Research On The Development Of The Daily Solar Radiation Estimation Models And Outdoor Design Radiation

To solve the problem of disparse distribution of solar radiation stations and serious lack of existing radiation data in China,this study proposes the whole process idea of"Completion of existing radiation station-Estimation of non-radiation station-Interpo-lation of non-station area".The"Estimation Model Based on Radiation Station-General Model of non-Radiation Station-Interpolation of Non-Station Area"is adopted to solves the problem of three kinds of radiation data complementation in turn:the absence of exsit-ing radiation data,the absence of radiation data and the absence of meteorological data in China.Furthermore,outdoor design radiation is the basic parameter for building load cal-culation,solar collector and photovoltaic panel area calculation.Due to the lack of radia-tion data,it is difficult to obtain long-term solar radiation data for outdoor design radiation demand.The traditional calculation method of building load in summer mainly uses the solar radiation on clear days of July 21.Influenced by the actual weather conditions,the actual solar radiation is less than solar radiation on clear day,which easily leads to large equipment capacity.The calculation of the capacity of solar energy system in winter and whole year mainly uses the statistical values of 12-month average daily radiation param-eters and annual total radiation parameters of short-term radiation data.Short-term data are difficult to characterize the long-term variation characteristics of radiation,which may lead to unreasonable calculation of equipment capacity.In view of the above problems,this study proposes statistical analysis methods for outdoor design radiation in different systems based on the 30-year long-term radiation data,which provides basic parameters for the design and calculation of solar energy utilization system.Specific research contents and results are as follows:?1?The solar radiation zoning is the precondition for the performance analysis of solar radiation estimation models and the establishment of general models.The traditional radiation zoning method based on radiation stations,which qualitatively divides the non-radiation stations into different climatic regions,is liable to cause the problem of misclas-sification in areas far away from the stations,especially in the boundary areas of the two zones.Based on k-means clustering and SVM-GA classification,a new two-step solar radiation zoning method is proposed.Firstly,using k-means clustering and taking monthly average solar radiation as evaluation index,the radiation stations are divided into five categories.Then,using SVM-GA,the conventional meteorological data of radiation stations are used as training set,and the conventional meteorological data of non-radiation stations are used as test set,and the non-radiation stations are divided into five categories.In this method,non-radiation stations are introduced into the zoning process,and the tra-ditional qualitative solar radiation zoning method is transformed into quantitative zoning method,which greatly improves the accuracy of radiation zoning.?2?Aiming at the problem of missing radiation data,the correlation between global radiation with sunshine and temperature in different radiation zones is analyzed by using the data of sunshine percentage,daily average temperature and daily range of 96 radiation stations,and the estimation model of total solar radiation based on sunshine percentage and temperature in China is constructed.Furthermore,this study analyzed the prediction accuracy of different models in different climatic zones,gave the most suitable total ra-diation estimation model for each climatic zone,and then constructed a general model for estimating total solar radiation in China,which supplemented the total solar radiation data of 564 non-radiation stations in China.?3?Diffuse estimation models can be classified into two categories:diffuse fraction and disffuse coefficient models.Generally speaking,the diffuse fraction model has higher accuracy,but it needs global radiation as input parameters;the diffuse coefficient model has lower accuracy,but it does not need global radiation as input parameters,so it has a wider range of applications.According to the global radiation and diffuse radiation data,meteorological stations can be divided into three categories:both global radiation and diffuse radiation data are available?52 stations?;only global radiation data are available?44 stations?and no radiation data are available?564 stations?.Based on the conventional meteorological data?including sunshine rate,temperature,relative humidity and rainfall?,the correlation between diffuse radiation with global solar radiation and conventional me-teorological parameters is analyzed,and the diffuse fraction model and diffuse coefficient model in China are developed.The missing data of diffuse radiation of 52 stations with both global radiation and diffuse radiation data are supplemented by the diffuse fraction model.Furthermore,the model accuracy of different models in different climatic zones is analyzed,the mechanism of the influence of meteorological parameters on diffuse ra-diation is revealed,and the order of the influence of different meteorological parameters on scattering radiation in different climatic zones is given.The most suitable model of diffuse estimation models in different climatic zones are given.The general model of diffuse fraction and diffuse coefficient models completes the diffuse radiation data of 44stations with global radiation data and 564 non-radiation stations in China.?4?Aiming at the problem of missing data in non-station area,this study uses the inverse distance weight interpolation method to obtain the spatial distribution grid data of radiation data,which completes the global radiation and diffuse radiation data in non-station area.Furthermore,in view of the problem that the traditional grid data of spatial distribution of solar radiation neglects the shielding effect of local topography on solar radiation,which leads to the overestimation value of solar radiation in some areas,this study calculates the spatial distribution of slope and slope direction in China by using DEM data with resolution of 90 m×90 m,and obtains local areas.For the spatial distri-bution data of the shielding factor of direct radiation,the grid data of the spatial distribu-tion of solar radiation in China are revised.?5?Aiming at the unreasonable problem of the traditional statistical analysis method of outdoor design radiation,based on the 30-year long-term solar radiation data,three statistical methods of outdoor design radiation are proposed in this study:based on the annual average of not guaranteeing 5 days of outdoor radiation in winter and summer,based on the monthly average of the minimum month of annual radiation,with the thresh-old of±0.20 MJm^-2d^-1.Average daily radiation parameters;based on the annual average annual total solar radiation parameters.Through comparative analysis,it is found that for the outdoor design radiation in summer design day,the statistical values in several cold and cold areas coincide with the recommended values,while in hot summer and cold winter,hot summer and warm winter and mild areas,the recommended values are gener-ally larger than the statistical values.The main reason is that most of the hot summer and cold winter,warm summer and warm winter and mild areas are in humid climatic areas.The recommended values underestimate the effect of humidity on solar radiation.For winter design days,the recommended values of the calculation method of sunny day solar radiation model recommended by relevant literatures are generally larger than the statis-tical values of this study.The main reason is that the design day in winter is usually cloudy and rainy,and the solar radiation is far less than that in sunny days.For the monthly average daily and annual total radiation parameters,the statistical values and recom-mended values differ greatly due to the different statistical data used.