The Design And Implementation Of Prediction-model-based Environmental Data Acquisition And Monitoring System

Through the comprehensive application of Internet of Things technology,Internet technology and machine learning algorithm,this paper designs and implements indoor and outdoor environmental data collection and monitoring system,which is convenient for users to intuitively understand the real-time status and historical changes of indoor and outdoor environment,and monitor existing data.The problem of predicting the function of single data is generally lacking in the system,and the application of the predictive model is proposed,which provides a reliable basis for further operation of the controllable variables to achieve environmental improvement.The system is realized by three parts: the acquisition end,the monitoring end and the prediction model.The environmental data collection end uses the Intel Galileo development version,the air quality sensor and the temperature and humidity sensor to collect indoor and outdoor environmental data,and completes the data transmission with the monitoring terminal through the wireless module;the environmental data monitoring terminal visualizes the received data.Display,and implement user operations with different permissions,mainly using the SSH backend framework and jQuery EasyUI front-end framework,using MySQL database for data storage,using Echart front-end plug-in to achieve real-time data and historical data chart display.The prediction model selects the indoor temperature as the prediction target,and uses Python's machine learning library Sklearn to analyze and process the collected environmental data and indoor temperature influence factors.The random forest algorithm and XGBoost algorithm are used to construct the prediction model separately.A comparative analysis of the forecasting effect and running time indicators,and selecting a model with better performance to apply to the system.The user can obtain the indoor temperature prediction value at a certain moment in the future,and use the prediction result as the basis for adjusting the environmental variables.This article provides users with a low-cost,easy-to-use environmental data-aware approach.The simple and effective function module reduces the accuracy loss of each link,and provides users with predictive functions and improvement methods for indoor environmental data.