Optimizing how wind turbines work with machine learning

October 14, 2014 by Ulrich Kreutzer
Wind parks produce their own air circulation dynamics. First row turbines get more wind than those in the middle. They also produce vortices that affect the performance of downstream turbines. Learning software can reduce these effects by optimizing rotor speeds and blade angles.

Machine learning helps make complex systems more efficient. Regardless of whether the systems in question are steel mills or gas turbines, they can learn from collected data, detect regular patterns, and optimize their own operations. Researchers at Siemens are demonstrating that continuous learning also allows wind turbines to increase their electricity output.

In his free time Volkmar Sterzing likes to work as a sailing instructor on Lake Starnberg south of Munich. A specialist in machine learning at Siemens Corporate Technology, Sterzing says that: "There are definitely parallels between sailing instruction and the machine learning process we use to optimize products." Whereas his pupils learn to understand the power of the wind and to intuitively know when and how they have to set their sails, Sterzing studies how such as wind turbines  can independently recognize regular patterns in collected data and thus learn how to optimize their operations.

Optimizing complex systems with machine learning

Siemens engineers have been studying machine learning for the past 25 years. "The associated processes offer many possibilities for making systems smarter and more efficient," says Professor Thomas Runkler, an expert for machine learning at Siemens Corporate Technology in Munich and a professor at the Technical University of Munich. "Siemens has used machine learning to optimize industrial facilities such as steel mills and gas turbines." Machine learning can also be used to reliably forecast the prices of energy and raw materials or to predict energy demand in entire regions.

Neural networks: Essential for machine learning

Machine learning specialist Volkmar Sterzing sees lots of similarities between learning how to improve the performance of wind turbines and sailboats.

A precondition for machine learning is the availability of computer systems that can learn from vast quantities of data and optimize their behavior accordingly. Among other things, Sterzing is studying how machine learning enables wind turbines to adjust themselves to fluctuating wind and weather conditions in order to increase their electricity output.

"Wind turbines optimize their output by comparing their operating data with weather data," says Sterzing. Sensors in and on such systems routinely record data regarding the direction and speed of the wind, temperatures, electric currents and voltages, as well as vibrations produced by major components such as the generator and the rotor blades. "Previously these sensor parameters were used only for remote maintenance and service diagnostics. But now they are also helping wind turbines generate more electricity," says Sterzing. Using , researchers at Siemens spent four years analyzing and modeling various dependencies and interrelationships. Neural networks are the key to successful machine learning in wind turbines. "Neural networks are computer models whose operations are similar to those of the human brain," explains Sterzing. They learn from examples, recognize patterns, and use past measurement data to make forecasts and ideal models regarding the future behavior of complex systems.

This is particularly applicable to wind turbines. On the basis of past measurement data, software calculates the optimal settings for various weather scenarios that involve a variety of factors such as sunshine duration, hazy conditions, and thunderstorms. The data is transmitted to the wind turbines' control units, which take it into account from then on as they adjust the functions. If familiar wind conditions arise, the control units immediately use the optimal settings that were ascertained as a result of . This can result in the adjustment of rotor blade angles, for example. "As a result, turbines become more and more efficient and produce more energy," says Sterzing.

But why should we optimize wind turbines? The answer is simple: Wind turbines don't always run at full capacity. They deliver far less energy than they potentially could – especially when the wind is weak or only blowing at half strength. "We want to increase their performance by a few percentage points in precisely these situations," Sterzing explains. This could further reduce the costs of wind energy in the future, and that's an important consideration, given that wind turbines are becoming more and more competitive compared to conventional energy generation systems.

The ALICE research project

Wind parks are improving their output thanks to learning systems that use archival data and current weather forecasts.

In the ALICE (Autonomous Learning in Complex Environments) research project, experts from Siemens, IdaLab GmbH, and the Machine Learning group at the Technical University of Berlin collected information on how to optimize wind turbines. The project, which was funded by Germany's Ministry of Education and Research, was completed in June 2014. Wind turbines demonstrated their learning abilities in field tests that were held as part of the project. In 2013 tests were conducted with up to eight turbines at two small offshore and onshore wind farms in Spain and Sweden. The wind power systems learned by using their own measurement data and achieved a noticeable increase in efficiency. However, it is still not quite clear by how much the efficiency increased, so additional larger-scale research projects are needed.

Thanks to sophisticated neural network-based algorithms developed by CT scientist Ralph Grothmann and his colleagues, Siemens wind turbines can learn from their own operating data, optimize their output, and predict upcoming conditions.

"Fall away!," Sterzing suddenly shouts out toward the lake. One of his pupils is sailing too close to the wind to reach the windward practice buoy. Even though he is sailing a shorter distance than the other pupils, it nonetheless takes him longer to get to the buoy because he is traveling more slowly. "The kids are learning how their boats behave in different weather conditions, and what they have to do in response—and our are doing something similar," Sterzing says.

Sterzing looks out over the lake. With great satisfaction, he observes that one of his pupils is now among the first to sail around the windward buoy. "Learning pays off," he says with a grin, thinking not only of sailing but also of Siemens. After all, Volkmar Sterzing's hobby is very closely connected to his profession.

“There are definitely parallels between sailing instruction and the machine learning process we use to optimize products," explains Volkmar Sterzing.

Explore further: Computer model optimizes wind farm

Related Stories

Computer model optimizes wind farm

July 25, 2011

A new software from Siemens will improve wind farms’ energy yields and extend their service life. When the wind causes the huge rotors to turn, it generates turbulence, which interferes with the operation of the wind ...

Using fluctuating wind power

March 25, 2013

Incorporating wind power into existing power grids is challenging because fluctuating wind speed and direction means turbines generate power inconsistently. Coupled with customers' varying power demand, many wind-farm managers ...

Automatic self-optimization of wind turbines

March 12, 2014

Siemens is "teaching" wind turbines how to automatically optimize their operation in line with weather conditions. The turbines are learning to use sensor data on parameters such as wind speed to make changes to their settings. ...

Recommended for you

Inferring urban travel patterns from cellphone data

August 29, 2016

In making decisions about infrastructure development and resource allocation, city planners rely on models of how people move through their cities, on foot, in cars, and on public transportation. Those models are largely ...

How machine learning can help with voice disorders

August 29, 2016

There's no human instinct more basic than speech, and yet, for many people, talking can be taxing. 1 in 14 working-age Americans suffer from voice disorders that are often associated with abnormal vocal behaviors - some of ...

Apple issues update after cyber weapon captured

August 26, 2016

Apple iPhone owners on Friday were urged to install a quickly released security update after a sophisticated attack on an Emirati dissident exposed vulnerabilities targeted by cyber arms dealers.

18 comments

Adjust slider to filter visible comments by rank

Display comments: newest first

julianpenrod
1 / 5 (3) Oct 14, 2014
I warned that converting wind energy into electricity meant moving air on the lee side of windmills had less energy. I was told by "science" devotees that I didn't understand conservation of energy. This article, at the top, admits that windmills in later ranks on wind farms are interacting with less energetic air. I warned that wind helps equalize temperature over wide areas and that wind farms interfere with this be removing the strength of moving air. I was laughed at by "science" devotees, but, now, it's admitted that air on the lee side of windmills is warmer than normal. The fact is that windmill and solar farms are causing climate change beyond even what "fossil fuels" would. No amount of "fossil fuels" ever changed the magnitude and direction of wind as much as wind farms or superheated the air and dust particles in it or prevented the formation of clouds the way solar farms do.
gkam
1 / 5 (3) Oct 14, 2014
Windmills grind grain. Electricity is produced by wind turbine-generators.

What we see is we need more dispersion of some sources. And they do not heat up the area, nor change the direction of the wind. As an former utility engineer, I will help you with your questions.

"Superheated the air"? Ridiculous. I should explain to you how it all works.
ekim
not rated yet Oct 14, 2014
Mountains, cities and trees also slow down the wind.
gkam
2.3 / 5 (3) Oct 14, 2014
No, mountains, and cities can funnel and channel the winds into higher speeds.
ekim
not rated yet Oct 14, 2014
No, mountains, and cities can funnel and channel the winds into higher speeds.

So can windmills in small localized areas, but the broad effect is the slowing of wind speed.
gkam
1 / 5 (3) Oct 14, 2014
"but the broad effect is the slowing of wind speed."

Yeah, that's the idea. It extracts much of the energy from that strip of wind for a while. What's the problem?
julianpenrod
1 / 5 (2) Oct 14, 2014
So easy to "prove" what you want by controlling the meanings of things. I didn't say windmills heat the air, but, but removing the velocity of moving air, they prevent areas from the heat equalizing effect of moving air masses. And it has been reported that temperatures downwind of windmills are warmer than normal. And the massive reflecting power of solar farm panels can both superheat particles of dust in the air and prevent cloud formation. And these effects are nowhere near, in nature or magnitude, what "fossil fuels" are accused of. Basically, things like the hate Mafia applying "1's" ratings solely out of viciousness, without legitimacy, and claims like gkam's that windmills slow wind "for a while" and that that has no effect erode respect for a plurality if not the majority of those who frequent PhysOrg's comment section.
gkam
1.3 / 5 (3) Oct 14, 2014
I did not say it had no effect. What is it with you folk? Have no science? I think you have little or no idea of the pollutants of fossil fuels or their gross amounts. Look up the amounts of particulates, the NOx, the radionuclides, from fossil plants.

Where is your source for heating the air? Ever been in a fossil plant of any kind?
julianpenrod
1 / 5 (3) Oct 14, 2014
gkam claims they did not say that slowing the speed of the wind has no effect. gkam said, "What's the problem?", and that means it has no effect. If something has an effect, that's the problem.
It was always obvious that winds farms, in reducing the moving air past them, impact effects like equalizing temperatures. But, if gkam needs to look because you cannot imagine that anyone who bought a ready made diploma from a diploma mill can figure things out, consider that, on April 30, 2012, Phys Org itself ran am article, "Wind farms lift the temperature in their region". They interfere with an impede the equalizing of temperature. gkam insists on saying that they heat the air, they prevent air from reaching a lower equilibrium.
And, the fact is, even "fossil fuels" didn't create manifestations like the massive turbulence and even redirecting of wind that wind farms can.
gkam
2 / 5 (4) Oct 15, 2014
I think some folk have no idea of scale. For every Btu of electricity put out by a fossil plant, two more go into the environment. For every Btu of electricity we get from a nuke plant, between three and four go into the environment. Tell me how many go into the environment as waste heat from a 5 MW wind turbine.
barakn
3 / 5 (2) Oct 16, 2014
Here's a short primer on how friction affects winds in and above the boundary layer. http://ww2010.atm...ric.rxml Looks to me like the added friction of wind farms could lead to faster collapse of pressure gradients.
ekim
not rated yet Oct 16, 2014
Here's a short primer on how friction affects winds in and above the boundary layer. http://ww2010.atm...ric.rxml Looks to me like the added friction of wind farms could lead to faster collapse of pressure gradients.

Wouldn't trees do the same thing?
ekim
not rated yet Oct 17, 2014
I think some folk have no idea of scale. For every Btu of electricity put out by a fossil plant, two more go into the environment. For every Btu of electricity we get from a nuke plant, between three and four go into the environment. Tell me how many go into the environment as waste heat from a 5 MW wind turbine.

Wouldn't the energy from a wind turbine already exist in the atmosphere prior to being harvested?
antialias_physorg
5 / 5 (1) Oct 17, 2014
but the broad effect is the slowing of wind speed

But you're not building windmills along broad swathes of land but at very specific points. And if there are points where a funnel effect increases windspeed, that's exactly where you put 'em.

Wouldn't trees do the same thing?

Yes. That's why they are planted along agricultural fields - to slow down wind and prevent soil erosion.
But you have to appreciate that air comes in layers, and that when wind speed is changed you also change air pressure. And when the pressure has nowhere to escape you get parts around the obstacles where the windspeed picks up noticeably.
Returners
1 / 5 (2) Oct 18, 2014
If you alternated altitude of the turbines, you could get more effectiveness out of the same area of land or ocean.

You could also alternate staggering by 1/3rd of 1 row width, and turbines in the prevailing wind direction will never interfere with one another.

8008
08008
8008
08008
8008
08008

notice how interferene along a diagonal is limited to no more than 2 consecutive, and iterference along veritical (prevailing) is limited to never 2 consecutive.

This is a better layout.
Returners
1 / 5 (3) Oct 18, 2014
More correctly drawn field:

8008008008
0000090000
0090000090
08008008008
0000090000
0000000000
8008008008
0000090000
0090000090
08008008008
000009
Assuming we break the offset into thirds, we can produce a scenario where interference is always less common than a square grid.

Offsetting by half doesn't work because that increases interference on diagonal winds.

"9's" are hypothetical extras where intereference is so unlikely compared to what the prevailing wind would be, that it is probably okay to add even more turbines. Notice how after all this, half the prevailing interference is removed, the diagonal interference is reduced overall, and you gain 8 more turbines, no more than half of which can ever be in interference at a time.

seanquallen
not rated yet Oct 21, 2014
I have to agree with gkam here, I think what's being missed is the idea of scale. Or maybe some basic fluid dynamics.

Yes, the streamwise velocity slows considerably (say, on the order of 50%) in the near wake of a turbine. That's the kinetic energy being extracted out of the wind. However, this slowing of wind takes place only in the circular rotor area. This is a very, very small volume compared to the ambient region surrounding it.
seanquallen
not rated yet Oct 21, 2014
No, mountains, and cities can funnel and channel the winds into higher speeds.

So can windmills in small localized areas, but the broad effect is the slowing of wind speed.


I fail to see, as all of us belong to the same planet and atmosphere, how turbines can actually be placed in "small localized areas". We're all at the bottom of an air ocean 100-some-odd miles deep.

And, as mentioned earlier, it is wind turbine. A windmill is a very different, obsolete, piece of machinery.

Please sign in to add a comment. Registration is free, and takes less than a minute. Read more

Click here to reset your password.
Sign in to get notified via email when new comments are made.