3 ms·

How exactly does this need machine learning? Naively I would assume that it's somewhat directly correlated with wind speed (and little more, maybe direction) an

by tobias2014 8y ago

How exactly does this need machine learning? Naively I would assume that it's somewhat directly correlated with wind speed (and little more, maybe direction) and a very simple model based on that data will give you good predictions.

sgillen 8y ago
I think determining the wind speed and direction 36 hours in advance is the tricky bit.
- joe_the_user 8y ago
  Weather forecasting is difficult problem but it's problem that as far as I know has been worked on intensely for quite a while. Weather forecasting is based on physics-based models that, so far as I know, are essentially correct with the fundamental barrier to prediction being a combination of random effects and the difficulty of solving the complex equations numerically. Deep learning involves training a heuristically approximate a system based on past data. It has been used to emulate various human-learning behaviors like recognizing images or recognizing good or bad position in games. Essentially, the visible successes are in "we don't know but can extrapolate roughly from data" (or we "know" but can't easily program it, in the case of image recognition). So I'd be surprised if deep learning do a good or better job in situation where we do know how things work, we already running algorithm which is correct - the situation with weather.
steve76 8y ago
The grid wants a dedicated supply, so you have to store it in something first, as the dynamo hooked up to the rotor varies. Wind speed, power load nearby, wind forecast, storage capacity ... Should I engage the alternator now, even if it's spinning? Could that use up more energy than if I let it spin on a flywheel and engage later? Also repairs. What should I fix, and when? Does it make sense just to keep it offline in the winter when consumption is low? Then you get all the pricing data available. It's also looking for variables not known. Perhaps there's a bad actor, and for whatever reason, when he's there, losses occur.

sgillen 8y ago

I think determining the wind speed and direction 36 hours in advance is the tricky bit.

joe_the_user 8y ago
Weather forecasting is difficult problem but it's problem that as far as I know has been worked on intensely for quite a while. Weather forecasting is based on physics-based models that, so far as I know, are essentially correct with the fundamental barrier to prediction being a combination of random effects and the difficulty of solving the complex equations numerically. Deep learning involves training a heuristically approximate a system based on past data. It has been used to emulate various human-learning behaviors like recognizing images or recognizing good or bad position in games. Essentially, the visible successes are in "we don't know but can extrapolate roughly from data" (or we "know" but can't easily program it, in the case of image recognition). So I'd be surprised if deep learning do a good or better job in situation where we do know how things work, we already running algorithm which is correct - the situation with weather.

steve76 8y ago

The grid wants a dedicated supply, so you have to store it in something first, as the dynamo hooked up to the rotor varies. Wind speed, power load nearby, wind forecast, storage capacity ... Should I engage the alternator now, even if it's spinning? Could that use up more energy than if I let it spin on a flywheel and engage later? Also repairs. What should I fix, and when? Does it make sense just to keep it offline in the winter when consumption is low? Then you get all the pricing data available. It's also looking for variables not known. Perhaps there's a bad actor, and for whatever reason, when he's there, losses occur.