Precise Control for Heating Supply to Households Based on Heating Load Prediction

  • Ruiting Wang
  • Fulin WangEmail author
  • Zhaohan Nan
  • Minjie Xiao
  • Aijun Ding
Conference paper
Part of the Environmental Science and Engineering book series (ESE)


Commonly, no control measures are mounted at the heating terminals in a household in district heating systems in China, so the problem of unbalanced heating pipe network causes problems of overheating, under-heating, energy loss, etc. At the same time, the mandatory installation of calorimeters in residential buildings, which aims to solve the inefficient energy-using behavior by charging heating cost according to heat amount used. However, charging heating cost according to heat amount was not successfully adopted by heating company. As a result, the aim of energy saving was not achieved, and giant investment was wasted as well. Aiming to solve these problems, this paper proposed a precise heating terminal control system utilizing the heating amounts measured by the calorimeter, which can solve the unbalance problem of heating supply and can precisely control the heat supplied to a household to meet the individual heating requirement of the household as well. The proposed control system takes advantage of the data predicting ability of artificial neural network (ANN) to predict heating requirement using the information of weather, desired indoor temperature, building envelope, etc. By comparing the predicted heating load with the heating amount measured by the calorimeter, the on/off of heating water valve is controlled to make the supplied heating amount match the predicted heating load. In this paper, the heating load prediction by ANN is described and the results show that the load prediction is accurate enough for the purpose of achieving precise control of heating amount supplied to the terminal users.


Precise heating control Heating system Heating load prediction Neural network 



This research is supported by Innovative Research Groups of the National Natural Science Foundation of China (grant number 51521005).


  1. 1.
    Lu, Y.: Heating charge of switching from floor area accordance to heating amount unsuccessful, People’s Daily, December 25 (2013)Google Scholar
  2. 2.
    Shi, L.: Application of artificial network to predict the hourly cooling load of an office building. In: IEEE 2009 International Joint Conference on Computational Sciences and Optimization, Sanya, Hainan, China 2009Google Scholar
  3. 3.
    Zhao, H., Magoulès, F.: A review on the prediction of building energy consumption. Renew. Sustain. Energy Rev. 16, 3586–3592 (2012)CrossRefGoogle Scholar
  4. 4.
    Ian, G., Yoshua, B., Aaaron, C.: Deep Learning, p. 196. MIT Press, Boston, USA 2016Google Scholar
  5. 5.
    Rafiq, M.Y., Bugmann, G., Easterbrook, D.J.: Neural network design for engineering applications. Comput. Struct. 79, 1541–1552 (2001)CrossRefGoogle Scholar
  6. 6.
    Pedregosa, et al.: Scikit-learn: machine Learning in Python. J. Mach. Learn. Res. 12, 2825–2830 (2011)MathSciNetzbMATHGoogle Scholar
  7. 7.
    DeST Developing Group: Building Environmental System Simulation and Analysis-DeST. China Architecture and Building Press, Beijing, China (2006)Google Scholar
  8. 8.
    Zhou, X., Hong, T.Z., Yan, D.: Comparison of HVAC system modeling in EnergyPlus, DeST and DOE-2.1E. In: Building Simulation, vol. 7, pp. 21–33 (2014)CrossRefGoogle Scholar
  9. 9.
    Fan, C., Xiao, F., Zhao, Y.: A short-term building cooling load prediction method using deep learning algorithms. Appl. Energy 195, 222–233 (2017)CrossRefGoogle Scholar

Copyright information

© Springer Nature Singapore Pte Ltd. 2020

Authors and Affiliations

  1. 1.Beijing Key Laboratory of Indoor Air Quality Evaluation and Control, School of ArchitectureTsinghua UniversityBeijingChina
  2. 2.Kechuang Jieneng Mechanical and Electrical Engineering Co. LtdYantaiChina

Personalised recommendations