CN103645422A - Online analyzing method for grid forced power oscillation caused by internal disturbance of power plant - Google Patents

Abstract

The invention discloses an online analyzing method for grid forced power oscillation caused by internal disturbance of a power plant and relates to the field of power system operation and control. The method is mainly used for judging whether the grid forced power oscillation is caused by internal periodical disturbance of the power plant. In actual operation of a grid, the forced power oscillation caused by generator disturbance is the most common oscillation mode, but whether grid oscillation is caused by the internal disturbance of the power plant where the grid is connected or by the disturbance at the grid end is difficult to judge directly through the characteristics of oscillation amplitude, frequency and the like. Therefore, the method is provided to identify measured data parameters of the power plant connected into grid nodes based on FFT (fast Fourier transform) and conduct energy flow phase difference analysis to judge whether the grid oscillation is caused by the internal disturbance of the power plant where the grid is connected. By means of the method, judgment can be conducted directly at a measuring terminal, and the method is simple in calculation and high in analyzing speed and has good actual application value.

Description

A kind of generating plant internal disturbance causes electrical network forced power oscillation on-line analysis

Technical field

The present invention relates to operation and control of electric power system field, more specifically based on FFT Fast Fourier Transform (FFT), carry out parameter identification, by what analyze power plant's access grid nodes, can flow phase differential, judge that online whether the oscillation of power of electrical network is caused by the intercycle property disturbance of access generating plant.

Background technology

Can there is relatively waving and cause persistent oscillation in the situation that lacking damping between rotor in electric system, surge frequency range is at 0.1～2.5Hz, therefore be called low-frequency oscillation under disturbance.Forced oscillation theory has good explanation to some actual generation oscillation of powers, more and more be subject to domestic and international experts and scholars' approval, this theory is thought and is subject to extraneously while continuing that periodically the frequency of power disturbance approaches system frequency when system, can cause significantly oscillation of power.Tang Yong is at < < electric power network technique > > magazine 2006,30 (10): the basic theory > > of the < < Forced Power Oscillation in Power System that 29-33 delivers; Yang Dongjun etc. are at < < Automation of Electric Systems > > magazine 2009,33(23): the low-frequency oscillation Analysis on Mechanism > > of the < < that 24-28 delivers based on WAMS metric data; And Yang Dongjun etc. is at < < Automation of Electric Systems > > magazine 2011,35(10): the < < synchronous generator asynchronous parallelizing that 99-103 delivers causes in the papers such as THE ANALYSIS OF FORCED POWER OSCILLATION IN POWER SYSTEM > >, by theory and electrical network actual case analysis, having proved for the most effective treatment measures of forced power oscillation is to find rapidly and excise disturbing source respectively.

WAMS is comprised of the synchronous phasor measurement unit based on GPS (GPS) (PMU) and communication system thereof, active power, reactive power, voltage, electric current, phase angle and the important switching signal of can be in wide area power system synchronous, high speed acquisition unit and operational outfit are a kind of instruments that can dynamic process of electrical power system be monitored and be analyzed.Several aspects such as the monitoring of the low-frequency oscillation that WAMS is electrical network, vibration event analysis and vibration prevention and inhibition provide new technological means.In addition the fault wave recording device that, has clock synchronous function also can be surveyed the situation of change of the various electric parameters in record fault front and back automatically, accurately and rapidly.The application of the measurement mechanism that these are advanced, has effectively improved fault analysis and the judgement of electrical network.

Electrical network forced power oscillation can be caused by the periodic disturbance of generator or load, and the electrical network forced power oscillation especially causing with generator in engineering reality is the most common.While there is low-frequency oscillation in electrical network, be difficult to directly judge that by features such as oscillation amplitude, frequencies whether vibration is because the intercycle property disturbance of access power plant causes, so cannot in time, effectively take the braking measure of vibrating.

Yang Dongjun etc. are at patent of invention " a kind of method for judging position of disturbance source for forced power oscillation in regional interconnected power grid " patent No.: ZL201110390520.4 and < < Automation of Electric Systems > > magazine 2012, 36(2): in the forced power oscillation disturbance source locating method > > paper of the < < that 26-30 delivers based on parameter identification, proposed a kind of by the oscillation data of grid branch being carried out to the disturbance source locating method that parameter identification solves the forced power oscillation of the direction of energy flow factor, main know-why is as follows:

For one machine infinity bus system, generator adopts second order classical model, and microvariations situation lower linear equation of rotor motion is:

$\left\{\begin{array}{c}M\frac{\mathrm{d\&Delta;\&omega;}}{\mathrm{dt}}=\mathrm{\&Delta;}{P}_M-\mathrm{\&Delta;}{P}_e-\mathrm{D\&Delta;\&omega;}\\ \frac{\mathrm{d\&Delta;\&delta;}}{\mathrm{dt}}={\mathrm{\&omega;}}_0\mathrm{\&Delta;\&omega;}\end{array}\right.---\left(1\right)$

In formula: M is inertia constant of a set, D is Generator Damping coefficient, Δ P _mfor power of disturbance variable quantity, Δ P _efor electric power variable quantity, Δ δ is rotor angle skew, and Δ ω is rotation speed change amount, ω ₀for reference frequency.

Formula (1) is carried out to first integral, with active power, flow out node direction for just, the energy function that can obtain system is:

V＝V _KE+V _PE＝V _M+V _D （2）

In formula:

Energetic function: $V_{\mathrm{KE}}=\&Integral;\mathrm{M\&Delta;}\stackrel{\&CenterDot;}{\mathrm{\&omega;}}\mathrm{\&Delta;\&omega;}{\mathrm{\&omega;}}_0\mathrm{dt};$

Potential-energy function: V _pE=∫ Δ P _eΔ ω ω ₀dt;

Execute disturbance energy function: V outward _m=∫ Δ P _mΔ ω ω ₀dt;

Damping work energy function: V _d=∫-D Δ ω Δ ω ω ₀dt;

In the steady-state process of forced power oscillation, forcing frequency and system frequency approximately equal, now execute disturbance acting outward and equate with system damping dissipation energy, V _m≈-V _d, kinetic energy and the potential energy of system are changed completely, V _kE≈-V _pE, system shows as undamped free oscillation.

The energy of executing disturbance acting injected system is outward to propagate in network by potential energy, branch road L in system _iji end potential-energy function be:

$V_{\mathrm{PEi}}\left(t\right)={\&Integral;}_0^t\mathrm{\&Delta;}{P}_{\mathrm{ij}}\mathrm{\&Delta;}{\mathrm{\&omega;}}_i{\mathrm{\&omega;}}_0\mathrm{dt}---\left(3\right)$

In formula: Δ P _ijfor branch power variable quantity, Δ ω _ifor branch road i end frequency variation.

In the steady-state process of forced power oscillation, each quantity of state all periodically changes with forcing frequency, establishes

Δ P wherein _mij, Δ ω _mibe respectively branch power, frequency change amplitude,

for branch power, frequency change initial phase, ω is forcing frequency; Substitution (3) Shi Ke get:

In formula:

$a=\frac{1}{4\mathrm{\&omega;}}\&CenterDot;\mathrm{\&Delta;}{P}_{\mathrm{mij}}\&CenterDot;\mathrm{\&Delta;}{\mathrm{\&omega;}}_{\mathrm{mi}}\&CenterDot;{\mathrm{\&omega;}}_0---\left(5\right)$

Executing disturbance injected system outward and in network, propagating the energy consuming is mainly that formula (6) embodies by the aperiodic component of formula (4), in literary composition, b is defined as " the direction of energy flow factor ", the direction of energy flow factor can characterize size and the direction of aperiodic component in potential energy.Definition potential energy flows out node i for just, flows into node i for negative, and when b > 0, potential energy trend flows to node j by node i, and when b < 0, potential energy trend flows to node i by node j.Thus, when branch road direction of energy flow factor b > 0, disturbing source is positioned at the region at branch road start node place; When b < 0, disturbing source is positioned at branch road terminal node region.

Patent " a kind of method for judging position of disturbance source for forced power oscillation in regional interconnected power grid " further proposes on the basis of the branch road direction of energy flow factor: by each region interconnection direction of energy flow factor is asked to algebraic sum, obtain the direction of energy flow factor in each region, direction of energy flow factor algebra and maximum region are disturbing source region.This patent is by the comprehensive analysis to the grid branch direction of energy flow factor, can solve the disturbing source position judgment problem of large-scale interconnected power system, comprise: the generator shown in accompanying drawing 1 to the point of electrical network to net oscillation problem, and shown in accompanying drawing 2,3, there is the roundabout complicated interconnected network disturbance source locating problem of hunting power.

In patent " a kind of method for judging position of disturbance source for forced power oscillation in regional interconnected power grid ", adopt Prony analytical approach as the parameter identification method of the direction of energy flow factor, Prony analytical approach is directly carried out signal analysis to the data waveform of record in time domain, actual measurement path is considered as to the linear combination of the sinusoidal signal (mode of oscillation) that some frequency is fixed, amplitude changes by exponential law, its model representation is:

Wherein: the number that n is mode of oscillation, for i mode of oscillation, A _0ifor oscillation amplitude; σ _ifor damping factor; ω _ifor hunting angle frequency,

for initial phase.Thereby problem is summarized as to the identification to each frequency, amplitude and ratio of damping.

From formula (6), the flow direction of Branch Potential Energy is by the positive and negative decision of the direction of energy flow factor, and the direction of energy flow factor positive and negative mainly by

determine.At this, establish

work as

have be direction of energy flow factor b > 0; When

time,

be direction of energy flow factor b < 0, therefore, by right analysis also can judge the flow direction of Branch Potential Energy.The present invention's definition

for flowing phase differential.Due to

the direction of Branch Potential Energy can only be characterized, the size of Branch Potential Energy can not be characterized, therefore, by flowing phase differential

can judge that hunting power is without the disturbing source position (as shown in Figure 1) in roundabout situation; And for existing the roundabout complex electric network disturbing source position judgment problem of hunting power still will complete by direction of energy flow factor b (as shown in accompanying drawing 2 and accompanying drawing 3).The present invention is by causing the on-line analysis of electrical network forced power oscillation to the property disturbance of generating plant intercycle, employing can be flowed phase differential and be judged, solve the generator shown in accompanying drawing 1 to the point of electrical network to net oscillation problem.Obviously, problems is adopted that to flow phase differential more easier than the direction of energy flow factor as criterion computing method.

The basis of FFT fast fourier transform algorithm is that supposition input signal is periodic function, can be decomposed into the component sum of integral multiple frequency, comprising constant DC component.One-period is the form that the function factorization of T becomes a unlimited trigonometric series sum:

In formula: ω ₁=2 π/T is the angular frequency of periodic function first-harmonic,

for the amplitude of each frequency content,

for the initial phase angle of each frequency content, n is frequency component number.

By orthogonality of trigonometric function, can be obtained the formula of each coefficient:

$c_0=\frac{1}{T}\underset{0}{\overset{T}{\&Integral;}}f\left(t\right)\mathrm{dt}---\left(10\right)$

$a_n=\frac{2}{T}\underset{0}{\overset{T}{\&Integral;}}f\left(t\right)\cos n{\mathrm{\&omega;}}_1\mathrm{tdt}---\left(11\right)$

$b_n=\frac{2}{T}\underset{0}{\overset{T}{\&Integral;}}f\left(t\right)\sin n{\mathrm{\&omega;}}_1\mathrm{tdt}---\left(12\right)$

By formula (11), formula (12) discretize, can obtain formula:

$a_n=\frac{2}{N}\underset{k=0}{\overset{N-1}{\mathrm{\&Sigma;}}}{f}_k\cos \frac{2\mathrm{\&pi;}}{N}\mathrm{nk}---\left(13\right)$

$b_n=\frac{2}{N}\underset{k=0}{\overset{N-1}{\mathrm{\&Sigma;}}}{f}_k\sin \frac{2\mathrm{\&pi;}}{N}\mathrm{nk}---\left(14\right)$

In formula: n=0,1 ..., (N/2)-1, N is sampling number in unit period.

Periodic function also can be expressed as the array configuration of exponential function:

$f\left(t\right)=\underset{-\&infin;}{\overset{+\&infin;}{\mathrm{\&Sigma;}}}{F}_n{e}^{\mathrm{jn}{\mathrm{\&omega;}}_{1}t},(n=0,\&PlusMinus;1,\&PlusMinus;2,...)---\left(15\right)$

$F_n=\frac{1}{T}\underset{0}{\overset{T}{\&Integral;}}f{\left(t\right)}_n{e}^{-\mathrm{jn}{\mathrm{\&omega;}}_{1}t}\mathrm{dt}---\left(16\right)$

$F_{-n}=\frac{1}{T}\underset{0}{\overset{T}{\&Integral;}}f{\left(t\right)}_n{e}^{\mathrm{jn}{\mathrm{\&omega;}}_{1}t}\mathrm{dt}---\left(17\right)$

Compare with Prony analytical approach, fft algorithm can not be analyzed damping factor and the attenuation of vibration, and this is for the Δ P asking in formula (6) _mijwith Δ ω _miparameter has impact.Yet employing can be flowed phase differential

the phase place that the criterion flowing to as oscillation energy only need to be analyzed branch road active power and frequency under leading oscillation frequency

with

and do not need to analyze Δ P _mijwith Δ ω _miparameter, therefore on parameter identification method, can select FFT is that Fast Fourier Transform (FFT) replaces Prony analytical approach.With respect to Time Domain Analysis such as Prony algorithms, FFT directly analyzes from frequency domain, there is better antijamming capability and computing velocity faster, especially for vibration transient phases free component, disturb in larger situation, adopt fft algorithm to carry out parameter identification and there is better accuracy, and can more early start judgement, shorten the judgement time.

In electrical network forced oscillation process, at vibration initial period, may have the divergent oscillation of long period, transient phases is shaken, as shown in Figure 4, if can realize disturbing source position judgment in vibration transient phases, to shortening the malfunction elimination time and preventing that vibration diffusion is significant.At the transient phases of vibration, free oscillation component Δ δ ₁with forced oscillation component Δ δ ₂exist simultaneously; Yet after the sufficiently long time, due to the existence of system damping, free oscillation will fade away, the motion of unique existence has been exactly forced oscillation, and it will, with frequencies omega without decline ground sustained vibration, enter the steady oscillation stage.

The load angle characteristic of vibration transient phases is:

Δδ(t)＝Δδ ₁(t)+Δδ ₂(t) （18）

In formula: Δ δ ₁(t) be characterized by the free oscillation component of system, Δ δ ₂(t) characterize forced oscillation component;

Wherein:

In formula: ξ is damping ratio, ω _nfor system natural mode shape, ω _dfor free oscillation frequency, A ₀for free oscillation component amplitude,

for free oscillation initial phase, ω is forced power oscillation frequency,

for initial phase corresponding to forced power oscillation frequency.

From formula (18)～(20), in the transient phases of forced oscillation, comprise the relevant information of forced oscillation, can obtain related data by parameter identification in theory.Owing to existing in transient phases free oscillation and forced oscillation component simultaneously, free oscillation component has interference effect to the identification of forced oscillation component, adopt fft algorithm to there is good antijamming capability, can improve to a certain extent the accuracy of parameter identification.

Summary of the invention

The object of this invention is to provide a kind of generating plant internal disturbance and cause electrical network forced power oscillation on-line analysis.The method to the mode of oscillation of electrical network (putting net mode of oscillation), judges whether the oscillation of power of electrical network causes owing to accessing the intercycle property disturbance of generating plant mainly for generating plant online.In the actual motion of electrical network, the forced power oscillation being produced by generator periodic disturbance is the most common a kind of oscillation form, but electrical network vibration is difficult to directly judge that by features such as oscillation amplitude, frequencies vibration is that still the disturbance due to electrical network end causes because access power plant internal disturbance causes.For this reason, the invention provides a kind of measurement data that accesses electric network terminal node by generating plant and judge that vibration produces the method for reason, the method can directly judge in measuring terminals, and can start data identification in the transient phases of vibration, have and calculate the feature easy, discrimination is high, analysis speed is fast and good actual application value.

In order to achieve the above object, the present invention adopts following technical scheme: when electrical network generation forced power oscillation, active power, frequecy characteristic that synchronous phasor measurement unit PMU based on FFTDui power plant access grid nodes or fault wave recording device are surveyed record carry out parameter identification, and try to achieve power plant access grid nodes can flow phase differential, by flowing oscillation of power that phase differential judges electrical network online whether because the intercycle property disturbance of accessed generating plant causes, implementation process as shown in Figure 5.

Generating plant internal disturbance causes an electrical network forced power oscillation on-line analysis, and the method comprises the following step:

A, in AC network, generating plant is connected branch road and is equiped with synchronous phasor measurement unit PMU or fault wave recording device with electrical network;

B, when the low-frequency oscillation of on-line monitoring monitoring of equipment to the leading oscillation frequency ω occurring in electrical network, by synchronous phasor measurement unit PMU or fault wave recording device, survey record generating plant is connected branch road grid side active power Δ P and frequency Δ ω data with electrical network, establishing active power, from electrical network, to flow to generating plant direction be positive dirction;

C, the Δ P and the Δ ω that adopt FFT fast Fourier transform analysis method to record survey carry out parameter identification, obtain n the frequency component ω of Δ P _piand corresponding initial phase

m the frequency component ω with Δ ω _{ω j}and corresponding initial phase

wherein, i ∈ 1,2,3 ..., n}, j ∈ 1,2,3 ..., m}, ω _pi, ω _{ω j}∈ (0.1Hz, 2.5Hz);

D, ask each frequency component of Δ P ω _pithe poor Δ ω of minimum frequency with the leading oscillation frequency ω of electrical network _p=min{| ω-ω _pi|| i=1,2 ..., n}, asks each frequency component of Δ ω ω _{ω j}the poor Δ ω of minimum frequency with the leading oscillation frequency ω of electrical network _ω=min{| ω-ω _{ω j}|| j=1,2 ..., m}; If frequency error threshold value is ε, if Δ ω _p< ε and Δ ω _ω< ε, can think the meritorious Δ P frequency component ω of the poor correspondence of minimum frequency _pileading oscillation frequency ω is identical with electrical network, and corresponding initial phase is

frequency Δ ω frequency component ω _{ω j}leading oscillation frequency ω is identical with electrical network, and corresponding initial phase is

if Δ ω _p> ε or Δ ω _ω> ε,, from new survey record active power Δ P and frequency Δ ω data, returns to step c;

If e

can flow phase differential is:

if

can flow phase differential is:

F, when time,

show that oscillation energy flows to generating plant from electrical network, vibration is caused by electrical network end, shows that vibration is caused by the property disturbance of generating plant intercycle; When

time, show that oscillation energy flows to electrical network from generating plant, vibration is caused by the property disturbance of generating plant intercycle.

The present invention has the following advantages:

1, compare with existing method, the present invention is the mode of oscillation to electrical network mainly for generating plant, put the mode of oscillation to net, can be directly synchronous phasor measurement unit PMU or fault wave recording device measurement data by terminal directly judge, do not need the comprehensive analysis through each branch road of electrical network, node, method is simple, efficiency of the practice is high.

2, because determination methods does not need oscillation amplitude and damping parameter as basis for estimation, so parameter identification can directly utilize frequency domain data analysis.The present invention adopts fft algorithm as parameter identification method, with respect to Time Domain Analysis such as Prony algorithms, there is better antijamming capability and computing velocity faster, especially for vibration transient phases free component, disturb larger, and mains frequency amplitude of variation less, disturb in more situation, adopt fft algorithm to carry out parameter identification and there is better accuracy, and can more early start judgement, shorten the judgement time.

3, the present invention adopts and can flow phase differential as basis for estimation, with respect to existing method, adopts the direction of energy flow factor as basis for estimation, and calculated amount is less, and computing velocity is faster, is more suitable for real-time online and calculates.

Accompanying drawing explanation

The mode of oscillation schematic diagram of point to net between Fig. 1 generating plant and electrical network.

The mode of oscillation schematic diagram of generating plant G1 to electrical network S, oscillation energy m1 flows to electrical network S from generating plant G1.

The interregional roundabout mode of oscillation schematic diagram of oscillation energy that exists of Fig. 2 two.

Mode of oscillation schematic diagram between regional power grid S1 and regional power grid S2, oscillation energy m1 flows to S1 from S2, and oscillation energy m2 flows to S2 from S1, and oscillation energy is roundabout flowing between S1 and S2.

The roundabout mode of oscillation schematic diagram of oscillation energy between the complicated multizone of Fig. 3.

S1, S2, mode of oscillation schematic diagram between S3 tri-regional power grids, oscillation energy m1 flows to S2 from S1, and oscillation energy m2 flows to S3 from S2, and oscillation energy m3 flows to S1 from S3.

Each stage schematic diagram of Fig. 4 forced power oscillation.

Forced power oscillation measured waveform figure, is divided into: transient phases, steady-state process, decling phase three phases, wherein approximately more than the 40 second duration of transient phases.

A kind of generating plant of Fig. 5 internal disturbance causes electrical network forced power oscillation on-line analysis process flow diagram.

Fig. 6 generating plant access electrical network schematic diagram.

Generating plant G1 access electrical network S, access node K1 is equiped with synchronous phasor measurement unit PMU or fault wave recording device.

Fig. 7 active power and hunting of frequency comparison of wave shape figure.

Embodiment

Below in conjunction with case study on implementation and accompanying drawing, the present invention is further illustrated.

Embodiment mono-

A, in AC network, generating plant is connected branch road and is equiped with synchronous phasor measurement unit PMU or fault wave recording device with electrical network;

B, when the low-frequency oscillation of on-line monitoring monitoring of equipment to the leading oscillation frequency ω occurring in electrical network, by synchronous phasor measurement unit PMU or fault wave recording device, survey record generating plant is connected branch road grid side active power Δ P and frequency Δ ω data with electrical network, establishing active power, from electrical network, to flow to generating plant direction be positive dirction;

C, the Δ P and the Δ ω that adopt FFT fast Fourier transform analysis method to record survey carry out parameter identification, obtain n the frequency component ω of Δ P _piand corresponding initial phase

m the frequency component ω with Δ ω _{ω j}and corresponding initial phase

wherein, i ∈ 1,2,3 ..., n}, j ∈ 1,2,3 ..., m}, ω _pi, ω _{ω j}∈ (0.1Hz, 2.5Hz);

D, ask each frequency component of Δ P ω _pithe poor Δ ω of minimum frequency with the leading oscillation frequency ω of electrical network _p=min{| ω-ω _pi|| i=1,2 ..., n}, asks each frequency component of Δ ω ω _{ω j}the poor Δ ω of minimum frequency with the leading oscillation frequency ω of electrical network _ω=min{| ω-ω _{ω j}|| j=1,2 ..., m}; If frequency error threshold value is ε, if Δ ω _p< ε and Δ ω _ω< ε, can think the meritorious Δ P frequency component ω of the poor correspondence of minimum frequency _pileading oscillation frequency ω is identical with electrical network, and corresponding initial phase is

frequency Δ ω frequency component ω _{ω j}leading oscillation frequency ω is identical with electrical network, and corresponding initial phase is if Δ ω _p> ε or Δ ω _ω> ε,, from new survey record active power Δ P and frequency Δ ω data, returns to step c;

If e

can flow phase differential is:

if

can flow phase differential is:

F, when

time,

show that oscillation energy flows to generating plant from electrical network, vibration is caused by electrical network end, shows that vibration is caused by the property disturbance of generating plant intercycle; When

time,

show that oscillation energy flows to electrical network from generating plant, vibration is caused by the property disturbance of generating plant intercycle.

Embodiment bis-

In the electric system shown in accompanying drawing 6, when electrical network S surveys record to low-frequency oscillation alarm, the survey record data analysis of application the present invention to the G1 of power plant access electrical network point K1, the intercycle property disturbance whether judgement vibrates by G1 caused, implementation step is as follows:

A, in the electric system shown in Fig. 6, generating plant G1 access electrical network S node K1 place is equiped with synchronous phasor measurement unit PMU;

B, in electrical network S, monitor the low-frequency oscillation of leading oscillation frequency 1.367Hz, the meritorious hunting power Δ P of the grid side of generating plant G1 and electrical network S tie point K1 and the online recorder data of frequency Δ ω are analyzed, record waveform comparison diagram as shown in Figure 7;

C, the meritorious hunting power and the frequency data that adopt fft analysis method to record survey carry out parameter identification, obtain 5 frequency component ω of Δ P _pi5 frequency component ω with Δ ω _{ω j}, and corresponding initial phase

with

Table 1 active power FFT parameter identification result

Table 2 frequency FFT parameter identification result

D, by ω _piand ω _{ω j}survey with electrical network S the leading oscillation frequency 1.367Hz contrast that record is arrived, obtain the most approaching with the leading oscillation frequency of electrical network and meet the frequency component of frequency error threshold value, i.e. ω in present case _p3and ω _{ω 3}frequency component.

E, will pass through ω _p3and ω _{ω 3}corresponding initial phase

with try to achieve and can flow phase differential and be:

F, due to interval, shows that oscillation energy flows to electrical network S from generating plant G1, and vibration is caused by the property disturbance of generating plant intercycle.

Embodiment tri-

In the electric system shown in accompanying drawing 6, when electrical network S surveys record to low-frequency oscillation alarm, the survey record data analysis of application the present invention to the G1 of power plant access electrical network point K1, the intercycle property disturbance whether judgement vibrates by G1 caused, implementation step is as follows:

A, in the electric system shown in Fig. 6, generating plant G1 access electrical network S node K1 place is equiped with synchronous phasor measurement unit PMU;

B, in electrical network S, monitor the low-frequency oscillation of leading oscillation frequency 0.62Hz, the meritorious hunting power Δ P of the grid side of generating plant G1 and electrical network S tie point K1 and the online recorder data of frequency Δ ω are analyzed, power reference direction flows to generating plant for just with electrical network;

C, the meritorious hunting power and the frequency data that adopt fft analysis method to record survey carry out parameter identification, obtain 5 frequency component ω of Δ P _pi5 frequency component ω with Δ ω _{ω j}, and corresponding initial phase

with

Table 1 active power FFT parameter identification result

Table 2 frequency FFT parameter identification result

D, by ω _piand ω _{ω j}survey with electrical network S the leading oscillation frequency 0.62Hz contrast that record is arrived, obtain the most approaching with the leading oscillation frequency of electrical network and meet the frequency component of frequency error threshold value, i.e. ω in present case _p2and ω _{ω 3}identical with electrical network dominant frequency.

E, will pass through ω _p2and ω _{ω 3}corresponding initial phase

with

try to achieve and can flow phase differential and be:

F, due to interval, shows that oscillation energy flows to generating plant G1 from electrical network S, and vibration is caused by grid side disturbance.

Claims (2)

1. generating plant internal disturbance causes an electrical network forced power oscillation on-line analysis, it is characterized in that, the method comprises the following step:

A, in AC network, generating plant is connected branch road and is equiped with synchronous phasor measurement unit PMU or fault wave recording device with electrical network;

B, when on-line monitoring monitoring of equipment is to the leading oscillation frequency occurring in electrical network

low-frequency oscillation, by synchronous phasor measurement unit PMU or fault wave recording device, survey record generating plant is connected branch road grid side active power with electrical network

and frequency

data, establishing active power, from electrical network, to flow to generating plant direction be positive dirction;

C, employing FFT fast Fourier transform analysis method arrive survey record

with

carry out parameter identification, obtain

n frequency component and corresponding initial phase with

m frequency component

and corresponding initial phase

; Wherein,

,

,

,

;

,

;

D, ask

each frequency component

with the leading oscillation frequency of electrical network

minimum frequency poor

, ask

each frequency component

with the leading oscillation frequency of electrical network

minimum frequency poor ; If frequency error threshold value is

if,

and

, can think the meritorious of the poor correspondence of minimum frequency

frequency component

with the leading oscillation frequency of electrical network identical, corresponding initial phase is

, frequency

frequency component

with the leading oscillation frequency of electrical network

identical, corresponding initial phase is

; If

or

, from new survey record active power and frequency

data, return to step c;

If e

, can flow phase differential and be:

if

, can flow phase differential and be: ;

F, when

time,

, showing that oscillation energy flows to generating plant from electrical network, vibration is caused by electrical network end; When

time,

, showing that oscillation energy flows to electrical network from generating plant, vibration is caused by the property disturbance of generating plant intercycle.

2. a kind of generating plant internal disturbance causes electrical network forced power oscillation on-line analysis according to claim 1, it is characterized in that: the frequency component number in described step c

, ; Frequency error threshold value in described steps d

.

Images