Research On Building Energy Consumption Prediction Based On Outdoor Meteorological Parameter Analysis

Outdoor meteorological parameters mainly affect the heating and cooling energy consumption of buildings,and the climatic characteristics of different regions have different effects on the HVAC system.Relevant research on accurate basic outdoor meteorological parameters is helpful to evaluate and optimize building energy-saving design schemes,realize building energy-saving and reduce CO₂ emissions.With the development of the data era,China is constantly improving the observation and statistics of outdoor meteorological data.However,there is the problem of incomplete solar radiation data recording,which has led to the lack of corresponding solar radiation data in many areas for selecting Typical Meteorological Year（TMY）and conducting building energy consumption simulation analysis.Building energy consumption prediction research based on the analysis of outdoor meteorological parameters is helpful to analyze the changes in building energy consumption,and formulating reasonable solutions for this is an urgent problem in this research.Therefore,this research mainly includes the following parts:（1）In the field of building energy conservation analysis,solar radiation data is particularly important.However,the number of solar radiation stations in China can’t match the general ground meteorological observation stations due to the high accuracy of equipment and instruments required to record the radiation data and the complexity of inspection and maintenance.Based on the actual meteorological data of solar radiation in China,TMY was selected for the regions with different recording accuracy（daily value,monthly value and no radiation）,and correlation analysis was made for the applicability of different regions.The research on hourly solar radiation data generation mainly adopts the statistical model method.On this basis,the neural network and the corresponding optimization algorithm are used to forecast the solar radiation data,and the correlation analysis is also done for the applicability of different regions.（2）Taking office buildings as a typical building,the TMY and hourly solar radiation data generated by each method were used to simulate building energy consumption,which verified the accuracy of the TMY selection method and various hourly solar radiation data prediction models proposed in this study when solar radiation was absent.And for the subsequent data experiment preparation for building energy consumption prediction and analysis based on outdoor meteorological parameters.（3）In order to explore the influence of outdoor meteorological parameters on building heating and cooling energy consumption,TMY and 30-year weather data of typical cities in Beijing and Guangzhou in the north and south regions were selected for energy consumption simulation.Taking a common fan-coil HVAC system as an example,the sensitivity analysis method was used to discuss the chiller COP,boiler BE and fresh air volume,and the influence factors of different design parameters of the system were determined.The correlation analysis method is used to analyze the meteorological factors affecting the heating and cooling energy consumption of various buildings,and the input of outdoor meteorological parameters to predict building energy consumption is determined through strong correlation.Through the comparison of temperature and energy consumption,the typical urban heating and cooling equilibrium temperature is determined,and a new"degree monthly temperature"index is calculated in combination with the monthly maximum and minimum temperature,which provides input for building energy consumption prediction and analysis to improve the prediction accuracy.（4）Through the analysis of the factors affecting the outdoor meteorological parameters of the main building energy consumption,intelligent algorithms such as BP,RBF and support vector machines in machine learning are used to predict building energy consumption,and the prediction accuracy is improved by the comparison of intelligent algorithms and the determination of new parameter indicators.Put forward a relatively complete forecasting scheme suitable for outdoor meteorological parameters to predict building energy consumption.