N-pod多品牌智能家居语音控制设备的设计与实现

人工智能养老机器人有助于解决当前日趋严峻的居家养老问题,但各品牌智能家居设备互不兼容,成熟的机器人产品又欠缺老人语音语义识别,缺乏自学习能力且控制响应速度慢、灵敏度不高。针对上述问题,采用Tornado框架和Home Assistant框架有效整合市场主流智能家居设备,在主机和人形机器人上构建N-pod主机,使用深度卷积神经网络实现语音识别,使用粒子群优化智能算法对采集的老人身体指标、居家环境信息实施分簇处理,形成语音控制多品牌智能家居设备的控制解决方案,达到千人千面的养老定制需求。经过实验和实际应用,该设备适用于居家养老人群,语音识别的正确率、控制指令的正确率和响应灵敏度均高于同类产品,具有较大的经济和社会效益。

Design and Implementation of N-pod Multi-brand Smart Home Voice Control Equipment

Artificial intelligence old-aged robot is helpful to solve the increasingly serious problem of homebased old-age care.However,various brands of smart home equipments are incompatible with each other,and mature robot products lack speech and semantic recognition for old people,lack self-learning ability,and have slow control response speed and low sensitivity.To solve the above problems,the Tornado framework and the Home Assistant framework were used to effectively integrate mainstream smart home devices in the market,N-pod host was built on the host and humanoid robots,deep convolutional neural networks was used to achieve speech recognition,and particle swarm optimization intelligent algorithms was used for clustering the collected elderly physical indicators and home environment information,a voice control solution for multi-brand smart home devices was formed which meets the needs of thousands of people for the elderly.Through experiment and practical application,the device is suitable for home-based elderly people,and the accuracy of voice recognition,the accuracy of control commands and the response sensitivity are higher than similar products,which have greater economic and social benefits.