Abstract: Aiming at the problems of high usage cost, low penetration rate, difficulties in intelligent interaction for the elderly and low information security in the currently marketed comprehensive smart home system, a voice control for smart home without networking is proposed using voice recognition, voice synthesis, infrared coding self-learning, wireless communication, double-layer protection mechanism, double mode switching, battery charging and discharging, voltage conversion and other technologies. The method has been tested to realize the voice control of smart home with the advantages of simple use, stable performance and low production cost.

Key words: speech recognition, speech synthesis, infrared coding self-learning, dual mode switching, double layer protection mechanism, wireless communication