Power load prediction method considering temperature fuzzification

Abstract

The invention discloses a power load prediction method considering temperature fuzzification, and the method comprises the steps: firstly collecting historical load data, historical temperature dataand related date data of a power grid, processing the historical load data, historical temperature data and related date data into a 15-dimensional feature vector, and dividing the 15-dimensional feature vector into a training data set and a test data set in proportion; establishing a three-layer long-short-term memory neural network, and performing iterative training on the three-layer long-short-term memory neural network through the training data set to obtain a power load prediction model; and finally, inputting prediction day data in the test data set into the power load prediction modelto obtain a power load prediction value. According to the invention, the short-term load can be accurately predicted by considering the load timing characteristics; and meanwhile, a membership function is used for fuzzifying the precise temperature to an interval of [0, 1] to serve as a prediction model feature vector input, so that the random influence caused by temperature variability and uncertainty in summer can be effectively reduced, and the generalization of the power load prediction model to the temperature is improved.

Description

technical field [0001] The invention belongs to the technical field of power load forecasting, and in particular relates to the design of a power load forecasting method considering temperature fuzziness. Background technique [0002] With the development of smart grid, the advanced measurement system (AMI) based on smart meters has been basically realized in actual projects. Smart meters can obtain 96 sets of data every day at intervals of 15 minutes, providing more reliable information for improving short-term load forecasting. However, in addition to the fact that the actual load change can be extracted from the load time requirement, the actual load is also largely affected by the temperature conditions. Therefore, in the summer when the temperature is changeable, the short-term load forecasting of electric power considering the influence of temperature is studied. method, which has important theoretical value and practical significance. [0003] At present, short-term ...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to solve the problem of how to effectively fuzzify the temperature variable in summer when the temperature is variable, so as to use deep learning to effectively predict the power load, and propose a power load forecasting method that considers temperature fuzzification

Images