Transformer substation partial discharge positioning method based on electromagnetic wave antenna array signal processing
Technical Field
The invention belongs to the technical field of insulation monitoring of high-voltage power equipment, and relates to a transformer substation partial discharge positioning method based on electromagnetic wave antenna array signal processing.
Background
Partial discharge is both a sign and an expression form of insulation degradation and a reason for further insulation degradation, so partial discharge monitoring becomes an effective means for timely finding insulation defects of electric power equipment and avoiding insulation breakdown faults.
The positioning of the partial discharge position is beneficial to formulating a targeted maintenance treatment scheme, the power failure time is reduced, and the maintenance efficiency is improved. The ultrahigh frequency electromagnetic wave has the advantages of strong anti-interference performance, high sensitivity, stable propagation speed and the like, and is a current hotspot problem when being applied to partial discharge positioning and fault diagnosis. A group of ultra wide band omnidirectional antenna arrays are arranged in the space of a whole substation of a transformer substation to receive electromagnetic wave signals radiated by partial discharge, the position of a partial discharge source is obtained by analyzing and processing the received electromagnetic wave signals, and preliminary research and verification are carried out on the feasibility of partial discharge positioning of the space of the transformer substation by the university of Strathclyde in the UK.
The propagation speed of electromagnetic waves in the air is stable and approximate to the light speed, so that signals received by a sensor array can be analyzed and processed to obtain a signal time delay sequence, and the accurate position of a partial discharge source is obtained by solving a positioning equation system based on the time delay sequence. In practical application, under the condition of low signal-to-noise ratio, the signal delay is difficult to calculate, and a large positioning error is brought, even the positioning cannot be carried out. The equation set to be solved by the positioning method based on the signal time delay sequence is nonlinear, the computer is complex to solve and large in calculation amount, and the measurement error and the time delay calculation error can cause no solution or uncertain solution of the equation set.
The objective of the antenna array side-to-side problem is to locate the signal source by determining the direction of arrival (DOA) of the signals of the two antenna arrays and using the intersection of straight lines in the two directions. The method proposed by Roy et al for obtaining the direction of arrival of a signal by using an invariant rotation technique (ESPRIT) estimates the signal parameters by using the invariant rotation technique, and obtains a generalized eigenvalue on a unit circle by decomposing the matrix beam to determine the direction of arrival of each signal. The method has the advantage of small calculation amount, and can effectively separate independent signal sources and noise sources. The array structure restricts the performance of the estimation of the direction of arrival, and Sarkar T K and the like research the lower bound of DOA estimation variance and prove that the performance of the L-shaped array is optimal.
Disclosure of Invention
The invention aims to overcome the defects of the prior art, and provides a transformer substation partial discharge positioning method based on the L-shaped antenna array signal processing and the idea of solving the signal arrival direction by using the rotation invariant technology.
The technical solution of the invention is as follows:
the invention is composed of an omnidirectional antenna, a built-in amplifier, a high-speed data sampling unit and a data processing unit (computer) which are installed in a transformer substation. The omnidirectional antenna is used for receiving electromagnetic waves excited by partial discharge, the bandwidth of the omnidirectional antenna used for the system is 300-2000M, the received signals are amplified by the preamplifier with the same bandwidth, and the signals are transmitted to the ultra-high speed data acquisition unit through the RF coaxial shielded cable; the ultra-high speed data acquisition unit synchronously acquires 4 paths of output signals, and the sampling frequency of each channel is more than 2 GS/s; the data processing unit processes the acquired 4 paths of digital signals by using the array signals to obtain the direction of arrival angle, and the accurate position of the partial discharge source is obtained by calculation.
2M-1 antennas are respectively distributed on the x axis and the y axis of a space three-dimensional coordinate system at equal intervals to form an L-shaped array, and K (K < M) statistically independent signals are assumed in the space and are respectively directed from two-dimensional angles
Reaches the antenna array S
kIs a signal source of S'
kIs S
kProjection onto the xoy plane, as shown in fig. 2.
Taking the first array element of the x-axis array (i.e. the array element at the origin position) as a reference point, the sampling length of the signal is N, and the output signals of the x-axis array and the y-axis array are respectively expressed as:
X=AxS+Nx (1)
Y=AyS+Ny (2)
wherein X, Y ∈ CM×NObserved signals of two arrays respectively, x is signal, Nx,Ny∈CM×NIs the observed noise of the array. A. thex,Ay∈CM×KThe matrix is a Vandermonde matrix which is composed of array manifold vectors of K signals, and the matrix is expressed by a formula (3) and a formula (4).
Wherein d isx,dyThe array element spacing of the x-axis array and the y-axis array is respectively, and lambda is the wavelength of the signal. The observed signals X and Y are blocked as follows:
<math>
<mrow>
<mi>X</mi>
<mo>=</mo>
<mfenced open='(' close=')'>
<mtable>
<mtr>
<mtd>
<msub>
<mi>X</mi>
<mn>1</mn>
</msub>
</mtd>
</mtr>
<mtr>
<mtd>
<msub>
<mover>
<mi>x</mi>
<mo>→</mo>
</mover>
<mi>M</mi>
</msub>
</mtd>
</mtr>
</mtable>
</mfenced>
<mo>=</mo>
<mfenced open='(' close=')'>
<mtable>
<mtr>
<mtd>
<msub>
<mover>
<mi>x</mi>
<mo>→</mo>
</mover>
<mn>1</mn>
</msub>
</mtd>
</mtr>
<mtr>
<mtd>
<msub>
<mi>X</mi>
<mn>2</mn>
</msub>
</mtd>
</mtr>
</mtable>
</mfenced>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>5</mn>
<mo>)</mo>
</mrow>
</mrow>
</math>
<math>
<mrow>
<mi>Y</mi>
<mo>=</mo>
<mfenced open='(' close=')'>
<mtable>
<mtr>
<mtd>
<msub>
<mi>Y</mi>
<mn>1</mn>
</msub>
</mtd>
</mtr>
<mtr>
<mtd>
<msub>
<mover>
<mi>y</mi>
<mo>→</mo>
</mover>
<mi>M</mi>
</msub>
</mtd>
</mtr>
</mtable>
</mfenced>
<mo>=</mo>
<mfenced open='(' close=')'>
<mtable>
<mtr>
<mtd>
<msub>
<mover>
<mi>y</mi>
<mo>→</mo>
</mover>
<mn>1</mn>
</msub>
</mtd>
</mtr>
<mtr>
<mtd>
<msub>
<mi>Y</mi>
<mn>2</mn>
</msub>
</mtd>
</mtr>
</mtable>
</mfenced>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>6</mn>
<mo>)</mo>
</mrow>
</mrow>
</math>
wherein
And
respectively
row 1 and row M signals of signal X,
and
respectively, the 1 st and mth row signals of signal Y. Matrix A
xAnd A
yThe corresponding blocks are as follows:
<math>
<mrow>
<msub>
<mi>A</mi>
<mi>x</mi>
</msub>
<mo>=</mo>
<mfenced open='(' close=')'>
<mtable>
<mtr>
<mtd>
<msub>
<mi>A</mi>
<mrow>
<mi>x</mi>
<mn>1</mn>
</mrow>
</msub>
</mtd>
</mtr>
<mtr>
<mtd>
<msub>
<mover>
<mi>a</mi>
<mo>→</mo>
</mover>
<mi>xM</mi>
</msub>
</mtd>
</mtr>
</mtable>
</mfenced>
<mo>=</mo>
<mfenced open='(' close=')'>
<mtable>
<mtr>
<mtd>
<msub>
<mover>
<mi>a</mi>
<mo>→</mo>
</mover>
<mrow>
<mi>x</mi>
<mn>1</mn>
</mrow>
</msub>
</mtd>
</mtr>
<mtr>
<mtd>
<msub>
<mi>A</mi>
<mrow>
<mi>x</mi>
<mn>2</mn>
</mrow>
</msub>
</mtd>
</mtr>
</mtable>
</mfenced>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>7</mn>
<mo>)</mo>
</mrow>
</mrow>
</math>
<math>
<mrow>
<msub>
<mi>A</mi>
<mi>y</mi>
</msub>
<mo>=</mo>
<mfenced open='(' close=')'>
<mtable>
<mtr>
<mtd>
<msub>
<mi>A</mi>
<mrow>
<mi>y</mi>
<mn>1</mn>
</mrow>
</msub>
</mtd>
</mtr>
<mtr>
<mtd>
<msub>
<mover>
<mi>a</mi>
<mo>→</mo>
</mover>
<mi>yM</mi>
</msub>
</mtd>
</mtr>
</mtable>
</mfenced>
<mo>=</mo>
<mfenced open='(' close=')'>
<mtable>
<mtr>
<mtd>
<msub>
<mover>
<mi>a</mi>
<mo>→</mo>
</mover>
<mrow>
<mi>y</mi>
<mn>1</mn>
</mrow>
</msub>
</mtd>
</mtr>
<mtr>
<mtd>
<msub>
<mi>A</mi>
<mrow>
<mi>y</mi>
<mn>2</mn>
</mrow>
</msub>
</mtd>
</mtr>
</mtable>
</mfenced>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>8</mn>
<mo>)</mo>
</mrow>
</mrow>
</math>
the following cross-covariance matrix is calculated:
<math>
<mrow>
<msub>
<mi>C</mi>
<mn>2</mn>
</msub>
<mo>=</mo>
<mi>E</mi>
<mo>[</mo>
<msub>
<mi>X</mi>
<mn>2</mn>
</msub>
<msubsup>
<mi>Y</mi>
<mn>1</mn>
<mi>H</mi>
</msubsup>
<mo>]</mo>
<mo>=</mo>
<msub>
<mi>A</mi>
<mrow>
<mi>x</mi>
<mn>1</mn>
</mrow>
</msub>
<msub>
<mi>Φ</mi>
<mi>x</mi>
</msub>
<msub>
<mi>R</mi>
<mi>s</mi>
</msub>
<msubsup>
<mi>A</mi>
<mrow>
<mi>y</mi>
<mn>1</mn>
</mrow>
<mi>H</mi>
</msubsup>
<mo>+</mo>
<msub>
<mi>N</mi>
<mn>2</mn>
</msub>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>10</mn>
<mo>)</mo>
</mrow>
</mrow>
</math>
<math>
<mrow>
<msub>
<mi>C</mi>
<mn>3</mn>
</msub>
<mo>=</mo>
<mi>E</mi>
<mo>[</mo>
<msub>
<mi>X</mi>
<mn>1</mn>
</msub>
<msubsup>
<mi>Y</mi>
<mn>2</mn>
<mi>H</mi>
</msubsup>
<mo>]</mo>
<mo>=</mo>
<msub>
<mi>A</mi>
<mrow>
<mi>x</mi>
<mn>1</mn>
</mrow>
</msub>
<msub>
<mi>R</mi>
<mi>s</mi>
</msub>
<msubsup>
<mrow>
<msub>
<mi>Φ</mi>
<mi>y</mi>
</msub>
<mi>A</mi>
</mrow>
<mrow>
<mi>y</mi>
<mn>1</mn>
</mrow>
<mi>H</mi>
</msubsup>
<mo>+</mo>
<msub>
<mi>N</mi>
<mn>3</mn>
</msub>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>11</mn>
<mo>)</mo>
</mrow>
</mrow>
</math>
<math>
<mrow>
<msub>
<mi>C</mi>
<mn>4</mn>
</msub>
<mo>=</mo>
<mi>E</mi>
<mo>[</mo>
<msub>
<mi>X</mi>
<mn>2</mn>
</msub>
<msubsup>
<mi>Y</mi>
<mn>2</mn>
<mi>H</mi>
</msubsup>
<mo>]</mo>
<mo>=</mo>
<msub>
<mi>A</mi>
<mrow>
<mi>x</mi>
<mn>1</mn>
</mrow>
</msub>
<msub>
<mi>Φ</mi>
<mi>x</mi>
</msub>
<msub>
<mi>R</mi>
<mi>s</mi>
</msub>
<mrow>
<msub>
<mi>Φ</mi>
<mi>y</mi>
</msub>
<msubsup>
<mi>A</mi>
<mrow>
<mi>y</mi>
<mn>1</mn>
</mrow>
<mi>H</mi>
</msubsup>
</mrow>
<mo>+</mo>
<msub>
<mi>N</mi>
<mn>4</mn>
</msub>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>12</mn>
<mo>)</mo>
</mrow>
</mrow>
</math>
wherein E [. C]Representing a mathematical expectation, NiI is 1,2,3,4 is the cross covariance matrix of the noise, Rs=E[SSH]∈CK×KIs a cross-covariance matrix of signals, R, formed by assuming that K signals are independent of each othersAs a diagonal matrix, phixAnd phiyFor the rotation matrix:
the construction matrix C ═ C1,C2,C3,C4)TAnd calculating RC=CCH. For matrix RCPerforming eigenvalue decomposition to obtain a signal subspace expressed by a matrix, wherein the maximum K eigenvalues are { f }kI K |, 1,2, …, K } to correspond to the feature value fkThe feature vector of (a) is a matrix composed of column vectors, and the block representation is:
<math>
<mrow>
<mi>E</mi>
<mo>=</mo>
<mfenced open='(' close=')'>
<mtable>
<mtr>
<mtd>
<msub>
<mi>E</mi>
<mn>1</mn>
</msub>
</mtd>
</mtr>
<mtr>
<mtd>
<msub>
<mi>E</mi>
<mn>2</mn>
</msub>
</mtd>
</mtr>
<mtr>
<mtd>
<msub>
<mi>E</mi>
<mn>3</mn>
</msub>
</mtd>
</mtr>
<mtr>
<mtd>
<msub>
<mi>E</mi>
<mn>4</mn>
</msub>
</mtd>
</mtr>
</mtable>
</mfenced>
<mo>=</mo>
<mfenced open='(' close=')'>
<mtable>
<mtr>
<mtd>
<msub>
<mi>A</mi>
<mrow>
<mi>x</mi>
<mn>1</mn>
</mrow>
</msub>
</mtd>
</mtr>
<mtr>
<mtd>
<msub>
<mi>A</mi>
<mrow>
<mi>x</mi>
<mn>1</mn>
</mrow>
</msub>
<msub>
<mi>Φ</mi>
<mi>x</mi>
</msub>
</mtd>
</mtr>
<mtr>
<mtd>
<msub>
<mi>A</mi>
<mrow>
<mi>x</mi>
<mn>1</mn>
</mrow>
</msub>
<msubsup>
<mi>Φ</mi>
<mi>y</mi>
<mi>H</mi>
</msubsup>
</mtd>
</mtr>
<mtr>
<mtd>
<msub>
<mi>A</mi>
<mrow>
<mi>x</mi>
<mn>1</mn>
</mrow>
</msub>
<msub>
<mi>Φ</mi>
<mi>x</mi>
</msub>
<msubsup>
<mi>Φ</mi>
<mi>y</mi>
<mi>H</mi>
</msubsup>
</mtd>
</mtr>
</mtable>
</mfenced>
<mi>T</mi>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>15</mn>
<mo>)</mo>
</mrow>
</mrow>
</math>
since the matrix T is reversible, Ω
yAnd
having the same characteristic value, Ω
xAnd phi
xHas the same characteristic value and is omega obtained by the formula (15)
yAnd Ω
xThe least squares solution of (c) is:
upper label
Represents the Moore-Penrose Pseudo inverse of the matrix.
The K eigenvalues of (a) are respectively:
the K eigenvalues of (a) are respectively:
k=1,2,…,K。
note the book
And
from the above analysis, u
kAnd v
kCan be respectively composed of matrix
And
the eigenvalues of (2) are solved. So that a signal S can be obtained
kThe two-dimensional direction of arrival angle of (d) is:
<math>
<mrow>
<msub>
<mi>θ</mi>
<mi>k</mi>
</msub>
<mo>=</mo>
<mi>arcsin</mi>
<msqrt>
<msubsup>
<mi>u</mi>
<mi>k</mi>
<mn>2</mn>
</msubsup>
<mo>+</mo>
<msubsup>
<mi>v</mi>
<mi>k</mi>
<mn>2</mn>
</msubsup>
</msqrt>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>18</mn>
<mo>)</mo>
</mrow>
</mrow>
</math>
compared with the prior art, the method does not need to calculate the time delay sequence of the signal, can reduce the requirement on the sampling rate of the acquisition system, can obtain the plane coordinate of the local discharge source by solving the intersection point of the straight lines in the two directions of arrival, namely solving the equation set of the first order of two elements, and avoids solving the nonlinear equation set.
Drawings
Fig. 1 is a schematic structural diagram of a radio frequency antenna array partial discharge detection apparatus according to the present invention.
Fig. 2 is a schematic diagram of an L-shaped antenna array.
Fig. 3 is a flowchart of the partial discharge detection apparatus according to the present invention.
Fig. 4 is a schematic diagram of a planar antenna array and a signal azimuth in an embodiment of the invention.
Fig. 5 shows a waveform and a frequency spectrum of a pulse current for simulation.
Fig. 6 is a schematic diagram (unit: m) of a simulated antenna array and a partially discharged source plane.
Fig. 7 is a simulated waveform of an electromagnetic wave signal received by an antenna array.
Fig. 8 is a field test antenna array layout, wherein (a) is a field test picture and (b) is an antenna coordinate system diagram.
Fig. 9 is a partial discharge signal waveform of a substation site.
Detailed Description
An embodiment of the present invention is given below with reference to the accompanying drawings. The present embodiment is implemented on the premise of the technical solution of the present invention, and a detailed implementation manner and a process are given, but the scope of the present invention is not limited to the following embodiments.
Examples
A planar antenna array composed of 4 antennas is placed in a substation, assuming that the projection of a local discharge source in the plane of the antenna array is P', firstly, an L-shaped array composed of antennas 1,2 and 4 is considered, a three-dimensional rectangular coordinate system is established by taking the antenna 1 as the origin of coordinates, and the xoy plane of the three-dimensional rectangular coordinate system is shown in FIG. 4.
This is the case when M is 2 and K is 1 in fig. 1, and the observed signal is a real signal. The following gives the local discharge source azimuth angle by calculation
Further obtaining the position of the partial discharge source:
step 1: c is calculated from the signals observed by the antennas 1,2, 4 according to the equations (9) to (12)1、C2、C3And C4And constructing a matrix: c ═ C1,C2,C3,C4)T;
Step 2: since K is 1, the matrix R is calculatedC=CCHThe eigenvector corresponding to the largest eigenvalue of (a) is the matrix E shown in equation (15).
And step 3: according to equation (16)
And
calculated to obtain
And
is a real number, i.e.:
u and v are obtained from formulas (19) and (20):
from equations (21, 22), and the two-dimensional direction angle of arrival of the signal
The calculation formula (17, 18) of theta is known as the azimuth angle
Independent of the wavelength of the signal, and theta is dependent on wavelength, discussed below by calculation
The method for positioning the partial discharge plane of the transformer substation is realized.
And 4, step 4: calculation of azimuth from equation (17)
By using the same method, considering the L-shaped array composed of the
antennas 3, 2, 4, and taking the
antenna 3 as the origin of coordinates, a rectangular coordinate system is established according to the dashed line in fig. 2, so as to obtain the azimuth angle of the local discharge source in the coordinate system x ' o ' y ', and the complementary angle is the azimuth angle in fig. 4
Let the coordinates of the antennas 1,2,3,4 in the coordinate system xoy be (0,0), (d) respectivelyx,0)、(dx,dy)、(0,dy) The projection P' of the partial discharge source in the xoy plane has coordinates (x, y).
By azimuth angle
And
and the coordinates of
antenna 1 and
antenna 3 may yield two linear equations:
and solving a linear equation system of two variables consisting of the formula (24) and the formula (25) to obtain the coordinates (x, y) of the projection P' of the local discharge source in the xoy plane, namely realizing the plane positioning of the local discharge source.
Simulation and test verification of method
In order to verify the accuracy and feasibility of the ultrahigh frequency array signal processing-based substation partial discharge positioning method, the partial discharge signals obtained by electromagnetic wave simulation software and the ultrahigh frequency electromagnetic wave signals actually measured by the substation are analyzed and processed respectively, and the position coordinates of the partial discharge source are calculated by using the method.
Simulation verification
Ansoft HFSS 13.0version software is adopted to perform simulation of partial discharge radiation electromagnetic waves. The simulation partial discharge source adopts Gaussian pulse with the amplitude of pulse current of 1A and the pulse width of 4ns, the current direction is the positive direction of the z axis, and the waveform diagram and the frequency spectrum analysis of the pulse current are shown in figure 5.
Establishing a three-dimensional coordinate system in a simulation substation space, wherein the position coordinate of a simulation partial discharge source is P (1.5,6,2), the coordinates of a radio frequency antenna array are respectively (0,0,0), (3,4,0) and (0,4,0), and the unit is as follows: m, assuming that the electromagnetic wave signal radiated by the partial discharge can reach the receiving antenna without a shielding straight line. A schematic plan view of the simulated antenna array and the partial discharge source is shown in fig. 6.
The results of the electromagnetic wave simulation software obtaining the waveforms of the electromagnetic wave signals received by the 4 antennas are shown in fig. 7. By using the method, the azimuth angle of the partial discharge source in fig. 4 can be obtained by analyzing and processing the simulated signal
And
75.3 deg. and 126.9 deg., respectively. Thereby obtaining a system of linear equations shown in formulas (26) and (27), and solving to obtain the plane position of the partial discharge source as (1.55, 5.93), unit: m, compared with the real plane position (1.5,6), the error is 8.6cm, and the requirement of positioning precision is met.
Verification of field test result of transformer substation
In order to verify the effectiveness of the algorithm under the condition of on-site interference of a transformer substation, an ultrahigh frequency signal actually measured by an antenna array partial discharge detection positioning system at a certain 500kV transformer substation is explained. The four ultrahigh frequency omnidirectional antennas of the system form a planar rectangular array, the length and the width of the array are respectively 1.58m and 1.16m, and the layout of the antenna array is shown in fig. 8. A coordinate system is established by taking the antenna 1 as a coordinate origin, and coordinates of the four antennas in the xoy plane are respectively (0,0), (1.58,1.16) and (0,1.16), and the unit is: and m is selected.
A simulated partial discharge source is arranged in the transformer substation, and the coordinate position is (-4,5, 5) m. The high-speed oscilloscope with the storage function is used as a signal acquisition system to synchronously acquire and store ultrahigh frequency electromagnetic waves radiated by the same discharge source and received by an antenna array, fig. 9 shows signal waveforms received by a group of 4 sensors synchronously acquired by the oscilloscope, the sampling rate is 5G, and the signal length is 2 mus (10000 sampling points).
By using the array signal processing method, the local discharge ultrahigh frequency signal acquired on site is analyzed and processed, and the azimuth angle of the partial discharge source in a coordinate system established by taking the
antenna 1 as the origin and the
antenna 3 as the origin can be obtained
And
126.2 deg. and 143.8 deg., respectively. Thereby obtaining a system of linear equations shown in formulas (26) and (27), and solving to obtain the plane position of the partial discharge source as (-3.83, 5.23), unit: and m, compared with the real plane position (-4,5), the error is 28.6cm, and the requirement of the substation partial discharge positioning precision is met. An antenna coordinate system (unit: m) is applied.