Power system short-term load prediction method based on deep learning

A short-term load forecasting, power system technology, applied in neural learning methods, forecasting, instruments, etc., can solve the problems of model generalization performance limitations, hindering forecasting accuracy, and limited ability to approximate complex functions, avoiding overfitting and localization. Minimal value problem, good information expression ability, the effect of improving accuracy

Inactive Publication Date: 2019-08-13
TIANJIN UNIV OF TECH & EDUCATION TEACHER DEV CENT OF CHINA VOCATIONAL TRAINING & GUIDANCE
View PDF3 Cites 5 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Benefits of technology

This patented technology uses advanced models like convolutional Neural Network (CNN) or recurrent neural network(RN). These techniques have excellent performance but they also suffer from issues such as over fitting with small data points during training process. To overcome these challenges, this innovation proposes utilizing Deep Learning for better prediction capabilities without requiring too much time on complex datasets.

Problems solved by technology

This patented technical problem addressed by this patents relates to improving the predictive capabilities of current models used during power systems' load forecasts due to limitations with approximating function complexity or poor data extraction from complicated loads sequences.

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
  • Power system short-term load prediction method based on deep learning
  • Power system short-term load prediction method based on deep learning
  • Power system short-term load prediction method based on deep learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0022] The present invention provides a short-term load forecasting method of a power system based on deep learning. The present invention will be described in detail below in conjunction with the accompanying drawings and embodiments.

[0023] The present invention aims at the problems of overfitting and local minima that are prone to occur in short-term load forecasting in related technologies. In the embodiment, the relevant influencing factors of short-term load forecasting are used as the input of the deep learning model, and the short-term load is predicted through deep learning. The solutions of the embodiments are described in detail below.

[0024] The deep learning model adopted in this embodiment is a DBN structure, which is a nonlinear network based on three hidden layers.

[0025] figure 1 It is an overall flow chart of the short-term load forecasting method for power systems based on deep learning, including an offline training forecasting model stage and an online

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 power system short-term load prediction method based on deep learning. The method comprises: firstly, obtaining historical loads, week types, holidays and festivals, meteorological information and other data; carrying out preprocessing such as normalization and coding to serve as input characteristics, learning the deep belief network formed by the stacked limited Boltzmann mechanism layer by layer by using a contrast divergence algorithm, extracting deep characteristics, and finally finely adjusting parameters of the whole deep belief network by using a back propagation algorithm, namely constructing the load prediction model by offline training; and carrying out online prediction on the load of the power system after 24 hours to obtain a predicted load. The short-term load prediction method improves the short-term load prediction precision of the power system.

Description

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

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
Owner TIANJIN UNIV OF TECH & EDUCATION TEACHER DEV CENT OF CHINA VOCATIONAL TRAINING & GUIDANCE
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap