作者：白青飞, 林永君, 杨凯, 李静
作者单位：华北电力大学 新能源电力系统国家重点实验室，河北 保定 071003
关键词：大气颗粒物;太阳辐射强度;k-means算法;rbf神经网络;光伏发电功率预测
摘要：
太阳辐射强度受大气颗粒物浓度影响显著，对光伏发电功率预测准确度的影响不容忽视。为提高太阳辐射强度以及光伏发电功率预测准确度，通过k-means算法对大颗粒物浓度以及相对湿度历史数据进行聚类，基于聚类数据利用径向基神经网络分别建立大气气溶胶光学厚度预测模型；在预测的大气气溶胶光学厚度基础上，结合双波段太阳辐射大气传输模型与倾斜面辐射模型，预测光伏电池板面接收到的太阳辐射强度；最后利用气温和预测光伏电池板面接收到的太阳辐射强度，结合光电转换模型实现光伏发电功率预测。通过实验验证，预测模型平均误差为6.07%，考虑大气颗粒物浓度对太阳辐射强度影响的光伏发电功率预测模型具有较高准确度。
photovoltaic power prediction considering the influence of atmospheric particles on irradiance
bai qingfei, lin yongjun, yang kai, li jing
state key laboratory of new energy power system, north china electric power university, baoding 071003, china
abstract: the intensity of solar radiation was significantly affected by the concentration of atmospheric particulate matter, and its impact on the accuracy of photovoltaic power generation forecasting cannot be ignored. in order to improve the prediction accuracy of solar radiation intensity and photovoltaic power, k-means algorithm was used to cluster the historical data of large particle concentration and relative humidity. based on the clustering data, radial basis function neural network is used to establish the prediction models of atmospheric aerosol optical thickness, based on the predicted aerosol optical thickness, the solar radiation intensity received by the photovoltaic panel was predicted by combining the two band solar radiation atmospheric transmission model and the tilted surface radiation model, finally, the solar radiation intensity received by the photovoltaic panel was predicted by using the air temperature and the photoelectric conversion model to realize the photovoltaic power prediction. through experimental verification, the average error of the prediction model is 6.07%, and the photovoltaic power generation power prediction model considering the influence of atmospheric particle concentration on solar radiation intensity has high accuracy.
keywords: atmospheric particulate matter;solar radiation intensity;k-means algorithm;rbf neural network;photovoltaic power generation prediction
2022, 48(8):117-123 收稿日期: 2021-05-23;收到修改稿日期: 2021-07-22
基金项目: 中央高校基本科研业务费专项资金资助（2019ms100）
作者简介: 白青飞(1996-)，男，山东滨州市人，硕士研究生，专业方向为光伏发电功率预测
参考文献
[1] 李国庆, 董存, 梁志峰, 等. 考虑光伏发电不确定性的跨省两级优化调度模型[j/ol]. 高电压技术: 1-14[2021-05-22]. https://doi.org/10.13336/j.1003-6520.hve.20201300.
[2] 陈炜, 任静, 武新芳, 等. 雾霾条件下光伏预测发电量的迭代优化与经济分析[j/ol]. 中国电力: 1-10[2021-04-11]. http://kns.cnki.net/kcms/detail/11.3265.tm.20210118.1039.004.html.
[3] 裴雨清, 李松丽, 孔华锋, 等. 雾霾影响下的光伏发电系统的可视化分析平台研究[j]. 计算机应用与软件, 2018, 35(6): 88-90 106
[4] 冯健, 牛海明. 空气质量监测数据驱动的光伏发电多尺度功率预测方法[j]. 热力发电, 2020, 49(6): 69-76
[5] 宋绍剑, 李博涵. 基于lstm网络的光伏发电功率短期预测方法的研究[j]. 可再生能源, 2021, 39(5): 594-602
[6] 孙辉, 冷建伟. 基于改进的semi boost天气聚类的cc-pso-dbn短期光伏发电预测[j]. 计算机应用与软件, 2020, 37(8): 103-109
[7] 刘卫亮, 刘长良, 林永君, 等. 基于aod估计的晴空太阳辐射强度滚动预测[j]. 太阳能学报, 2019, 40(7): 1797-1804
[8] 邵丝露, 孙艳玲, 马振兴, 等. 基于融合modis aod和lme模型的京津冀地区pm2.5季节浓度模拟[j]. 天津师范大学学报(自然科学版), 2020, 40(5): 54-62
[9] 金囝囡, 杨兴川, 晏星, 等. 京津冀及周边maiac aod和pm2.5质量浓度特征及相关性分析[j/ol]. 环境科学: 1-18[2021-05-19]. https://doi.org/10.13227/j.hjkx.202009200.
[10] 万和跃, 董晓波, 刘思瑶, 等. 基于飞机和modis观测的华北地区气溶胶标高分析[j]. 气象科学, 2016, 36(5): 655-660
[11] 刘倩颖, 阮应君, 时翔, 等. 基于k-means聚类与bp神经网络算法的办公建筑逐时电负荷预测[j]. 热能动力工程, 2018, 33(3): 138-144
[12] 徐晴, 周超, 赵双双, 等. 基于机器学习的短期电力负荷预测方法研究[j]. 电测与仪表, 2019, 56(23): 70-75
[13] christian a g. rest2: high-performance solar radiation model for cloudless-sky irradiance, illuminance, and photosynthetically active radiation-validation with a benchmark dataset[j]. solar energy, 2007, 82(3): 272-285
[14] christian a g. direct solar transmittance and irradiance predictions with broadband models. part i: detailed theoretical performance assessment[j]. solar energy, 2003, 74(5): 355-379
[15] zhong x, kleissl j. clear sky irradiances using rest2 and modis[j]. solar energy, 2015, 116: 144-164
[16] 王恩宇, 郭强, 张学友, 等. 倾斜面太阳辐射量计算方法研究[j]. 热科学与技术, 2019, 18(1): 35-41
[17] li d, zhang c, yu y, ning, et al. a novel way for improving the overall photoelectric conversion efficiency of dye sensitized tio2 solar cells[j]. procedia computer science, 2018, 131: 1336-1345
[18] 廖天军, 张辉. 光伏电池与分光耦合系统的热学与电学特性[j]. 激光与光电子学进展, 2020, 57(19): 57-62
