research news no. 4 – 2010 intelligent networking of wind farms · 2020-05-10 · intelligent...

Fraunhofer PressPhone: +49 89 [email protected]/press

Intelligent networking of wind farmsThe EU member states have passed a resolution requiring one fifth oftheir electricity needs to be derived from renewable energy sources by2020, with wind farms covering a large part of this requirement. Resear-chers have succeeded in interconnecting large wind farms in clusters.

Savvy injection moldingWith the help of neural networks, in which complex algorithms areused to monitor critical process steps, engineers are paving the way forzero-defect production in the area of metal powder injection molding.The gain for manufacturers is less waste combined with time savings.

A smart way to charge upElectromobility makes sense only if car batteries are charged using elec-tricity from renewable energy sources. But the supply of green electricityis not always adequate. An intelligent charging station can help, by ad-apting the recharging times to suit energy supply and network capacity.

Tracking down rustDamage to concrete bridges caused by rust can have fatal consequen-ces, at worst leading to a total collapse. Now, researchers have develo-ped an early-warning system for rust. Sensor-transponders integratedin the concrete allow the extent of corrosion to be measured.

Detecting defects with wind and waterBridges, aircraft and wind turbines are in constant movement. Naturalforces and pedestrians all create vibrations. Previously, time-consumingtests were needed to determine how building components would re-act to vibrations. Now, researchers have developed a simpler method.

Multimedia search without detoursFinding a particular song or video is often no easy matter. Manually as-signed metadata may be incorrect, and the unpacking of compresseddata can slow up the search. DIVAS, a multimedia search engine, usesdigital fingerprints to reliably locate what you are looking for.

Smart career planningWho’ll be the best fit for the company? While human resources ex-perts often find it difficult to identify suitable candidates or evaluateexisting employees’ talents and skills, assessment centers undoubtedlyprovide a useful tool for appraising competences.

No. 4 – 2010

1

2

3

4

5

6

7

Research News

Fraunhofer Institute for WindEnergy and Energy SystemTechnology IWESBranch KasselKönigstor 5934119 Kassel, GermanyPress contact:Uwe KrengelPhone +49 561 7294-345Fax: +49 561 [email protected]

© RED Electrica de Espana (REE)

In future, Fraunhofer's Wind Farm Cluster ManagementSystem will enable Spanish grid operator REE to steer windfarms from its central control station.

Picture in color and printing quality: www.fraunhofer.de/press

Intelligent networking of wind farmsThe future belongs to renewable energies. Wind energy is one of themain forms of renewable energy experiencing a boom. According tothe German Wind Energy Association, 952 new facilities were installedin Germany in 2009. The German Renewable Energy Federation pre-dicts that 47 percent of total electricity consumption in 2020, estima-ted at 595 terawatt hours, will be covered by renewable energysources, with wind energy accounting for one quarter.

»Wind on the Grid«, one of the EU's biggest network integration pro-jects, has just been completed. A group of European industrial compa-nies and research organizations teamed up to investigate how windfarms on the Iberian peninsula could be safely integrated in the Euro-pean electricity grid on a large scale. To assist grid operators in the cap-ture, control and forecasting of wind energy, the Fraunhofer Institutefor Wind Energy and Energy System Technology IWES has made avail-able both its Wind Farm Cluster Management System (WCMS) and itsWind Power Management System, adding new functions to both. Theresearchers used these software packages to integrate five wind farmsin Portugal with a total capacity of 204 megawatts and six in Spainwith a total capacity of 107 megawatts in the power grid. The integra-tion was achieved in real-time tests under a variety of weather condi-tions. »We used the WCMS to link the scattered wind farms in a clus-ter, enabling them to be controlled by the central control station of thePortuguese and Spanish power utilities respectively. While the WCMSkeeps both the frequency and the voltage of the electricity grid con-stant, thus ensuring safe operation, Fraunhofer's Wind Power Man-agement System forecasting software uses artificial neural networks tocalculate expected wind energy on the basis of predicted weather pat-terns,« explains Dr. Kurt Rohrig, department head at the Kassel branchof the IWES. Individual wind farms are subject to wide fluctuations inoutput The more wind farms that can be interconnected in a cluster,the easier it is to balance out the fluctuations caused by variations inwind force, from gusty winds to totally calm conditions. And the higherthe number of installed facilities, the lower the energy price. »The priceof electricity produced using wind energy is currently seven cents perkilowatt-hour; by 2025 it should be around four cents,« says Rohrig.

Fraunhofer researchers are negotiating with the Portuguese grid oper-ator with the aim of integrating their software in the latter's controlsystem. And Kurt Rohrig is convinced: »In the long term, wind farmswill replace traditional power plants.«

Research NewsTopic 1

No. 4 – 2010

For further information:Dr.-Ing. Kurt RohrigPhone +49 561 [email protected]

Fraunhofer Institute for Manufacturing and AdvancedMaterials IFAMWiener Strasse 1228359 Bremen, GermanyPress contact:Martina OhlePhone +49 421 5665-404Fax +49 421 [email protected]

© Fraunhofer IFAM

IFAM researchers inspecting components produced usingmetal injection molding.


Savvy injection moldingThe metal components used in the hinges of spectacle frames, surgicalinstruments or artificial heart valves are often very small. For someyears now, manufacturers of components with complex geometries ofthis type have relied on a special production process: metal injectionmolding. Things can occasionally go awry during production, and thenit is often impossible to detect defects until after sintering, the final stepin the process chain, by which time it is too late to correct the defect.

Now, researchers at the Fraunhofer Institute for Manufacturing andAdvanced Materials IFAM are working towards achieving zero-defectproduction. Their idea is that, at any time during the molding process,the system should be able to monitor all parameters – such as weight,pressure and temperature – and to deliver a verdict on the quality ofthe component. »In this way, errors, dimensional inaccuracies and de-fects such as cracks, warps or cavities can be detected on line,« ex-plains IFAM project manager Dr. Thomas Hartwig. »This will allow themanufacturer to respond immediately by changing the relevant set-tings.« In the long run the system can, if required, even be pro-grammed to alter the parameters fully automatically. The necessarytechnical support is provided by a neural network developed for metalinjection molding (MIM) in a joint effort by the IFAM engineers and al-gorithmica technologies, a private company. »The neural network isbased on highly complex algorithms,« says Hartwig. »Its advantageover existing solutions is that it is self-learning.« After a mandatory ini-tial training phase it can interpret all the measured data in the system,detecting correlations between them that would be impossible to findwithout the network. All information of relevance to the manufacturercan be given by the process control system, for instance the finalweight of the component if the pressure or temperature is changed ata given step in the process.

»Our goal with neural networks is to reduce reject rates by at least 50percent,« says Hartwig. »This represents a huge cost saving for manu-facturers because the raw materials are so expensive. Until now, com-panies often had to reject large numbers of components in the firstfew days before quality requirements could be met again.« Anotheradvantage of neural networks is that they could eventually make quali-ty checks superfluous and could also be deployed in other types of se-ries production such as die-casting in the light-metal industry. Havingsuccessfully produced a test component with the aid of neural net-works, the researchers are now looking for industrial partners.


No. 4 – 2010

For further information:Dr. Thomas HartwigPhone +49 421 2246-156Fax +49 421 [email protected]

Fraunhofer Institute for Solar Energy Systems ISEHeidenhofstrasse 279110 Freiburg, GermanyPress contact:Karin SchneiderPhone +49 761 4588-5150Fax +49 761 [email protected]

© Fraunhofer ISE

Customers can enter their requirements via a touchscreenon the charging station. Costs and energy readings can alsobe displayed during the recharging process.


A smart way to charge upGermany aims to have one million electric vehicles – powered by ener-gy from renewable sources –on the road by 2020. And, within tenyears, the German environment ministry expects »green electricity« tomake up 30 percent of all power consumed. Arithmetically speaking, itwould be possible to achieve CO2-neutral electromobility. But, in reali-ty, it is a difficult goal to attain. As more and more solar and wind en-ergy is incorporated in the power grid, the proportion of electricity thatcannot be controlled by simply pressing a button is on the increase. Inaddition, there is a growing risk that the rising number of electric vehi-cles will trigger extreme surges in demand during rush hour.

»What we need is a smart grid that carries information in addition topower,« says Dominik Noeren of the Fraunhofer Institute for Solar En-ergy Systems ISE. The structure of the grid has to change from a pushsystem based on energy demand to a pull system based on productionoutput. In Noeren's opinion, »electric cars are best equipped to meetthis challenge.« Introduced in large numbers, they have the capacity tostore a lot of energy. On average, a car is parked for at least 20 hoursout of 24. That is more than enough time to recharge them when thewind picks up or the demand for electricity is low.

Developed by Fraunhofer researchers, the »smart« charging station isa device that enables electric vehicles to recharge when the systemload is low and the share of energy from renewable resources is high.In this way, load peaks can be avoided and the contribution of solarand wind power fully exploited. »For us, it is important that end con-sumers are completely free to decide when they want to recharge. Wedo not want them to suffer any disadvantages from the controlledrecharging of their vehicles' batteries,« Noeren emphasizes. That's whyhe favors electricity rates that adapt to the prevailing situation in thepower grid – ones that are more expensive in periods of peak demandand particularly cheap when there is a surfeit of renewable energy. Theperson using the »smart« charging station could then choose betweenrecharging immediately or opting for a cheaper, possibly longer,recharging time. If they go for the second option, all they need to do isenter the time when their vehicle has to be ready to drive again. Thecharging station takes care of everything else, calculating the costs andcontrolling the recharging process. Via the display the user can track theprogress of recharging and also see the costs incurred and the amountof energy used. The experts will be presenting their charging device atthe Hannover Messe from April 19 through 23 (Hall 27, Stand K55).


No. 4 – 2010

For further information:Dominik NoerenPhone +49 761 4588-5455Fax +49 761 [email protected]

Fraunhofer Institute for Micro-electronic Circuits and Systems IMSFinkenstraße 6147057 Duisburg, GermanyPress contact:Dipl.-Ing. Martin van AckerenPhone +49 203 3783-130Fax +49 203 [email protected]

© Fraunhofer IMS

The passive sensor-transponder continuously monitors theprotective coating of concrete bridges, allowing corrosiondamage to be detected at an early stage.


Tracking down rustConcrete bridges have to be strong enough to cope with a wide vari-ety of different impacts: frost, heavy traffic and emissions all take theirtoll on these structures. And then there are the various types of roadsalt used in winter to combat icy roads. The most common of these issodium chloride, which is deployed in large amounts on Germany'sroads. When the ice thaws, these salts break down into their ioniccomponents that penetrate the concrete, destroying its five-centimeterthick protective alkaline layer. Any salt that leaches through to the steelrods used to reinforce the concrete pad will cause them to rust, result-ing in structural damage. The result is cracks. In a worst-case scenariothe bridge itself could collapse.

Until now there have been no effective tests to determine how deepthe ions have penetrated the concrete and what damage they have al-ready caused. Current practice is time-consuming and involves con-struction workers hammering on the reinforced concrete in search ofcavities, which are a sure sign of corrosion damage. But experts at theFraunhofer Institute for Microelectronic Circuits and Systems IMS inDuisburg have now hit upon a more reliable and cost-effective methodfor detecting rust corrosion at an early stage. With a new sensor-transponder they can continuously measure and monitor how deepthe ions have penetrated the concrete. While the sensor was devel-oped by the building materials testing facility in Braunschweig (MPABraunschweig), the integrated passive wireless transponder system isthe work of IMS researchers. The sensor itself is crisscrossed by veryfine iron wires, laid down at even distances. »If the dissolved saltsreach the iron wires, these begin to corrode and break. The number ofdefective iron wires is an indicator of the extent of corrosion and thedepth to which the concrete's protective layer has been penetrated.This allows us to determine when the next repair work needs to be car-ried out,« explains Frederic Meyer, a researcher at the IMS. Thetransponder transmits the measured data by wireless to the readingdevice carried by the construction workers. »Our transponder does notget the energy it needs to measure the corrosion from a battery, butfrom a magnetic field. This means it does not need to be replaced andcan remain within the concrete structure permanently,« says Meyer.

The first field tests are already underway, with the sensor-transponderbeing integrated and put through its paces in a test bridge constructedby the MPA Braunschweig. The researchers will be exhibiting a prototypeat the EURO ID trade fair, to be held in Cologne from May 4 through 6.


No. 4 – 2010

For further information:Frederic MeyerPhone +49 203 3783-193Fax +49 203 [email protected]

Fraunhofer Institute for Structural Durability and System Reliability LBFBartningstrasse 4764289 Darmstadt, GermanyPress contact:Anke Zeidler-FinselPhone +49 6151 705-268Fax +49 6151 [email protected]

© Fraunhofer LBF

Will the support of the wind turbine withstand wind andweather? LBF researchers are using environmental impactsto determine the structural health of building components.


Detecting defects with wind and water A narrow iron bridge is perched high above the river. A pedestrian on itcan feel quite clearly how it vibrates the moment another person setsfoot on the bridge or rides onto it on a bicycle. To ensure that peoplecan cross the bridge in safety, time-consuming methods are employedto determine what condition the structure is in and whether it is liableto collapse. »The vibration response of bridges is often tested using gi-gantic drop hammers,« says Dr. Dirk Mayer of the Fraunhofer Institutefor Structural Durability and System Reliability LBF in Darmstadt. DirkMayer and his team are currently investigating means of employingnatural phenomena such as wind and water to carry out an automaticanalysis of a structure's vibrational behavior. Thanks to Output OnlyModal Analysis, time-consuming and expensive tests with drop ham-mers could soon be a thing of the past.

»Continuous vibration monitoring is important not only for bridges,but also for ships, wind turbines or airplanes,« explains the Fraunhoferexpert. »In the case of aircraft, electrodynamic shakers integratedunder the wing surfaces are used to detect defects.« In contrast toconventional testing methods, the scientists are using neither ham-mers nor shakers – instead, their only tools are environmental impacts.In order to deploy the wind's capacity to cause vibrations, the expertsattached two acceleration sensors to a model wind turbine. Two com-puters were used to process the data measured. »It was necessary tohave two because of the huge amount of data that quickly accumu-lates at sampling rates as low as 100 per second,« explains Dirk Mayer.»While one computer preprocesses the signals and compresses thedata, the second carries out another evaluation and determines theresonant frequency. If the frequency diminishes over time, it means thatthe stiffness of the structure is decreasing. Testers know what the rigidityor damping readings signify for their particular components or ma-chines. »In our test we can determine whether the support of the modelwind turbine is defective. In the case of bridges this method makes itpossible to determine the condition of structural components by evaluat-ing the vibrations caused by pedestrians over a period of several weeks.«

Going forward, the researchers would like to offer this procedure as aservice. They are also considering how unwanted vibrations could beactively dampened. The Fraunhofer experts will be presenting this newmethod of analysis using a model aircraft at the joint Adaptronicsstand at the Hannover Messe (Hall 2, Stand D26). The eventual aim isto monitor airliners during operation.


No. 4 – 2010

For further information:Dr.-Ing. Dirk MayerPhone +49 6151 [email protected]

Fraunhofer Institute for Digi-tal Media Technology IDMTEhrenbergstr. 3198693 Ilmenau, GermanyPress contact:Julia EdlingPhone +49 3677 467-310Fax +49 3677 [email protected]

Multimedia search without detoursFingerprints at the scene of a crime can betray the perpetrator. Anddigital fingerprints can help to locate video and audio files – a partyvideo on the Internet, an advertising spot in a television program, orthe original version of a film extract in the broadcaster's archive. Fin-gerprints in music files can provide information on its tempo, genre orthe proportion of rhythmic instruments. Those used in videos may con-tain information on scene changes, camera movements and thebrightness of the image. In contrast to manually captured metadata,automatically generated fingerprints are never ambiguous. In manyarchives – for example those of television broadcasters – they are gen-erated when the file is initially stored. This differs from the Internet,where fingerprints have to be recreated with every search. For this pur-pose, it was previously necessary to decompress compressed files. Butthat is no longer the case with the multimedia search engine DIVAS,short for direct video & audio content search engine. DIVAS finds thefile you are looking for without needing to first unpack the entiremedia archive.

The search engine is suitable not only for searches on the Internet andin archives, but also for monitoring TV programs, for example to checkwhether a contractually agreed advertisement has been broadcast. Asthe software can take »fingerprints« of compressed files, it is fasterthan comparable search engines. »Our series of tests with MP3 filesshowed that search times can be halved,« reports Prof. Gerald Schullerof the Fraunhofer Institute for Digital Media Technology IDMT in Ilme-nau. Together with Matthias Gruhne he was responsible for develop-ing the audio component of DIVAS, which covers the extraction of fin-gerprints and the subsequent search. Both of these function even withmusic files that are recorded from a loudspeaker using a cell phone.»Even if the sound has been severely distorted, our methods make itpossible to clearly identify and classify the song,« explains Schuller.

The fingerprints are stored in MPEG-7 format, the ISO standard formultimedia data. DIVAS can be used to detect similarities between dif-ferent video or audio contents as well as for topic-related data search-es, for instance to find songs of a certain genre. The software can han-dle audio and video formats such as MP3, AAC, H.264 and MPEG-2.DIVAS was the brainchild of an EU-wide joint project in which sevenpartners from seven different countries took part. They included Bel-gian television broadcaster BETV, which plans to use the multimediasearch engine in its own archive.


No. 4 – 2010

For further information:Prof. Dr.-Ing. Gerald SchullerPhone +49 3677 467-355Fax +49 3677 [email protected]

Matthias GruhnePhone +49 3677 467-203Fax +49 3677 [email protected]

Fraunhofer Institute for Digi-tal Media Technology IDMTEhrenbergstr. 3198693 Ilmenau, GermanyPress contact:Julia EdlingPhone +49 3677 467-310Fax +49 3677 [email protected]

© Fraunhofer IDMT

The EDMedia digital platform facilitates interactive profes-sional training to ensure effective human resource development.


Smart career planningHowever, assessment centers are only useful if they are tailored to theindividual needs of a specific company. A one-size-fits-all approach isinappropriate, since standardized tests simply cannot reflect the pre-cise structures and working practices of a particular firm. Researchersfrom the Fraunhofer Institute for Digital Media Technology IDMT in Il-menau have now released a program that equips companies with atime-saving and cost-effective tool for carrying out objective assess-ments of their workforce.

The EDMedia Assessment Center is a new addition to the IDMT’s Educa-tional Media (EDMedia) functional learning environment – a learning con-tent management system for intuitive, individual computer-based learn-ing in which users can search for content in accordance with theirpersonal abilities and learning preferences, and choose whether to workalone or in virtual study groups as they exploit the multimedia learningtool. The EDMedia Assessment Center works in a similar way: Multi-level,individualized and standards-based tests completed online at the interac-tive user interface allow human resources experts to divide their employ-ees into groups according to competency profiles and direct their subse-quent professional development in keeping with their respective needs.

Dr.-Ing. Fanny Klett, head of the Data Representation and Interfacesbusiness unit in the IDMT, says: »The EDMedia Assessment Centerserves as an infrastructure for competency management within a com-pany and helps devise personal career paths for individual employees.«The standardized software makes it possible for companies to pool re-sources to develop tests – not only across different locations, but alsoacross company boundaries – and allows the system to connect withother EDMedia functions or external programs. The Assessment Cen-ter is also eminently suitable for multinational firms, as it takes into ac-count country-specific data protection regulations.

Klett continues: »Competency Management systems are already thenorm in all U.S. firms.« Since October 2009, she and her team at theIDMT have been supporting the U.S. Advanced Distributed Learning(ADL) Initiative in an effort to advance and standardize educational andcompetency technologies on the international scene as well. She isfirmly convinced of the benefits of ADL technologies and predicts:»These new, technologically-advanced competency management toolswill soon prove their worth and become indispensable aids to humanresources planning in German companies too.«


No. 4 – 2010

For further information:Dr.-Ing. Fanny Bastianova-KlettPhone +49 3677 467-360 Fax +49 3677 [email protected]

research news no. 4 – 2010 intelligent networking of wind farms · 2020-05-10 · intelligent...

Documents