CAJ | 학술논문

基于获取建筑能耗数据并对其进行准确的规划和配置，从而实现建筑节能，需要建立高效的建筑能耗预测模型。本文针对灰箱模型中的RC热网络建筑能耗预测模型进行综述研究。以典型的RC热网络模型为例，分析其建模思想以及模型推导方法。总结RC热网络建筑能耗预测模型传递矩阵的五个推导步骤，并对核心的模型参数优化算法：序列二次规划法、共轭梯度法和遗传算法三种算法进行分析。为提高模型计算速度和精度，数据融合和模型降阶方法成为模型的主要发展方向。本文通过RC热网络建筑能耗预测模型应用实例展示了其省时、高效、准确的预测性能，并且指出了其输入参数不稳定和预测形式单一的缺点。最后，展望了RC热网络建筑能耗预测模型的发展趋势和研究挑战。
기우획취건축능모수거병대기진행준학적규화화배치，종이실현건축절능，수요건립고효적건축능모예측모형。본문침대회상모형중적RC열망락건축능모예측모형진행종술연구。이전형적RC열망락모형위례，분석기건모사상이급모형추도방법。총결RC열망락건축능모예측모형전체구진적오개추도보취，병대핵심적모형삼수우화산법：서렬이차규화법、공액제도법화유전산법삼충산법진행분석。위제고모형계산속도화정도，수거융합화모형강계방법성위모형적주요발전방향。본문통과RC열망락건축능모예측모형응용실례전시료기성시、고효、준학적예측성능，병차지출료기수입삼수불은정화예측형식단일적결점。최후，전망료RC열망락건축능모예측모형적발전추세화연구도전。
In order to obtain the data of building energy consumption and make accurate planning and configuration , it is significant to establish building energy consumption forecasting model for the energy saving. RC thermal network building energy consumption forecasting model, typical gray-box model (combination of physics based and data-driven), approaches are focused in this review. A typical RC thermal network model is taken as an example to analyze the modeling idea and modeling method. This paper summarizes the five processes of the design about the transfer matrix of RC thermal network building energy consumption forecasting model. Moreover, the core model parameter optimization algorithms are systematically described, including sequential quadratic programming, conjugate gradient method and genetic algorithm. There are two ways to improve the model calculation speed and precision. Incorporating with data fusion schemes and model reduction are presented. These two aspects are also the main development direction of the model. However, RC thermal network building energy consumption forecasting model also have superiorities in its applications, especially the excellent performance in efficient and accurate prediction. This paper also points out its instability of input parameters and existing problems of single forecast form. Finally, with summarizing the research hot issues, the challenges and the developing trend of RC thermal network building energy consumption forecasting model are analyzed.