Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

A short-term wind power prediction method based on double-time sequence feature learning

A wind power prediction and feature learning technology, applied in complex mathematical operations, biological neural network models, instruments, etc., to achieve the effects of improving prediction accuracy, learning ability, and robustness

Pending Publication Date: 2019-05-21
TIANJIN UNIV
View PDF8 Cites 8 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] The power generated by wind turbines is mainly related to wind speed and wind direction. Affected by weather changes, wind speed and wind direction have great mutations in time series. How to accurately model based on historical wind power data is still a very challenging problem. question

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • A short-term wind power prediction method based on double-time sequence feature learning
  • A short-term wind power prediction method based on double-time sequence feature learning
  • A short-term wind power prediction method based on double-time sequence feature learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0033] A short-term wind power prediction method based on dual time series feature learning, the method is based on the dual time series feature learning neural network model (BTFLNN), the key points of its prediction technology are: 1) effective data denoising; 2) sufficient time series features extract.

[0034] The main design and analysis process of the entire BTFLNN model can be divided into the following four parts: 1) data selection and conversion; 2) data denoising and principal component selection; 3) dual time series feature learning of data and wind power regression analysis; 4) BTFLNN model prediction results analysis.

[0035] The method includes the following steps:

[0036] 101: Construct a training set and a test set, and simultaneously convert the original data into labeled data;

[0037] 102: Use the singular spectrum analysis method to de-dry and select the principal components of the labeled data;

[0038] 103: Construct a dual time-series feature learni...

Embodiment 2

[0042] The following is combined with specific examples, mathematical formulas, Figure 1-Figure 3 , the scheme in embodiment 1 is further introduced, see the following description for details:

[0043] 1. Data selection and conversion

[0044] The experimental data set in the embodiment of the present invention comes from a wind farm in Chenzhou City, Hunan Province, China. All data of fans No. 26 to No. 50. Specifically, the installed capacity of these 25 wind turbines is 2000KW, and the time resolution of the data is 10 minutes. The collected data contains five variables, namely: average temperature outside the engine room, reactive power, active power, wind speed and wind direction . Since the reactive power is an irrelevant variable (not beneficial to the design and realization of the model), this variable is removed, leaving the remaining four relevant variables. In addition, in actual scenarios, it is a very common phenomenon that wind turbines are disconnected from...

Embodiment 3

[0084] Combine below Figure 4-Figure 8 The scheme in embodiment 1 and 2 is carried out feasibility verification, see the following description for details:

[0085] This experiment uses two evaluation indicators commonly used in data prediction experiments: mean absolute error (NMAE) and root mean square error (NRMSE) to verify the prediction performance of the dual temporal feature learning neural network model BTFLNN. Among them, NMAE is a linear evaluation index, and NRMSE is a secondary evaluation index, which is more sensitive to errors. The calculation methods of NMAE and NRMSE are described as follows (the lower the value, the better the model effect):

[0086]

[0087]

[0088] Among them, p inst is the installed capacity of the fan, N is the length of the time series data, x' n is the predicted value, x n is the actual measured value.

[0089] Figure 4 It shows the corresponding NMAE and NRMSE results when the input data of the model takes different leng...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a short-term wind power prediction method based on double-time sequence feature learning, and the method comprises the following steps: building a training set and a test set,and converting original data into labeled data at the same time; Adopting a singular spectrum analysis method to perform de-noising and principal component selection on the original wind power data; Constructing a double-time-sequence feature learning neural network model composed of a local time sequence learning module and two long-short-term memory networks, and obtaining local wind power dataat different moments according to the input of the neural network model; And the neural network model outputs the double-time sequence characteristics processed by one local time sequence learning module and two long and short term memory networks through a full connection layer, and performs final regression analysis to obtain a to-be-predicted wind power value at the t + 1 moment at the t moment. According to the method, through principal component selection and multi-scale time sequence characteristic learning of original data, accurate prediction of the power generation power of the singlefan of the wind power plant is finally realized.

Description

technical field [0001] The present invention relates to the field of computer data forecasting, and in particular to a short-term wind power forecasting method based on dual time-series feature learning. The present invention realizes a single fan in a wind field by performing noise data elimination, feature selection, and time-series feature learning on historical wind power data. wind power forecast. Background technique [0002] Due to the increasingly serious energy crisis and environmental pollution problems, countries around the world have increased the development of green energy to achieve energy conservation and emission reduction. As an important growth point of green energy, wind energy has increased its cumulative global installed capacity by approximately 14 times in the past 15 years, reaching 539.123GW at the end of 2017. However, there are certain uncertainties in both time and space in the process of wind power generation, which will have a great impact on ...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): G06N3/04G06F17/18G06Q50/06
CPCY04S10/50
Inventor 孙美君李攀王征
Owner TIANJIN UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com