Predictive capability testing and sensitivity analysis of a model for building energy efficiency
building energy modelling, energy consumption assessment, model sensitivity analysis, model predictive capability
Building energy modelling presents a good tool for estimating building energy consumption. Different modelling approaches exist in literature comprising white-box/physical/calculation-based models, black-box/statistical/measurement-based models or hybrid models combining the former two. Our work presented in this paper deals with a calculation-based quasi-steady-state model for building energy consumption based on the ISO 13790 standard and its implementation in MATLAB/Octave. The model is also well compared to the ISO 52016 standard updating ISO 13790. The model predictive capability is confirmed against both EnergyPlus dynamic simulator results and calculation results of a commercially available relevant tool used as benchmarks. Machine learning techniques are applied to a large dataset of simulated data and a sensitivity analysis is presented narrowing down to the most influential model parameters.
Tsinghua University Press
G. Kalogeras, S. Rastegarpour, C. Koulamas et al. Predictive capability testing and sensitivity analysis of a model for building energy efficiency. Build Simul, 2020, 13(1): 33–50.