Research On Monitoring Method Of Average Indoor Temperature In Smart Heating Buildings

With the gradual development of smart heating and the emergence of smart heating buildings,how to correctly and accurately obtain room temperature information is one of the important issues in achieving smart building heating.Due to the uneven temperature distribution inside the room and the uncertainty of the temperature field distribution,the layout of the traditional room temperature sensor is limited by space.Because of wiring and technical requirements,the temperature measured by the sensor can only reflect a certain space around the location of the sensor The average temperature information cannot accurately reflect the actual temperature of the room.This article investigates and summarizes the types,functions and usage scenarios of different room temperature sensors,and makes a basic summary of the use of sensor equipment.By means of field investigation,the installation situation of different types of room temperature sensors in several buildings in Harbin was recorded and sorted out.Six representative rooms were selected for on-site indoor thermal environment measurement and analysis during the heating period,and the measured data provided verification for the numerical simulation experiments in this article.Orthogonal experimental design was used to select 6 factors that have an impact on the indoor thermal environment,including orientation,outer wall thickness,window-to-wall ratio,floor,presence or absence of cold gables,and solar radiation.The CFD software Flo Vent was used to analyze 16 types of work.Simulate the situation to obtain the indoor temperature field distribution.On this basis,extract the different near-wall air temperature distribution data and do the corresponding processing,and use the data visualization to convert the plane temperature data into pictures,and then use the perceptual hash algorithm to calculate the similarities between them.The degree of difference is analyzed,and three error designs are used in data processing to analyze the priority and degree of influence of these influencing factors.In order to study the relationship and correlation between the measurement point position of the room and the average indoor temperature under common conditions,set basic working conditions,change the influencing factors such as orientation,solar radiation,outer wall thickness and window-to-wall ratio,and design 10 sets of values for different working conditions In the simulation experiment,different air temperature distributions near the wall are obtained,and the data format is standardized through the bi-cubic spline interpolation,and the Pearson correlation analysis is performed on the temperature and the average temperature of each measurement point.Use machine learning methods to establish a room temperature measurement model,compare and analyze the influence of neural network type,time series length,input parameter selection and training set data volume on the accuracy of the room temperature measurement model establishment,and give a suitable room temperature measurement model program.