应用PSO-RBF神经网络预测太阳能PV/T系统的热、电性能  

Prediction of thermal and electrical performance of solar PV/T system using PSO-RBF neural network

作　　者：何迪 王聪聪 陈红兵 孙俊辉 高雪宁 王传岭 马卓越 He Di;Wang Congcong;Chen Hongbing;Sun Junhui;Gao Xuening;Wang Chuanling;Ma Zhuoyue(School of Environment and Energy Engineering,Beijing Municipal Key Lab of HVAC,Beijing University of Civil Engineering and Architecture,Beijing 100044,China;China Construction Sixth Engineering Bureau Crop.,Ltd,Tianjin 300012,China;Tong Yuan Design Group Co.,Ltd,Ji'nan 250024,China)

机构地区：[1]北京建筑大学环境与能源工程学院供热供燃气通风及空调工程北京市重点实验室,北京100044 [2]中国建筑第六工程局有限公司,天津300012 [3]同圆设计集团股份有限公司,山东济南250024

出　　处：《可再生能源》2024年第4期455-463,共9页Renewable Energy Resources

基　　金：北京市科技计划项目(KM202010016012)。

摘　　要：为准确预测太阳能光伏光热(Solar Photovoltaic/Thermal,PV/T)系统的热、电性能,文章利用PSO(Particle Swarm Optimization)算法优化了RBF(Radial Basis Function)神经网络,并基于此方法建立了太阳能PV/T系统性能的仿真预测模型,与基于未优化RBF神经网络建立的预测模型进行了对比分析。同时,搭建了太阳能PV/T实验平台,通过云平台采集实验数据用于上述模型。研究结果表明:使用PSO算法优化后的RBF神经网络模型相较于未优化模型预测精度提高了20%,预测稳定性提高了30%,拟合优度R值有所提升。基于PSO-RBF神经网络建立的预测模型可精确预测太阳能PV/T系统的热、电性能。In order to accurately predict the thermal and electrical performance of solar photovoltaic/thermal(PV/T)systems,this study utilized the Particle Swarm Optimization(PSO)algorithm to optimize the Radial Basis Function(RBF)neural network.Based on this method,a simulation prediction model for the performance of solar PV/T systems was established and compared with a prediction model based on an unoptimized RBF neural network.Additionally,this research built a solar PV/T experimental platform and collected experimental data using a cloud platform for the aforementioned model.The research results indicate that the RBF neural network model optimized using the PSO algorithm exhibits better prediction accuracy compared to the unoptimized RBF neural network model.The optimized RBF neural network model demonstrates a20%improvement in prediction accuracy and a 30%increase in prediction stability compared to the unoptimized model.The goodness of fit,as indicated by the R-value,is also improved compared to the unoptimized model.The prediction model established based on the PSO-RBF neural network can accurately predict the thermal and electrical performance of solar PV/T systems.

关 键 词：PV/T RBF神经网络 PSO算法 模拟预测 

分 类 号：TK5199[动力工程及工程热物理—热能工程]