嵌入式无线语音控制系统的设计与实现

发布时间：2018-03-06 03:28

本文选题：嵌入式系统　切入点：无线控制　出处：《郑州大学》2016年硕士论文　论文类型：学位论文

【摘要】：让机器具有“五官”能力在这个信息化的时代已逐渐成为现实,计算机技术及传感器技术的快速发展和广泛应用,使得机器能够“感触”世界和执行人类的一系列控制命令。让机器识别人类的语言,“听懂”人类的语音并流畅的做出应答,不仅是智能工业的一个重要研究领域,也是智能生活人性化的一个研究方向。德国工业4.0和中国制造2025的提出,标志着智能工业和创新发展渐成趋势,无线语音控制作为智能工业的重要一环,显示出了重要的研究价值。机器能否对语音进行有效识别是系统实现无线语音控制的关键,目前,语音识别的研究已有近七十年,取得了长足发展。本文对语音识别、无线控制及嵌入式系统应用进行了研究:运用MATLAB仿真了语音信号的频域分析、时域分析、倒谱域分析等特性,在神经元基础理论和BP算法的基础上,探究了语音识别的基础理论和数学模型;结合无线通信系统的实时应用,选取NRF24L01+芯片对数据信号进行了无线发送与接收;利用SPCE061A芯片的语音辨识功能,实现了特定人语音识别的信号输入;结合智能工业背景和实际应用需要,选取STC89C52芯片搭建了无线控制系统的执行电路,以期实现无线语音微控制系统在工业控制上的应用。在嵌入式无线语音控制系统的研究中,机器对语音的识别是研究重点,对语音信号的识别,涉及语音库、预处理、特征参数、特征变换及识别的模型建立。论文从语音信号的基本处理和算法模型出发,对语音识别基本原理展开研究。其次论文研究了无线通信系统中无线传输问题,选取NRF24L01+芯片搭建无线系统,系统通过SPI串口实现无线通信。论文基于SPCE061A和STC89C52实现硬件电路的设计,基于unSP IDE2.0.0和Keil uVision4 IDE开发环境实现软件程序的仿真。基于无线语音微控系统的整体搭建测试,对系统进行了验证。
[Abstract]:In this information age, the rapid development and wide application of computer technology and sensor technology have enabled the machine to have the capability of "five features". Enabling machines to "feel" the world and execute a series of human control commands. It is not only an important area of research in the intelligent industry that machines recognize human languages, "understand" human speech and respond fluently. It is also a research direction of humanization of intelligent life. The proposal of German industry 4.0 and Chinese manufacture 2025 marks the gradual trend of intelligent industry and innovation. Wireless voice control is an important part of intelligent industry. It shows the important research value. Whether the machine can recognize the speech effectively is the key of the system to realize the wireless speech control. At present, the research of the speech recognition has been nearly 70 years, and has made great progress. The application of wireless control and embedded system is studied. The characteristics of frequency domain analysis, time domain analysis and cepstrum domain analysis of speech signal are simulated by MATLAB, based on the basic theory of neuron and BP algorithm. This paper explores the basic theory and mathematical model of speech recognition, selects NRF24L01 chip to send and receive the data signal wireless, and makes use of the speech recognition function of SPCE061A chip, combines the real-time application of wireless communication system, selects the NRF24L01 chip to carry on the wireless transmission and the reception to the data signal. The signal input of specific speech recognition is realized, and the executive circuit of wireless control system is built by selecting STC89C52 chip according to the background of intelligent industry and practical application. In the research of embedded wireless speech control system, the recognition of speech by machine is the focal point of the research. The recognition of speech signal involves speech corpus, preprocessing, and so on. Based on the basic processing and algorithm model of speech signal, the basic principle of speech recognition is studied in this paper. Secondly, the problem of wireless transmission in wireless communication system is studied. NRF24L01 chip is selected to build the wireless system, and the system realizes wireless communication through SPI serial port. The hardware circuit is designed based on SPCE061A and STC89C52. The simulation of software program is realized based on unSP IDE2.0.0 and Keil uVision4 IDE development environment, and the system is verified based on the whole building and testing of wireless voice microcontrol system.
【学位授予单位】：郑州大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TN912.3;TP273

【相似文献】