基于粒子群优化算法的地埋管换热性能预测研究  

Prediction of heat transfer performance of ground heat exchanger based on particle swarm optimization algorithm

作　　者：魏善明 贾寒冰 曹嘉琪 田彦法 毕栋威 聂玉朋 WEI Shanming;JIA Hanbing;CAO Jiaqi;TIAN Yanfa;BI Dongwei;NIE Yupeng(Shandong Geological and Mineral Engineering Investigation Institute(Shandong Geological and Mineral Exploration and Development Bureau 801 Hydrogeology and Engineering Geology Brigade),Jinan 250013,China;Shandong Jianzhu University,Jinan 250101,China;Shandong Institute of Product Quality Inspection,Jinan 250101,China;Shandong Huake Planning and Architectural Design Co.,Ltd.,Liaocheng 252000,China)

机构地区：[1]山东省地矿工程勘察院(山东省地质矿产勘查开发局八〇一水文地质工程地质大队),山东济南250013 [2]山东建筑大学,山东济南250101 [3]山东省产品质量检验研究院,山东济南250101 [4]山东华科规划建筑设计有限公司,山东聊城252000

出　　处：《区域供热》2025年第1期107-116,共10页District Heating

摘　　要：为了对聊城某地源热泵工程换热器的换热性能进行预测,利用BP神经网络(BPNN)模型和粒子群优化算法(PSO)建立了地埋管换热性能预测模型。通过皮尔逊相关性分析确定影响地埋管出口温度的关键参数,进而预测换热器性能。利用粒子群算法对BPNN模型进行优化,以提升学习效率和收敛速度。模型预测结果显示,BPNN和PSO-BPNN均能有效预测地埋管出口温度,PSO-BPNN在预测精度和稳定性方面更优,其均方根误差(RMSE)为0.1073,平均相对误差(MRE)为0.0049,决定系数(R2)为0.96757,均优于BPNN模型的相应指标。通过建立预测模型,可以对地埋管换热器的出口温度进行短期预测,进而预测换热器在未来一段时间内的性能表现,并提前做出调整,确保地源热泵系统稳定、高效地运行。通过这种方式,可以最大限度地发挥地源热泵系统的性能,同时避免因温度波动导致的性能下降或系统不稳定。In order to predict the heat transfer performance of a ground source heat pump proj ect in Liaocheng,a model for predicting the heat transfer performance of ground heat exchangers is established by using BP neural network(BPNN)model and particle swarm optimization(PSO)algorithm.Through Pearson correlation analysis,the key parameters affecting the outlet temperature of the ground heat exchanger are determined,and then the performance of the heat exchanger is predicted.The particle swarm optimization algorithm is used to optimize the BPNN model to improve the learning efficiency and convergence speed.The prediction results show that both BPNN and PSO-BPNN can effectively predict the outlet temperature of ground heat exchanger,and PSO-BPNN is better in terms of prediction accuracy and stability,with RMSE of 0.1073,MRE of 0.0049,and R2 of 0.96757.All of them are better than the corresponding indexes of BPNN model.By establishing the prediction model,the outlet temperature of the ground heat exchanger can be predicted in the short term,and then the performance of the heat exchanger in the future can be predicted,and the adjustment can be made in advance to ensure the stable and efficient operation of the ground source heat pump system.In this way,the performance of the ground source heat pump system can be maximized while avoiding performance degradation or system instability due to temperature fluctuations.

关 键 词：地源热泵 BP神经网络 粒子群优化算法 地埋管换热器换热性能 

分 类 号：TU83[建筑科学—供热、供燃气、通风及空调工程] TP18[自动化与计算机技术—控制理论与控制工程]