Skip to main content
SpringerLink
Log in

Enhancing source domain availability through data and feature transfer learning for building power load forecasting

  • Research Article
  • Advances in Modeling and Simulation Tools
  • Published:
Building Simulation Aims and scope Submit manuscript

Abstract

During the initial phases of operation following the construction or renovation of existing buildings, the availability of historical power usage data is limited, which leads to lower accuracy in load forecasting and hinders normal usage. Fortunately, by transferring load data from similar buildings, it is possible to enhance forecasting accuracy. However, indiscriminately expanding all source domain data to the target domain is highly likely to result in negative transfer learning. This study explores the feasibility of utilizing similar buildings (source domains) in transfer learning by implementing and comparing two distinct forms of multi-source transfer learning. Firstly, this study focuses on the Higashita area in Kitakyushu City, Japan, as the research object. Four buildings that exhibit the highest similarity to the target buildings within this area were selected for analysis. Next, the two-stage TrAdaBoost.R2 algorithm is used for multi-source transfer learning, and its transfer effect is analyzed. Finally, the application effects of instance-based (IBMTL) and feature-based (FBMTL) multi-source transfer learning are compared, which explained the effect of the source domain data on the forecasting accuracy in different transfer modes. The results show that combining the two-stage TrAdaBoost.R2 algorithm with multi-source data can reduce the CV-RMSE by 7.23% compared to a single-source domain, and the accuracy improvement is significant. At the same time, multi-source transfer learning, which is based on instance, can better supplement the integrity of the target domain data and has a higher forecasting accuracy. Overall, IBMTL tends to retain effective data associations and FBMTL shows higher forecasting stability. The findings of this study, which include the verification of real-life algorithm application and source domain availability, can serve as a theoretical reference for implementing transfer learning in load forecasting.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Similar content being viewed by others

Abbreviations

CNN:

convolutional neural network

CV-RMSE:

coefficient of variation-root mean square error

DANN:

domain antagonism neural network

DTW:

dynamic time warping

ED:

Euclidean distance

FBMTL:

feature-based multi-source transfer learning

GAN:

generative adversarial network

IBMTL:

instance-based multi-source transfer learning

JSD:

Jensen-Shannon divergence

LSTM:

long short-term memory

MAPE:

mean absolute percentage error

MK-MMD:

multi-core maximum mean discrepancy

MMD:

maximum mean discrepancy

RNN:

recurrent neural network

SC:

Spearman coefficient

References

  • Ahmad T, Chen H (2019). Deep learning for multi-scale smart energy forecasting. Energy, 175: 98–112.

    Article  Google Scholar 

  • Antoniadis A, Gaucher S, Goude Y (2023). Hierarchical transfer learning with applications to electricity load forecasting. International Journal of Forecasting, https://doi.org/10.1016/j.ijforecast.2023.04.006

  • Antonopoulos I, Robu V, Couraud B, et al. (2020). Artificial intelligence and machine learning approaches to energy demand-side response: A systematic review. Renewable and Sustainable Energy Reviews, 130: 109899.

    Article  Google Scholar 

  • Dai W, Yang Q, Xue G, et al. (2007). Boosting for transfer learning. In: Proceedings of the 24th International Conference on Machine learning, Corvalis, OR, USA.

  • Fan C, Sun Y, Xiao F, et al. (2020). Statistical investigations of transfer learning-based methodology for short-term building energy predictions. Applied Energy, 262: 114499.

    Article  Google Scholar 

  • Fan C, Lei Y, Sun Y, et al. (2022). Data-centric or algorithm-centric: Exploiting the performance of transfer learning for improving building energy predictions in data-scarce context. Energy, 240: 122775.

    Article  Google Scholar 

  • Fang X, Gong G, Li G, et al. (2021). A general multi-source ensemble transfer learning framework integrate of LSTM-DANN and similarity metric for building energy prediction. Energy and Buildings, 252: 111435.

    Article  Google Scholar 

  • Fang X, Gong G, Li G, et al. (2023). Transferability investigation of a Sim2Real deep transfer learning framework for cross-building energy prediction. Energy and Buildings, 287: 112968.

    Article  Google Scholar 

  • Freund Y, Schapire RE (1995). A decision-theoretic generalization of on-line learning and an application to boosting. In: Proceedings of the 2nd European Conference on Computational Learning Theory, Barcelona, Spain.

  • Gao Y, Ruan Y, Fang C, et al. (2020). Deep learning and transfer learning models of energy consumption forecasting for a building with poor information data. Energy and Buildings, 223: 110156.

    Article  Google Scholar 

  • Gu Q, Dai Q, Yu H, et al. (2021). Integrating multi-source transfer learning, active learning and metric learning paradigms for time series prediction. Applied Soft Computing, 109: 107583.

    Article  Google Scholar 

  • Guo Z, Zhou K, Zhang X, et al. (2018). A deep learning model for short-term power load and probability density forecasting. Energy, 160: 1186–1200.

    Article  Google Scholar 

  • Hochreiter S, Schmidhuber J (1997). Long short-term memory. Neural Computation, 9: 1735–1780.

    Article  Google Scholar 

  • IRENA (2023). Energy transition outlook. International Renewable Energy Agency. Available at https://www.irena.org/Energy-Transition/Outlook

  • Johannesen NJ, Kolhe M, Goodwin M (2019). Relative evaluation of regression tools for urban area electrical energy demand forecasting. Journal of Cleaner Production, 218: 555–564.

    Article  Google Scholar 

  • Kang HJ, Rhee EK (2012). A development of energy load prediction equations for multi-residential buildings in Korea. Journal of Asian Architecture and Building Engineering, 11: 383–389.

    Article  Google Scholar 

  • Kebede AA, Kalogiannis T, Van Mierlo J, et al. (2022). A comprehensive review of stationary energy storage devices for large scale renewable energy sources grid integration. Renewable and Sustainable Energy Reviews, 159: 112213.

    Article  Google Scholar 

  • Kiruthiga D, Manikandan V (2023). Levy flight-particle swarm optimization-assisted BiLSTM + dropout deep learning model for short-term load forecasting. Neural Computing and Applications, 35: 2679–2700.

    Article  Google Scholar 

  • Kumar S, Shukla AK, Muhuri PK (2021). Anomaly based novel multi-source unsupervised transfer learning approach for carbon emission centric GDP prediction. Computers in Industry, 126: 103396.

    Article  Google Scholar 

  • Li A, Xiao F, Fan C, et al. (2021). Development of an ANN-based building energy model for information-poor buildings using transfer learning. Building Simulation, 14: 89–101.

    Article  Google Scholar 

  • Li C, Li G, Wang K, et al. (2022a). A multi-energy load forecasting method based on parallel architecture CNN-GRU and transfer learning for data deficient integrated energy systems. Energy, 259: 124967.

    Article  Google Scholar 

  • Li J, Zhang C, Zhao Y, et al. (2022b). Federated learning-based short-term building energy consumption prediction method for solving the data silos problem. Building Simulation, 15: 1145–1159.

    Article  Google Scholar 

  • Li K, Wei B, Tang Q, Liu Y (2022c). A data-efficient building electricity load forecasting method based on maximum mean discrepancy and improved TrAdaBoost algorithm. Energies, 15: 8780.

    Article  Google Scholar 

  • Liu H, Tang Y, Pu Y, et al. (2022). Short-term load forecasting of multi-energy in integrated energy system based on multivariate phase space reconstruction and support vector regression mode. Electric Power Systems Research, 210: 108066.

    Article  Google Scholar 

  • Liu L, Wang J, Li J, et al. (2023). An online transfer learning model for wind turbine power prediction based on spatial feature construction and system-wide update. Applied Energy, 340: 121049.

    Article  Google Scholar 

  • Lopez Santos M, Diaz Garcia S, Garcia-Santiago X, et al. (2023). Deep learning and transfer learning techniques applied to short-term load forecasting of data-poor buildings in local energy communities. Energy and Buildings, 292: 113164.

    Article  Google Scholar 

  • Lu Y, Tian Z, Zhou R, et al. (2021). A general transfer learning-based framework for thermal load prediction in regional energy system. Energy, 217: 119322.

    Article  Google Scholar 

  • Lu Y, Wang G, Huang S (2022a). A short-term load forecasting model based on mixup and transfer learning. Electric Power Systems Research, 207: 107837.

    Article  Google Scholar 

  • Lu F, Zhou G, Liu Y, et al. (2022b). Ensemble transfer learning for cutting energy consumption prediction of aviation parts towards green manufacturing. Journal of Cleaner Production, 331: 129920.

    Article  Google Scholar 

  • Lu H, Wu J, Ruan Y, et al. (2023). A multi-source transfer learning model based on LSTM and domain adaptation for building energy prediction. International Journal of Electrical Power & Energy Systems, 149: 109024.

    Article  Google Scholar 

  • Ma J, Cheng JCP, Jiang F, et al. (2020). A bi-directional missing data imputation scheme based on LSTM and transfer learning for building energy data. Energy and Buildings, 216: 109941.

    Article  Google Scholar 

  • Ma H, Wang H, Yan Z, et al. (2022). Evaluating peak-regulation capability for power grid with various energy resources in Chinese urban regions via a pragmatic visualization method. Sustainable Cities and Society, 80: 103749.

    Article  Google Scholar 

  • Mocanu E, Nguyen PH, Kling WL, et al. (2016). Unsupervised energy prediction in a Smart Grid context using reinforcement cross-building transfer learning. Energy and Buildings, 116: 646–655.

    Article  Google Scholar 

  • Peng C, Tao Y, Chen Z, et al. (2022). Multi-source transfer learning guided ensemble LSTM for building multi-load forecasting. Expert Systems with Applications, 202: 117194.

    Article  Google Scholar 

  • Qian F, Gao W, Yang Y, et al. (2020). Potential analysis of the transfer learning model in short and medium-term forecasting of building HVAC energy consumption. Energy, 193: 116724.

    Article  Google Scholar 

  • Ribeiro M, Grolinger K, ElYamany HF, et al. (2018). Transfer learning with seasonal and trend adjustment for cross-building energy forecasting. Energy and Buildings, 165: 352–363.

    Article  Google Scholar 

  • Ruan Y, Ma J, Meng H, et al. (2023). Potential quantification and impact factors analysis of energy flexibility in residential buildings with preheating control strategies. Journal of Building Engineering, 78: 107657.

    Article  Google Scholar 

  • Sun S, Chen H (2021). Data-driven sensitivity analysis and electricity consumption prediction for water source heat pump system using limited information. Building Simulation, 14: 1005–1016.

    Article  Google Scholar 

  • Wu D, Wang B, Precup D, et al. (2020). Multiple kernel learning-based transfer regression for electric load forecasting. IEEE Transactions on Smart Grid, 11: 1183–1192.

    Article  Google Scholar 

  • Yue W, Liu Q, Ruan Y, et al. (2022). A prediction approach with mode decomposition-recombination technique for short-term load forecasting. Sustainable Cities and Society, 85: 104034.

    Article  Google Scholar 

  • Zhang L, Chen Z, Zhang X, et al. (2023). Challenges and opportunities of machine learning control in building operations. Building Simulation, 16: 831–852.

    Article  Google Scholar 

  • Zhao B, Cheng C, Zhang G, et al. (2022). An instance and feature-based hybrid transfer model for fault diagnosis of rotating machinery with different speeds. IEEE Transactions on Instrumentation and Measurement, 71: 1–12.

    Google Scholar 

  • Zhu C, Ni J, Yang Z, et al. (2022). Bandgap prediction on small thermoelectric material dataset via instance-based transfer learning. Computational and Theoretical Chemistry, 1217: 113872.

    Article  Google Scholar 

Download references

Acknowledgements

This research was supported by the National Key Research and Development Program of China (No. 2023YFC3807102).

Author information

Authors and Affiliations

  1. College of Mechanical and Energy Engineering, Tongji University, Shanghai, 200092, China

    Fanyue Qian, Yingjun Ruan, Hua Meng & Tingting Xu

  2. United Automotive Electronic Systems Co., Ltd, Shanghai, 201206, China

    Huiming Lu

Authors
  1. Fanyue Qian
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Yingjun Ruan
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Huiming Lu
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Hua Meng
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Tingting Xu
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding authors

Correspondence to Yingjun Ruan or Huiming Lu.

Ethics declarations

The authors have no competing interests to declare that are relevant to the content of this article.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Qian, F., Ruan, Y., Lu, H. et al. Enhancing source domain availability through data and feature transfer learning for building power load forecasting. Build. Simul. 17, 625–638 (2024). https://doi.org/10.1007/s12273-023-1087-0

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s12273-023-1087-0

Keywords

Search

Navigation