气象条件对光伏电站日污秽损失率的影响分析

ANALYSIS OF INFLUENCE OF METEOROLOGICAL CONDITIONS ON DAILY POLLUTION LOSS RATE OF PV POWER STATIONS

光伏发电是利用半导体的光生伏特效应将太阳能直接转变为电能。空气中的灰尘在光伏组件表面沉积,不仅会降低透射到光伏组件上的光能,还会影响光伏组件的散热,从而降低光伏电站的系统效率,进而影响光伏电站的发电量。利用遗传算法与反向传播(BP)神经网络相结合(下文简称为“遗传算法BP神经网络”)的方式建模,通过输入风速、风向、相对湿度及环境PM10浓度,分析气象条件对光伏电站的日污秽损失率的影响。测试结果表明:利用遗传算法BP神经网络对光伏电站日污秽损失率进行预测的误差可以满足精度要求,相比单纯使用人工神经网络计算的误差可减小5.9%。随着灰尘的累积,日污秽损失率随光伏组件清洗周期的延长呈线性增长,光伏组件表面年均洁净率呈线性下降。遗传算法BP神经网络能很好地预测气象环境导致的污秽损失率,通过气象条件参数可以对当地建设的光伏电站的污秽损失率进行预估。

PV power generation uses the PV effect of semiconductors to directly convert solar energy into electrical energy. The deposition of dust in the air on the surface of PV modules will not only reduce the light energy transmitted to the PV modules,but also affect the heat dissipation of PV modules,thereby reducing the system efficiency of PV power stations,then affecting the power generation of PV power stations. This paper uses the combination of genetic algorithm and BP neural network(hereinafter referred to as“genetic algorithm BP neural network”) to model,analyze the impact of meteorological conditions on the daily pollution loss rate of the PV power station by inputting the wind speed,wind direction,relative humidity and ambient PM10concentration. The test results show that the error of using the genetic algorithm BP neural network to predict the daily pollution loss rate of the PV power station can meet the accuracy requirements,and the error is reduced by 5.9% compared with the simple artificial neural network calculation. With the accumulation of dust,the daily pollution loss rate increases linearly with the growth of the cleaning cycle of PV modules,and the annual cleaning rate of PV modules surface decreases linearly. The genetic algorithm BP neural network can well predict the pollution loss rate caused by the meteorological environment,and the pollution loss rate of locally built PV power stations through the meteorological condition parameters.