IEEE Power & Energy Magazine - May/June 2022 - 23

The accuracy reported in one paper could be for a
specific area during a certain period, which may not
be generalized to other jurisdictions.
network in 2001. In the 2010s, a series of three energy forecasting
competitions, known as Global Energy Forecasting
Competitions (GEFComs), was organized, attracting
hundreds of participants worldwide. These competitions
also stimulated interest from the industry, resulting in
many other smaller competitions organized by power
companies. In the past, good models might be buried
by many other mediocre papers. Currently, good models
have many opportunities to show up in the leaderboards
of various competitions.
Before the 2010s, load forecasting, or energy forecasting
in general, was a minority subject in power engineering.
Energy forecasters did not have a professional group as
their home. In 2011, the IEEE Working Group on Energy
Forecasting was formed by the IEEE Power & Energy Society.
The group actively promoted energy forecasting by
organizing the three GEFComs, delivering panel sessions
at major power and energy conferences, teaching tutorials,
and so forth. In 2019, the International Institute of Forecasters
started the Section on Water, Energy, and Environment
(SWEET) to bring together forecasters interested in
issues around water, energy, and environment. Today, load
forecasters can easily find a welcoming community to network
with other forecasters.
Together
Traditionally, the load forecasting job is done by power
system operators and planners in power companies. However,
the aforementioned load forecasting competitions indicate
winning contributions from internationally recognized
experts who are not power system operators or planners.
Their mastery of techniques from fields such as statistics,
econometrics, and computer science allowed them to develop
models that proved superior to the competition.
Newcomers to the field of load forecasting often ask,
" What is the most accurate technique for load forecasting? "
The short and sweet answer should be " None. " No
technique is superior to all other methods in load forecasting.
What matters in load forecast accuracy depends more
on the mastery of applying the technique than the technique
itself. For example, an ANN expert and OLS regression
specialist in a competition may eventually obtain similarly
accurate forecasts, while their forecasts are likely to
be better than the forecasts developed by people without
in-depth experience applying any specific technique or
tuning the models.
may/june 2022
To accelerate the improvement of load forecasting
practices, talented people with experience beyond power
companies deserve consideration. As listed in Table 2, the
winning teams of those five competitions came from eight
different organizations. Only one of the eight organizations
is a power company. The other seven include a technology
start-up, a weather company, and five universities. Load
forecasting is an interdisciplinary subject. Building industry-academia
partnerships as well as collaborations among
the energy and other communities, such as meteorology
and data science, may provide new insights and lead to
further breakthroughs. By working together, better models
may be realized.
For Further Reading
G. Gross and F. D. Galiana, " Short-term load forecasting, "
Proc. IEEE, vol. 75, no. 12, pp. 1558-1573, Dec. 1987, doi:
10.1109/PROC.1987.13927.
R. Ramanathan, R. Engle, C. W. Granger, F. VahidAraghi,
and C. Brace, " Short-run forecasts of electricity
loads and peaks, " Int. J. Forecasting, vol. 13, no. 2, pp. 161-
174, 1997, doi: 10.1016/S0169-2070(97)00015-0.
A. Khotanzad, R. Afkhami-Rohani, and D. Maratukulam,
" ANNSTLF-artificial neural network short-term load forecaster-generation
three, " IEEE Trans. Power Syst., vol. 13,
no. 4, pp. 1413-1422, Nov. 1998, doi: 10.1109/59.736285.
B.-J. Chen, M.-W. Chang, and C.-J. lin, " Load forecasting
using support vector Machines: A study on EUNITE
competition 2001, " IEEE Trans. Power Syst., vol. 19, no.
4, pp. 1821-1830, Nov. 2004, doi: 10.1109/TPWRS.2004.
835679.
T. Hong, P. Pinson, S. Fan, H. Zareipour, A. Troccoli, and
R. J. Hyndman, " Probabilistic energy forecasting: Global
energy forecasting competition 2014 and beyond, " Int. J.
Forecasting, vol. 32, no. 3, pp. 896-913, Jul./Sep. 2016, doi:
10.1016/j.ijforecast.2016.02.001.
T. Hong and S. Fan, " Probabilistic electric load forecasting:
A tutorial review, " Int. J. Forecasting, vol. 32, no. 3, pp.
914-938, Jul./Sep. 2016, doi: 10.1016/j.ijforecast.2015.11.011.
Biographies
Tao Hong is with the University of North Carolina at Charlotte,
Charlotte, North Carolina, 28223, USA.
Pu Wang is with Duke Energy Corporation, Charlotte,
North Carolina, 28202, USA.
p&e
ieee power & energy magazine
23
IEEE Power & Energy Magazine - May/June 2022

Table of Contents for the Digital Edition of IEEE Power & Energy Magazine - May/June 2022
