Research On Personalized Recommendation Method Of Dynamic Supply Air Mode In Indoor Thermal Environment

In recent years,studies on air-conditioning system dynamic air supply,personalized air supply and other air supply strategies were conducted with the numerical simulation method.Compared with the supply air mode of the traditional air-conditioning system,it could effectively lower air-conditioning operation energy consumption and provide a good indoor thermal comfort environment.However,there are limits for the regulation mode of the dynamic supply air mode of indoor air conditioning system,and it is not good for the optimization and recommendation effect of personalized supply air mode.Therefore,it is difficult to achieve the ideal effect of providing indoor thermal environment and energy-saving of air conditioning system.To achieve one better recommendation effect,this study structures personalized supply air mode.The main contents are as follows.Firstly,the physical simulation model of the actual building space was established,and the simulation cases designed by design-expert software were conducted numerical simulation with CFD software.To validate the reliability of the CFD simulation,there was a related analysis with the data such as indoor air temperature and indoor air velocity between CFD simulation and field experiment.Secondly,based on the ventilation performance data and environmental parameters collected from the simulation experiment,the surface model and multiple linear regression model were used to design two supply air modes,respectively.Considering the energy-saving of air-conditioning system,the response surface model and multiple linear regression model method were used to train two dynamic supply air modes,respectively.For multiple linear regression model,it is more reasonable to build dynamic supply air mode through the error of analysis.Based on the prediction data of the linear programming model,the TOPSIS method and TOPSIS method based on PMV strategy were used to determine the optimal comprehensive ventilation performance for different outdoor temperatures.And the basic requirement of the indoor air conditioning system is to meet the thermal comfort requirement of the human body.According to the thermal comfort requirement and the above-mentioned analysis,it was more reasonable to use TOPSIS method based on PMV strategy.Finally,based on the ventilation performance of decision-making and its corresponding environmental parameters,BP neural network algorithm was used to establish the personalized recommendation model of dynamic supply air mode.The input neurons of neural network were outdoor temperature and metabolic rate,and the output neurons were supply vane angle,supply air temperature,supply air rate,PMV,LMAA,draft rate or energy consumption.According to the user’s metabolic rate,the recommended model could recommend the best comprehensive ventilation performance and the corresponding supply air parameters.The actual indoor air conditioning system was varied as the outdoor environment varied when the thermal requirement of occupant was achieved.