An intelligent way to predict the building thermal needs: ANNs and optimization

作者：

Highlights：

• Artificial intelligence to forecast the thermal loads of the building.

• An alternative method to solve the traditional building thermal balance.

• Identification of fundamental inputs to reduce the computation time.

• The high reliability of the results guaranteed by a deep statistical analysis of errors.

• Performance validation of the neural model according to ASHRAE guidelines.

摘要

•Artificial intelligence to forecast the thermal loads of the building.•An alternative method to solve the traditional building thermal balance.•Identification of fundamental inputs to reduce the computation time.•The high reliability of the results guaranteed by a deep statistical analysis of errors.•Performance validation of the neural model according to ASHRAE guidelines.

论文关键词：Artificial Neural Network,Thermal energy demand,Forecast method,Sensitivity analysis,Statistical error analysis

论文评审过程：Received 28 January 2021, Revised 5 October 2021, Accepted 23 November 2021, Available online 3 December 2021, Version of Record 8 December 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2021.116293