Genetic algorithm for building envelope calibration

Germán Ramos Ruiz

doi:10.1016/j.apenergy.2016.01.075

2016-04-01

PRE

Germán Ramos Ruiz *

Carlos Fernández Bandera

T. Gómez-Acebo

Ana Sánchez-Ostiz Gutiérrez

DOI：10.1016/j.apenergy.2016.01.075

原文链接

原文求助

摘要

En 中文

Buildings today represent 40% of world primary energy consumption and 24% of greenhouse gas emissions. In our society there is growing interest in knowing precisely when and how energy consumption occurs. This means that consumption measurement and verification plans are well-advanced. International agencies such as Efficiency Valuation Organization (EVO) and International Performance Measurement and Verification Protocol (IPMVP) have developed methodologies to quantify savings. This paper presents a methodology to accurately perform automated envelope calibration under option D (calibrated simulation) of IPMVP - vol. 1. This is frequently ignored because of its complexity, despite being more flexible and accurate in assessing the energy performance of a building. A detailed baseline energy model is used, and by means of a metaheuristic technique achieves a highly reliable and accurate Building Energy Simulation (BES) model suitable for detailed analysis of saving strategies. In order to find this BES model a Genetic Algorithm (NSGA-II) is used, together with a highly efficient engine to stimulate the objective, thus permitting rapid achievement of the goal. The result is a BES model that broadly captures the heat dynamic behaviour of the building. The model amply fulfils the parameters demanded by ASHRAE and EVO under option D. (C) 2016 Elsevier Ltd. All rights reserved.

Keyword:

Calibration

Energy simulation

Parametric analysis

Sensitivity analysis

Genetic algorithm (NSGA-II)

Energy savings