A regional power grid load prediction method based on user electricity consumption behaviors

Abstract

The invention relates to the technical field of power systems, in particular to a regional power grid load prediction method based on user electricity consumption behaviors, which comprises the following steps of A) importing historical load data; B) carrying out normalization processing on the load data; C) constructing a sample; D) constructing a neural network; E) calculating the residual errorof the node of the last layer of the neural network; F) calculating node residual errors of all layers of the neural network except the first layer; G) obtaining a cost function value; H) obtaining the expression of the cost function with respect to the connection weight value; I) obtaining an expression of the cost function with respect to the bias; J) obtaining a connection weight value and a bias value with the minimum cost function value. The method has the substantive effects that the dispersion degree of subsequent data processing is reduced through the normalization processing of the load data, the precision of a calculation result is improved, and the accuracy of a prediction result can be improved through modeling of an improved neural network which is more suitable for power grid data characteristics.

Description

technical field [0001] The invention relates to the technical field of electric power systems, in particular to a method for forecasting loads of regional power grids based on user power consumption behaviors. Background technique [0002] The power load forecast can provide reference for the daily operation of the regional power grid and the formulation of dispatching plans. According to the results of the power load forecast, the dispatcher can reasonably coordinate the output distribution of each power plant, maintain the balance between supply and demand, and ensure the safety of the power grid. Start-stop arrangement, reduce redundant generator reserve capacity value, and reduce power generation cost. Before the popularization of computers, load forecasting workers did not have the conditions to use software package modeling, rulers, charts and personal experience are the main tools of load forecasting. For example, the look-up table method is a method in which load fo...

AI Technical Summary

Problems solved by technology

[0004] The technical problem to be solved by the present invention is: the technical problem that the accuracy of the short-term load forecasting structure of the current power grid is not high

Images