#### Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Options

# Electrical grid

Started Electrical grid.

• Options
1.
edited December 2010

We need a lot of info on this subject, eventually. At one point I wanted to learn a lot about the "smart grid", since it seems like it could use a lot of sophisticated mathematics. But I've gotten distracted by other projects!

Comment Source:Excellent! I added a bit more information. We need a lot of info on this subject, eventually. At one point I wanted to learn a lot about the "smart grid", since it seems like it could use a lot of sophisticated mathematics. But I've gotten distracted by other projects!
• Options
2.

Added some more. Not sure if a smart grid would need sophisticated maths or sophisticated software.

Comment Source:Added some more. Not sure if a smart grid would need sophisticated maths or sophisticated software.
• Options
3.

My understanding is that smart grid proposals are attempts to send signals through the grid to users in advance of events like bursts of renewable electricity or sudden demands from certain sectors (the "boil the kettle after TV programme X" problem), and have users (both domestic and industrial) interpret those signals in order to make best use of the currently available energy. This is both a general prediction and a prediction-with-control problem. However, because a lot of the entities involved are not described by parametric functions the modelling is done by simulations with "rule-based agents" in order to discern "emergent behaviour".

Comment Source:My understanding is that smart grid proposals are attempts to send signals through the grid to users in advance of events like bursts of renewable electricity or sudden demands from certain sectors (the "boil the kettle after TV programme X" problem), and have users (both domestic and industrial) interpret those signals in order to make best use of the currently available energy. This is both a general prediction and a prediction-with-control problem. However, because a lot of the entities involved are not described by parametric functions the modelling is done by simulations with "rule-based agents" in order to discern "emergent behaviour".
• Options
4.

My understanding is that the signals are about current grid conditions rather than in advance. Eg, when all the kettles go on, the fridges and car battery chargers turn themselves off.

Comment Source:My understanding is that the signals are about current grid conditions rather than in advance. Eg, when all the kettles go on, the fridges and car battery chargers turn themselves off.
• Options
5.
edited December 2010

I'm not sure how much of what I said is based on plans for the future rather than current near-term deployment: most of my knowledge comes from preparing for a job interview a couple of months ago for a university research project on developing SmartGrid technology. They were certainly looking at doing very large scale agent simulation for learning rules about emergent behaviours. The slightly worrying thing was that they didn't really have, or have plans to obtain in future, any real volume of real-world energy demand data to validate the prediction and control rules they were planning to learn.

Of course, that research may not translate into actual deployment.

Comment Source:I'm not sure how much of what I said is based on plans for the future rather than current near-term deployment: most of my knowledge comes from preparing for a job interview a couple of months ago for a university research project on developing SmartGrid technology. They were certainly looking at doing very large scale agent simulation for learning rules about emergent behaviours. The slightly worrying thing was that they didn't really have, or have plans to obtain in future, any real volume of real-world energy demand data to validate the prediction and control rules they were planning to learn. Of course, that research may not translate into actual deployment.