← 返回
光伏发电技术 ★ 5.0

基于未来三天逐时辐射预测的PV/T集热器多种运行策略研究

Multiple operational strategies investigations of the PV/T collectors based on 3 days ahead hourly radiation prediction

作者 Shiqian Donga · Yanqiang Dib · Yafeng Gaoa · He Longc · Zhixuan Fana · Jingxuan Guana · Lijun Hand · Yingming Wangd
期刊 Applied Energy
出版日期 2025年1月
卷/期 第 377 卷
技术分类 光伏发电技术
相关度评分 ★★★★★ 5.0 / 5.0
关键词 Hybrid operational strategies were investigated based on the WOA algorithm.
语言:

中文摘要

摘要 PV/T集热器是一种先进的光伏光热一体化技术,能够实现太阳能全光谱利用。PV/T集热器的能量收益不仅依赖于精密的设备设计,还与运行调控条件密切相关。本文基于未来三天逐时水平面太阳总辐射预测模型,结合鲸鱼优化算法(WOA),对混合运行策略进行了分析。结果表明,为期三天的变倾角运行策略效果不理想。当地面反射率从0.4变化到0.95时,月累计辐射量相对提高了8%;地面反射率每增加5%,平均可带来0.45%的性能提升。反射板长边与宽边的最优比例分别为0.9和0.6,可使热能增益和电能增益分别提高8.79%和3.85%。相较于无反射板的固定式PV/T系统,采用石墨材料与梯形流道结构的新型PV/T系统在多模式混合运行策略下,热能和电能增益相对提升分别达到17.76%和13.61%,相当于在2.08 m²有效面积上额外获得39.5 MJ的热能和7.8 kWh的电能。最后,针对未来三天逐时辐射预测的最佳模型为LSTM-Attention模型,其训练集和测试集的决定系数R²分别达到96.8%和94.0%。本文的研究成果为PV/T集热器的运行优化提供了参考依据,尤其适用于户用型农村PV/T系统。

English Abstract

Abstract The PV/T collectors are the advanced photovoltaic and photothermal integration technology, realizing full-spectrum utilization. The energy profitability of PV/T collectors not only depends on sophisticated equipment but also significantly correlates with the operational conditioning. Hybrid operational strategies are analyzed with the WOA algorithm based on 3 days ahead hourly prediction model for horizontal total solar radiation. The 3-day variable inclination strategy is unsatisfactory. The monthly cumulative radiation increases by 8 % relatively with the ground reflectivity varying from 0.4 to 0.95. And, the 5 % addition of ground reflectivity leads to an average improvement of 0.45 %. The optimal long-side and wide-side ratios of the reflector are 0.9 and 0.6 with an improvement of 8.79 % and 3.85 % in thermal and electrical gain. Compared to fixed PV/T without a reflector, the multi-hybrid strategy can achieve a relative improvement of 17.76 % and 13.61 % in thermal and electrical gain for the novel PV/T with graphite and trapezoidal fluid channels, equivalent to 39.5 MJ and 7.8 kWh with a 2.08 m2 effective area. Finally, the best model is LSTM-Attention with the R2_train and R2_test of 96.8 % and 94.0 % for 3 days ahead hourly radiation prediction. The contribution of this paper provides references on the operational optimization of PV/T collectors, especially for household rural PV/T systems.
S

SunView 深度解读

该PV/T光热光伏一体化运行优化研究对阳光电源SG系列光伏逆变器及iSolarCloud平台具有重要参考价值。研究基于3天辐照预测的多混合策略可使热增益和电增益分别提升17.76%和13.61%,这与阳光电源MPPT优化技术和预测性维护理念高度契合。LSTM-Attention预测模型(R²达94%)可集成至iSolarCloud平台,实现户用光伏系统的智能倾角调节、地面反射率优化及反射板配置决策,提升发电效率。该技术尤其适用于农村户用场景,可与阳光电源户用逆变器及储能系统协同,通过算法优化运行策略,最大化全光谱利用效率,为智能运维提供数据支撑。