Paper Conference

Proceedings of eSim 2010: 6th Conference of IBPSA-Canada

     

DAILY LOAD PROFILE CLUSTERING: A TOOL FOR SIMULATION CALIBRATION

Michaël Fournier, Karine Lavigne, Eng Eng

Abstract: This paper presents a clustering algorithm of building's electrical load profiles and its custom made visualization tool. The algorithm groups daily load profiles based on their shape in order to produce a reduced set of typical profiles. The visualization tool shows these building's typical profiles and their distribution throughout the year. Sample cases are given to demonstrate the diagnostic capability of the clustering approach for the calibration of building energy simulations. The first case shows how the observed typical daily load profile for a specific time period can be used as an aggregated operation schedule indicator. The approach is also used to validate the transition from heating to cooling dominated thermal response of the building. Another example exposes higher than predicted night consumption which suggests an improper temperature setback. This analysis tool, allowing the comparison of simulated to measured data, has been implemented in a DOE2.1-based simulation software called SIMEB.
Pages: 64 - 71
Paper:
esim2010_3A1_064_071