Research And Implementation Of Big Data Analysis And Intelligent Control Algorithm For Convalescent Housing

In recent years,the proportion of China’s elderly population continues to grow,and the problem of"empty nest elderly"living alone for the aged is prominent,which increases the social pressure.The smart pension model is the application of new information technology,relying on the Internet of things and big data,which is also a new direction to deal with the pension problem in China^[1].At present,most of the home-based elderly care is focused on the theme of intelligent monitoring,but lack of research on the elderly as the main body of smart home,the design of intelligent control system to meet the needs of elderly users.Sensors are deployed in mobile terminals and elderly care equipment.The elderly are connected with the living environment.The Internet of things can sense the physical signs and behavior dynamics of the elderly in real time to help the elderly to complete the action purpose and difficult behavior.In this paper,the wireless sensor network is constructed to collect the data and basic physiological parameters of the elderly at home.The acquired behavior information is cleaned,fused and stored.Through the improved k-means algorithm based on self-organizing neural network and the association rule algorithm with time constraint factor,the behavior analysis model of the elderly is established to assist the elderly to complete the intelligent equipment The collected physiological parameters are used to make a preliminary judgment on the health status of the elderly.To realize the comprehensive state perception of the elderly at home,improve the control of the elderly on smart home,and ensure the safety of home.The main points of this study are as follows:1.Through the research of various data mining and behavior analysis algorithms,select the data mining algorithm which is consistent with the actual needs.2.By comparing the advantages and disadvantages of Zig Bee,sub-g,Wi Fi and other common wireless communication technologies,build a hybrid network of sub-g and Wi Fi to collect and upload relevant data for the elderly at home.3.The improved k-means algorithm based on SOM is used for double-layer clustering of behavior data,supplemented by association rule algorithm with time constraint,to realize the analysis and prediction of elderly behavior data,and establish the habit behavior pattern of the elderly.4.Using B/S software architecture,the front-end part of the code implementation uses Java Script,Vue and other technologies,and the server uses Java and SSM（Spring+spring MVC+mybaits）framework to develop corresponding functions.