基于全天空图像的短时云图预测和太阳法向辐照度趋势预测.pdfVIP

浙江大学硕士学位论文摘要

实验表明．本文所提出全天空序列云图预测模型可以预测长时的全天空图像，并且模

型表现性能良好。同时使用预测得到的全天空云图可以完成相应DNI值的提取从而实现

DNI的预测。实验表明，预测的DNI变化趋势与测量值基本一致，且少数位置本文基于云

图的预测结果比测量值更加真实，符合预期目标。

关键词：塔式太阳能光热发电系统；全天空序列图像；云图预测；太阳中心定位；DNI趋

势预测：

浙江大学硕士学位论文Abstract

Abstract

Withtheofand

sciencedemandforisalso

technology，mankind’S

developmentenergy

akindofcleantakesanof

increasing．Asenergy,solarenergyupincreasingenergy

proportion

sourcesduetoitsandothersolar

renewable，environmental

protectionadvantages．Thetower-type

thermalhastheofconcen订ation

powergenerationsystemadvantageslargeratio，largecapacity,

ithascommercialhasreceivedmoreand

highefficiency,etc．．SOimportantandmore

prospects

research．Thesolarthermalusessolarreceivertofocusthe

tower-typepowergenerationsystem

directnormalthesolarabsorb

irradiance(DNI)onsystemreceivertoandconverttocomplete

occlusionandofcloudsarethemainfactorsthatcausethedirect

powergeneration．Thedeparture

normalirradianceofthesun．Whenthecloudentersthedirect

changesuddenlyfield，thenormal

irradiancereceivedtheheliostatfieldthe

bychangessuddenly,andtemperaturedecreases，which

causestheheatmediumtoandcauseoreven

storagesolidifymaypipeblockageshutd