摘要
光伏组件覆雪易造成发电损耗、组件变形、内部破坏等问题,是影响光伏电站冬季正常运行的重要因素之一。目前国内外对于光伏组件覆雪多采用人工清扫或任其自然消融,但人工清扫费时费力,而自然消融过程中造成的光伏发电损失难以评估。为此,文中首先对冰雪条件下光伏发电功率预测中的关键环节与参数进行了详细分析;随后对覆雪状态下光伏组件发电功率损失原理进行分析,进而搭建光伏组件覆雪实验平台,得出不同冰雪类型、厚度下的发电功率损失;最后建立基于蝗虫算法优化支持向量回归机的预测模型,并通过与4种不同模型预测对比,验证所提预测模型的有效性。该方法有助于对冰雪条件下光伏发电功率进行预测,可为光伏电站冬季运维提供参考。
Snow coating on photovoltaic module is likely to cause such problems as power generation loss,module deformation and internal damage,which is one of the important factors affecting the normal operation of photovoltaic power plants in winter.At present,the snow coating on photovoltaic modules is mostly cleaned manually or allowed to melt naturally.However,the manual cleaning is laborious,and the loss of photovoltaic power generation caused by natural melt is difficult to be assessed.For this reason,in this paper the key links and parameters of photovoltaic power prediction under snow coating conditions was analyzed firstly in detail.Then,the principle of power loss of photovoltaic modules under snow coating condition was analyzed,and a photovoltaic module power generation experimental platform was built to obtain the generation power loss under different types and thickness of snow.Finally,a prediction model based on locust algorithm optimized support vector regression machine was set up,and the effectiveness of the proposed prediction model was verified by comparing with four different models.This method is helpful to predict photovoltaic power generation under snow coating conditions,and can provide reference for operation and maintenance of photovoltaic power station in winter.
作者
李民
杨暑森
李科锋
朱永灿
LI Min;YANG Shusen;LI Kefeng;ZHU Yongcan(Shaanxi Province Institute of Water Resource and Electric Power Investigation and Design,Xi’an 710001,China;School of Electronics and Information,Xi’an Polytechnic University,Xi’an 710048,China)
出处
《高压电器》
CAS
CSCD
北大核心
2023年第9期250-257,共8页
High Voltage Apparatus
基金
陕西省水利科技项目(2019slkj-05)
陕西省重点研发计划项目(2021GY-306)。
关键词
光伏组件
覆雪
功率预测
蝗虫算法
photovoltaic module
snow coating
power prediction
locust algorithm
