Short-term photovoltaic power prediction method based on self-encoder deep learning model

A technology of deep learning and autoencoder, which is applied in instrumentation, design optimization/simulation, calculation, etc., to achieve better prediction performance and better prediction performance

Pending Publication Date: 2022-06-28
STATE GRID JIBEI ELECTRIC POWER COMPANY LIMITED CHENGDE POWER SUPPLY
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The deep learning model based on VAE (Variational Auto-Encoder) has excellent performance in time series modeling and nonlinear approximation feature extraction, which is expected to improve the prediction accuracy, but there is no solution for short-term photovoltaic power prediction in the existing technology

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 photovoltaic power prediction method based on self-encoder deep learning model
  • Short-term photovoltaic power prediction method based on self-encoder deep learning model
  • Short-term photovoltaic power prediction method based on self-encoder deep learning model

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0061] Provides short-term photovoltaic power prediction methods based on autoencoder deep learning models, including:

[0062] S1, build a variational autoencoder model VAE;

[0063] S2, short-term photovoltaic power prediction is performed based on the variational autoencoder model VAE.

[0064] As a preferred embodiment, the construction of the variational autoencoder VAE is based on deep learning technology, which has a good ability in automatically learning relevant features of embedded data and can be used to predict photovoltaic power output. Variational Autoencoders Model (VAE) is a class of generative-based techniques that can efficiently and automatically extract information from data in an unsupervised manner. A desirable property of VAEs is their ability to reduce input dimensionality and are very effective for approximating complex data distributions using stochastic gradient descent. Compared with traditional autoencoders, VAE uses an adjustment mechanism durin...

specific Embodiment

[0096] 1. Photovoltaic power station data

[0097] In this embodiment, data from a solar photovoltaic power station in Guizhou is used to verify the performance of the deep learning model prediction method. The real-time monitoring irradiance and actual power curve of photovoltaic power station are as follows: Figure 5 (a) and Figure 5 (b). The time resolution of the photovoltaic power plant is 15 minutes. The power station will be fully put into operation in July 2021, with an installed capacity of 50MW, and the data collected will be from August 1, 2021 to October 16, 2020.

[0098] Image 6 (a)-(d) depict the irradiance, ambient temperature, component surface temperature and corresponding DC output of the power station on a certain day. Observe how the average irradiance, ambient temperature, module temperature and DC power first increase and then decrease from sunrise to sunset. Generally, module temperature increases with temperature and irradiation.

[0099] 2. ...

Embodiment 2

[0120] A short-term photovoltaic power prediction system based on an autoencoder deep learning model, including:

[0121] a model building block for building a variational autoencoder model VAE; and

[0122] A power prediction module, configured to perform short-term photovoltaic power prediction based on the variational autoencoder model VAE.

[0123] The present invention also provides a memory that stores a plurality of instructions, and the instructions are used to implement the method described in the first embodiment.

[0124] like Figure 10 As shown, the present invention also provides an electronic device, comprising a processor 301 and a memory 302 connected to the processor 301, the memory 302 stores a plurality of instructions, the instructions can be loaded by the processor and is executed, so that the processor can execute the method described in the first embodiment.

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 photovoltaic power prediction method based on an auto-encoder deep learning model. The method comprises the following steps: S1, constructing a variational auto-encoder model VAE; and S2, performing short-term photovoltaic power prediction based on the variational auto-encoder model VAE. The S1 comprises the following steps: S11, constructing an encoder and a decoder of the variational auto-encoder model VAE; s12, obtaining the encoder through an approximate posterior qtheta (zx), and obtaining the decoder through a likelihood p phi (xz); s13, constructing and calculating a loss function to train feature extraction of the VAE; and S14, minimizing a loss function L (theta, phi) by using the training observation value to obtain parameters of the encoder and the decoder. The invention also discloses a short-term photovoltaic power prediction system based on the self-encoder deep learning model, electronic equipment and a computer readable storage medium.

Description

technical field [0001] The invention belongs to the technical field of smart power, and in particular relates to a short-term photovoltaic power prediction method based on an autoencoder deep learning model. Background technique [0002] Accurate modeling and prediction of solar output in photovoltaic (PV) systems is crucial for the safe operation of new power systems built with new energy sources. Accurate photovoltaic power prediction can reduce the impact of photovoltaic power generation uncertainty on the power grid, and improve the power quality and the penetration level of photovoltaic systems. The output of photovoltaic power generation has high volatility and intermittency. The accurate modeling and prediction of the output power of photovoltaic systems can effectively improve the effective operation of the energy management system and the grid-connected dispatch management level of the power system. [0003] The photovoltaic power generation power prediction mainly...

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): G06F30/27G06F113/04G06F119/02G06F119/06
CPCG06F30/27G06F2113/04G06F2119/02G06F2119/06
Inventor 袁绍军郭金智毕圆圆尹兆磊张宝华丁然周迎伟陈晨刘震宇刘嗣萃于立强白明辉杨慢慢张柏杨段明慧赵磊
Owner STATE GRID JIBEI ELECTRIC POWER COMPANY LIMITED CHENGDE POWER SUPPLY
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