Distant-view 220kV power grid networking method and system based on neural network

Abstract

The invention discloses a distant-view 220kV power grid networking method and system based on a neural network. The method comprises the following steps of: inputting the five parameters including theshort-circuit current of the 500kV bus of a transformer substation, the 220kV district power plant rated active power, the 220kV district power plant access power grid voltage level, and the planningtransformer substation parallel operation main transformer number into a preset BP neural network model; enabling the preset BP neural network model to perform networking according to the parametersand output a short-circuit current value corresponding to a networking mode; and when the short-circuit current value is smaller than or equal to the maximum preset short-circuit current, updating thenetworking mode corresponding to the short-circuit current value to a feasible networking mode library. According to the method, a detailed circuit model does not need to be constructed when the short-circuit current is calculated, the working process is simplified, the short-circuit current calculation efficiency and the result accuracy are improved, and the method and system can be used for long-range 220kV network networking planning.

Description

technical field [0001] The invention relates to the technical field of power grid networking, in particular to a neural network-based method and system for the networking of the Vision 220kV power grid. Background technique [0002] At present, for the networking method that uses short-circuit calculation and checks the upper limit of short-circuit current, short-circuit current calculation is the key. Among them, BPA short-circuit current calculation is usually divided into two basic methods based on power flow and grid-based. Both calculation modes first need to build a complex large-scale grid grid and complete the filling of many component parameters. It is suitable for existing or 5-10 The network frame within the year; for the longer-term network frame, the applicability is poor. The calculation of long-term short-circuit current in power system planning and design generally adopts simplified or approximate calculations, and assumptions need to be made: for example, i...

AI Technical Summary

Problems solved by technology

Among them, BPA short-circuit current calculation is usually divided into two basic methods based on power flow and grid-based. Both calculation modes first need to build a complex large-scale grid grid and complete the filling of many component parameters. It is suitable for existing or 5-10 In-year grids; for longer-term grids, the applicability is poor

The calculation of long-term short-circuit current in power system planning and design generally adopts simplified or approximate calculations, and assumptions need to be made: for example, it is assumed that there is no load before the fault, and the per unit value of all node voltages before the fault is equal to 1; assuming that the resistance of each component of the system is slightly Ignore it, and only calculate the periodic component of the fundamental frequency of the short-circuit current; assuming that the short-circuit current of the busbar on the high-voltage side of a 500kV substation provided by the large power grid is close to the switch limit, the bearing capacity of the operating mode of the 220kV grid in extreme cases is measured, etc., so there are often parameter inconsistencies. Certain situations, including uncertain power supply reactance parameters, uncertain line lengths, etc., lead to inaccurate calculation results; at the same time, the aforementioned assumptions are uncertain or there are many combinations, which will lead to complex short-circuit current calculations and low efficiency. In the end, the networking scheme has a large blindness

Images