基于大数据分析的城市热网负荷综合预测研究

Comprehensiveresearch on prediction of urban heat network load based on big data analysis

随着智慧城市与行业改革的推进,城市供热需要新一轮技术革新。针对国家节能减排号召,实现信息化、高效化供热,需要建立智慧供热平台,以实时监测热网运行,并预测未来热网负荷。采用LSTM(长短期记忆神经网络)深度学习算法,对湖北某地级市气象、热网负荷等数据进行大数据分析,建立预测模型并进行一系列调参优化。利用Python计算机语言编写程序,实现对未来热网负荷预测,并按重要性对数据各项特征进行分析并排名。综上建立智慧供热平台界面,以显示实时数据以及预测数据。

With the advancement of smart city and industry reform,a new round of technological innovation is required in urban heating.Meanwhile,In response to the national call for energy saving and emission reduction and to realize information and efficient heat supply,a smart heat supply platform is needed to monitor heat network operation in real time and predict future heat network load.In this paper,LSTM(Long Short-term Memory)is used to make big data analysis on meteorological and heat network load data of a prefecture-level city in Hubei province,establish prediction model and conduct a series of parameter tuning optimization.A program is written in Python to forecast future heat network load and analyze the ranking of data by their characteristic importance.In summary,a smart heating platform interface is established to display real-time data and forecast data.