Smart Home System Design Based On Home Assistant Indoor Air Monitoring And Fall Recognition

With the advent of 5G era,which makes the interconnection of everything possible,and along with the rise of artificial intelligence,indoor living environment starts to move from traditional electrical equipment to home automation and intelligence,smart home gradually becomes the focus of people’s discussion.Smart home systems can improve people’s living environment,facilitate the control of indoor electronic devices,collect indoor environmental information,and ensure the safety of indoor living environment.Bad home conditions can cause various respiratory diseases,and some unpredictable emergencies also seriously threaten people’s life safety,such as the use of water,electricity,gas and unpredictable falling behavior.Therefore,in order to meet the human demand for efficient and safe home living conditions,and to address the shortcomings of smart home environment monitoring products and the use of smart home devices in the current market,a smart home system that combines indoor environment monitoring with fall recognition detection is designed.First,the thesis based on the main control board Raspberry Pi 4B and microcontroller ESP32 to complete the formaldehyde,carbon dioxide,temperature,humidity,PM2.5,PM10,TVOC,PM10,TVOC in real time,and realize the real-time transmission of environmental monitoring data through the open-source Home Assistant smart home system,and set the automatic trigger of the system to give early warning notification when the value of environmental parameters involving safety exceeds the threshold value,and realize the synchronization of information from the web terminal and the mobile terminal.The information on the web and cell phone is synchronized with the warning.At the same time,in view of the current demand of indoor environment monitoring and the problems of smart home in the market,and considering the increase of aging and the increase of empty nesters in China,this paper designs a fall recognition detection based on monitoring indoor environment.The paper proposes a fall recognition detection algorithm based on Google’s open source posture prediction model Move Net,which improves on the original posture recognition algorithm to increase the accuracy of unpredictable fall detection.The algorithm is deployed with the main control board Raspberry Pi 4B as the hardware experimental platform,collecting image data through the camera,feeding the fall detection results into the Home Assistant smart home system,thus combining the smart home system with the fall recognition detection and effective automatic alarm,and finally transmitting the fall recognition image information and message warning to the client in the first place.After completing the requirements of the above functions,the thesis has tested the application functions of the smart home system developed based on Home Assistant,and the test results can achieve the expected effect of the design of this thesis,and the functions and indicators can meet the design function requirements of this thesis,and at the same time reduce the design and development cost,which has certain landing application value and development for the future design and development of smart home system.The test results can meet the functional requirements of this thesis.