| 基于LSTM算法的严寒地区办公建筑过渡季室内温度预测模型构建 |
| |
| 引用本文： | 殷青,张岩,韩昀松.基于LSTM算法的严寒地区办公建筑过渡季室内温度预测模型构建[J].低温建筑技术,2019,41(3):8-12. |
| |
| 作者姓名： | 殷青 张岩 韩昀松 |
| |
| 作者单位： | 哈尔滨工业大学建筑学院,哈尔滨150001;寒地城乡人居环境科学与技术工业和信息化部重点实验室,哈尔滨150001;哈尔滨工业大学建筑学院,哈尔滨150001;寒地城乡人居环境科学与技术工业和信息化部重点实验室,哈尔滨150001;哈尔滨工业大学建筑学院,哈尔滨150001;寒地城乡人居环境科学与技术工业和信息化部重点实验室,哈尔滨150001 |
| |
| 基金项目： | 国家重点研发计划课题项目 |
| |
| 摘 要： | 预测控制方法作为改善用户舒适性和降低建筑能耗的重要手段之一,其关键是室内温度预测模型的建立。由于寒地过渡季建筑的室内温度受室外温度影响较大,本文引入了长短时记忆网络(LSTM)这一算法,该算法能学习室内温度和室外温度的内在联系,预测室内温度在室外温度影响下的变化,同时可以有效处理具有滞后性、时序性的物理量关系。通过在哈尔滨某高校教学楼办公室的实测数据,建立LSTM室内温度预测模型,训练并对预测结果进行评估。结果表明,该算法具有较高的准确度,拟合优度达到了98.7%。
|
| 关 键 词： | 严寒地区 过渡季 机器学习 LSTM算法 室内温度 预测模型 |
AN INDOOR TEMPERATURE PREDICTION MODEL FOR OFFICE BUILDING IN TRANSITION SEASON IN SEVERE COLD REGION BASED ON LSTM |
| |
| Affiliation: | (School of Architecture,Harbin Institute of Technology,Harbin 150001,China;Key Laboratory of Cold Region Urban and Rural Human Settlement Environment Science and Technology,Ministry of Industry and Information Technology,Harbin 150001,China) |
| |
| Abstract: | The predictive model of indoor temperature is essential for the predictive control to improve indoor comfort and reduce building energy consumption.Since the indoor temperature of buildings in severe cold region during the transition season is greatly affected by outdoor temperature,this paper adopts the long-short-time memory network(LSTM)algorithm,which can learn the internal relationship between indoor temperature and outdoor temperature,and predict the influence of indoor temperature on outdoor temperature.It also can effectively deal with hysteresis and time series problem.Using the measured data of the office building of a university in Harbin,we established the LSTM indoor temperature prediction model,trained and evaluated the prediction.The results show that the prediction has the accuracy of 98.7%. |
| |
| Keywords: | severe cold region transition season machine-learning LSTM indoor temperature predictive model |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
