CN104020370A - Transformer internal fault diagnosis method based on virtual parameter change monitoring - Google Patents

Abstract

The invention provides a transformer internal fault diagnosis method based on virtual parameter change monitoring, specifically comprising the following steps: setting a time interval to obtain active power, reactive power, current and transformer tap position of each side of a transformer of current time section from a SCADA database; determining active and reactive power flow sides of the transformer and calculating total active loss and total reactive loss of the transformer; combining current of each side, calculating virtual leakage reactance and resistance of each side of the transformer by a recursion least square method and recording; monitoring a virtual leakage reactance and resistance changing curve after transformer tap changes; and when a current virtual parameter changes within a period of time and exceeds a definite value and it lasts long enough, it is considered that internal faults might happen to the transformer. According to the invention, measured data requirement is lowered and internal fault online diagnosis of a transformer becomes a convenient and feasible technology. Thus, a transformer with internal faults is timely found, and loss of a power grid is minimized.

Description

Based on virtual parameter, change the power transformer interior fault diagnostic method monitoring

Technical field

The invention belongs to power system device fault diagnosis field, relate more specifically to utilize SCADA data, based on to the tracing and monitoring of transformer virtual resistance and Leakage Reactance variable quantity, realize the inline diagnosis to power transformer interior fault.

Background technology

Conventional monitoring of equipment main website, for example, patrol fibrillar center, to the evaluation of running state of transformer, is that the non-electric quantity indexs such as temperature based on transformer, pressure, oil gas constituent analysis realize.These methods depend on invests corresponding secondary private volumn measurement equipment to transformer primary equipment, not only need extra great amount of investment, also may exist equipment that the difficult problem in transformation is installed, especially for the transformer station having built up, the measurement equipment that installs these non-electric quantities additional has considerable influence to normal operation.Patrol the monitoring of equipment main websites such as fibrillar center and can be online from SCADA, obtain the measurement informations such as the electric current, voltage of each side of transformer, meritorious, idle, load tap changer position, and because these measurements are all the same equipment that is positioned at same station, so the simultaneity of these measurements is substantially secure.Based on these electric parameters measurement informations, can obtain the parameter information such as resistance, leakage reactance of transformer.These parameters do not change when normally operation of equipment, external fault and excitation surge current, therefore, when on-line monitoring is during to parameter generation significant changes such as the resistance of transformer, capacitor, leakage reactances, often mean device interior generation insulation damages, the faults such as winding deformation short circuit.Therefore pass through the variation of the parameter of real time monitoring equipment, can find in time the internal fault of transformer equipment.Thereby save the investment of measuring equipment and avoid transformer that the difficulty of non electrical quantities measurement equipment is installed.

In addition, traditional transformer parameter recognition methods based on electric parameters, based on transformer model more accurately, such as considering no-load voltage ratio, energizing loop impedance, phase-splitting identification etc., its objective is transformer resistance and the leakage reactance calculated closer to actual.Yet, grid dispatching center than the SCADA data that are easier to obtain in, the integral power factor that conventionally only has a second step voltage, Current magnitude measurement and calculate according to three phase power, the transformer accurately that is difficult to accordingly to calculate ordinary meaning divides phase parameter.In order to make transformer online diagnosing technique of support shaft be convenient to apply in the Surveillance center of any routine, the fault diagnosis of therefore only utilizing SCADA data to complete transformer inside is very important.For this reason, this project has proposed a forward-order current amplitude based on SCADA, the comprehensive method of meritorious and the virtual leakage inductance of idle calculating three-phase transformer and resistance; By calculating and statistics to virtual leakage inductance and resistance in a period of time, identify the ANOMALOUS VARIATIONS of virtual leakage inductance and resistance, may there is internal fault in prompting operations staff transformer.Owing to not needing accurately to calculate leakage inductance and the resistance value of transformer, therefore reduced the requirement to measurement data, be convenient to the extensive enforcement of this method.In addition, in same transformer station, RTU is that the simultaneity of SCADA metric data can be too not poor, also for the enforcement of the inventive method with effectively utilize advantage is provided.

Summary of the invention

The object of the invention is to utilize the SCADA data that are easy to obtain in daily electrical network monitors, realize the inline diagnosis to transformer equipment internal fault.Come from the SCADA data of RTU with respect to the Dynamic Phasors data of instantaneous value recorder data and PMU, there is synchronization accuracy not high, the Time Density of data is little, there is no the feature such as voltage or current phase information (except three-phase is amounted to the consideration of power power factor) accurately.The true resistance and the leakage reactance that based on SCADA data, are conventionally difficult to accurately calculate transformer.Propose herein to utilize statistical method to find the ANOMALOUS VARIATIONS of transformer resistance and leakage reactance, do not need to accurately calculate resistance and the leakage reactance of transformer, reduced the requirement to measurement data.

The concrete steps of this invention are as follows:

1. each side three-phase of transformer that for example, obtains discontinuity surface when up-to-date with stage time interval second (2 seconds) from SCADA database is always idle, each each side electric current, load tap changer position mutually of meritorious, three-phase always;

2. the maximum side of the total active power of the Three-Phase Transformer of take is active power inflow side P _in, all the other sides are active power outflow side P _outthereby, obtain the total meritorious P of consumption on transformer _loss; The three winding step-down transformers of take are example, and the active power consuming on this transformer is:

P _loss＝P _in-P _out1-P _out2

3. the maximum side of the total reactive power absolute value of the Three-Phase Transformer of take is reactive power inflow side Q _in, all the other sides are that reactive power flows out side Q _outthereby, obtain the total reactive power consumption Q on transformer _loss; The three winding step-down transformers of take are example, and the reactive power consuming on this transformer is:

Q _loss＝Q _in-Q _out1-Q _out2

4. with P _loss, Q _lossform and measure vectorial d=[P _loss, Q _loss] ^t, the current SCADA effective value of each each side electric current of phase of take measures as basis and forms state vector u, for three winding step-down transformers, and order

$u_k=\left[\begin{array}{cccccc}{I}_{\mathrm{in}\_k}^2& -I_{\mathrm{out}1\_k}^2& -I_{\mathrm{out}2\_k}^2& & 0& \\ & 0& & I_{\mathrm{in}\_k}^2& -I_{\mathrm{out}\_1k}^2& -I_{\mathrm{out}2\_k}^2\end{array}\right],$ Wherein k represents a, b, c phase,

u＝[u _au _bu _c]

With each each side virtual resistance of phase of transformer and virtual leakage reactance, form parameter vector w to be solved, for three winding step-down transformer w _k=[R _{in_k}, R _{out1_k}, R _{out2_k}, X _{in_k}, X _{out1_k}, X _{out2_k}] ^t, w=[w _aw _bw _c] ^t

This tittle meets following linear equation,

$\left[\begin{array}{c}{P}_{\mathrm{loss}}\\ Q_{\mathrm{loss}}\end{array}\right]=d=\mathrm{uw}=\left[\begin{array}{ccc}{u}_a& u_b& u_c\end{array}\right]{\left[\begin{array}{ccc}{w}_a& w_b& w_c\end{array}\right]}^T=u_a{w}_a+u_b{w}_b+u_c{w}_c$

Wherein

$u_k{w}_k=\left[\begin{array}{cccccc}{I}_{\mathrm{in}\_k}^2& -I_{\mathrm{out}1\_k}^2& -I_{\mathrm{out}2\_k}^2& & 0& \\ & 0& & I_{\mathrm{in}\_k}^2& -I_{\mathrm{out}\_1k}^2& -I_{\mathrm{out}2\_k}^2\end{array}\right]\left[\begin{array}{c}{R}_{\mathrm{in}\_k}\\ R_{\mathrm{out}1\_k}\\ R_{\mathrm{out}2\_k}\\ X_{\mathrm{in}}\\ X_{\mathrm{out}1\_k}\\ X_{\mathrm{out}2\_k}\end{array}\right]$

In formula, following table k represents that abc is separate.

5. with recurrent least square method, the parameter phasor that solves above-mentioned equation at each measurement time section i, specifically comprises following sub-step:

5.1) each virtual impedance of transformer and recurrent least square method parameter in parameter phasor are carried out to initialization

A) for three winding step-down transformers, initial each each side resistance reactance of phase is

w _k,-1＝[0.01,0.01,0.01，0.1,0.1,0.1] ^T；

w _-1＝[w _a,-1w _b,-1w _c,-1] ^T

B) to w _{k ,-1}in each initialization virtual parameter set trust-factor, obtain thus diagonal matrix P _{k ,-1}, its diagonal element is comprised of each trust-factor, to the dimension of three-winding transformer one phase, is 6 * 6; If establish each trust-factor, be 0.1,

$P_{k,-1}=\left(\begin{array}{cccccc}0.1& 0& 0& & & \\ 0& 0.1& 0& & 0& \\ 0& 0& 0.1& & & \\ & & & 0.1& 0& 0\\ & 0& & 0& 0.1& 0\\ & & & 0& 0& 0.1\end{array}\right)$

$P_{-1}=\left(\begin{array}{ccc}{P}_{a,-1}& 0& 0\\ 0& P_{b,-1}& 0\\ 0& 0& P_{c,-1}\end{array}\right)$

C) initialization forgetting factor λ, can be taken as 0.99.

5.2) press following formula solution matrix g _i,

$g_i=\frac{{\mathrm{\λ}}^{-1}{P}_{i-1}{u}_i}{1+{\mathrm{\λ}}^{-1}{u}_i^T{P}_{i-1}{u}_i}$

5.3) by following formula, solve vectorial w _i, i.e. the virtual resistance of each each side of transformer and virtual Leakage Reactance, and count historical data base

$w_i=w_{i-1}+g_i[d_i-u_i^T{w}_{i-1}]$

5.4) press following formula solution matrix P _i, be next SCADA or state estimation moment T _i+1solving linear combination coefficient prepares

$P_i={\mathrm{\λ}}^{-1}[P_{i-1}-g_i{u}_i^T{P}_{i-1}].$

6. the load tap changer of usining variation or program initial start each virtual parameter curve after 60 seconds, as efficient pressure swing device virtual parameter observation curve, are estimated the inaccurate impact causing to reduce initial parameter.After having load tap changer change events, regenerate efficient pressure swing device virtual parameter observation curve.

7. get the mean value of the estimated result of initial 120 seconds of effective observation curve of each virtual parameter as the current benchmark virtual parameter of each side winding of transformer.

8. virtual impedance ANOMALOUS VARIATIONS judgement: for virtual resistance parameter, when the virtual resistance parameter of effective observation curve after 120 seconds surpasses current benchmark virtual resistance 20%, and continue more than 600 seconds, think virtual resistance generation ANOMALOUS VARIATIONS; For virtual Leakage Reactance, when the virtual Leakage Reactance of effective observation curve after 120 seconds surpasses the virtual leakage reactance 15% of current benchmark, and continue more than 600 seconds, think virtual leakage reactance generation ANOMALOUS VARIATIONS.

9. when finding transformer side winding virtual resistance or Leakage Reactance generation ANOMALOUS VARIATIONS, think that transformer may exist internal fault, to user, send warning information, user is to this transformer in prompting, and the branch road that especially virtual resistance and leakage reactance change overhauls.

The statistical method of utilizing that the present invention proposes is found the ANOMALOUS VARIATIONS of transformer resistance and leakage reactance, does not need to accurately calculate resistance and the leakage reactance of transformer, has reduced the requirement to measurement data; In addition the simultaneity that the RTU of same transformer station measures can be too not poor for method of the present invention provides advantage yet, this just makes the SCADA data that measure based on RTU of widespread use in engineering, can be in transformer parameter identification, internal fault of electric generator inline diagnosis is become for convenience of feasible technology, thereby large-scale application likely, find that there is in time the transformer of internal fault, reduce grid loss.

Accompanying drawing explanation

Below in conjunction with drawings and the specific embodiments, the present invention is further described in more detail.

Fig. 1 is the power transformer interior fault diagnostic method process flow diagram based on virtual parameter variation monitoring.

Embodiment

The proposed by the invention power transformer interior fault diagnostic method of realizing based on RTU/SCADA data can be applicable to dispatching center, patrols the equipment state supervision module in fibrillar center or transformer station.Raw data can directly be picked up from RTU, also can from the real-time storehouse of SCADA or history library, obtain or forward.Although require to be the data of getting discontinuity surface when same, owing to having adopted the method for statistical method and virtual parameter, therefore not strict to the simultaneity requirement of data, there is very strong fault-tolerance.Conventionally as long as the same time profile data in the SCADA database that main website clock in Qu Yi dispatching center is benchmark just can reach requirement.In addition, due to each side of transformer RTU metric data, normally take the same clock of same transformer station is benchmark, or even same measuring equipment, therefore the error that its simultaneity is caused by clock reference difference is less, be subject to the impact of propagation time delay identical, being therefore conducive to this algorithm obtains diagnosis effect more accurately.

For the selection of the type of metric data, completely according to the collection convention of SCADA.For SCADA, the data of delivering to dispatching center main website on it are three-phase current, three-phase phase voltage and three-phase line voltage normally, and three phases active power sum and three phase reactive power sum and load tap changer positional information.Therefore, the present invention, chooses each side three phases active power sum of transformer and three phase reactive power sum and three-phase current for basis, builds virtual parameter recognition methods.Rather than adopt single-phase power or positive sequence to measure.

In order to monitor in real time transformer equipments all in compass of competency, the application server of EMS or set up independent equipment monitor server take in real time about 2 seconds be the cycle, based on recursive algorithm, calculate the virtual parameter that is respectively monitored transformer, and add up, when finding that virtual parameter and history value have larger variation, remind operations staff to carry out finer diagnosis to corresponding transformer.Virtual parameter is that to take the resistance of transformer and leakage reactance be with reference to building, and these two kinds of parameters do not have large variation under normal operation or during external fault conventionally, therefore can be used as the criterion of power transformer interior fault.Meanwhile, due to only, by the variation discovery fault of these two parameters, therefore can only consider to affect the main state variables of its variation, and needn't accurately calculate its actual value, thereby improve feasibility and the usability of monitoring algorithm.

Realize the method flow diagram of this invention and see accompanying drawing 1, concrete steps are as follows:

Step 1: each side three-phase of transformer that for example, obtains discontinuity surface when up-to-date with stage time interval second (2 seconds) from SCADA database is always idle, each each side electric current, load tap changer position mutually of meritorious, three-phase always;

Step 2: the maximum side of the total active power of the Three-Phase Transformer of take is active power inflow side P _in, all the other sides are active power outflow side P _outthereby, obtain the total meritorious P of consumption on transformer _loss; The three winding step-down transformers of take are example, and the active power consuming on this transformer is:

P _loss＝P _in-P _out1-P _out2

Step 3: the maximum side of the total reactive power absolute value of the Three-Phase Transformer of take is reactive power inflow side Q _in, all the other sides are that reactive power flows out side Q _outthereby, obtain the total reactive power consumption Q on transformer _loss; The three winding step-down transformers of take are example, and the reactive power consuming on this transformer is:

Q _loss＝Q _in-Q _out1-Q _out2

Step 4: with P _loss, Q _lossform and measure vectorial d=[P _loss, Q _loss] ^t, the current SCADA effective value of each each side electric current of phase of take measures as basis and forms state vector u, for three winding step-down transformers, and order

$u_k=\left[\begin{array}{cccccc}{I}_{\mathrm{in}\_k}^2& -I_{\mathrm{out}1\_k}^2& -I_{\mathrm{out}2\_k}^2& & 0& \\ & 0& & I_{\mathrm{in}\_k}^2& -I_{\mathrm{out}\_1k}^2& -I_{\mathrm{out}2\_k}^2\end{array}\right],$ Wherein k represents a, b, c phase,

u＝[u _a u _b u _c]

With each each side virtual resistance of phase of transformer and virtual leakage reactance, form parameter vector w to be solved, for three winding step-down transformer w _k=[R _{in_k}, R _{out1_k}, R _{out2_k}, X _{in_k}, X _{out1_k}, X _{out2_k}] ^t, w=[w _aw _bw _c] ^t

This tittle meets following linear equation,

$\left[\begin{array}{c}{P}_{\mathrm{loss}}\\ Q_{\mathrm{loss}}\end{array}\right]=d=\mathrm{uw}=\left[\begin{array}{ccc}{u}_a& u_b& u_c\end{array}\right]{\left[\begin{array}{ccc}{w}_a& w_b& w_c\end{array}\right]}^T=u_a{w}_a+u_b{w}_b+u_c{w}_c$

Wherein

$u_k{w}_k=\left[\begin{array}{cccccc}{I}_{\mathrm{in}\_k}^2& -I_{\mathrm{out}1\_k}^2& -I_{\mathrm{out}2\_k}^2& & 0& \\ & 0& & I_{\mathrm{in}\_k}^2& -I_{\mathrm{out}\_1k}^2& -I_{\mathrm{out}2\_k}^2\end{array}\right]\left[\begin{array}{c}{R}_{\mathrm{in}\_k}\\ R_{\mathrm{out}1\_k}\\ R_{\mathrm{out}2\_k}\\ X_{\mathrm{in}}\\ X_{\mathrm{out}1\_k}\\ X_{\mathrm{out}2\_k}\end{array}\right]$

In formula, following table k represents that abc is separate.

Step 5: with recurrent least square method, the parameter phasor that solves above-mentioned equation at each measurement time section i, specifically comprises following sub-step:

5.1) each virtual impedance of transformer and recurrent least square method parameter in parameter phasor are carried out to initialization,

A) for three winding step-down transformers, initial each each side resistance reactance of phase is

w _k,-1＝[0.01,0.01,0.01，0.1,0.1,0.1] ^T；

w _-1＝[w _a,-1 w _b,-1 w _c,-1] ^T

B) to w _{k ,-1}in each initialization virtual parameter set trust-factor, obtain thus diagonal matrix P _{k ,-1}, its diagonal element is comprised of each trust-factor, to the dimension of three-winding transformer one phase, is 6 * 6; If establish each trust-factor, be 0.1,

$P_{k,-1}=\left(\begin{array}{cccccc}0.1& 0& 0& & & \\ 0& 0.1& 0& & 0& \\ 0& 0& 0.1& & & \\ & & & 0.1& 0& 0\\ & 0& & 0& 0.1& 0\\ & & & 0& 0& 0.1\end{array}\right)$

$P_{-1}=\left(\begin{array}{ccc}{P}_{a,-1}& 0& 0\\ 0& P_{b,-1}& 0\\ 0& 0& P_{c,-1}\end{array}\right)$

C) initialization forgetting factor λ, can be taken as 0.99.

5.2) press following formula solution matrix g _i,

$g_i=\frac{{\mathrm{\λ}}^{-1}{P}_{i-1}{u}_i}{1+{\mathrm{\λ}}^{-1}{u}_i^T{P}_{i-1}{u}_i}$

5.3) by following formula, solve vectorial w _i, i.e. the virtual resistance of each each side of transformer and virtual Leakage Reactance, and count historical data base

$w_i=w_{i-1}+g_i[d_i-u_i^T{w}_{i-1}]$

5.4) press following formula solution matrix P _i, be next SCADA or state estimation moment T _i+1solving linear combination coefficient prepares

$P_i={\mathrm{\λ}}^{-1}[P_{i-1}-g_i{u}_i^T{P}_{i-1}].$

Step 6: the load tap changer of usining variation or program initial start each virtual parameter curve after 60 seconds, as efficient pressure swing device virtual parameter observation curve, are estimated the inaccurate impact causing to reduce initial parameter.After having load tap changer change events, regenerate efficient pressure swing device virtual parameter observation curve.

Step 7: get the mean value of the estimated result of initial 120 seconds of effective observation curve of each virtual parameter as the current benchmark virtual parameter of each side winding of transformer.

Step 8: virtual impedance ANOMALOUS VARIATIONS judgement: for virtual resistance parameter, when the virtual resistance parameter of effective observation curve after 120 seconds surpasses current benchmark virtual resistance 20%, and continue more than 600 seconds, think virtual resistance generation ANOMALOUS VARIATIONS; For virtual Leakage Reactance, when the virtual Leakage Reactance of effective observation curve after 120 seconds surpasses the virtual leakage reactance 15% of current benchmark, and continue more than 600 seconds, think virtual leakage reactance generation ANOMALOUS VARIATIONS.

Step 9: when finding transformer side winding virtual resistance or Leakage Reactance generation ANOMALOUS VARIATIONS, think that transformer may exist internal fault, to user, send warning information, user is to this transformer in prompting, and the branch road that especially virtual resistance and leakage reactance change overhauls.

Claims (8)

1. the power transformer interior fault diagnostic method change monitoring based on virtual parameter, it is characterized in that, the method is the virtual parameter of calculating transformer only, and the ANOMALOUS VARIATIONS of utilizing statistical method to pass through virtual parameter is found the internal fault of transformer, and it comprises following steps:

Step 1): with total always idle, each each side electric current, the load tap changer position mutually of meritorious, three-phase of each side three-phase of transformer that second, stage time interval obtained discontinuity surface when up-to-date from SCADA database;

Step 2): the maximum side of the total active power of the Three-Phase Transformer of take is active power inflow side P _in, all the other sides are active power outflow side P _outthereby, obtain the total meritorious P of consumption on transformer _loss; The three winding step-down transformers of take are example, and the active power consuming on this transformer is:

P _loss＝P _in-P _out1-P _out2

Step 3): the maximum side of the total reactive power absolute value of the Three-Phase Transformer of take is reactive power inflow side Q _in, all the other sides are that reactive power flows out side Q _outthereby, obtain the total reactive power consumption Q on transformer _loss; The three winding step-down transformers of take are example, and the reactive power consuming on this transformer is:

Q _loss＝Q _in-Q _out1-Q _out2

Step 4): with P _loss, Q _lossform and measure vectorial d=[P _loss, Q _loss] ^t, the current SCADA effective value of each each side electric current of phase of take measures as basis and forms state vector u, for three winding step-down transformers, and order

wherein k represents a, b, c phase,

u＝[u _a u _b u _c]

With each each side virtual resistance of phase of transformer and virtual leakage reactance, form parameter vector w to be solved, for three winding step-down transformer w _k=[R _{in_k}, R _{out1_k}, R _{out2_k}, X _{in_k}, X _{out1_k}, X _{out2_k}] ^t, w=[w _aw _bw _c] ^t

This tittle meets following linear equation,

Wherein

In formula, following table k represents that abc is separate;

Step 5): with recurrent least square method, at each measurement time section i, solve the parameter phasor w of described equation;

Step 6): with load tap changer, change or program initial start T ₁each virtual parameter curve after second, as efficient pressure swing device virtual parameter observation curve, is estimated the inaccurate impact causing to reduce initial parameter,, after having load tap changer change events, regenerates efficient pressure swing device virtual parameter observation curve;

Step 7): the initial T that gets effective observation curve of each virtual parameter ₂the mean value of the estimated result of second is as the current benchmark virtual parameter of each side winding of transformer;

Step 8): virtual impedance ANOMALOUS VARIATIONS judgement: for virtual resistance parameter, as effective observation curve T ₂virtual resistance parameter after second surpass current benchmark virtual resistance number percent number reach α and more than, and lasting T ₃when second above, think virtual resistance generation ANOMALOUS VARIATIONS; For virtual Leakage Reactance, as effective observation curve T ₂virtual Leakage Reactance after second surpass the virtual leakage reactance number percent of current benchmark number reach β and more than, and lasting T ₃when second above, think virtual leakage reactance generation ANOMALOUS VARIATIONS;

Step 9): when finding transformer side winding virtual resistance or Leakage Reactance generation ANOMALOUS VARIATIONS, think that transformer may exist internal fault, to user, send warning information, user is to this transformer in prompting, and the branch road that especially virtual resistance and leakage reactance change overhauls.

2. the power transformer interior fault diagnostic method change monitoring based on virtual parameter according to claim 1, is characterized in that described step 1) in a second stage time interval get 2 seconds.

3. the power transformer interior fault diagnostic method change monitoring based on virtual parameter according to claim 1, is characterized in that described step 5) in while solving the recurrent least square method parameter of the parameter phasor w of discontinuity surface i specifically comprise the following steps:

(1) each virtual impedance of transformer and recurrent least square method parameter in parameter phasor are carried out to initialization:

A) for three winding step-down transformers, initial each each side resistance reactance of phase is

w _k,-1＝[0.01,0.01,0.01，0.1,0.1,0.1] ^T；

w _-1＝[w _a,-1 w _b,-1 w _c,-1] ^T

B) to w _{k ,-1}in each initialization virtual parameter set trust-factor, obtain thus diagonal matrix P _{k ,-1}, its diagonal element is comprised of each trust-factor, to the dimension of three-winding transformer one phase, is 6 * 6; If establish each trust-factor, be 0.1,

C) initialization forgetting factor λ gets 0.99;

(2) press following formula solution matrix g _i,

(3) by following formula, solve vectorial w _i, i.e. the virtual resistance of each each side of transformer and virtual Leakage Reactance, and count historical data base

(4) press following formula solution matrix P _i, be next SCADA or state estimation moment T _i+1solving linear combination coefficient prepares

4. the power transformer interior fault diagnostic method monitoring that changes based on virtual parameter according to claim 1, is characterized in that described step 6) middle time parameter T ₁get 60 seconds.

5. the power transformer interior fault diagnostic method monitoring that changes based on virtual parameter according to claim 1, is characterized in that described step 7) middle time parameter T ₂get 120 seconds.

6. the power transformer interior fault diagnostic method monitoring that changes based on virtual parameter according to claim 1, is characterized in that described step 8) middle time parameter T ₃get 600 seconds.

7. the power transformer interior fault diagnostic method change monitoring based on virtual parameter according to claim 1, is characterized in that described step 8) in number percent count α and get 20%.

8. the power transformer interior fault diagnostic method change monitoring based on virtual parameter according to claim 1, is characterized in that described step 8) in number percent count β and get 15%.