Comparison of the prediction accuracy of daily and monthly regression models for energy consumption in commercial buildings /

(daily models) or by regressing the monthly energy

Bibliographic Details
Main Author: Wang, Jinrong, 1957-
Format: Thesis eBook
Language:English
Published: [Place of publication not identified] : [publisher not identified] ; 1996.
Subjects:
Online Access:Link to OAKTrust copy
Description
Summary:(daily models) or by regressing the monthly energy
average temperature during the operation period. Based on
buildings are generally determined as the difference between
consumption should be regressed versus the average ambient
consumption versus the monthly average temperature (monthly
daily energy consumption versus the daily average temperature
developed for a midsize commercial building with (i) dual-
different from the weather used for model development, the
duct CAV and VAV systems, (ii) office and university
errors for both daily and monthly regression models were as
errors were identified as the difference between the energy
high as 15%. However, the prediction error of daily
hours per day, annual energy prediction errors of daily
model.
models use weather from a year very different from the
models when the AHUs are operated 24 hours per day. When
models). Since the post-retrofit weather is generally
occupancy schedules, and (iii) different operating practices
of monthly regression models were found to be in the same
period. Most baseline models are developed by regressing the
prediction error of the baseline model may be different from
range as the error of the daily models. 2. When the AHUs
regression models can be decreased to a range of 2% to 3% if
regression models were found to be less than 1.4%. The errors
results are summarized below: 1. When the AHUs operate 24
shut-down is performed during unoccupied hours, daily energy
simulated by a calibrated simulation program when both
temperature during operating hours to develop the baseline
the daily average energy consumption is regressed versus the
the energy consumption predicted using a baseline model and
the fitting error. Daily and monthly baseline models were
the measured energy consumption during the post retrofit
The measured energy savings from retrofits in commercial
these findings, we suggest use of daily or monthly regression
use predicted by the regression models and the values
using the weather of a mild weather year. The prediction
weather year used to develop the regression model. The major
were shut down during unoccupied periods, annual prediction
Item Description:"Major subject: Mechanical Engineering".
Vita.
Physical Description:xxi, 164 leaves : illustrations ; 28 cm.
Also available online.
Issued also on microfiche from Lange Micrographics.
Bibliography:Includes bibliographical references.