提出一种基于改进蜣螂优化算法（IDBO）的双向时间卷积网络（BiTCN）和双向长短期记忆网络（BiLSTM）结合自注意力机制（SA）的光伏组件积灰预测模型,用于对光伏组件透光率进行有效预测,再基于透光率与积灰密度的拟合公式,实现光伏组件的积灰预测。预测模型利用BiTCN提取局部特征、BiLSTM捕捉全局时间依赖关系,并通过SA优化特征权重分配,以提高预测精度。同时,使用IDBO优化模型超参数,进一步提升预测性能。通过搭建物联网采集平台获取光伏组件透光率及相关环境数据制作数据集用于模型仿真分析。实验结果显示,所提模型的均方根误差为0.0126,平均绝对误差为0.007,预测精度达到96.26%,整体优于其他算法,提升了光伏组件积灰预测的精度。
Abstract
This study proposes an innovative dust accumulation prediction model for photovoltaic （PV） modules through the integration of an improved dung beetle optimization algorithm （IDBO） with a hybrid neural network architecture combining a bidirectional temporal convolutional network （BiTCN）, a bidirectional long short-term memory （BiLSTM）, and self-attention （SA）. The developed model achieves accurate prediction of PV module transmittance, subsequently enabling dust density estimation through established transmittance-dust density correlation formulas. The architecture strategically employs BiTCN for local feature extraction, BiLSTM for capturing global temporal dependencies, and SA for dynamic feature weighting optimization, thereby enhancing overall prediction precision. Furthermore, the IDBO algorithm is implemented for hyperparameter optimization to maximize model performance. To validate the approach, we establish an IoT-based data acquisition platform that collects comprehensive PV module transmittance data and associated environmental parameters, forming a robust dataset for model training and evaluation. Experimental results demonstrate superior performance metrics with root mean square error （RMSE） of 0.0126, mean absolute error （MAE） of 0.007, and prediction accuracy reaching 96.26%, significantly outperforming benchmark algorithms. This advancement enables more precise dust accumulation forecasting.
关键词
光伏组件 /
预测模型 /
积灰 /
双向长短期记忆网络 /
注意力机制
Key words
photovoltaic modules /
prediction model /
dust deposition /
bi-directional short-term memory network /
attention mechanism
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] MAKA A O M, ALABID J M. Solar energy technology and its roles in sustainable development[J]. Clean energy, 2022, 6(3): 476-483.
[2] KAZEM H A, CHAICHAN M T.Experimental analysis of the effect of dust’s physical properties on photovoltaic modules in Northern Oman[J]. Solar energy, 2016, 139: 68-80.
[3] JAVED W, WUBULIKASIMU Y, FIGGIS B, et al.Characterization of dust accumulated on photovoltaic panels in Doha, Qatar[J]. Solar energy, 2017, 142: 123-135.
[4] KAZEM H A.Impact of long-term dust accumulation on photovoltaic module performance: a comprehensive review[J]. Environmental science and pollution research, 2023, 30(57): 119568-119593.
[5] RUSĂNESCU C O, RUSĂNESCU M, ISTRATE I A, et al. The effect of dust deposition on the performance of photovoltaic panels[J]. Energies, 2023, 16(19): 6794.
[6] 吕玉坤, 周庆文, 魏子安, 等. 积灰对光伏组件性能影响的试验研究及数值模拟[J]. 电源学报, 2024, 22(2): 175-182.
LYU Y K, ZHOU Q W, WEI Z A, et al.Experimental study and numerical simulation of influence of ash deposition on performance of photovoltaic modules[J]. Journal of power supply, 2024, 22(2): 175-182.
[7] WANG Z H, XU Z G.Optimal cleaning schedule of photovoltaic module[C]//2018 IEEE International Conference on Electric Machine and Equipment (IEEM). IEEE, 2018: 1637-1641.
[8] 高瑜, 孔晓龙, 王寅清, 等. 基于改进SSA-SVM光伏板积灰预测及功率影响研究[J]. 西安科技大学学报, 2023, 43(2): 398-405.
GAO Y, KONG X L, WANG Y Q, et al.Prediction of ash depositon and power impact of photovoltaic panels based on improved SSA-SVM[J]. Journal of Xi’an University of Science and Technology, 2023, 43(2): 398-405.
[9] 薛阳, 燕宇铖, 贾巍, 等. 基于改进灰狼算法优化长短期记忆网络的光伏功率预测[J]. 太阳能学报, 2023, 44(7): 207-213.
XUE Y, YAN Y C, JIA W, et al.Photovoltaic power prediction model based on IGWO-LSTM[J]. Acta energiae solaris sinica, 2023, 44(7): 207-213.
[10] 丁婷婷, 杨明, 于一潇, 等. 基于误差修正的短期风电功率集成预测方法[J]. 高电压技术, 2022, 48(2): 488-496.
DING T T, YANG M, YU Y X, et al.Short-term wind power integration prediction method based on error correction[J]. High voltage engineering, 2022, 48(2): 488-496.
[11] 欧阳静, 秦龙, 王坚锋, 等. 基于PCA-ShapeDTW-QWGRU的分布式光伏集群短期功率预测[J]. 太阳能学报, 2024, 45(5): 458-467.
OUYANG J, QIN L, WANG J F, et al.Short-term power prediction for distributed PV clusters based on PCA-ShapeDTW-QWGRU[J]. Acta energiae solaris sinica, 2024, 45(5): 458-467.
[12] LIU W, LIU Q, LI Y L.Ultra-short-term photovoltaic power prediction based on modal reconstruction and BiLSTM-CNN-Attention model[J]. Earth science informatics, 2024, 17(3): 2711-2725.
[13] 杨建, 常学军, 姚帅, 等. 基于WT-CNN-BiLSTM模型的日前光伏功率预测[J]. 南方电网技术, 2024, 18(8): 61-69, 79.
YANG J, CHANG X J, YAO S, et al.Day-ahead photovoltaic power forecasting based on WT-CNN-BiLSTM model[J]. Southern power system technology, 2024, 18(8): 61-69, 79.
[14] 韩素敏, 赵国帅, 尚志豪, 等. 基于改进DBO优化BiLSTM的IGBT老化预测模型[J]. 电子测量技术, 2024, 47(1): 46-54.
HAN S M, ZHAO G S, SHANG Z H, et al.IGBT aging prediction model based on improved DBO optimization BiLSTM[J]. Electronic measurement technology, 2024, 47(1): 46-54.
[15] ZHANG D D, CHEN B A, ZHU H Y, et al.Short-term wind power prediction based on two-layer decomposition and BiTCN-BiLSTM-attention model[J]. Energy, 2023, 285: 128762.
[16] 欧阳福莲, 王俊, 周杭霞. 基于改进迁移学习和多尺度CNN-BiLSTM-Attention的短期电力负荷预测方法[J]. 电力系统保护与控制, 2023, 51(2): 132-140.
OUYANG F L, WANG J, ZHOU H X.Short-term power load forecasting method based on improved hierarchical transfer learning and multi-scale CNN-BiLSTM-Attention[J]. Power system protection and control, 2023, 51(2): 132-140.
[17] WAN L T, ZHAO L Q, XU W S, et al.Dust deposition on the photovoltaic panel: a comprehensive survey on mechanisms, effects, mathematical modeling, cleaning methods, and monitoring systems[J]. Solar energy, 2024, 268: 112300.
[18] 高瑜, 康兴国, 周少迪, 等. 基于改进BES-LSSVM光伏组件积灰预测[J]. 太阳能学报, 2023, 44(6): 213-219.
GAO Y, KANG X G, ZHOU S D, et al.Ash deposition prediction of photovoltaic modules based on improved BES-LSSVM[J]. Acta energiae solaris sinica, 2023, 44(6): 213-219.
[19] 官燕玲, 张豪, 闫旭洲, 等. 灰尘覆盖对光伏组件性能影响的原位实验研究[J]. 太阳能学报, 2016, 37(8): 1944-1950.
GUAN Y L, ZHANG H, YAN X Z, et al.Experimental study of influence of dust cover on performance of PV module[J]. Acta energiae solaris sinica, 2016, 37(8): 1944-1950.
[20] LI Z Y, XU X P.L2-BiTCN-CNN: spatio-temporal features fusion-based multi-classification model for various Internet applications identification[J]. Computer networks, 2024, 243: 110298.
[21] 吴嘉俊, 苏春, 张玉茹. 基于双重自注意力机制和长短时记忆网络的剩余寿命预测[J]. 系统工程与电子技术, 2024, 46(6): 1986-1994.
WU J J, SU C, ZHANG Y R.Remaining useful life prediction based on double self-attention mechanism and long short-term memory network[J]. Systems engineering and electronics, 2024, 46(6): 1986-1994.
[22] XUE J K, SHEN B.Dung beetle optimizer: a new meta-heuristic algorithm for global optimization[J]. The journal of supercomputing, 2023, 79(7): 7305-7336.
基金
国家自然科学基金（61861007）; 贵州省教育厅创新群体项目（黔教合KY字[2021]012）; 贵州省科技支撑计划（黔科合支撑[2023]一般411）; 贵州省科技支撑计划（黔科合支撑[2023]一般412）
PDF(7633 KB)
