Electric power meteorological load predicating method based on cluster screening and neural network

Abstract

The invention discloses an electric power meteorological load predicating method based on cluster screening and neural network. The method comprises the steps of extracting original data from electricpower data, simplifying the original data by a manner of combining a clustering algorithm and a main component analysis method, performing standard processing, inputting the standardized meteorological data and load data into the neural network for training; after training, predicating processing output for obtaining predicated load data, performing calculation and determining the predication precision of the predicated load data, and continuously adjusting an inner parameter of obtaining a neural network predication model and predicating the load data of each day in a to-be-predicated time period. According to the method of the invention, influence of the meteorological data to load fluctuation is sufficiently considered; the data scale is sufficiently considered; through the clusteringalgorithm and the main component analysis method, load and meteorological load volume are simultaneously reduced. Through a defined predication precision calculation formula, the presented combined algorithm combination model has advantages of ensuring high predication precision of the neural network model and improving predication efficiency and predication precision.

Description

technical field [0001] The invention relates to a load forecasting method, in particular to a power meteorological load data forecasting method based on cluster screening and neural network. Background technique [0002] The power system is composed of the power network and power users. Its function is to provide reliable and standard electric energy to various users of the power system as economically as possible, so as to meet the requirements of various users at any time, that is, to meet the load requirements. However, at present, electric energy cannot be stored in a large amount, which requires that the power generation of the system should keep a dynamic balance with the change of the system load at any time. Stablize. The acquisition of future load changes of the system is achieved through load forecasting, so the power system load forecasting has developed and become an important research field in engineering science and an important content in power system automat...

AI Technical Summary

Problems solved by technology

[0007] 1. Most of the existing technologies do not fully consider the impact of meteorological data on load fluctuations, and do not combine meteorological factors with load parameters

[0008] 2. Even if meteorological factors are considered in the load forecasting of the prior art, the prediction efficiency will be low due to the huge amount of meteorological data and load data

[0009] 3. Even if meteorological factors are considered in the load forecasting of the prior art, the prediction accuracy will be low due to the huge meteorological data and load data

[0010] 4. In the existing load forecasting algorithm, the simplified processing of data is only for load data or meteorological data, and does not simplify the processing of load data and meteorological data at the same time

Images