建筑物热负荷预测及控制方案研究的中期报告.docx

建筑物热负荷预测及控制方案研究的中期报告 【摘要】 本中期报告主要针对建筑物热负荷预测及控制方案研究进行了深入研究。首先，对建筑物热负荷的含义、计算方法及影响因素进行了详细介绍和分析，梳理了相关研究现状。其次，对热负荷预测方法进行了探讨，包括传统的经验公式法、统计学方法和神经网络等方法。针对神经网络方法，对其模型结构及训练方法进行了详细阐述。最后，对建筑物热负荷控制方案进行了探讨，分析了传统空调系统和地源热泵系统的优缺点，并提出了基于智能控制的新型控制方案。 【关键词】 建筑物；热负荷；预测；控制方案；神经网络 【Abstract】 This mid-term report mainly focuses on the study of building thermal load prediction and control schemes. Firstly, the meaning, calculation method and influencing factors of building thermal load were introduced and analyzed in detail, and the relevant research status was sorted out. Secondly, the prediction methods of thermal load were discussed, including traditional empirical formula method, statistical methods and neural network methods. For neural network methods, the model structure and training methods were elaborated in detail. Finally, the building thermal load control schemes were discussed. The advantages and disadvantages of traditional air conditioning system and ground source heat pump system were analyzed, and a new intelligent control-based control scheme was proposed. 【Keywords】 building; thermal load; prediction; control scheme; neural network