio/video channels,but now uncompressed video can not be transferred.Because the technology is mature overseas, and many kinds of in-vehicle products adopt MOST interfaces. But we are dropped behind inthis field. Studying the technology is meaningful to the industry ofautomotive electronics.Now the control of in-vehicle electronics equipment uses the traditionalhuman-machine interface which is based on our eyes and hands. It isinconvenient and unsafe to the driver. Using speech interface, we are releaseour eyes and hands from controlling the in-vehicle equipment. Therefore,this way of human-machine interface is very safe and convenience. And ituse less space than any other interface ways. Speech is the most natural wayof interacting between human beings. With the development of speechrecognition and speech synthesis, speech interface will take the place of theway of interface using eyes and hands. The interface based on speech willlead the human-machine interface which is used to in-vehicle electronicequipment.There are four nodes in the system: audio node, amplifier node, controlnode and speech interface node. The audio node that provides audio data forthe system is consisting of audio coder, CD-ROM and CD-ROM controller.The amplifier node which consists of audio decoder and speaker is the audiointerface to users. That means the user can hear the music which come fromCD-ROM. By pressing its keys, the control node can control CD-ROMwhich is part of audio node. Speech interface node can control CD-ROM byspeech commands which will control the CD-ROM to play, stop, previousand next. The four nodes make up a MOST network which is ring topology.They distribute at different position on the MOST network.In the MOST system, we use the following circuit:Audio node: MOST network transceiver OS8104, microcontroller89S51, CD-ROM controller, CD-ROM, coder CS8414.Amplifier node: MOST network transceiver OS8104, microcontroller89S51, speaker, decoder CS4334. 89S51, keys.Speech interface node: MOST network transceiver OS8104,microcontroller 89S51, speech chip RSC300, program memory 29E020,data memory 24C65, microphone.With respect to the ISO communication model, MOST networktransceiver OS8104 chip supports the Data-Link layer. The microcontroller89S51 supports the Network layer through the Application layer. Thesoftware programs of MOST nodes include node initial program, speechtraining program, speech recognition program, control data send/receiveprogram, CD-ROM controller program, simulating I2C interface programand serial port transferring program. Node initial program will lead the nodeto work. Speech recognition program convert the speech commands tocontrol data. Control data send program will send the control data to theother nodes. CD-ROM controller program will make CD-ROM to executeinstructions which is from control data. Simulating I2C interface program isused to communicate microcontroller and MOST network transceiver. Serialport transferring program is used to debug other programs and watch overthe speed of them.We realize the control of MOST network equipment based on speechinterface. MOST network is a new way to solve the problem that transferaudio data in vehicle multimedia system. Speech interface is a kind ofhuman-machine interface which is safe and convenience. We have learnedthe principle how to transfer all kinds of data by realizing MOST networkfunction. By analyzing speech recognition technology, we convert thespeech commands to MOST control data. Synchronous data can betransferred safely and speedily in real time. At the same time, control datawill be transferred to the given node. Both the two kinds of data transferringcan be at the same time and can not interfere with each other. This paper which is about MOST network will become a foundation in theory andpractice. In the future, we may consider the video data transfer technologyand realize MOST audio/video network.更多还原