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

Ultra-short-term power load prediction method and device based on machine learning

A technology of electric load and machine learning, which is applied in the direction of load forecasting, machine learning, and circuit devices in the AC network, can solve the problem of low forecasting accuracy, achieve the effects of ensuring forecasting efficiency, improving forecasting accuracy, and avoiding time-consuming effects

Pending Publication Date: 2022-01-21
STATE GRID JIANGXI ELECTRIC POWER CO LTD RES INST +1
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] Aiming at the problem that the power load is affected by the complex environment with strong randomness and low prediction accuracy, this invention considers the influence of weather and holidays to construct data features and establishes a model through data mining and machine learning algorithms, and proposes a machine learning-based ultra-short-term power Load forecasting method and device

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
  • Ultra-short-term power load prediction method and device based on machine learning
  • Ultra-short-term power load prediction method and device based on machine learning
  • Ultra-short-term power load prediction method and device based on machine learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0029] The present invention is further explained in detail below.

[0030] refer to figure 1 , an ultra-short-term power load forecasting method based on machine learning, the steps are as follows:

[0031] The first step is to normalize the original data;

[0032] Normalize the raw data using the peak method, the formula is as follows:

[0033]

[0034] Where: X i represents the original load, X max means X i the maximum value of Indicates the normalized value, and i represents the serial number.

[0035] The second step is to use the variational mode decomposition (VMD) method to process the load data, and decompose the highly volatile original load data into multiple relatively stable intrinsic mode function (IMF) components;

[0036] The parameters of variational mode decomposition (VMD) have a great influence on the decomposition effect, so the important parameters should be determined before decomposition. First, according to previous research results and pr...

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 an ultra-short-term power load prediction method based on machine learning. The method comprises the following steps: carrying out normalization operation on original data; processing load data by using a variational mode decomposition method, and decomposing original load data with relatively strong volatility into a plurality of relatively stable intrinsic mode function components; clustering each intrinsic mode function component by using a Kmeans method to obtain a new load sequence; performing feature construction by considering the influence of influence factors on the power load; and predicting the new load sequence by using an XGBoost model, and then superposing and reconstructing predicted values to obtain a final prediction result. According to the method, the influence of factors such as weather, holidays and the like on data feature construction is considered, and the advantages of VMD, Kmeans and XGBoost methods are combined, so that the ultra-short-term load prediction precision can be improved.

Description

technical field [0001] The invention relates to a machine learning-based ultra-short-term power load forecasting method and device, and belongs to the technical field of power data processing. Background technique [0002] Power load forecasting is one of the technical prerequisites to ensure the safe and stable operation of power systems, and ultra-short-term load forecasting is of great significance for real-time security analysis and real-time economic dispatch. With the rapid development of the current economy, the demand for electricity in various industries is getting higher and higher, and the load types are becoming more and more complex. The power load is affected by factors such as temperature, humidity, and holidays, and its randomness is strong. Therefore, improving the accuracy of load forecasting is a current research main goal of . [0003] Research on load forecasting models at home and abroad is mainly divided into time series models, machine learning and d...

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): G06Q10/04G06K9/62G06Q50/06H02J3/00G06N20/00
CPCG06Q10/04H02J3/003G06Q50/06G06N20/00G06F18/23213
Inventor 欧阳文华常乐蒙天骐徐在德安义戚沁雅
Owner STATE GRID JIANGXI ELECTRIC POWER CO LTD RES INST
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