基于灰色多元线性回归的上海某办公建筑空调能耗预测模型

Air Conditioning Systems Energy Consumption Prediction Model of Office Buildings Based on Grey-multiple Linear Regression Theory

能耗预测可以为建筑管理人员提供空调系统未来运营期间的能耗额以及增长趋势,以便制定相应的节能策略。主要讨论空调能耗的影响因素及各因素的影响程度,将灰色模型和多元线性回归模型结合,基于灰色多元线性回归建立空调能耗预测模型。首先,以上海某办公建筑为例,通过灰色模型对空调能耗的异常能耗值及室外天气温度、湿度的观测值进行修正,降低数据的随机波动性,得到灰色模型修正数据;然后,计算影响因素与空调能耗额之间的相关性系数,讨论各因素对空调能耗的影响程度;最后,利用灰色模型修正数据建立灰色多元线性回归空调能耗预测模型,并将空调能耗预测值和实际值相对比,结果表明,构建的灰色多元线性回归预测模型符合多元线性回归建模要求,并具有较高的能耗预测精度。

Energy consumption prediction model can provide building managers with energy consumption value and growth trend during the futural operation,so as to facilitate managers to make relevant energy-saving strategies.This paper mainly discusses the influence degree of influencing factors on air conditioning system and establishes the air conditioning energy consumption prediction model based on the combination of grey model and multiple linear regression model.Firstly,the grey model is used to modify the air conditioning energy consumption value of an office building in Shanghai and the observation value of outdoor weather temperature and humidity in 2018,which can reduce the random volatility of collected data.Secondly,calculating correlation coefficient between the influencing factors and air conditioning energy consumption,in order to explore the influence degree of the energy influencing factors.Finally,the grey multiple linear regression energy consumption prediction model is established by using the emendated data.This paper analyzes the influence degree of influencing factors on air conditioning energy consumption,and established grey multiple linear regression prediction model meets the requirements of multiple linear regression modelling and obtains high accuracy of energy consumption prediction.