Estimating unknown parameters of a building stock using a stochastic-deterministic-coupled approach

Interest in urban building energy modeling as a tool to evaluate effective management of urban building energy and carbon reduction policies is increasing. Since it takes a lot of time to model all buildings in the city with the lack of information about buildings, archetypes have been used. However, the deterministic method using archetypes does not reflect uncertainty with the disadvantage of not presenting design alternatives. To overcome such problems, we proposed an urban building energy modeling method using a stochastic-deterministic-coupled approach. In this paper, the identifiability of unknown parameters in a building stock was verified by analyzing the effect of the meta-model accuracy on the estimation of unknown parameters in the proposed urban building energy model. In addition, the possibility of using energy conservation measure analysis of the proposed model was evaluated. As a result, the posterior distribution of unknown input parameters was distorted when using the meta-model instead of the original building energy model in Bayesian calibration for building stock. However, the distorted posterior distribution could evaluate the effect of energy conservation measures within 0.71% errors. (c) 2021 Elsevier B.V. All rights reserved.

Automated summary

Interest in urban building energy modeling for evaluating effective urban energy management and carbon reduction policies is rising. Researchers proposed an urban building energy modeling method using a stochastic-deterministic-coupled approach to overcome limitations of using archetypes. The study verified the identifiability of unknown parameters in a building stock and evaluated the possibility of energy conservation measure analysis in the proposed model.