基于物联网的智能家居语言输入端控制系统设计

Design of control system for smart home language input terminal based on Internet of things

采用云智控智能家居系统控制家居设备时,通过遥控器传输红外码控制家用红外设备,遥控器不具备语音控制功能,仅能手动控制,智能控制效果差。设计基于物联网的智能家居语言输入端控制系统,用户通过感知层中的激光传感器模块采集家居现场感知数据后,采用语言输入端APP将用户语音指令通过ZigBee无线数据传输模块传输至网络层的云端识别模块中进行指令识别,下达指令至应用层中的STM32控制板响应相应的控制程序;通过梅尔频率倒谱系数提取智能家居语言信号特征参数,采用DTW算法将语言信号特征参数变换成语言信号特征向量,依据该向量识别语言音节提取关键词,实施相匹配的控制命令,完成智能家居声控。实验结果表明,所设计系统用于智能家居声控时,语言指令识别率高达92%、控制准确率均值为99.46%、整体控制耗时仅有1.1 s,且用户体验满意度100分,智能控制效果佳。

When using Yunzhi intelligent home control system to control home equipment,the home infrared equip-ment is controlled by transmitting infrared code through the remote control.The remote control does not have voice con-trol function and can only be manually controlled and intelligent control effect is poor.Thus,a intelligent home language input control system based on the Internet of Things is designed.After users collect home field perception data through laser sensor module in the perception layer,The language input APP is used to transmit the users’voice instruction through ZigBee wireless data transmission module to cloud recognition module of network layer for instruction recogni-tion,and the STM32 control board in the application layer is issued to respond to the corresponding control program.The language signal feature parameters are transformed into language signal feature vectors by using Merle frequency in-version coefficient,and the language signal feature parameters are transformed into language signal feature vectors by u-sing DTW algorithm.The experimental results show that when the designed system is used for intelligent home voice control,the recognition rate of language instructions is as high as 92%,the average control accuracy rate is 99.46%,the overall control time is only 1.1s,and the user experience satisfaction is 100 points,which prove that the intelligent control effect is good.