Research On Intelligent Modeling Method Of Building Indoor Thermal Environment

Heating is common for most parts of the north in winter,and people have higher and higher demand for comfortable and warm environment,therefore,the comfort and the decrease of energy consumption are the hotspot in the field of intelligent building research.Indoor temperature is the most intuitive evaluation factors for indoor comfort.At present,for most parts of the north in winter,the indoor thermal environment is mainly adjusted by heating,which cannot meet the requirements of people for comfort,and high energy consumption is included.Parameters that influence indoor thermal environment have characteristics of nonlinear,strong coupling and environmental complex,so it is difficult to establish the system model of indoor thermal environment and achieve the intelligent control of thermal comfort.Based on the requirements of comfort,energy-saving and health of indoor environment,the modeling of indoor thermal environment is studied in this paper.The research contents of this paper are as follows:(1)According to references,analyze the domestic and foreign research situation of indoor thermal environment and comfort building.Solutions for problems that training time of the model is too long,meteorological data is difficult to collect,and data inadequate exist are proposed in this paper.(2)For the problems that the overmuch complex indoor thermal environment parameters and data collection difficulty,an indoor thermal environment parameters data collection system is established to collect the heating data of indoor thermal system and conduct the simulation experiment,which solved the problem of data inadequacy and inaccuracy.(3)For the problem that the severe energy waste of heating in the north winter,this paper established an indoor spatial dynamic thermal model based on the prior knowledge of building thermal environment,the data collected in the field and the existing neural network modeling methods of building interior space.The BP neural network model,that takes the heat generated in the hot area as the input,the indoor temperature value as the output and the microclimate parameters as the interference,solved the problem of energy wasting for overmuch indoor heating.(4)The average thermal sensation index(Predicted Mean Vote,PMV)was selected as the evaluation index to characterize the thermal response of human body,the data acquisition platform was set up,the data were collected,and the PMV value was calculated by PMV formula.Facing the indoor thermal environment,PMV value intelligent prediction model is established based on BP neural network and PSO-BP neural network,which takes indoor temperature,humidity,radiation temperature and wind speed as inputs,and PMV value as output.The feasibility of PMV intelligent prediction model is verified by a large number of simulation experiments.