基于语音识别实现智能家居的控制—系统整体整合与调参.docxVIP

PAGE 20 基于语音识别实现智能家居控制 ——系统整体整合与调参 摘 要 为配合各种终端设备较高的智能程度和集成程度，舍弃以往使用按键及菜单操作对设备进行控制的方式，让检索信息和操作命令的方式变得更加便捷，语音识别技术逐渐进入人们的视线。智能家居不仅能够为人类生活提供方便快捷的服务，而且能够在需要的时候提供更为人性化的服务，从而使得智能家居在现代化发展领域占据非常重要的地位，因此，将智能家居与语音识别技术结合变成了人们的迫切需求。本文着重从语音识别的模型设计及信号的传输与控制等方面介绍了本次基于语音识别实现智能家居控制的研究。 语音识别的过程可分为模式识别和匹配两个模块，本设计语音识别部分运用人工神经元网络模型，对捕获的语音信号进行初步的判断处理，将音频和文字数据进行格式转化，并且将语音信号特征点进行提取，在进行初步判断基础上建立和训练模型，得到相应语音识别的模板，等待下一次捕获语音时，将获取到的语音信号与之前的语音信号比对判断之后，结合建立的相关函数及相关算法模型，获取到匹配率最高的模板，并得到一个较高的识别率。 除语音识别模块外，此次设计还有语音录入模块及信号输出控制模块，用以完整实现通过语音识别实现控制智能家居的最终目的。语音输入模块通过调用麦克风录音进行语音录入，信号输出控制模块则是通过网络通信发送信号到另一设备来实现语音控制。 关键词：语言识别，神经网络，信息传输，训练模型，智能控制 Control of Smart Home Based on Speech Recognition Abstract In order to cooperate with the higher intelligence and integration of various terminal equipment, the way of using buttons and menus to control the equipment is abandoned, and the way of retrieving information and operating commands becomes more convenient. Speech recognition technology gradually enters peoples attention. Smart home can not only provide convenient and fast services for human life, but also provide more humanized services when needed, which makes smart home occupy a very important position in the field of modern development. Therefore, the combination of smart home and speech recognition technology has become an urgent need of people. This paper mainly introduces the research of smart home control based on speech recognition from the aspects of the model design of speech recognition, signal transmission and control. The process of speech recognition can be divided into two modules: pattern recognition and matching. In the speech recognition part of this design, the artificial neural network model is used to preliminarily judge and process the captured speech signal, transform the format of the audio and text data, extract the feature points of the speech signal, establish and train the model based on the preliminary judgment, and get the corresponding speech recognition For other templates, wh