• Title/Summary/Keyword: 데이터 드리븐 분석법

Search Result 1, Processing Time 0.015 seconds

Comparing Methodology of Building Energy Analysis - Comparative Analysis from steady-state simulation to data-driven Analysis - (건물에너지 분석 방법론 비교 - Steady-state simulation에서부터 Data-driven 방법론의 비교 분석 -)

  • Cho, Sooyoun;Leigh, Seung-Bok
    • KIEAE Journal
    • /
    • v.17 no.5
    • /
    • pp.77-86
    • /
    • 2017
  • Purpose: Because of the growing concern over fossil fuel use and increasing demand for greenhouse gas emission reduction since the 1990s, the building energy analysis field has produced various types of methods, which are being applied more often and broadly than ever. A lot of research products have been actively proposed in the area of the building energy simulation for over 50 years around the world. However, in the last 20 years, there have been only a few research cases where the trend of building energy analysis is examined, estimated or compared. This research aims to investigate a trend of the building energy analysis by focusing on methodology and characteristics of each method. Method: The research papers addressing the building energy analysis are classified into two types of method: engineering analysis and algorithm estimation. Especially, EPG(Energy Performance Gap), which is the limit both for the existing engineering method and the single algorithm-based estimation method, results from comparing data of two different levels- in other words, real time data and simulation data. Result: When one or more ensemble algorithms are used, more accurate estimations of energy consumption and performance are produced, and thereby improving the problem of energy performance gap.