Study On Typical Meteorological Year Generation Method In South China Sea

In the process of architectural design,it is necessary to judge whether the building meets the energy-saving standard,usually with the simulation software to simulate the energy consumption,to carry out building energy consumption simulations typically require a whole year of hourly data on the related meteorological parameters.For longterm meteorological data,it should be typical and include a full year data.Most of the islands and reefs in the South China Sea belong to extremely hot and humid climate zone,buildings are exposed to high temperature,high humidity,and high salt environment throughout the year,so it consumes a lot of energy.However,there is no meteorological data that can be used to simulate building energy consumption.Therefore it is necessary to conduct comparative studies on typical annual meteorological methods in the South China Sea.Firstly,this paper obtains meteorological data of the South China Sea from NOAA and TRMM,convert the obtained meteorological data into a readable format,we obtain a higher precision radiation prediction model and model accuracy through machine learning stochastic forest algorithm deal with nonlinear regression relation.Training data is 0.963,test data 0.743,this model complemented the 16-year total radiation data of 8 stations in the South China Sea and analyzed the meteorological characteristics of the South China Sea,it was found that each island and reef has islands-scale meteorological laws.Secondly,This paper aims at 7 sites with relatively complete data in the South China Sea,respectively employs Sandia,Danish and Festa-Ratto method,make each site obtain three typical meteorological annual results.As for three results of each site,by evaluating monthly average,standard deviation and difference values of analysis and comparison of indicator meteorological parameters(air temperature,dew point temperature,wind speed,and horizontal total daily radiation),proceed adaptability evaluation of typical meteorological annual algorithm in the South China Sea,synthesized various indicators evaluation results,it found that the festa-ratto algorithm was more comprehensive than the Sandia and Danish methods,it is an applicable algorithm to generate the typical meteorological year in the South China Sea.Finally,this paper analyzes the radiation hourly model and the conventional meteorological parameters hourly method,hourly convert the total radiation data select and use C-P&R statistical model,conduct direct dispersion separation with Gompertz function,The cubic spline interpolation method is selected to hourly convert other data.Through Python language,separately programmed the C-P&R statistical model,the Gompertz function model,and the cubic spline interpolation method,batch process the selected typical meteorological year data,realized the chronology of each meteorological data and analyzed its characteristics.Respectively carried out analysis of meteorological parameters and energy consumption simulation for Sansha City,Manila Control Group,and Taiping Island and Sparta Bay Bay Control Groups,it was found that the data of representative meteorological parameters of typical meteorological years in the two control groups were significantly different,as well the difference between total energy consumption and air conditioning load throughout the year is also large.Therefore,For the accuracy of building energy consumption simulation in the South China Sea,the typical meteorological year for each site in the South China Sea was selected for the simulation of building energy consumption.The work is very necessary and superior to the previous use of nearby site data.It is suggested that the typical meteorological year data generated in this paper can be used to simulate building energy consumption in the South China Sea.