Research On Thermal Comfort Of Smart Home Based On Adaptive Method

The intelligent environment measurement and control system is an important branch of the smart home.At present,most of the intelligent environmental measurement and control systems are through real-time monitoring of the indoor environmental factors,and then adjust the indoor environmental factors with human instruction to achieve a comfortable living condition.Indoor living environment is the most frequent living environment in people's life.A good indoor environment can make people relax,concentrate and improve their work efficiency.On the contrary,the cold and hot indoor environment can make people restless,difficult to meditate,and even affect physical and mental health.Therefore,the study of the comfort of indoor environment is becoming more and more important.This paper mainly focuses on the research of the environmental thermal comfort in the intelligent home environment measurement and control system,introduces the shortage of the current environmental measurement and control system to control the thermal comfort of the environment.According to the traditional control strategy,this paper studies the thermal comfort of the environment based on the adaptive method,and introduces the interior thermal comfort.The capacity,including the six main factors of thermal comfort and thermal comfort evaluation method,selected the thermal comfort evaluation model used in the Research Institute,studied the influence of environmental variables on the environmental thermal comfort to select the system control variables and the system control mode,and set up a theory for the establishment of the thermal comfort control system.Basics?The research on the selected thermal index PMV evaluation model,because the model involves many parameters,the equation is complex,the iterative type is very obvious,and some parameters can not be measured directly and directly.It can not meet the requirements of different people for thermal comfort.So in view of these defects,the relationship between neural network and thermal comfort evaluation is established,and the application of traditional BP network in comfort evaluation is discussed,and the simulation model of comfort evaluation and experimental verification are established.Because the traditional feedforward neural network(BP,RBF,etc.)has no good adaptive ability in the various indoor environment,it can not be used well in the on-line real-time measurement and control system.Therefore,the adaptive neural network algorithm ART2 is applied to the indoor thermal comfort evaluation modeling,and the network has good online learning.According to the characteristics of real-time response and automatic recognition,and in view of the shortcomings of the ART2 network,the improved ART2 algorithm has been proposed to make the network have good prediction ability.The experimental results show that the ART2 network has achieved good prediction accuracy and convergence time.Finally,the results of these experimental models are discussed and analyzed,and the advantages and disadvantages of different network prediction models in training time and model accuracy are analyzed,and the feasibility analysis of the application in the comfort measurement and control system is also analyzed.