Research On The Prediction Of Intelligent Home Environment Based On Raspberry Pi

In recent years,the penetration rate of smart home products is getting higher and higher with the development of smart home.Many scholars have entered this field and have conducted research on intelligent control,monitoring systems,hardware components,and algorithms used.Mainly,the Existing research results of smart home are mostly detection and control systems,real-time detection and intelligent control.Some environmental variables such as ambient light intensity provide immediate feedback for control.However,some environmental variables such as temperature require a certain amount of time for control feedback.Aiming at the problem of hysteresis of the adjustment of environmental variables,a method of predicting the smart home environment based on Raspberry Pi was proposed.This article predicts the light environment such as the light intensity and thermal environment such as temperature and humidity in the living room,by analyzes the changing trend and law of the predicted environmental factors to determine the basic forecasting model direction.Based on the analysis of common classical prediction algorithm models,and analyze their characteristics,advantages and disadvantages,we choose a prediction-based Granger causality test which is suitable for sequences.Then combining the Granger causality test model with polynomial regression analysis model,making it more suitable for prediction of home environment.This method takes the raspberry pi as the carrier,uses multi sensors to collect real time data of home environment,and gives a prediction conclusion on the changing trend of home environment.In this paper,the operation flow and technical principle of the whole method are expounded,and the prediction effect of this method is verified and compared.In this method,firstly,collect the historical data of home environment’s various factors,then analysis the historical data by improved Granger causality test,Granger causality test is a kind of causal analysis algorithm based on prediction.We improve the Granger causality test and combine the estimation model with the polynomial regression model.Then,referring to the predicted causal relationship between the environmental factors obtained based on the improved causality test,we get the regression equation of each home environment factor.Finally,in the actual prediction,the regression equation obtained is used as the basis of prediction.The parameters of household environment are collected in real time by sensors,and the prediction results are calculated based on the data obtained,and the prediction conclusions are given.At the end of the article,the method proposed in this paper is verified.After testing this method,good prediction effect can be achieved in the actual home environment prediction.