POWER GENERATION PREDICTION SYSTEM AND A METHOD THEREOF

• US 20190012598A1
• Filed: 07/05/2018
• Published: 01/10/2019
• Est. Priority Date: 07/04/2017
• Status: Active Grant

• Current Assignee: Taiwan Power Company (Government of Taiwan)
• Original Assignee: Taiwan Power Company (Government of Taiwan)
• Inventors: CHANG, CHIH-JUNG, CHOU, LI-FEN, LEU, YIH-GUANG
• Granted Patent: US 10,643,128 B2
• Time in Patent Office: Days
• CPC Class Codes: G05B 13/027 using neural networks only
• CPC Class Codes: G06N 3/045 Combinations of networks
• CPC Class Codes: G06N 3/08 Learning methods
• CPC Class Codes: G06N 3/082 modifying the architecture, e.g. adding, deleting or silencing nodes or connections
• CPC Class Codes: G06Q 50/06 Energy or water supply
• CPC Class Codes: H02J 2203/20 Simulating, e g planning, reliability check, modelling or computer assisted design [CAD]
• CPC Class Codes: H02J 2300/24 of photovoltaic origin
• CPC Class Codes: H02J 2300/28 The renewable source being wind energy wind motors F03D
• CPC Class Codes: H02J 2310/12 The local stationary network supplying a household or a building
• CPC Class Codes: H02J 3/003 Load forecast, e.g. methods or systems for forecasting future load demand
• CPC Class Codes: H02J 3/004 Generation forecast, e.g. methods or systems for forecasting future energy generation
• CPC Class Codes: H02J 3/38 Arrangements for parallely feeding a single network by two or more generators, converters or transformers
• CPC Class Codes: H02J 3/381 Dispersed generators
• CPC Class Codes: Y02E 10/56 Power conversion systems, e.g. maximum power point trackers
• CPC Class Codes: Y02E 10/76 Power conversion electric or electronic aspects
• CPC Class Codes: Y02E 60/00 Enabling technologies; Technologies with a potential or indirect contribution to GHG emissions mitigation
• CPC Class Codes: Y04S 40/20 Information technology specific aspects, e.g. CAD, simulation, modelling, system security

Abstract

A power generation prediction system using a first neural network and a second neural network is provided, and the first neural network is connected to the second neural network. The first neural network receives first input data, and generates the amount prediction data according to the first input data. The first input data is used to determine amount prediction data, and the amount prediction data is used to determine power generation prediction data. The second neural network receives the amount prediction data, and calculates the power generation prediction data according to the amount prediction data. When a device in a selected area is deteriorated or reinstalled, the second neural network is fine-tuned and trained again.

Claims

1. A power generation prediction system, applicable to an area where a plurality of devices are installed for power generation with a power generation conversion curve, and the power generation prediction system comprising:

a first computing device configured to receive first input data, and use a first neural network to generate amount prediction data according to the first input data, wherein the first input data comprises meteorological data and used to determine the amount prediction data, and the amount prediction data is used to determine power generation prediction data;

a second computing device connected to the first computing device and configured to receive the amount prediction data, and use a second neural network to perform an approximation function to calculate the power generation prediction data according to the amount prediction data; and

a third computing device connected to the second computing device, and configured to receive second input data, and use a third neural network to generate a power generation module parameter prediction data according to the second input data, wherein the second input data is used to determine the power generation module parameter prediction data, and the power generation module parameter prediction data is used to correct the power generation prediction data;

wherein the second computing device configured to receives the power generation module parameter prediction data, and uses the second neural network to perform the approximation function to calculate the power generation prediction data according to the amount prediction data and the power generation module parameter prediction data;

wherein when the power generation conversion curve is affected by deterioration or reinstallation of at least one of the plurality of devices in the area, the second neural network is fine-tuned and trained again to update the approximation function thereof.

2. The power generation prediction system according to claim 1, wherein the area is used for solar power generation, and the first input data comprises multiple pieces of solar irradiance observation data and solar irradiance calculation data of the area, and the second input data comprises meteorological data and multiple pieces of temperature observation data, and the power generation module parameter prediction data comprises temperature prediction data, and the plurality of devices comprise a pyranometer and multiple components of solar power generation equipment.

3. The power generation prediction system according to claim 2, wherein the solar irradiance calculation data is determined by operations of:

determining a shape, a size and a location of the area;

searching multiple reference parameters of a known solar power facility in the area, and obtaining a location and an effective range of the known solar power facility in the area;

according to the reference parameters of the known solar power facility in the area, obtaining the solar irradiance observation data of the known solar power facility;

calculating an overlapping area between the area and the effective range of the known solar power facility in the area;

according to the overlapping area between the area and the effective range of the known solar power facility in the area, calculating a weight of the known solar power facility; and

according to the weight of the known solar power facility and the solar irradiance observation data in the area, calculating the solar irradiance calculation data of the area.

4. The power generation prediction system according to claim 2, wherein the solar irradiance calculation data is determined by operations of:

determining a shape, a size and a location of the area;

searching multiple reference parameters of a known solar power facility in the area, and obtaining a location and an effective range of the known solar power facility in the area;

according to the multiple reference parameters of the known solar power facility in the area, obtaining the solar irradiance observation data of the known solar power facility;

according to a location of the known solar power facility in the area and the location of the area, correcting the solar irradiance observation data of the known solar power facility in the area;

calculating an overlapping area between the area and the effective range of the known solar power facility in the area;

according to the overlapping area between the area and the effective range of the known solar power facility in the area, calculating a weight of the known solar power facility; and

according to the weight of the known solar power facility in the area and the corrected solar irradiance observation data, calculating the solar irradiance calculation data of the area.

5. The power generation prediction system according to claim 4, wherein the solar irradiance observation data of the known solar power facility in the area is corrected according to a height of the known solar power facility in the area and a height of the area.

6. The power generation prediction system according to claim 1, wherein the area is used for wind power generation, and the first input data comprises the meteorological data and wind amount observation data, and the plurality of devices comprises a wind power generator and multiples components of a wind meter.

7. A power generation prediction method, applicable to an area where a plurality of devices are installed for power generation with a power generation conversion curve, and the power generation prediction method comprising:

processing first input data by using a first neural network to generate amount prediction data, wherein the first input data comprises meteorological data, the first input data is used to determine the amount prediction data, and the amount prediction data is used to determine power generation prediction data;

processing the amount prediction data by using a second neural network to perform an approximation function to calculate the power generation prediction data, wherein the second neural network is connected to the first neural network; and

processing second input data by using a third neural network to generate a power generation module parameter prediction data, wherein the second input data is used to determine the power generation module parameter prediction data, and the power generation module parameter prediction data is used to correct the power generation prediction data;

wherein the amount prediction data and the power generation module parameter prediction data are processed by the second neural network according to the approximation function to calculate the power generation prediction data;

wherein when the power generation conversion curve is affected by deterioration or reinstallation of at least one of the plurality of devices in the area, the second neural network is fine-tuned and trained again to update the approximation function thereof.

8. The power generation prediction method according to claim 7, wherein the area is for solar power generation, the first input data comprises multiple pieces of solar irradiance observation data and a solar irradiance calculation data of the area, the second input data comprises the meteorological data and multiple pieces of temperature observation data, the power generation module parameter prediction data comprises temperature prediction data, and the plurality of devices comprise a pyranometer and multiple components of solar power generation equipment.

9. The power generation prediction method according to claim 8, wherein the solar irradiance calculation data is determined by operations of:

determining a shape, a size and a location of the area;

searching multiple reference parameters of the known solar power facility in the area, and obtaining a location and an effective range of the known solar power facility in the area;

according to the multiple reference parameters of the known solar power facility in the area, obtaining the solar irradiance observation data of the known solar power facility;

calculating an overlapping area between the area and the effective range of the known solar power facility in the area;

according to the overlapping area between the area and the effective range of the known solar power facility in the area, calculating a weight of the known solar power facility; and

according to the weight of the known solar power facility and the solar irradiance observation data in the area, calculating the solar irradiance calculation data of the area.

10. The power generation prediction method according to claim 8, wherein the solar irradiance calculation data is determined by operations of:

determining a shape, a size and a location of the area;

searching multiple reference parameters of the known solar power facility in the area, and obtaining a location and an effective range of the known solar power facility in the area;

according to the multiple reference parameters of the known solar power facility in the area, obtaining the solar irradiance observation data of the known solar power facility;

according to a location of the known solar power facility in the area and a location of the area, correcting the solar irradiance observation data of the known solar power facility in the area;

calculating an overlapping area between the area and the effective range of the known solar power facility in the area;

according to the overlapping area between the area and the effective range of the known solar power facility in the area, calculating a weight of the known solar power facility; and

according to the weight of the known solar power facility in the area and the corrected solar irradiance observation data, calculating the solar irradiance calculation data of the area.

11. The power generation prediction method according to claim 10, wherein the solar irradiance observation data of the known solar power facility in the area is corrected according to a height of the known solar power facility in the area and a height of the area.

12. The power generation prediction method according to claim 7, wherein the area is used for wind power generation, and the first input data comprises the meteorological data and wind amount observation data, and the plurality of devices comprise a wind meter and multiple components of a wind power generator.