Least square support vector machine power consumption prediction method based on adaptive genetic algorithm

Abstract

The invention provides a least square support vector machine power consumption prediction method based on an adaptive genetic algorithm, and the method comprises the steps: collecting historical powerconsumption data from a certain enterprise, and carrying out the denoising processing through a Gaussian filter; decomposing the data after noise filtering into a plurality of basic mode components and margins by adopting an empirical mode decomposition method, and normalizing the basic mode components and the margins; secondly, predicting the normalized components and margins by adopting a leastsquare support vector machine, and optimizing two parameters of the least square support vector machine by adopting a crossover probability and mutation probability adaptive genetic algorithm to findan optimal parameter combination; and finally, obtaining a prediction result by adopting a least square support vector machine model. The method has high prediction precision for the power consumption of the enterprise, and is an efficient power consumption prediction method.

Description

technical field [0001] The invention belongs to the technical field of power consumption forecasting methods, in particular to a least squares support vector machine power consumption forecasting method based on an adaptive genetic algorithm. Background technique [0002] With the deepening of the reform of the electricity sales side and the liberalization of the electricity sales market, the number of electricity sales companies and the forms of participating in market transactions have continued to increase. At present, my country's electricity sales companies are still in the initial stage of development, and their development model is still in the formation stage. As the "spokesperson" of the electricity market reform, electricity sales companies, in addition to market-oriented electricity transactions and providing users with personalized energy services In addition to reducing the cost of comprehensive energy use, it also needs to assume the role of a "firewall" to supp...

AI Technical Summary

Problems solved by technology

[0006] As mentioned above, the existing electricity consumption forecasting methods often use forecasting models to directly predict them, and the prediction accuracy is not high. In order to obtain more information about the time series of electricity consumption, the predecessors often used BP Neural network analysis method, but it has the disadvantages of high redundancy and large deviation of results; while genetic algorithm optimizes parameters, the crossover probability and mutation probability often do not change. Therefore, a method based on adaptive genetic algorithm and The method of least squares support vector machine combination forecasting electricity consumption is of great significance

Images