System and method for wind generation forecasting

Patent metadata

Field	Value
Publication number	US-9460478-B2
Application number	US-201314109586-A
Country	US
Kind code	B2
Filing date	Dec 17, 2013
Priority date	Dec 17, 2012
Publication date	Oct 4, 2016
Grant date	Oct 4, 2016

Abstract

Systems and methods for managing electricity of an electrical supply grid electrically connected to a wind farm are disclosed. Some systems and methods may include identifying relationships among a plurality of turbines in the wind farm with a minimum spanning tree, calculating a power output relationship among the plurality of turbines with the minimum spanning tree, creating a finite state space Markov chain forecast model for the plurality of turbines in the wind farm, predicting a power output of the wind farm with the finite state space Markov chain forecast model, and modifying at least one of a generation of electricity and a distribution of electricity based on the predicted power output of the wind farm. Also disclosed are systems and methods for predicting the power output of a wind farm.

First claim

What is claimed is:

1. A method of managing electricity of an electrical supply grid electrically connected to a wind farm, the method comprising:
identifying relationships among a plurality of turbines in the wind farm with a minimum spanning tree, wherein the minimum spanning tree separates the plurality of wind turbines into a plurality of classes, identifies a parent turbine for each turbine within a particular class, and guarantees that each turbine in a particular class is linked to a root turbine of that class;
calculating a power output relationship among the plurality of turbines with the minimum spanning tree;
creating a finite state space Markov chain forecast model for the plurality of turbines in the wind farm;
predicting a power output of the wind farm with the finite state space Markov chain forecast model; and
modifying at least one of a generation of electricity and a distribution of electricity based on the predicted power output of the wind farm.

2. The method according to claim 1 , further comprising forming a probability distribution of aggregate power generation of the wind farm.

3. The method according to claim 1 , further comprising measuring real-time wind farm electricity generation.

4. The method according to claim 3 , wherein the predicting a power output of the wind farm comprises at least one of a distributional forecast and a point forecast.

5. The method according to claim 1 , wherein the predicted power output of the wind farm is predicted for a period about 10 minutes into the future.

6. The method according to claim 1 , wherein the predicted power output of the wind farm is a period of time about 5 minutes into the future.

7. The method according to claim 1 , wherein the predicted power output of the wind farm is a period of time about one hour into the future.

8. The method according to claim 1 , wherein the creation of the finite state space Markov chain forecast utilizes ramp trend information.

9. The method according to claim 1 , wherein the creation of the finite state space Markov chain forecast model comprises historical data comprising at least one of a historic data of wind turbine power output and a historic data of wind speed.

10. The method according to claim 1 , wherein a plurality of finite state space Markov chains are created for an epoch.

11. The method according to claim 1 , wherein a plurality of finite state space Markov chains are created for each month in a year.

12. A system for managing electricity of an electrical supply grid electrically connected to a wind farm, the system comprising:
a processor configured to be in electrical communication with a wind farm power output sensor, wherein the processor is configured to:
identify relationships among a plurality of turbines in the wind farm with a minimum spanning tree, wherein the minimum spanning tree separates the plurality of wind turbines into a plurality of classes, identifies a parent turbine for each turbine within a particular class, and guarantees that each turbine in a particular class is linked to a root turbine of that class;
calculate a power output relationship among the plurality of turbines with a minimum spanning tree;
create a finite state space Markov chain forecast model for the plurality of turbines in the wind farm;
predict a power output of the wind farm with the finite state space Markov chain forecast model; and
determine, based on the predicted power output of the wind farm, whether at least one of a generation of electricity and a distribution of electricity should be modified.

13. The system according to claim 12 , wherein the system is in electrical communication with a meteorological tower collocated with a turbine of the wind farm.

14. The system according to claim 12 , wherein the system is in electrical communication with a class sensor.

15. The system according to claim 14 , wherein the class sensor signals least one of a turbine height, a turbine manufacturer, and a turbine model.

16. The system according to claim 12 , further comprising a power grid sensor that detects a current electrical power of the electrical grid.

Assignees

• Zhang Junshan
• He Miao
• Yang Lei
• Vittal Vijay
• Univ Arizona State

Inventors

• Zhang Junshan
• He Miao
• Yang Lei
• Vittal Vijay

Classifications

• H02J3/004

  Generation forecast, e.g. methods or systems for forecasting future energy generation · CPC title

• H02J2103/30

  Simulating, planning, modelling, reliability check or computer assisted design [CAD] of electric power networks · CPC title

• H02J2101/28

  Wind energy · CPC title

• G06Q10/04

  Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem" (market predictions or forecasting for commercial activities G06Q30/0202) · CPC title

• G06Q50/06 Primary

  Energy or water supply · CPC title