User side load probability prediction method based on time domain convolutional neural network

Abstract

The invention discloses a user side load probability prediction method based on a time domain convolutional neural network, and the method comprises the steps: firstly collecting the historical load data of a user and the weather type and temperature value in a corresponding time range, and then carrying out the preprocessing, so as to meet the input demands of the time domain convolutional neuralnetwork; then constructing a time-domain convolutional neural network as a load probability prediction model, and enabling the time-domain convolutional neural network to accurately predict the loadprobability through model training; and finally, predicting load data acquired by the user in real time.

Description

technical field [0001] The invention belongs to the technical field of power load forecasting, and more specifically, relates to a user-side load probability forecasting method based on a time-domain convolutional neural network. Background technique [0002] With the widespread application of advanced metering infrastructure in modern power systems, detailed load curves and individual power consumption behaviors at the user end can be easily obtained. However, the explosive growth of data exceeds the capacity of traditional load forecasting systems. Specifically, even though a large amount of data can be collected, it cannot be effectively analyzed and utilized. In recent years, with the development of artificial intelligence (AI) technology, rich and fine-grained electricity consumption data can be processed, analyzed and utilized. Load forecasting of individual customers provides utilities with the opportunity to fine-tune the management of the grid, resulting in better ...

AI Technical Summary

Problems solved by technology

However, the existing short-term load forecasting methods for individual users based on RNN and CNN have model structural defects, which make it difficult to deal with severe practical challenges: in the face of high-frequency model training and online update of massive load data, the most basic requirements of the forecasting model are Be as reliable as possible while performing efficient computations

Images