CN116520073A - Fault positioning method for power supply system of submarine observation network - Google Patents
Fault positioning method for power supply system of submarine observation network Download PDFInfo
- Publication number
- CN116520073A CN116520073A CN202310260482.3A CN202310260482A CN116520073A CN 116520073 A CN116520073 A CN 116520073A CN 202310260482 A CN202310260482 A CN 202310260482A CN 116520073 A CN116520073 A CN 116520073A
- Authority
- CN
- China
- Prior art keywords
- node
- leakage current
- nodes
- fault
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 43
- 239000013598 vector Substances 0.000 claims abstract description 102
- 230000002159 abnormal effect Effects 0.000 claims abstract description 88
- 239000011159 matrix material Substances 0.000 claims abstract description 29
- 239000013535 sea water Substances 0.000 claims abstract description 22
- 238000012937 correction Methods 0.000 claims description 16
- 238000005259 measurement Methods 0.000 claims description 12
- 238000001514 detection method Methods 0.000 claims description 7
- 230000036541 health Effects 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 5
- 230000007704 transition Effects 0.000 claims description 4
- 238000011160 research Methods 0.000 claims description 3
- 238000010586 diagram Methods 0.000 description 11
- 230000004807 localization Effects 0.000 description 10
- 230000000694 effects Effects 0.000 description 8
- 238000004891 communication Methods 0.000 description 5
- 230000002547 anomalous effect Effects 0.000 description 3
- 238000004422 calculation algorithm Methods 0.000 description 3
- 241000258957 Asteroidea Species 0.000 description 2
- 101100499229 Mus musculus Dhrsx gene Proteins 0.000 description 2
- 238000013459 approach Methods 0.000 description 2
- 238000012512 characterization method Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 238000004088 simulation Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 1
- 239000002131 composite material Substances 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 230000016507 interphase Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/08—Locating faults in cables, transmission lines, or networks
- G01R31/088—Aspects of digital computing
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/08—Locating faults in cables, transmission lines, or networks
- G01R31/081—Locating faults in cables, transmission lines, or networks according to type of conductors
- G01R31/086—Locating faults in cables, transmission lines, or networks according to type of conductors in power transmission or distribution networks, i.e. with interconnected conductors
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/50—Testing of electric apparatus, lines, cables or components for short-circuits, continuity, leakage current or incorrect line connections
- G01R31/52—Testing for short-circuits, leakage current or ground faults
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
- Y04S10/52—Outage or fault management, e.g. fault detection or location
Abstract
A power supply system fault positioning method of a submarine observation network comprises the steps of constructing a node admittance matrix containing seawater nodes, calculating and analyzing leakage current vectors under different fault scenes, and constructing adjacent threshold leakage current direction similarity vectors according to threshold leakage current characteristics so as to identify nodes with abnormal data; and then carrying out data interpolation based on spatial correlation on the abnormal data, and accurately positioning the fault position by analyzing the relation between the threshold leakage current and the measured data.
Description
Technical Field
The invention relates to a technology in the field of submarine power grid control, in particular to a fault positioning method for a submarine observation network power supply system.
Background
The short circuit fault in the power supply system of the submarine observation network has the advantages of high maintenance cost and long maintenance period, and the rapid and accurate fault positioning is a key for ensuring the reliable and economic operation of the submarine observation network. And because the quality of the data collected by the submarine junction box is poor, the electrical faults in the submarine power grid are more difficult to locate and repair. First, marine environments and biological, natural disasters and human activities can pose a threat to electrical sensors, resulting in the occurrence of corrupted data. Second, subsea photoelectric composite cables are typically encapsulated by cables and communication fibers. Electrical faults may damage the optical fibers, causing cascading failures in the communications network, resulting in large-scale missing data. In addition, the trunk node simplifies its internal structure including a measurement module and a communication module in order to improve reliability and economy, but the shore base station cannot acquire current data of the trunk node. Therefore, achieving accurate positioning of faults in a power supply system of a submarine observation network with limited measurement data and abnormal data is a key to be solved.
The existing power distribution network fault positioning technology is characterized in that a plurality of intelligent bodies are arranged at each switch and load node of a power distribution network feeder line (comprising a main line and each branch line), and the risk of interphase short-circuit faults of different types of sections and circuit lines of different power distribution network running modes is early warned and accurately positioned by judging the leakage current of each phase. However, in the prior art, the influence of abnormal data involved in the submarine observation process is not considered, and a main node of a submarine observation network power supply system is not provided with a communication module, so that special conditions such as current data of a main cable and the like cannot be obtained generally, and the current submarine fault detection field is still blank.
Disclosure of Invention
The invention provides a power supply system fault positioning method for a submarine observation network, which aims at the defects in the prior art, adopts abnormal data detection based on adjacent leakage current direction similarity vectors to accurately identify abnormal nodes by analyzing the characteristics of threshold leakage current, accurately corrects the abnormal data by data interpolation based on spatial correlation, and determines the fault position according to the relation between the threshold leakage current and the interpolation data.
The invention is realized by the following technical scheme:
the invention provides a method for positioning faults of a power supply system of a submarine observation network, which comprises the steps of calculating and analyzing leakage current vectors under different fault scenes by constructing a node admittance matrix containing seawater nodes, and constructing adjacent threshold leakage current direction similarity vectors according to threshold leakage current characteristics so as to identify nodes with abnormal data; and then, carrying out data interpolation based on spatial correlation on the abnormal data, and accurately positioning the fault position by analyzing the relation between the threshold leakage current and the measured data.
Preferably, the invention performs case simulation under different fault scenarios by establishing a system model in the PSCAD to verify the effectiveness of the method.
The leakage current vector is as follows: according to kirchhoff's current law, the sum of the currents flowing into and out of the nodes should be equal to zero. The leakage current of the nodes at two sides of the fault can be caused by the short circuit current generated by the fault, and the leakage current of each node is calculated and recorded in a vector form, namely the leakage current vector.
The threshold leakage current is as follows: in theory, only the nodes on two sides of the fault line will generate leakage current due to short-circuit current, and the leakage current of the other nodes should be zero. However, in practical engineering application, since the leakage currents of all nodes are not equal to zero due to the influence of the measurement errors of the equipment, a corresponding threshold value needs to be set to determine whether the leakage currents are zero in theory. The leakage current generated by the nodes affected by the fault or abnormal data exceeds the threshold value, and is not in the range of measurement errors, namely, the leakage current exceeds the threshold value.
The adjacent threshold leakage current direction similarity vector is as follows: according to the characteristic that abnormal nodes can cause adjacent nodes to generate threshold leakage current in the same direction, the number of adjacent nodes of a certain node and the leakage current direction are analyzed, and one index for detecting the abnormal nodes is defined.
The invention relates to a system for realizing the method, which comprises the following steps: a leakage current vector calculation unit, an abnormal data detection unit, an abnormal data correction unit, and a fault location unit, wherein: the leakage current vector calculation unit constructs a node admittance matrix containing seawater nodes, a voltage vector and a current vector according to system parameters, measured voltages of all nodes, output currents of all bank base stations and information of electrode equivalent impedance, multiplies the node admittance matrix containing seawater nodes by the voltage vector, and finally adds the current vector to obtain a leakage current vector containing the leakage current of each node; the abnormal data detection unit can acquire information of a node incidence matrix, a total number vector of adjacent nodes and a leakage current direction vector according to the node admittance matrix and the leakage current vector, multiply the node incidence matrix with the leakage current direction vector, add the total number vector of the adjacent nodes to obtain an adjacent threshold value leakage current direction similarity vector, and can determine that abnormal data exists in a corresponding node according to an element equal to 0 in the vector; the abnormal data correction unit is used for interpolating the abnormal data by utilizing the health voltage data of the adjacent nodes according to the system parameters and the information of the voltage vectors, replacing the corresponding abnormal data with the interpolated data to update the voltage vectors, and recalculating the leakage current vectors to update the leakage current vectors; the fault locating unit deduces and calculates the fault locating distance and the fault transition resistance by analyzing the relation between the threshold leakage current in the updated leakage current vector and the system parameter and the measured voltage according to the system parameter, the updated leakage current vector and the updated voltage vector.
Technical effects
According to the invention, after the seawater loop is considered and the seawater node is added, the fault positioning error is reduced by 1% as a whole. The identification of 100% precision and the effective correction of the abnormal data can be realized for a plurality of existing abnormal data nodes, and the error of the fault location after correction is reduced by 63.9% on average. Compared with the existing data interpolation method, the fault positioning accuracy is improved by 52.6 percent compared with the Lagrange interpolation method, 36.9 percent compared with the Newton interpolation method, and 16.83 percent compared with the KNN method.
Drawings
FIG. 1 is a schematic diagram of a typical subsea negative DC system;
fig. 2 is a schematic diagram of a section of a main cable in different scenes;
in the figure: (a) is normal operation, (b) is open circuit fault, and (c) is short circuit fault;
FIG. 3 is a simplified pictorial illustration of a portion of a negative DC system;
FIG. 4 is a flow chart diagram illustration of a fault localization method based on leakage current vectors;
FIG. 5 is a schematic diagram of an embodiment application system model;
FIG. 6 is a schematic diagram of a distribution of short-circuit fault leakage current vectors without abnormal data;
FIG. 7 is a schematic diagram of the distribution of open fault leakage current vectors without abnormal data;
FIG. 8 is a diagram showing data comparison before and after interpolation;
FIG. 9 is a graph showing the distribution of leakage current vectors before and after interpolation when abnormal data exists;
FIG. 10 is a diagram showing the comparison of fault location accuracy before and after data interpolation for different fault resistances;
FIG. 11 is a graph showing the distribution of leakage current vectors before and after interpolation in the presence of missing data;
FIG. 12 is a diagram showing the comparison of fault location accuracy before and after data interpolation for different fault resistances;
FIG. 13 is a graph showing the distribution of leakage current vectors before and after interpolation in the presence of abnormal data and missing data;
FIG. 14 is a diagram showing the comparison of fault location accuracy before and after data interpolation for different fault resistances;
fig. 15 is a schematic diagram of the effect of the embodiment.
Detailed Description
As shown in fig. 1, in the subsea negative direct current network, power supply is realized by adopting electrodes and seawater to form a loop. Most fault localization documents of submarine observation networks ignore seawater loops, which actually have an effect on the localization accuracy. Seawater is a uniform, good conductive medium, approximating an equipotential body in the vast sea. Thus, the sea water may act as an auxiliary node of the subsea observation network.
After the seawater nodes are added, all branch currents can be obtained through calculation through voltage data of the branch unit nodes and the seawater nodes and equivalent resistances. The measurement of the branch current is no longer necessary. Thus, voltage error is the only factor to consider. Each branch unit node forms a power supply loop with the shore base station through a seawater node.
In theory, when the system is operating normally leak =0. The leakage current vector is in fact not zero due to the effect of measurement errors. The leakage current error is determined by the voltage data measurement error, the main cable admittance and the node branch current. The branch current does not need to be measured after the seawater node is added. The leakage current error at node j isWherein: epsilon Ileak,j Is the leakage current error of node j; epsilon n Is the measurement error of the node n voltage sensor.
For the submarine negative direct current network, the maximum number of adjacent nodes is three due to the structural limitation of the branch units, namely, the leakage current error can be calculated by three groups of cable admittances and voltage errors at most. In order to ensure that zero leakage current cannot be misjudged, the threshold value of the zero leakage current should be equal to the maximum possible leakage current error, specifically: t (T) Ileak ＝n·Y s ·ε max Wherein: n is the maximum number of adjacent branching unit nodes; y is Y s Is the admittance value of the main cable; epsilon max Is the maximum measurement error of all node voltage sensors.
Since the main cable is much longer than the branch cables, the focus of this embodiment is how to accurately locate faults on the main cable. For a faulty node its leakage current will be above the threshold (maximum error), i.e. the more threshold leakage current.
As shown in fig. 2 and fig. 4, this embodiment relates to a method for positioning faults of a power supply system of a submarine observation network, including:
step 1, calculating a leakage current vector, which specifically comprises the following steps:
1.1 A node admittance matrix, a voltage vector and a current vector can be obtained according to the system parameters, respectively. The leakage current vector which can be obtained based on kirchhoff's current law is specifically:wherein: i leak Is the vector of leakage currents for each node and Y is the node admittance matrix. I leak,n Is the leakage current at node n; u (U) n Is the voltage at node n; i SS Is the current of the shore base station, I SN Is the current of the sea water node.
1.2 Taking node j and node k as research objects, when no abnormal data is considered, the following three different scenes exist:
a) As shown in fig. 2 (a), the leakage current of the node j is when the system is operating normallyWherein: i leak,j Is the leakage current of node j; y is Y ji The cable admittance between node j and node i; g ji Is the cable conductance magnitude between node j and node i.
b) As shown in FIG. 2 (b), when an open circuit fault occurs between node j and node k, the actual conductance between the two nodes will change, and the leakage current at node j will become variableAnd can obtain two threshold leakage currents G jk ·(U j -U k )＝I leak,j ＝-I leak,k ≠0。
c) As shown in fig. 2 (c), when a short-circuit fault occurs, the fault is added as a new node between node j and node k. Leakage current of node j becomesAnd two threshold leakage currents can be obtained>
Thus, leakage current vector characteristics for three scenarios without considering abnormal data are shown in table 1.
TABLE 1 characterization of leakage current vector in the absence of anomalous data
And 2, detecting abnormal data, namely identifying abnormal nodes by checking the similarity of leakage current directions of adjacent nodes.
The more threshold leakage current generated by the fault can be used for fault localization. Whereas the more threshold leakage current generated by the anomaly data only interferes with fault localization, i.e., the pseudo more threshold leakage current, as shown in fig. 3. Since the cable distance between each node is typically tens of kilometers, a fault will only affect the sensors and communication devices of its nearby nodes. Three typical anomaly nodes are shown in Table 2.
TABLE 2 characterization of leakage current vector in the presence of anomalous data
The leakage current from the branching unit to the junction box is defined as positive. Since the errors caused by the affected electrical sensors are unpredictable, the anomaly data may be greater or less than the true data, i.e., the deviation is positive or negative.
The step 2 specifically comprises the following steps:
2.1 From the node admittance matrix and the leakage current vector, a leakage current direction vector T can be obtained lcd And a neighboring node total number vector N an Calculating the similarity of adjacent leakage current directionsWherein: s is S alcd For adjacent leakage currentsA directional similarity; s is S alcd,1 Abnormal data for identifying positive deviation, S alcd,2 Abnormal data for identifying a negative bias; m is M ni Is a node association matrix in which the elements +.>T lcd Is the leakage current direction of each node, wherein the element +.>Wherein: m is M ni,jk Representative matrix M ni The j-th row and the k-th column of the element; t (T) lcd,n Representative vector T lcd Corresponding to the nth row of the plurality of elements.
2.2 When matrix S alcd If there is an element equal to 0, the threshold leakage current direction is the same for all adjacent nodes of the corresponding node of the element, i.e. the corresponding node is a potential abnormal node. Removing the identified abnormal node from the node admittance matrix and the voltage current vector, and recalculating the leakage current vector sum S alcd To identify more potential outlier nodes.
2.3 If no more abnormal nodes can be identified, the leakage current vector still does not match the normal operation characteristics in table 1, and then two adjacent abnormal nodes with the same deviation exist. By traversing T lcd All the adjacent nodes with the same elements in the network, combining the found adjacent nodes into a node, updating the node admittance matrix, and recalculating the S alcd The remaining outlier nodes are identified.
Step 3, abnormal data interpolation, namely correcting abnormal data by utilizing health data of adjacent nodes, specifically comprising the following steps:
3.1 If the abnormal node x is not adjacent to any other pseudo-fault node, the power supply of the node x to the connection box is stopped, and the abnormal data is corrected by using an interpolation equation, specifically:wherein: u (U) re Interpolation voltage for pseudo fault node, R xi Is the resistance between node x and node i.
3.2 After the data interpolation is realized in the step 3.1), if the abnormal node still has the threshold leakage current, the abnormal node is adjacent to the fault point, and the interpolation equation of the data of the single adjacent node is utilized for correction, specifically:
wherein: node x is an outlier node and node z is one of the neighboring healthy nodes.
3.3 When two adjacent abnormal nodes (node x and node y) exist, the correction is performed through interpolation of an equation set, specifically:wherein: u (U) re,x Is the interpolated data for node x.
3.4 After the data interpolation is realized in the step 3.3), the abnormal node still has the threshold leakage current, and the fault point is adjacent to one of the abnormal nodes, so that a correction equation set is further providedWherein: node z is one of healthy neighboring nodes of node y, U z Is the voltage at node z; r is R zi Is the resistance between node z and node i.
Step 4, fault location, namely, passing through a locating device of the shore base station according to the updated node leakage current I leak,j ,I leak,k The accurate fault location specifically includes:
4.1 Calculating the fault position according to the relation between the threshold leakage current and the measurement data, specifically: according to the resistance from node j to fault pointGet the fault distance +.>And fault resistanceWherein: r is R jf Is the electricity from node j to the point of failureResistance size; i leak,j Is the leakage current magnitude of node j, - (I) leak,j +I leak,k ) Is the current through the fault resistor.
4.2 Calculating a fault distance set by using the steady-state data after the fault in the time domain, and taking an arithmetic average value as a final result to improve the precision of fault positioning.
Through specific practical experiments, a sea-king star system model is established in PSCAD, as shown in figure 5. The maximum measurement error of the voltage sensor in the starfish system is +/-0.7%. The shortest cable length between two nodes is 100km, i.e. the maximum cable admittance is 0.0476S. The leakage current threshold was set to + -0.9A considering that the voltage level of the starfish system was 10 kV. The fault location error in this embodiment is an absolute error.
Faults of 100 omega, 500 omega, 1k omega, 2k omega, 3k omega resistances are respectively arranged on the cable between the node 2 and the node 30, and the distance between the fault and the node 2 is 20km. As shown in fig. 6, the distribution of the leakage current is represented using a polar scattergram. The leakage current gradually increases from the center of the circle to the outer circle, and each circle radius represents a different leakage current value. All leakage currents in the figure are absolute values.
The distribution of leakage current vectors is shown in fig. 6. It can be seen from the leakage current vector profile that there are only two threshold-crossing leakage currents, i.e. the short-circuit fault characteristics described in table 1 are met. The effect on the accuracy of the fault location method after adding seawater nodes is shown in table 3. After considering the sea water nodes, the average fault location error can be reduced by 1%.
TABLE 3 comparison of fault location accuracy before and after considering seawater nodes
An open circuit fault was placed on the cable between node 29 and node 30, the distance between the fault point and node 29 being 10km.
The original leakage current vector distribution is shown in fig. 7. Only two nodes 29 and 30 have threshold leakage currents of equal absolute value. And from the leakage current vector, the leakage currents of the two nodes are exactly opposite numbers, which are matched with the open-circuit fault characteristics described in table 1. Therefore, the fault locating method can accurately locate the open circuit fault when no abnormal data exists.
A short-circuit fault was set on the cable between node 6 and node 41 and it was simulated that there was damage data for node 6, where the distance between the fault and node 41 was 30km.
The voltage data pairs before and after data interpolation are shown in fig. 3-8, for example. And randomly extracting 10 groups of steady-state voltage data after faults, and respectively marking the steady-state voltage data as 1-10. There is a deviation of about + -200V of the pre-interpolation data from the real data, which may result in a larger error in the threshold leakage current. After data interpolation, the voltage deviation can be reduced to 5V, namely the influence on the threshold leakage current is reduced.
The leakage current vector is calculated using the raw data and the interpolated data, respectively, as shown in fig. 9. From the diagonal columns in the figure, it can be obtained that there are multiple threshold leakage currents, i.e. there are abnormal nodes in the system. Firstly, the abnormal data is interpolated, after the interpolation, 3 threshold-crossing leakage currents still exist in the transverse line column, the threshold-crossing leakage currents are not matched with the characteristics in the normal operation in the table 1, and the fault is adjacent to the abnormal node. After interpolation is carried out on the abnormal data, the leakage current vector is matched with the characteristics of the short circuit fault, as shown by a solid column. The fault that is available after the data interpolation is completed is located between node 6 and node 41.
The fault location accuracy pairs before and after data interpolation are shown in fig. 10. The average fault localization error before data interpolation was 39%, as indicated by the short dashed line. As can be seen from fig. 8, the value of the threshold leakage current is affected by the pseudo-threshold leakage current before data interpolation. And the calculated fault distance for the threshold leakage current is subject to a larger error after the impact, as shown in fig. 9. And after data interpolation, the fault location error is reduced by at least 17%.
A short circuit fault was set on the cable between node 16 and node 17 and it was simulated that there was missing data at node 15, the fault to node 16 distance was 10km.
There are 5 threshold leakage currents in the system shown by the diagonal columns in fig. 11, i.e., there are potential outlier nodes. Missing data interpolation is achieved, and the interpolated leakage current vector is matched with the short circuit fault signature, so that a fault is located between node 16 and node 17. Since the missing data is typically set to 0 in the simulation, there is a larger deviation between the missing data and the real data than the corrupted data, which will result in the generation of a larger pseudo-threshold leakage current. There may be missing data from graph node 5 to node 7. And the fault distance is determined by the proportional relation between the two threshold leakage currents. Before data interpolation, a set of pseudo-threshold leakage currents (nodes 14, 15, 16, 35) are caused by missing data, which is approximately 10 times greater than the other threshold leakage currents (node 17). As shown by the dashed line in fig. 12, the fault distance calculated with the affected threshold leakage current will be close to the fault cable length. The fault location error is about 94% before data interpolation, and the error can be reduced by about 90% after data interpolation is realized.
A short-circuit fault was set between node 9 and node 44 and it was simulated that node 8 had corrupted data and node 43 had missing data, with the fault being 30km from node 9.
The leakage current vector distribution is shown in fig. 13. The raw data shown in red bar indicates that there are 7 threshold leakage currents, which means that there is bad data for more than one node. Nodes 7-9 may have anomalous data and nodes 42-44 may have missing data. After data interpolation, the leakage current vector matches the short circuit fault signature and clearly shows that the fault is located between nodes 9 and 30.
The fault location accuracy pairs before and after data interpolation for different transition resistances are shown in fig. 14. According to the fault distance result after data interpolation, the method can accurately identify and interpolate the abnormal data under the condition that different types of abnormal data exist at the same time. As shown by the short dashed line, the fault localization error is about 67% before the data interpolation. The proposed data interpolation method can reduce the error to 3%, as indicated by the long-dashed line in the figure.
Simulating that 5 nodes simultaneously have bad data of different types, wherein nodes 27, 40 and 41 have bad data and nodes 25 and 29 have missing data. Short circuit faults are set between nodes 28 and 29 (fault 1), nodes 6 and 41 (fault 2), respectively, and open circuit faults are set between nodes 24 and 25 (fault 3).
The threshold leakage currents under different fault scenarios were calculated and the results are shown in table 4. The threshold leakage currents in faults 1 and 3 are affected by missing data due to the fact that they areMismatch with the fault signature in table 2, results in difficulty in determining the actual fault location. Whereas the threshold leakage current in fault 2 is affected only by the damage data, it is possible to +.>Matching the described short circuit fault signature. However, the threshold leakage current is deviated by 4.47A before and after data interpolation, which also results in a large fault location distance error. From the fault 3 condition, the threshold leakage current after data interpolation approaches to be the opposite number, and the sum of 41.01A and-41.06A is smaller than the threshold of + -0.9A, i.e. meets the characteristics of an open circuit fault. The fault location results in different scenarios are shown in table 5. The provided interpolation method can realize the correction of abnormal data under different fault scenes, and can reduce the fault positioning error to 3.2% on average.
TABLE 4 threshold leakage currents under different fault conditions
TABLE 5 short circuit fault location results when 5 pseudo-fault nodes exist
Existing fault location methods typically delete poor quality data rather than perform data interpolation. This is a reasonable approach considering redundancy of land grid data. However, cascading failures are common in subsea observation networks, which will lead to large-scale generation of abnormal data. From the previous case, data interpolation is critical to subsea fault localization. Conventional data interpolation methods are typically based on statistical interpolation and regression algorithms. The comparison results of the method with the classical Lagrange interpolation method and the Newton interpolation method are shown in table 6, and the fault scene is adopted and the transition resistance is set to be 1000 ohms.
The result shows that after the data is interpolated by adopting the Lagrange interpolation method or the Newton interpolation method, a large error still exists in fault positioning. For failure scenario 1, the error of newton method was 31.1% and the error of lagrangian method was 25.9%. For failure scenario 2, the error of newton method is 79.3% and the error of lagrangian method is 53.1%. The data correction effect of the invention is obviously superior to that of the traditional data interpolation method, the error in the fault 1 scene is only 3.25%, and the error in the fault 2 scene is only 2.91%. In the case of research, the invention can improve the fault location precision by at least 23%.
The reason reasoning is as follows: both interpolation-based data correction methods require health data in the time domain. However, the abnormal data may continuously exist on the node, resulting in poor effect of the interpolation method. In contrast, the proposed data interpolation method relies only on health data of neighboring nodes and on topology relations. This spatial correlation method based on circuit theory will not be positively influenced by the information in the time domain.
TABLE 6 comparison of fault location accuracy under different data correction methods
The existing anomaly data detection method is combined with a proximity algorithm (KNN, K-Nearest Neighbor) and compared with the present invention. As shown in fig. 3-15, when there is abnormal data on the node P. K healthy node data of the adjacent node P are taken and a weighted average is calculated, and the abnormal data on the node P is replaced by the healthy node data. Since the branch unit node has only 3 interfaces, one node can only connect to at most 3 other nodes, setting K to 3. The weight coefficient is the length of the adjacent node connected cable and adopts the same fault scene as the above.
The comparison result is shown in FIG. 15. From the results, the average fault localization accuracy of the combined algorithm was 18.32%, and the average fault localization accuracy of the present invention was 1.49%. In this example, the invention improves fault location accuracy by at least 8.9% because machine learning methods like KNN are driven based on data that do not take into account the hidden relationship between leakage current vector, node voltage and cable admittance.
The foregoing embodiments may be partially modified in numerous ways by those skilled in the art without departing from the principles and spirit of the invention, the scope of which is defined in the claims and not by the foregoing embodiments, and all such implementations are within the scope of the invention.
Claims (5)
1. A method for positioning faults of a power supply system of a submarine observation network is characterized by comprising the steps of calculating and analyzing leakage current vectors under different fault scenes by constructing a node admittance matrix containing seawater nodes, and constructing adjacent threshold leakage current direction similarity vectors according to threshold leakage current characteristics so as to identify nodes with abnormal data; then, carrying out data interpolation based on spatial correlation on the abnormal data, and accurately positioning the fault position by analyzing the relation between the threshold leakage current and the measured data;
the adjacent threshold leakage current direction similarity vector is as follows: according to the characteristic that abnormal nodes can cause adjacent nodes to generate threshold leakage current in the same direction, the number of adjacent nodes of a certain node and the leakage current direction are analyzed, and one index for detecting the abnormal nodes is defined.
2. The method for positioning faults of a power supply system of a submarine observation network according to claim 1, comprising the following steps:
step 1, calculating a leakage current vector, which specifically comprises the following steps:
1.1 Respectively obtaining a node admittance matrix, a voltage vector and a current vector according to system parameters; the leakage current vector which can be obtained based on kirchhoff's current law is specifically:wherein: i leak Is a vector composed of leakage currents of each node, and Y is a node admittance matrix; i leak,n Is the leakage current at node n; u (U) n Is the voltage at node n; i SS Is the current of the shore base station, I SN Is the current of the sea water node;
1.2 Taking the node j and the node k as research objects, and obtaining leakage current vector characteristics of different scenes when abnormal data are not considered;
step 2, detecting abnormal data, namely identifying abnormal nodes by checking the similarity of leakage current directions of adjacent nodes, wherein the method specifically comprises the following steps:
2.1 From the node admittance matrix and the leakage current vector, a leakage current direction vector T can be obtained lcd And a neighboring node total number vector N an Calculating the similarity of adjacent leakage current directionsWherein: s is S alcd The similarity of adjacent leakage current directions; s is S alcd,1 Abnormal data for identifying positive deviation, S alcd,2 Abnormal data for identifying a negative bias; m is M ni Is a node association matrix in which the elements +.>T lcd Is the leakage current direction of each node, the elements thereinWherein: m is M ni,jk Representative matrix M ni The j-th row and the k-th column of the element; t (T) lcd,n Representative vector T lcd N th row ofCorresponding elements;
2.2 When matrix S alcd If the element equal to 0 exists in the node, the threshold leakage current direction of all adjacent nodes of the corresponding node of the element is the same, namely the corresponding node is a potential abnormal node; removing the identified abnormal node from the node admittance matrix and the voltage current vector, and recalculating the leakage current vector sum S alcd To identify more potential outlier nodes;
2.3 If more abnormal nodes cannot be identified, the leakage current vector is still not matched with the normal operation characteristics in the table 1, and two adjacent abnormal nodes with the same deviation exist at the moment; by traversing T lcd All the adjacent nodes with the same elements in the network, combining the found adjacent nodes into a node, updating the node admittance matrix, and recalculating the S alcd Identifying remaining outliers;
step 3, abnormal data interpolation, namely correcting abnormal data by utilizing health data of adjacent nodes, specifically comprising the following steps:
3.1 If the abnormal node x is not adjacent to any other pseudo-fault node, the power supply of the node x to the connection box is stopped, and the abnormal data is corrected by using an interpolation equation, specifically:wherein: u (U) re Interpolation voltage for pseudo fault node, R xi Is the resistance between node x and node i;
3.2 After the data interpolation is realized in the step 3.1), if the abnormal node still has the threshold leakage current, the abnormal node is adjacent to the fault point, and the interpolation equation of the data of the single adjacent node is utilized for correction, specifically:wherein: node x is an abnormal node and node z is one of the adjacent healthy nodes;
3.3 When two adjacent abnormal nodes (node x and node y) exist, the correction is performed through interpolation of an equation set, specifically:wherein: u (U) re,x Is the interpolation data of node x;
3.4 After the data interpolation is realized in the step 3.3), the abnormal node still has the threshold leakage current, and the fault point is adjacent to one of the abnormal nodes, so that a correction equation set is further providedWherein: node z is one of healthy neighboring nodes of node y, U z Is the voltage at node z; r is R zi Is the resistance between node z and node i;
step 4, fault location, namely, passing through a locating device of the shore base station according to the updated node leakage current I leak,j ,I leak,k The accurate fault location specifically includes:
4.1 Calculating the fault position according to the relation between the threshold leakage current and the measurement data, specifically: according to the resistance from node j to fault pointGet the fault distance +.>And fault resistanceWherein: r is R jf The resistance from the node j to the fault point; i leak,j Is the leakage current magnitude of node j, - (I) leak,j +I leak,k ) Is the current flowing through the fault resistor;
4.2 Calculating a fault distance set by using the steady-state data after the fault in the time domain, and taking an arithmetic average value as a final result to improve the precision of fault positioning.
3. The method for positioning faults of a power supply system of a submarine observation network according to claim 2, wherein the obtaining of the leakage current vector characteristics of different scenes without considering abnormal data is that:
a) Node j leakage current is when the system is operating normallyWherein: i leak,j Is the leakage current of node j; y is Y ji The cable admittance between node j and node i; g ji Is the cable conductance between node j and node i;
b) When an open circuit fault occurs between node j and node k, the actual conductance between the two nodes will become G' jk At this time, the leakage current of the node j becomesSubstituting leakage current vector to calculate I leak,j And obtain two threshold leakage currents G jk ·(U j -U k )＝I leak,j ＝-I leak,k ≠0；
c) When a short circuit fault occurs, the fault is added between the node j and the node k as a new node; leakage current of node j becomesSubstituting the leakage current to obtain
Leakage current vector characteristics for three scenarios when anomaly data is not considered include: node number 2 with threshold leakage current at open fault, characteristic of threshold leakage currentThe number of nodes having a threshold leakage current at the time of a short-circuit fault is 2, the characteristic of the threshold leakage current is ∈ ->
4. The method for locating faults of a power supply system of a submarine observation network according to claim 2, wherein the identifying abnormal nodes is: when the number of nodes with the pseudo-threshold leakage current in the shore base station nodes is 2, the pseudo-threshold leakage current meets the following conditionsWhen the number of nodes having pseudo-threshold leakage current among the end branch unit nodes is 2, the pseudo-threshold leakage current satisfies +.>When the number of nodes having pseudo-threshold leakage current among the branch unit nodes is 3 or 4, the pseudo-threshold leakage current satisfies +.>Wherein: the leakage current from the branching unit to the junction box is positive; since the errors caused by the affected electrical sensors are unpredictable, the anomaly data may be greater or less than the true data, i.e., the deviation is positive or negative.
5. A system for implementing the method for locating faults in a power supply system of a submarine observation network according to any of claims 1 to 4, comprising: a leakage current vector calculation unit, an abnormal data detection unit, an abnormal data correction unit, and a fault location unit, wherein: the leakage current vector calculation unit constructs a node admittance matrix containing seawater nodes, a voltage vector and a current vector according to system parameters, measured voltages of all nodes, output currents of all bank base stations and information of electrode equivalent impedance, multiplies the node admittance matrix containing seawater nodes by the voltage vector, and finally adds the current vector to obtain a leakage current vector containing the leakage current of each node; the abnormal data detection unit obtains information of a node incidence matrix, a total number vector of adjacent nodes and a leakage current direction vector according to the node admittance matrix and the leakage current vector, multiplies the node incidence matrix by the leakage current direction vector, adds the total number vector of the adjacent nodes to obtain an adjacent threshold value leakage current direction similarity vector, and determines that abnormal data exists in a corresponding node according to an element equal to 0 in the vector; the abnormal data correction unit is used for interpolating the abnormal data by utilizing the health voltage data of the adjacent nodes according to the system parameters and the information of the voltage vectors, replacing the corresponding abnormal data with the interpolated data to update the voltage vectors, and recalculating the leakage current vectors to update the leakage current vectors; the fault locating unit deduces and calculates the fault locating distance and the fault transition resistance by analyzing the relation between the threshold leakage current in the updated leakage current vector and the system parameter and the measured voltage according to the system parameter, the updated leakage current vector and the updated voltage vector.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310260482.3A CN116520073B (en) | 2023-03-17 | 2023-03-17 | Fault positioning method for power supply system of submarine observation network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310260482.3A CN116520073B (en) | 2023-03-17 | 2023-03-17 | Fault positioning method for power supply system of submarine observation network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116520073A true CN116520073A (en) | 2023-08-01 |
| CN116520073B CN116520073B (en) | 2024-01-30 |
Family
ID=87405470
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310260482.3A Active CN116520073B (en) | 2023-03-17 | 2023-03-17 | Fault positioning method for power supply system of submarine observation network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116520073B (en) |
Citations (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2011052890A2 (en) * | 2009-10-30 | 2011-05-05 | 한국전력공사 | System and method for detecting location of a fault in an undersea cable |
| CN102066956A (en) * | 2008-04-14 | 2011-05-18 | 努沃尔特股份有限公司 | Electrical anomaly detection method and system |
| US20140021962A1 (en) * | 2012-07-20 | 2014-01-23 | Viper Subsea Technology Limited | Subsea Deployed Apparatus and Method |
| KR101449706B1 (en) * | 2014-01-24 | 2014-10-15 | 이영돈 | Distributing board for realtime segmented sensing fault, harmonic wave and grounding resistance of power line and having a brocking and recovering function |
| CN105301447A (en) * | 2015-11-10 | 2016-02-03 | 上海交通大学 | Flexible DC power transmission system cable monopolar fault range finding method |
| CN106443360A (en) * | 2016-11-11 | 2017-02-22 | 国网新疆电力公司巴州供电公司 | Incidence matrix-based distribution network traveling wave fault localization method |
| CN106600452A (en) * | 2016-11-11 | 2017-04-26 | 国网新疆电力公司巴州供电公司 | Time analysis matrix and clustering analysis-based power distribution network traveling wave fault location method |
| CN106773993A (en) * | 2017-01-04 | 2017-05-31 | 国网山东省电力公司菏泽供电公司 | Field data acquiring terminal and system |
| CN206223910U (en) * | 2016-11-29 | 2017-06-06 | 国网冀北电力有限公司张家口供电公司 | Overhead transmission line earial drainage is monitored on-line and accurate-location device |
| CN108732464A (en) * | 2018-05-23 | 2018-11-02 | 同济大学 | Submarine observation network extra large cable on-line fault diagnosis and localization method |
| CN110261706A (en) * | 2019-07-22 | 2019-09-20 | 西南交通大学 | A kind of electric transmission line fault detection method based on neighborhood distance |
| CN110940894A (en) * | 2019-12-16 | 2020-03-31 | 国网北京市电力公司 | Electric shock fault detection method based on instantaneous amplitude-phase sum characteristics of leakage current |
| CN111563626A (en) * | 2020-05-09 | 2020-08-21 | 山东大学 | Power system prediction auxiliary state estimation method and system |
| CN112799898A (en) * | 2021-01-08 | 2021-05-14 | 北京科技大学 | Interconnection system fault node positioning method and system based on distributed fault detection |
| CN113295967A (en) * | 2021-06-24 | 2021-08-24 | 长沙理工大学 | Cable insulation state monitoring method under multi-characteristic quantity synergistic effect |
| CN113672658A (en) * | 2021-07-06 | 2021-11-19 | 国网宁夏电力有限公司电力科学研究院 | Power equipment online monitoring error data identification method based on multiple correlation coefficients |
| CN114184232A (en) * | 2021-12-10 | 2022-03-15 | 重庆大学 | Multi-parameter integrated monitoring system for power transmission line |
| CN114755530A (en) * | 2022-04-20 | 2022-07-15 | 西南交通大学 | Robust fault positioning method for power transmission line |
| CN115656687A (en) * | 2022-11-04 | 2023-01-31 | 烟台哈尔滨工程大学研究院 | Method for identifying and positioning underwater cable moisture fault |
-
2023
- 2023-03-17 CN CN202310260482.3A patent/CN116520073B/en active Active
Patent Citations (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102066956A (en) * | 2008-04-14 | 2011-05-18 | 努沃尔特股份有限公司 | Electrical anomaly detection method and system |
| WO2011052890A2 (en) * | 2009-10-30 | 2011-05-05 | 한국전력공사 | System and method for detecting location of a fault in an undersea cable |
| US20140021962A1 (en) * | 2012-07-20 | 2014-01-23 | Viper Subsea Technology Limited | Subsea Deployed Apparatus and Method |
| KR101449706B1 (en) * | 2014-01-24 | 2014-10-15 | 이영돈 | Distributing board for realtime segmented sensing fault, harmonic wave and grounding resistance of power line and having a brocking and recovering function |
| CN105301447A (en) * | 2015-11-10 | 2016-02-03 | 上海交通大学 | Flexible DC power transmission system cable monopolar fault range finding method |
| CN106443360A (en) * | 2016-11-11 | 2017-02-22 | 国网新疆电力公司巴州供电公司 | Incidence matrix-based distribution network traveling wave fault localization method |
| CN106600452A (en) * | 2016-11-11 | 2017-04-26 | 国网新疆电力公司巴州供电公司 | Time analysis matrix and clustering analysis-based power distribution network traveling wave fault location method |
| CN206223910U (en) * | 2016-11-29 | 2017-06-06 | 国网冀北电力有限公司张家口供电公司 | Overhead transmission line earial drainage is monitored on-line and accurate-location device |
| CN106773993A (en) * | 2017-01-04 | 2017-05-31 | 国网山东省电力公司菏泽供电公司 | Field data acquiring terminal and system |
| CN108732464A (en) * | 2018-05-23 | 2018-11-02 | 同济大学 | Submarine observation network extra large cable on-line fault diagnosis and localization method |
| CN110261706A (en) * | 2019-07-22 | 2019-09-20 | 西南交通大学 | A kind of electric transmission line fault detection method based on neighborhood distance |
| CN110940894A (en) * | 2019-12-16 | 2020-03-31 | 国网北京市电力公司 | Electric shock fault detection method based on instantaneous amplitude-phase sum characteristics of leakage current |
| CN111563626A (en) * | 2020-05-09 | 2020-08-21 | 山东大学 | Power system prediction auxiliary state estimation method and system |
| CN112799898A (en) * | 2021-01-08 | 2021-05-14 | 北京科技大学 | Interconnection system fault node positioning method and system based on distributed fault detection |
| CN113295967A (en) * | 2021-06-24 | 2021-08-24 | 长沙理工大学 | Cable insulation state monitoring method under multi-characteristic quantity synergistic effect |
| CN113672658A (en) * | 2021-07-06 | 2021-11-19 | 国网宁夏电力有限公司电力科学研究院 | Power equipment online monitoring error data identification method based on multiple correlation coefficients |
| CN114184232A (en) * | 2021-12-10 | 2022-03-15 | 重庆大学 | Multi-parameter integrated monitoring system for power transmission line |
| CN114755530A (en) * | 2022-04-20 | 2022-07-15 | 西南交通大学 | Robust fault positioning method for power transmission line |
| CN115656687A (en) * | 2022-11-04 | 2023-01-31 | 烟台哈尔滨工程大学研究院 | Method for identifying and positioning underwater cable moisture fault |
Non-Patent Citations (5)
| Title |
|---|
| AHMAD SALEHI DOBAKHSHARI 等: "A Wide-Area Scheme for Power System Fault Location Incorporating Bad Data Detection", IEEE TRANSACTIONS ON POWER DELIVERY, vol. 30, no. 2, XP011576531, DOI: 10.1109/TPWRD.2014.2352336 * |
| WEI LIU 等: "Research on Insulation Detection and Fault Location System of Ship Power System", 2021 IEEE INTERNATIONAL CONFERENCE ON ADVANCES IN ELECTRICAL ENGINEERING AND COMPUTER APPLICATIONS (AEECA) * |
| 于伟经;李智刚;孙凯;秦宝成;李默竹;冯迎宾;: "海底观测网电能管理控制系统研究", 机械设计与制造, no. 05 * |
| 张希鹏 等: "EMTR理论在电力系统线路故障测距中的应用基础Ⅰ:理论部分", 电网技术, vol. 44, no. 3 * |
| 耿坤 等: "基于PCA/LSTM的海底观测网电力系统供电海缆故障定位", 海洋技术学报, vol. 39, no. 3 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116520073B (en) | 2024-01-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wang et al. | Online tracking of transmission-line parameters using SCADA data | |
| Syahputra | A neuro-fuzzy approach for the fault location estimation of unsynchronized two-terminal transmission lines | |
| CN109142986B (en) | Fault positioning method for alternating current distribution line | |
| Lee et al. | A synchrophasor-based fault location method for three-terminal hybrid transmission lines with one off-service line branch | |
| Usman et al. | Fault classification and location identification in a smart distribution network using ANN | |
| Jamali et al. | A Fast and accurate fault location method for distribution networks with DG using genetic algorithms | |
| Zhang et al. | A fault locating method for multi-branch hybrid transmission lines in wind farm based on redundancy parameter estimation | |
| CN115128395A (en) | Voltage sag source positioning method and device | |
| CN116520073B (en) | Fault positioning method for power supply system of submarine observation network | |
| Ghaedi et al. | Modified WLS three-phase state estimation formulation for fault analysis considering measurement and parameter errors | |
| KR101988670B1 (en) | Device for state estimation of power distribution system | |
| Shiroei et al. | A new algorithm for fault location on transmission lines | |
| Mokhlis et al. | Fault location estimation for distribution system using simulated voltage sags data | |
| Sun et al. | Distribution fault location using graph neural network with both node and link attributes | |
| CN112710924B (en) | Single-phase earth fault line selection method based on steady-state variable quantity before and after fault | |
| CN105977932B (en) | A kind of online computing system of power distribution network locality protection associated domain and its method | |
| CN110658419B (en) | Micro-grid fault positioning method based on incomplete information | |
| CN114371364A (en) | Short-circuit fault judgment method based on load end positive sequence voltage amplitude variation | |
| Mokhlis et al. | Locating fault using voltage sags profile for underground distribution system | |
| Asbery et al. | Electric transmission system fault identification using artificial neural networks | |
| Huang et al. | A Leakage-current Vectors Fault Location Method for Undersea Observation Networks with Poor Data Quality | |
| CN112162164A (en) | Cable life prediction system based on neural network | |
| Kumar et al. | Fault prediction in transmission line using GOOSE protocol and Artificial Neural Network | |
| Pérez-Pinacho et al. | Identification of electric transmission line deterioration by an adaptive observer | |
| Jain et al. | Soft computing based fault detection & classification for transmission lines |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |