|
|
|
Good
morning!
|
|
I am
very happy to have a chance today to help you in the IES community move
faster to meet some of the important challenges in front of us. In my view,
the technologies we are working on here might well decide whether the human
species manages to avoid extinction in the coming decades. This is not an
exaggeration; I will explain a bit today – and I hope I will help.
|
|
The electric power area – from motors and
chips to global energy economics – is a highly crossdisciplinary area in
itself. Major opportunities are being lost in the world today because of the
gaps in communication just within this area. In power electronics,
especially, we need to understand the real needs of the customers – the world
energy system – in order to anticipate and meet their needs in a more
proactive way. We need to do this, in order to get ahead of the curve and not
always lag behind, playing catch-up with our competitors.
|
|
But the neural network field is also a
large cross-disciplinary area, and most of what you read in the general
literature is not quite correct. There are new and powerful tools now
available, and there are also some pitfalls that are not heavily advertized.
They apply to all kinds of “prediction” and “control” tasks, including some
we think of as pattern recognition, data mining, planning, scheduling, state
estimation, sensor fusion, data compression, etc., etc.
|
|
Today, I will try to give you an overview
of how these two large areas come together. Because time is limited, I won’t
be able to give you all the equations you need to implement everything I talk
about – but I will tell you where to look for more detail and for a few of
the working examples. Most of my slides will come from two sources:
|
|
www.ieeeusa.org/policy/energy_strategy.ppt
and www.eas.asu.edu/~nsfadp.
|
|
I can
send you a few additional papers by email upon request.
|
|
The energy policy slides contain text
explanations. You can see the text in powerpoint by clicking on “Notes” under
“View,” or printing with the “notes” option in the print options window.
|
|
|