A Power Quality Prediction Method for Distribution Networks with Distributed Generation Based on Deep Learning Model

Abstract

A power quality prediction method based on a deep learning model for a distributed power distribution network, including: acquisition of power quality historical data, including power quality influencing factor variable data and power quality index data; power quality historical data preprocessing, including data normalization Normalization, time series conversion, training data and evaluation data segmentation; power quality long short-term memory neural network, that is, LSTM prediction model determination, including LSTM prediction model construction, parameter initialization, LSTM prediction model determination based on training data; evaluation data-based Performance evaluation of prediction model; prediction of system power quality index in the future period. The advantages of the present invention are: 1. The power quality of the distribution network containing DG is effectively predicted; 2. Higher precision can be obtained in processing time series prediction problems; 3. Each characteristic input variable in the prediction model is fully considered influence.

Description

technical field [0001] The invention relates to a method for predicting the power quality of a distribution network containing distributed power sources based on a deep learning model, belonging to the fields of electrical engineering and power quality. Background technique [0002] As a product entering the market, electric energy should also emphasize quality just like other commodities. The index items that reflect the status of power quality mainly include voltage deviation, harmonics, voltage flicker and fluctuations, voltage dips and interruptions, frequency deviations, harmonics and inter-harmonics, and three-phase voltage unbalance. With the development of green energy and smart grid technology, it will become a future trend that distributed generators (DG) such as photovoltaic power generation and wind power generation will be widely connected to the distribution network. They are a useful supplement to the traditional distribution network. However, due to its larg...

AI Technical Summary

Problems solved by technology

[0005] The present invention aims to overcome the problem that the potential power quality problems of the system cannot be effectively sensed, warned and dealt with due to the insufficient power quality prediction ability of the existing DG-containing distribution network, and provides a power quality prediction of the DG-containing distribution network based on the LSTM deep learning model method

Images