Short-period forecast method for power load

Abstract

The invention relates to a short-period forecast method for a power load. The method comprises the following steps of (1) acquiring a historical power load data sequence, and calculating a hurst indexH of the historical power load data sequence; (2) building a fractional Brownian motion model according to the hurst index H; (3) performing global optimization on the hurst index H in the fractionalBrownian motion model to obtain optimal value Hgbest of the hurst index, and further obtaining a fractional Brownian motion optimization model; and (4) forecasting power load data by the fractional Brownian motion optimization model. Compared with the prior art, the method has the advantages that high-accuracy forecast is performed on short-period d instable power load data.

Description

technical field [0001] The invention relates to a power load forecasting method, in particular to a short-term power load forecasting method. Background technique [0002] Power load forecasting is an important part of power system operation and an important content of power dispatching. According to power dispatching, power system operators can determine the running time of the grid and reduce potential losses. Therefore, accurate power load forecasting helps operators grasp the future power development trend and better dispatch the grid. [0003] At present, there are many traditional power load methods, gray model (GM) is widely used in power load forecasting, however, the accuracy of power load forecasting is often affected by many factors, and the GM exponential growth law cannot deal with these factors, so that a reasonable prediction effect. Autoregressive Integrated Moving Average (ARIMA) has been successfully used to estimate electricity demand, but cannot handle ...

AI Technical Summary

Problems solved by technology

[0003] At present, there are many traditional power load methods, gray model (GM) is widely used in power load forecasting, however, the accuracy of power load forecasting is often affected by many factors, and the GM exponential growth law cannot deal with these factors, so that a reasonable The forecast effect of

Autoregressive Integrated Moving Average (ARIMA) has been successfully used to estimate electricity demand, but cannot handle multivariate ARIMA and heteroscedasticity related issues

In recent years, the error backpropagation (BP) algorithm has been widely used in power forecasting, but it has the disadvantages of slow convergence and easy to fall into local space.

The use of artificial neural networks for power load forecasting guarantees accuracy, however, artificial neural networks also have drawbacks, such as the problem of overfitting and the need for large training instances

Support Vector Regression Machine (SVM) has better prediction results than neural network, but has a complicated calculation process

[0004] At present, the particle swarm optimization (PSO) algorithm is often used to solve optimization problems. The particle swarm optimization algorithm has an inherent shortcoming. Because the particle swarm optimization algorithm and the moving speed of each particle in the search space search its own optimal and global optimal Therefore, the particle swarm optimization algorithm is easy to fall into the local optimal solution

Images