Power load prediction method

Abstract

The invention discloses a power load forecasting method, which includes the following steps: acquiring historical daily data and forming an evaluation matrix M nm , m ij Expressed in the evaluation matrix M nm The index value of the j-th item index of the i-th evaluation object; the evaluation matrix M nm Perform linear transformation and normalization processing to obtain the normalization matrix S nm ;Calculation matrix S nm The similar day feature weight of each index in the index; calculate the correlation degree between the forecast date and the historical date; sort the evaluation objects according to the correlation degree from large to small, and select the first W group of evaluation objects as the training samples of the prediction algorithm; use The genetic algorithm improves the forecasting algorithm, optimizes the weights and thresholds of the forecasting algorithm, calculates the optimal weights and thresholds for training; inputs the forecast date information into the optimized forecasting algorithm, and denormalizes the output value After that, the power load value of the forecast day is obtained. The invention can predict the power load of the next day according to the relevant data of the historical day.

Description

technical field [0001] The invention relates to the technical field of power distribution, in particular to a power load forecasting method. Background technique [0002] With the continuous development of my country's electric power industry and the continuous improvement of people's living standards, the demand for electric energy in all walks of life is increasing. Due to the simultaneous production and consumption of electric energy, and the consumption is a random process, the storage cost of the grid is high, and accurate and reasonable electric energy distribution can meet the different needs of various users, so that the generator sets and transmission lines can run safely with the longest life. Short-term power load forecasting is the daily work of the power dispatching department, including forecasting daily load forecasting (maximum load forecasting, average load forecasting), power load forecasting at 96 o’clock (5-minute records) on the forecasting day, and the ...

AI Technical Summary

Problems solved by technology

[0003] In the existing technology, short-term load forecasting is mostly based on historical load data to do mathematical statistics, and the influence factors of power load forecasting are insufficiently utilized; some algorithms take into account meteorological factors such as temperature, do not comprehensively analyze and consider various influencing factors, and do not choose a suitable forecasting model ;Modern artificial intelligence algorithms are mostly used in BP neural networks, and there are also shortcomings in the selection of initial weights and other uncertainties, such as overfitting, and the prediction accuracy and practicability cannot meet the requirements at the same time.

Images