The Control Platform Design Of Intelligent Home Based On Voice Interaction

Voice recognition technology is a kind of technology which is used to translate the voice signals into the computer language.As research continues,a series of developments on voice recognition technology have been gotten and the research results have been applied in many aspects such as mobile application,wearable device,the Internet of things.The intelligent home,based on the Internet of things,is one of them.This subject focuses on the problems that the control system of the intelligent home,developed by the home appliance enterprises,are not only expensive,but also only can be used in single grand products.In this thesis,the former home appliances avoid an elimination and play an important role in intelligent home system with the help of the control platform of intelligent home based on voice interaction.This thesis focuses on the simulation of the voice recognition and the platform design of intelligent residential system based on voice interaction.In this thesis,the voice recognition algorithm is simulated by Matlab.The methods to detect the endpoint of voice signal combined the short-term amplitude with short-time zero-crossing rate.According to the characteristics of human audio system,we can figure out the Mel Frequency Cepstral Coefficients(MFCC)in each frame of voice as soon as the effective voice signal is detected.Then,Dynamic Time Warping(DTW)is used to make a comparison between the voice signal and the feature template,and finally the voice recognition results can be get.Further more,the hardware and software design of intelligent residential system based on voice interaction have been accomplished,including security module,lighting control,environmental monitoring and infrared transceiver module.People can monitor or control all the subsystems of the intelligent home through the voice instructions and it makes much sense of the promotion of voice interaction technique used in the intelligent housing system.