Unlock instant, AI-driven research and patent intelligence for your innovation.

Short-term load prediction method and system based on VDM decomposition and LSTM improvement

A short-term load forecasting, power load technology, applied in forecasting, neural learning methods, biological neural network models, etc., can solve problems such as manually defining wavelet bases, inability to adapt, etc., to improve computing speed, reduce complexity, and improve fitting. effect of effect

Active Publication Date: 2021-06-01
NANJING INST OF TECH
View PDF6 Cites 4 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Then use LSSVM modeling for low-frequency components and LSTM modeling for high-frequency components. In the process of network optimization, APSO is used to optimize network parameters. In the process of wavelet transformation, we need to manually define wavelet bases, and in the process of processing local inability to adapt

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
  • Short-term load prediction method and system based on VDM decomposition and LSTM improvement
  • Short-term load prediction method and system based on VDM decomposition and LSTM improvement
  • Short-term load prediction method and system based on VDM decomposition and LSTM improvement

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0055] The present invention will be specifically described below in conjunction with the accompanying drawings.

[0056] as attached figure 1 Shown: A short-term load forecasting method based on VDM decomposition and LSTM improvement, including the following steps:

[0057] Step 1: Obtain the power load data of the area to be predicted in the latest certain period of time, and the data interval is the preset time, so as to form the original power load data set.

[0058] In this step, the predicted area may be a station area, a township, etc. where load forecasting is required. Each load data in the original data set for prediction in the present invention includes time and the electric load value corresponding to the time.

[0059] Step 2: The missing data in the original data set obtained in step 1 is filled with the nearest neighbor method to form a complete power load data set.

[0060] In the original power load data set, the power load data at some time points is miss...

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 load prediction method and system based on VDM decomposition and LSTM improvement. The method comprises the steps of data cleaning, intelligent prediction, multi-objective optimization and comprehensive evaluation. filling missing values by adopting a nearest neighbor method, and decomposing nonlinear and fluctuating load data into a plurality of relatively stable subsequences by adopting a VDM technology; optimizing and recombining a plurality of relatively stable subsequences into three sequence components through adoption of an approximate entropy method; on the basis, performing LST network in-model training on the three decomposed main components; and in the iteration process, adding an attention mechanism to allocate different weights to parameters with different characteristics, and in the gradient solution process, adding a locust algorithm to accelerate the solution process and avoid a local optimal solution. The short-term load prediction provided by the invention is well adapted to the condition that each component of the power load is complexly formed.

Description

technical field [0001] The invention relates to the technical field of electric power system and electric power market, in particular to a short-term load forecasting method and system based on VDM decomposition and LSTM improvement. Background technique [0002] Power load forecasting plays an important role in power grid planning and operation. Accurate load forecasting can not only provide guidance for power planning, but also ensure reliable operation of the power system, reduce costs and ensure power grid security. The accuracy of power load forecasting is directly related to the balance of supply and demand of the power grid, and also affects the operating cost of the power grid. Power load forecasting not only provides a guarantee for the safe and economical operation of the power system, but also provides an important basis for scheduling and power supply planning in the market environment. [0003] Due to the characteristics of the power demand side, there are thou...

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
Patent Type & Authority Applications(China)
IPC IPC(8): G06Q10/04G06Q50/06G06N3/00G06N3/08
CPCG06Q10/04G06Q50/06G06N3/006G06N3/08Y04S10/50
Inventor 欧阳孟可沈卫康成徽石凯
Owner NANJING INST OF TECH