IEEE Power & Energy Magazine - May/June 2022 - 16

In 1956, a group of scientists participated in
the Dartmouth Summer Research Project on Artificial Intelligence,
widely recognized as the birth of AI as a research field.
control rooms today, either manually just like in the pre-PC
era, or in an automated fashion via computer programs with
sophisticated algorithms.
Some load forecasters also figured out other creative
ways. For example, a forecaster might take a bicycle ride
across the system territory every morning. Upon returning
to the office, he weighed his shirt, which carried much sweat
during a hot and humid day or little sweat during a chilly
and windy day. A heavier shirt indicated a higher peak load
for the day and vice versa. This physical method relied on
a human body to incorporate the effect of weather on the
system load. Today, we can capture weather effects using
comprehensive models with various weather variables.
AI's Inception, Two Waves,
and Two Winters
Long before the Middle Ages, Chinese, Indian, and Greek
philosophers and mathematicians had already begun to
investigate formal reasoning. In the 1930s, several mathematicians
and logicians attempted to formalize the notion of
computability, that is, the ability to solve a problem effectively.
In the 1940s, scientists started to investigate whether
machines could think like humans. A notable contributor was
Alan Turing, who published a series of papers on the topic of
intelligent machinery from the late 1940s to the early 1950s.
On the other hand, the foundation of biological neural networks
was built in the late 19th century by philosophers and
psychologists such as Alexander Bain and William James. Bain
and James independently proposed that interactions among
neurons within the brain result in thoughts and body activities.
The studies of mathematical logic and psychology eventually
led to the inception of AI. In 1943, neurophysiologist Warren
Sturgis McCulloch and logician Walter Pitts implemented the
first artificial neurons that modeled the basic logic units of a
brain. A collection of these artificial neurons is called an Artificial
Neural Network (ANN), which is inspired by the biological
neural networks that constitute animal brains. Figure 2
shows a simple three-layer, feedforward ANN, which has two
neurons in the input layer, three in the hidden layer, and one in
the output layer. In 1956, a group of scientists participated in
1879: Thomas Edison invented incandescent lightbulb
1882: Construction of Pearl Street Station
(a) Load Forecasting Events
1895: Niagara Falls Power Plant was built
1902: Willis Carrier invented air conditioner
1944: Henry Dryar's paper " The Effect of Weather on
the System Load " was published
1945: ENIAC was made
1930s:
Mathematicians and
logicians attempted to
formalize the notion
of computability
1940s:
Scientists started to
investigate whether a
machines can think
like humans
1943: First artificial neuron
1948: Intelligent machinery, by Alan Turing
1956: Dartmouth Summer Research Project
(b) AI Events
figure 1. The major events related to load forecasting and AI. ENIAC: The Electronic Numerical Integrator and Computer.
16
ieee power & energy magazine
may/june 2022
1974:
Backpropagation from Paul Werbos'
dissertation
1969: Perceptrons: An Introduction to Computational
Geometry, by Marvin Minsky and Seymour Papert
was published

IEEE Power & Energy Magazine - May/June 2022

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