Js. Haberl et Te. Bousaada, PROCEDURES FOR CALIBRATING HOURLY SIMULATION-MODELS TO MEASURED BUILDING ENERGY AND ENVIRONMENTAL DATA, Journal of solar energy engineering, 120(3), 1998, pp. 193-204
This paper discusses procedures for creating calibrated building energ
y simulation programs. It begins with reviews of the calibration techn
iques that have been reported in the previous literature and presents
new hourly calibration methods including a temperature bin analysis to
improve hourly x-y scatter plots, a 24-hour weather-daytype bin analy
sis to allow for the evaluation of hourly temperature and schedule dep
endent comparisons, and a 52-week bin analysis to facilitate the evalu
ation of long-term trends. In addition, architectural rendering is rev
iewed as a means of verifying the dimensions of the building envelope
and external shading placement as seen by the simulation program. Seve
ral statistical methods are also presented that provide goodness-of-fi
t indicators, including percent difference calculations, mean bias err
or (MBE), and the coefficient of variation of the root mean squared er
ror (CV(RMSE)). The procedures are applied to a case study building lo
cated in Washington, D.C. where nine months of hourly whole-building e
lectricity data and site-specific weather data were measured and used
with the DOE-2.1D building energy simulation program to test the new t
echniques. Simulations that used the new calibration procedures were a
ble to produce an hourly MBE of -0.7% and a CV(RMSE) of 23.1% which co
mpare favorably with the most accurate hourly neural network models (K
reider and Haberl, 1994a, b).