Research On Home Appliance User Behavior Prediction Technology Based On Weighted Markov Chain And Deep Learning

In recent years,intelligent home furnishing equipment is more and more popular,reflecting people's higher and higher requirements on the quality of life,household appliance industry closely related to people's home furnishing life needs to seek new technological breakthroughs to meet the requirements of the market.In university-enterprise cooperation project of user behavior to predict water heater as the breakthrough point of intelligent home appliances research,electric water heater is used as user daily behavior data as the basis,the history of the Markov chain model and machine learning algorithm is adopted to predict user behavior,which can be represented by electric water heater according to the forecast results of decision-making control,intelligent home appliances equipment improving user experience at the same time,reduce energy consumption.This paper studies the prediction technology and application of household appliance user behavior.The main contents and innovations of this paper are as follows:(1)Based on the actual behavior data of smart home appliance users in enterprises,the user behavior characteristics obtained from the analysis are taken as the reference benchmark,and the users' behavior behaviors are graded randomly according to the regularity of user behaviors,and the corresponding user behavior data is generated through simulation.(2)Aiming at the problem that traditional rule-based user behavior prediction methods have been unable to meet the needs of users,combined with the development status of singlemachine intelligent home appliances,markov model prediction method based on statistical principles was adopted to predict user behavior.Combination directly using the weighted markov chain to predict user behavior,puts forward the weighted markov chain model and hidden markov model of combination forecast method.In the use of hidden markov model to forecast the user behavior,combined with similar behavior data filtering way to select user behaviour on the similar historical data.The last to have different behavior characteristics of user behavior data to forecast experiment,verified the reliability of the combination model prediction method is put forward in this paper.(3)According to the actual user behavior due to individual differences in the scene,interfere with each other,and all kinds of emergencies,the traditional user behavior prediction model meets the requirement of user behavior prediction accuracy problems.Combining with the development of intelligent home appliances based on cloud server status quo,the prediction method based on deep learning to predict user behavior research.In the research of deep learning user behavior prediction algorithm,two models of deep feedforward network and long and short time memory network are compared.The experimental results show that the effect of predicting user behavior based on long and short time memory network model is better.(4)Based on the research of user behavior prediction algorithm,the research results are applied to the intelligent water heater products developed by enterprises based on user behavior,and the actual energy-saving effect of the products developed is verified according to German VDE certification standard and European energy efficiency certification standard.The results show that the intelligent water heater products with user behavior prediction technology have obvious energy saving effect and have successfully passed the certification of the above two energy saving standards.