IPC분류정보
| 국가/구분 |
United States(US) Patent
등록
|
| 국제특허분류(IPC7판) |
|
| 출원번호 |
US-0393703
(2003-03-19)
|
|
발명자
/ 주소 |
- Barnes, David L.
- Juarez, Ruben
- Wade, John
|
| 출원인 / 주소 |
|
| 대리인 / 주소 |
|
| 인용정보 |
피인용 횟수 :
33 인용 특허 :
0 |
초록
▼
Methods and apparatus for forecasting an energy output of a wind farm. A method for forecasting an energy output of a wind farm includes maintaining a data base of wind patterns, each wind pattern being associated with an energy output that the wind farm produces. The method includes receiving a cur
Methods and apparatus for forecasting an energy output of a wind farm. A method for forecasting an energy output of a wind farm includes maintaining a data base of wind patterns, each wind pattern being associated with an energy output that the wind farm produces. The method includes receiving a current wind pattern. The method includes searching the data base for a wind pattern that matches the current wind pattern. The method includes calculating a forecast energy output that the wind farm will produce in response to the current wind pattern, the calculation being based, when there is a matching wind pattern in the knowledge base, on the energy output associated with the matching wind pattern.
대표청구항
▼
1. A method for forecasting an energy output of a wind farm, the method comprising:maintaining a database of wind patterns, each wind pattern being associated with an energy output that the wind farm produces;receiving a current wind pattern;searching the database for a wind pattern that matches the
1. A method for forecasting an energy output of a wind farm, the method comprising:maintaining a database of wind patterns, each wind pattern being associated with an energy output that the wind farm produces;receiving a current wind pattern;searching the database for a wind pattern that matches the current wind pattern; andcalculating a forecast energy output that the wind farm will produce in response to the current wind pattern, the calculation being based, when there is a matching wind pattern in the database, on the energy output associated with the matching wind pattern.2. The method of claim 1, wherein maintaining a database includes receiving and storing wind patterns and energy outputs, the method further comprising:identifying received wind patterns that are erroneous, the identifying being based on one of a range test, a relational test, and a trend test;identifying received energy outputs that are erroneous; the identifying being based on one of a range test, a relational test, and a trend test; andremoving from the database the wind patterns and energy outputs identified as erroneous.3. The method of claim 1, further comprising:determining how many wind patterns to associate with each energy output.4. The method of claim 3, wherein wind patterns in the database are categorized into categories, each category having one energy output that is associated with all wind patterns of the category; the method further comprising:when there is no wind pattern in the database that matches the current wind pattern, identifying the category to which the current wind pattern belongs and adding the current wind pattern to the database as an additional wind pattern of the category identified.5. The method of claim 3, wherein a neural network determines how many wind patterns to associate with each energy output, the method further comprising:adjusting the neural network to change a resolution at which the neural network can distinguish different wind patterns, the resolution determining whether the current wind pattern matches with one of the wind patterns in the database.6. The method of claim 5, wherein:the neural network is a variation of the Perceptron, the variation including threshold knobs that determines the resolution; andadjusting the neural networks includes adjusting the threshold knobs.7. The method of claim 5, further comprising:receiving an actual energy output the wind farm produces in response to the current wind pattern, wherein the adjusting is based on the actual energy output.8. The method of claim 7, wherein:adjusting includes applying Bayesian logic.9. The method of claim 1, further comprising:receiving an actual energy output the wind farm produces in response to the current wind pattern; andupdating a probability that the wind farm will produce the forecast energy output.10. The method of claim 9, wherein:updating includes applying Bayesian logic.11. The method of claim 1, further comprising:receiving information from an upwind sentry;detecting a trend from the received information;updating the current wind pattern based on the trend; andsearching the database for a wind pattern that matches the updated current wind pattern.12. The method of claim 1, further comprising:when there is no wind pattern in the database that matches the current wind pattern, identifying a wind pattern that is closest to matching the current wind pattern, wherein the calculation of forecast energy output is based on the energy output associated with the wind pattern identified as being closest to matching the current wind pattern.13. A computer program product, tangibly stored on machine-readable medium, for forecasting an energy output of a wind farm, the product includes instructions operable to cause a processor to:maintain a database of wind patterns, each wind pattern associated with an energy output that the wind farm produces;receive a current wind pattern;search the database for a wind pattern that matches the current wind pattern;calculate a forecast energy output that the wind farm will produce in response to the current wind pattern, the calculation being based, when there is a matching wind pattern in the database, on the energy output associated with the matching wind pattern.14. A system for forecasting an energy output of a wind farm, comprising:instruments operable to measure a current wind pattern over a wind farm;a neural network having a knowledge base of wind patterns and energy outputs, each wind pattern being associated with one energy output; anda network operable to send the current wind pattern to the neural network, wherein the neural network is operable to identify a wind pattern in the knowledge base that matches the current wind pattern and, furthermore, calculate a forecast energy output using the energy output associated with the wind pattern that matches the current wind pattern.15. The system of claim 14, further comprising:instruments operable to measure the energy output the wind farm actually produces in response to the current wind pattern, wherein the network is operable to send the measured energy output to the neural network, and wherein the neural network is operable to adjust, based on the energy output the wind farm actually produces, a probability that the forecast energy output is the same as the energy output that the wind farm actually produces.16. The system of claim 14, wherein:the knowledge base is a variation of the Perceptron.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.