CN100541213C - Electric network fault self-adapting diagnostic method based on element correlation analysis - Google Patents

Abstract

Electric network fault self-adapting diagnostic method based on element correlation analysis, it is characterized in that to fault indices function of each power failure element definition in the district that has a power failure, this function has represented towards the output quantity of other power failure elements degree of support to this element fault, according to value identification fault element and the fault type of fault indices function. The inventive method can be according to group move behavior and the electric parameters situation of change of switch, protection; the multiple failure of correct identification large-scale electrical power system; has stronger fault-tolerance; and execution speed is fast in order to be applicable to real-time stabilized control system; can diagnose the multiple failure in the same tripping operation district, be applicable to the fault distinguishing of large-scale electrical power system control system.

Description

Electric network fault self-adapting diagnostic method based on element correlation analysis

Technical field

The invention belongs to the power grid control field, the present invention relates to more precisely a kind of electrical network multiple failure self-adapting diagnostic method based on element correlation analysis, be applicable to the fault diagnosis of large-scale electrical power system, and can be used as the failure criterion method of stabilizing control system.

Background technology

When catastrophe failure that some dispatcher can't react occurs in power system, require stability control device to distinguish quickly and accurately abort situation, the as early as possible normal operation of recovery system of adopting an effective measure. Traditional stability control device fault distinguishing method generally comes failure judgement element and fault type according to protection action and circuit breaker trip information; failure criterion is set usually in advance; the change in topology that can not adapt to electrical network needs again artificially to revise after power network wiring changes. And traditional stability control device is difficult to the complicated multiple failure of correct identification.

Summary of the invention

The objective of the invention is to seek a kind of method for diagnosing faults of energy adaptive network change in topology; the method should be able to be according to group move behavior and the electric parameters situation of change of switch, protection; the multiple failure of correct identification large-scale electrical power system; has stronger fault-tolerance; and execution speed is fast in order to be applicable to real-time stabilized control system, therefore the present invention proposes the electric network fault self-adapting diagnostic method based on element correlation analysis.

The inventive method general thought is as follows: in a power supply interrupted district that comprises a plurality of elements, if 1 element breaks down, then this fault should be able to reflect from the information such as the tripping operation of element around it and overload. Fault indices function of each power failure element definition in the district if give to have a power failure, this function have represented towards the output quantity of other power failure elements degree of support to this element fault, correctly identify fault element and fault type according to the value of fault indices function.

Specifically, the inventive method is described below: all elements in the power supply interrupted district are numbered forming element incidence matrix Y, the element Y of Y_ijBe defined as follows:

Y _ij=Y (i, j)=1, i ≠ j, and element i is adjacent with element j

Y _ij=Y (i, j)=1, i=j, or i ≠ j and element i and element j are non-conterminous

The fault indices function definition of element is as follows:

$F\left(i\right)=\underset{j}{\mathrm{\Σ}}{x}_{\mathrm{ji}}(i=\mathrm{1,2},...,N_{\mathrm{off}})$

N wherein_offBe power failure parts number, x_jiBe the support situation of element j to the element i fault, if j supports i fault, then x_ji=1, otherwise, x_ji＝0。

After calculating the fault indices of all elements, according to the related situation of each element, by the fault indices of each element relatively, ask for the maximum point of fault indices function, if element i₀Fault indices function F (i) be extreme point, then an element i₀It is a fault element. If in the same tripping operation zone, have a plurality of extreme points, then for multiple failure has occured. If same extreme point comprises a plurality of elements, then the element that comprises of this point all is possible trouble point. By analyzing the distribution situation of F (i), can obtain final diagnostic result.

The beneficial effect of the inventive method is as follows: actual proof the method execution speed is fast, and zmodem can be diagnosed the multiple failure in the same tripping operation district. Fault distinguishing applicable to the large-scale electrical power system control system.

The specific embodiment

Embodiment: the adaptive failure of northern Suzhou, Jiangsu stabilizing control system is differentiated

This system is comprised of main website, substation, execution (the cutting machine) three-decker of standing, and purpose is the potential safety hazard behind the defence regional power grid generation multiple failure. In the system and device failure criterion, adopt the electric network fault self-adapting diagnostic method based on element correlation analysis of the present invention.

Generate online overload circuit associated circuitry bug list in the main website of system:

Sequence number	The overload circuit	Associated circuitry 1	Associated circuitry 2	Associated circuitry 3	Associated circuitry 4	Associated circuitry 5	Associated circuitry 6	Associated circuitry 7	Associated circuitry 8	Associated circuitry 9	Associated circuitry 10	Associated circuitry 11	Associated circuitry 12	Associated circuitry 13	Associated circuitry 14	Associated circuitry 15	Associated circuitry 16
																		1	Two upper 5235
2	On the nasal mucus 5236
																		3	Fort two 5233
4	Fort nasal mucus 5234
																		5	Two branchs of a river 5243
6	The nasal mucus branch of a river 5244
																			。。。。。

In the table space be fault correlation function F (i) at the line computation refresh data, when main website has detected the overload event and occurs, determine control mode by the inquiry to the upper table degree of association.

The present invention is illustrated according to the preferred embodiment, should be appreciated that above-described embodiment does not limit the present invention in any form, and the technical scheme that the similar variation of all employings obtains all drops within protection scope of the present invention.

Claims (1)

1, based on the electric network fault self-adapting diagnostic method of element correlation analysis, it is characterized in that to fault indices function of each power failure element definition in the district that has a power failure, this function has represented towards the output quantity of other power failure elements degree of support to this element fault, value identification fault element and fault type according to the fault indices function, specifically, the inventive method is as follows:

All elements in the power supply interrupted district are numbered forming element incidence matrix Y, the element Y of Y_ijBe defined as follows:

Y _ij=Y (i, j)=1, i ≠ j, and element i is adjacent with element j;

Y _ij=Y (i, j)=1, i=j, or i ≠ j and element i and element j are non-conterminous,

The fault indices function definition of element is as follows: $F\left(i\right)=\underset{j}{\mathrm{\Σ}}{x}_{\mathrm{ji}}$ (i＝1，2，...，N _off)，

N wherein_offBe power failure parts number, x_jiBe the support situation of element j to the element i fault, if j supports i fault, then x_ji=1, otherwise, x_ji=0, calculate the fault indices of all elements after, according to the related situation of each element, by the fault indices of each element relatively, ask for the maximum point of fault indices function, if element i₀Fault indices function F (i) be extreme point, then an element i₀It is a fault element. If in the same tripping operation zone, have a plurality of extreme points, then for multiple failure has occured. If same extreme point comprises a plurality of elements, then the element that comprises of this point all is possible trouble point.