Building cooling energy consumption prediction with a hybrid simulation Approach: Generalization beyond the training range

Kwok Wai Mui, Manoj Kumar Satheesan, Ling Tim Wong

Research output: Journal article publicationJournal articleAcademic researchpeer-review

3 Citations (Scopus)

Abstract

Greenhouse gas emissions associated with energy consumption in buildings contribute significantly to climate change. In subtropical regions, a well-designed building envelope can effectively reduce the energy demand for space cooling. This study proposes a hybrid simulation approach that can be applied to diverse building types to estimate the annual cooling energy consumption. The proposed model, which is based on Bayesian regularization, demonstrates good generalization ability to predict energy consumption for residential as well as healthcare buildings. In the study, a genetic algorithm was employed to optimize the model parameters for obtaining the minimum or maximum envelope heat gain. It is recommended through this study that by adopting a combination, namely, (i) design parameters resulting in minimum envelope heat gain, (ii) coupling of an air change rate of 9 h−1 and recirculation ratio of 50 %, (iii) lowering lighting power density from 13 W/m2 to 7.3 W/m2, would prove to be an efficient strategy that balances energy and infection control for a general inpatient ward. The proposed generalized hybrid simulation approach can be a helpful tool for building systems engineers to formulate design guidelines or renovation plans to reduce energy consumption and thus greenhouse gas emissions.

Original languageEnglish
Article number112502
JournalEnergy and Buildings
Volume276
DOIs
Publication statusPublished - 1 Dec 2022

Keywords

  • Annual cooling energy prediction
  • Generalization
  • Healthcare facilities
  • Hybrid EP-ANN model
  • Optimization

ASJC Scopus subject areas

  • Civil and Structural Engineering
  • Building and Construction
  • Mechanical Engineering
  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Building cooling energy consumption prediction with a hybrid simulation Approach: Generalization beyond the training range'. Together they form a unique fingerprint.

Cite this