CN121164829A - Method and System for Online Monitoring and Fault Source Location Diagnosis of Sheath Grounding Circulation Current - Google Patents
Method and System for Online Monitoring and Fault Source Location Diagnosis of Sheath Grounding Circulation CurrentInfo
- Publication number
- CN121164829A CN121164829A CN202511706181.4A CN202511706181A CN121164829A CN 121164829 A CN121164829 A CN 121164829A CN 202511706181 A CN202511706181 A CN 202511706181A CN 121164829 A CN121164829 A CN 121164829A
- Authority
- CN
- China
- Prior art keywords
- sheath
- energy
- phase
- time
- gradient
- 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
Landscapes
- Supply And Distribution Of Alternating Current (AREA)
- Locating Faults (AREA)
Abstract
本发明提供一种护层接地环流在线监测与故障源定位诊断方法及系统，涉及电力系统故障诊断技术领域，包括通过获取多物理场时域信号并提取特征，构建护层电气连通性拓扑图并进行分区，建立能量耗散时空场分布识别疑似故障区域，分析电压‑环流响应关系，最终通过李雅普诺夫指数判断故障类型并定位故障源。本发明能够准确监测电缆护层接地环流状态，实现故障源的精确定位与类型识别，提高电缆运行可靠性。
This invention provides a method and system for online monitoring and fault source location diagnosis of cable sheath grounding current, relating to the field of power system fault diagnosis technology. The method includes acquiring multi-physics field time-domain signals and extracting features, constructing and partitioning a topological map of the sheath's electrical connectivity, establishing a spatiotemporal distribution of energy dissipation to identify suspected fault areas, analyzing the voltage-circulating current response relationship, and finally determining the fault type and locating the fault source using the Lyapunov exponent. This invention can accurately monitor the state of cable sheath grounding current, achieve precise fault source location and type identification, and improve cable operational reliability.
Description
Technical Field
The invention relates to the technical field of fault diagnosis of power systems, in particular to a method and a system for online monitoring and fault source positioning diagnosis of sheath grounding circular flow.
Background
The power cable is used as important power transmission equipment in a power system, and the safe and stable operation of the power cable is important for guaranteeing the power supply. The sheath ground circulation refers to the current generated in the closed loop formed by the sheath of cable, the abnormal change of which is often a precursor to cable failure. With the scale expansion and complexity improvement of the power grid, the protection layer grounding annular flow monitoring and fault source positioning technology becomes an important means for guaranteeing the safe operation of the cable.
At present, the cable sheath grounding loop monitoring technology mainly comprises an off-line detection mode and an on-line monitoring mode. The online monitoring realizes real-time data acquisition, but still faces a plurality of challenges in the aspects of accurate fault source positioning and fault type identification.
The prior art has the defects and shortcomings that the traditional sheath loop current monitoring technology is based on single physical quantity, lacks multi-physical field coupling analysis capability, cannot effectively capture multi-scene relevance of loop current abnormality, cable temperature change, voltage disturbance and the like, and is low in fault diagnosis accuracy, the traditional fault positioning method generally adopts a simplified model, the cable sheath is regarded as a homogeneous conductor, and non-uniformity degradation possibly occurring in the long-term operation process of the cable sheath is ignored, so that fault source positioning accuracy is limited, especially for long-distance cable systems with complex terrains and multiple grounding points, the current fault type identification is based on empirical rules or static parameter threshold judgment, lacks a nonlinear analysis framework supported by dynamic system theory, is difficult to distinguish transient fluctuation and continuous fault development trend of sheath loop current abnormality, and is easy to cause misjudgment or missed judgment, and timeliness and pertinence of fault processing are affected.
Disclosure of Invention
The embodiment of the invention provides a method and a system for online monitoring and fault source positioning diagnosis of sheath grounding circular flow, which can solve the problems in the prior art.
In a first aspect of the embodiments of the present invention, a method for online monitoring and fault source positioning diagnosis of a sheath grounding loop is provided, including:
Acquiring multi-physical field time domain signals of the cable sheath at a plurality of grounding points, and extracting a circulation amplitude characteristic, a circulation phase characteristic, a temperature gradient evolution characteristic and a voltage disturbance characteristic;
constructing a sheath electrical connectivity topological graph based on a phase gradient tensor of a loop phase characteristic and a transfer attenuation factor of a loop amplitude characteristic, performing spectrum decomposition on the sheath electrical connectivity topological graph, screening effective eigenmodes, and dividing the sheath into a plurality of electrical homogeneous sections based on a space gradient abrupt change characteristic of the effective eigenmodes;
Performing product accumulation integration on the temperature gradient evolution characteristic sequence and the circulating amplitude characteristic sequence to construct energy dissipation space-time field distribution, calculating an energy concentration degree through an energy duty entropy value, and calibrating a space-time region in which the energy concentration degree exceeds a preset entropy value threshold and the spatial gradient is in a converging mode as a suspected fault source region;
Aiming at the suspected fault source region, analyzing the association response relation between the voltage disturbance characteristic and the loop amplitude characteristic to obtain a voltage-loop dynamic response curved surface;
and carrying out phase space reconstruction on the voltage-loop dynamic response curved surface to calculate the Lyapunov index, judging the fault type, and binding and positioning the fault source position by combining the electromagnetic field boundary condition of the fault type.
In an alternative embodiment, constructing the sheath electrical connectivity topology map based on the phase gradient tensor of the loop phase characteristic and the transfer attenuation factor of the loop amplitude characteristic includes:
Performing space-time coupled differential operation on the circulation phase characteristics, calculating phase increment vectors between adjacent grounding points, performing tensor decomposition according to the axial component, the circumferential component and the time evolution component of the sheath, and constructing a phase gradient tensor field;
Eigenvalue decomposition is carried out on the phase gradient tensor field, a main eigenvector and a secondary eigenvector of each grounding point are extracted, and the ratio is calculated to determine the phase propagation anisotropism degree of each grounding point;
carrying out multipath propagation tracking on the circulation amplitude characteristics, calculating the accumulation attenuation of the circulation amplitude, and carrying out weighted fusion normalization on the accumulation attenuation of different propagation paths to obtain a transmission attenuation factor;
Establishing a connectivity graph structure taking grounding points as nodes, marking the grounding points with the phase propagation anisotropy degree larger than a preset judging threshold value as anisotropic nodes, respectively establishing strong connecting edges and weak connecting edges based on a main eigenvector and a secondary eigenvector, establishing conventional connecting edges for other grounding points according to the gradient direction of a phase gradient tensor field, and assigning transfer attenuation factors to the corresponding connecting edges to form differential edge weights to obtain a weighted directed graph;
and calculating the relative error of the sum of the inflow edge weights and the sum of the outflow edge weights of each node in the weighted directed graph, and when the relative error exceeds a preset flow conservation threshold value, reversely tracing and identifying an abnormal propagation path along a strong connecting edge, removing the abnormal edge and the associated node, and generating the sheath electric connectivity topological graph.
In an alternative embodiment, spectrally decomposing the sheath electrical connectivity topology map, screening for effective eigenmodes, and dividing the sheath into a plurality of electrical homogeneity segments based on spatial gradient abrupt characteristics of the effective eigenmodes comprises:
expressing the sheath electric connectivity topological graph as a graph Laplace matrix, and performing spectrum decomposition operation to obtain a plurality of eigenvectors ordered according to eigenvalue sizes;
screening eigenvectors with eigenvalues in a preset spectrum energy concentration interval from the eigenvectors as effective eigenvalues, calculating the amplitude distribution of each effective eigenvalue at each grounding point position of the protective layer, and carrying out weighted superposition on the effective eigenvalue amplitude of each grounding point position to construct the response space field of the electrical modal of the protective layer;
Carrying out spatial gradient calculation on the sheath electric mode response space field, extracting gradient change rate of the sheath electric mode response space field along the extension direction of the sheath, and marking the extreme point position of the gradient change rate as a section boundary candidate point;
And carrying out electric continuity verification on each section boundary candidate point, calculating the amplitude difference degree of effective eigenmodes at two sides of the section boundary candidate points, and when the amplitude difference degree is larger than a preset mode mutation threshold value, confirming that the section boundary candidate points are section boundary points, and dividing the protective layer into a plurality of electric homogeneous sections according to the protective layer range between adjacent section boundary points.
In an alternative embodiment, integrating the temperature gradient evolution feature sequence with the circulating amplitude feature sequence to build up the energy dissipation spatiotemporal field distribution comprises:
Temperature acquisition is carried out on the grounding points in each electrical homogeneous section, the temperature change rate of each grounding point in a continuous time window is calculated, and the temperature change rate is subjected to space difference operation along the axial direction of the protective layer to obtain a temperature gradient evolution characteristic sequence of each electrical homogeneous section;
Extracting circulation amplitude measurement data of each grounding point in each electrical homogeneous section, and performing time sequence arrangement according to a time window to obtain a circulation amplitude characteristic sequence of each electrical homogeneous section;
performing time alignment on the temperature gradient evolution characteristic sequence and the circulating amplitude characteristic sequence, calculating the product of the temperature gradient evolution characteristic sequence and the circulating amplitude characteristic sequence in each time window, and performing accumulated integration along the time dimension to obtain the local energy dissipation density of each electric homogeneity section in each time window;
And carrying out three-dimensional mapping on the local energy dissipation density of each electrical homogeneity section in each time window according to the space coordinates of the grounding point and the time sequence coordinates of the time window, and constructing a three-dimensional tensor structure on the axial space dimension of the sheath, the circumferential space dimension of the sheath and the time evolution dimension to form energy dissipation space-time field distribution.
In an alternative embodiment, calculating the energy concentration through the energy duty entropy value, and calibrating the space-time area with the energy concentration exceeding the preset entropy threshold and the spatial gradient in the convergence mode as the suspected fault source area includes:
extracting local energy dissipation density of each electric homogeneous section in each time window, and calculating the energy ratio of the local energy dissipation density to the total energy dissipation density of all electric homogeneous sections;
the energy duty ratio is used as probability weight to carry out entropy calculation, global energy dispersity of each time window is obtained, and the energy concentration degree is determined through the difference value between the maximum dispersity and the global energy dispersity;
extracting a time evolution sequence of the energy concentration degree, and identifying a time window in which the energy concentration degree is mutated;
Calculating the spatial gradient of the energy concentration degree between each electrical homogeneity section and the adjacent electrical homogeneity section, and calibrating the electrical homogeneity section of the spatial gradient presenting aggregation increasing mode and the corresponding time window as a suspected fault source region, wherein the energy concentration degree exceeds a preset entropy threshold value.
In an alternative embodiment, performing phase space reconstruction calculation on the voltage-loop dynamic response curved surface to obtain Lyapunov exponent, judging the fault type, and combining the electromagnetic field boundary condition constraint of the fault type to locate the fault source position comprises:
extracting sampling point data of the voltage-loop dynamic response curved surface on a time sequence, and selecting an embedding dimension and a time delay parameter to reconstruct the sampling point data in a phase space so as to obtain a phase space track;
Calculating the maximum Lyapunov index of the phase space track to obtain a numerical variation curve, determining a stability characterization vector through zero residence time statistics and piecewise fitting slope analysis, and judging the fault type according to the directed distance from the stability characterization vector to the feature space judgment hyperplane;
And searching a preset electromagnetic field boundary condition constraint library according to the fault type, acquiring corresponding electromagnetic field boundary condition constraints, screening energy density extreme points meeting the electromagnetic field boundary condition constraints in energy dissipation space-time field distribution, and calibrating space position coordinates corresponding to the energy density extreme points as fault source positions.
In an alternative embodiment, calculating the maximum lyapunov exponent of the phase space trajectory to obtain a numerical variation curve, determining a stability characterization vector by zero residence time statistics and piecewise fitting slope analysis, and determining the fault type according to the directed distance from the stability characterization vector to the feature space discrimination hyperplane includes:
calculating the maximum Lyapunov exponent of the phase space track to obtain a numerical variation curve of the maximum Lyapunov exponent on a time sequence;
Performing zero detection on the numerical variation curve, counting residence time distribution on two sides of the zero point, extracting statistical moments of the residence time distribution, determining a symbol stability metric, performing piecewise linear fitting on the numerical variation curve, extracting slope symbol sequences and slope amplitude sequences of all fitting segments, calculating the conversion entropy of the slope symbol sequences and the energy concentration degree of the slope amplitude sequences to determine a time domain evolution metric, and determining a stability characterization vector based on the symbol stability metric and the time domain evolution metric;
Projecting the stability characterization vector in a feature space, presetting a discrimination hyperplane corresponding to each fault type in the feature space, calculating the directional distance from the stability characterization vector to each discrimination hyperplane, determining a fault type area corresponding to the stability characterization vector according to the sign and the amplitude of the directional distance, and determining the fault type.
In a second aspect of the embodiments of the present invention, there is provided a system for online monitoring and fault source location diagnosis of a sheath grounding loop, including:
the first unit is used for acquiring multi-physical-field time domain signals of the cable sheath at a plurality of grounding points and extracting a circulation amplitude characteristic, a circulation phase characteristic, a temperature gradient evolution characteristic and a voltage disturbance characteristic;
The second unit is used for constructing a sheath electrical connectivity topological graph based on a phase gradient tensor of a loop phase characteristic and a transfer attenuation factor of a loop amplitude characteristic, carrying out spectrum decomposition on the sheath electrical connectivity topological graph, screening effective eigenmodes, and dividing the sheath into a plurality of electrical homogeneity sections based on the space gradient abrupt change characteristics of the effective eigenmodes;
the third unit is used for carrying out product accumulation integration on the temperature gradient evolution characteristic sequence and the circulation amplitude characteristic sequence to construct energy dissipation space-time field distribution, calculating an energy concentration degree through an energy duty entropy value, and calibrating a space-time region which exceeds a preset entropy value threshold and has a space gradient in a convergence mode as a suspected fault source region;
A fourth unit, configured to analyze, for the suspected fault source region, an association response relationship between the voltage disturbance characteristic and the circulation amplitude characteristic, to obtain a voltage-circulation dynamic response curved surface;
And the fifth unit is used for carrying out phase space reconstruction calculation on the voltage-loop dynamic response curved surface, judging the fault type, and positioning the fault source position by combining the electromagnetic field boundary condition constraint of the fault type.
In a third aspect of an embodiment of the present invention, there is provided an electronic device including:
a processor;
A memory for storing processor-executable instructions;
wherein the processor is configured to invoke the instructions stored in the memory to perform the method described previously.
In a fourth aspect of embodiments of the present invention, there is provided a computer readable storage medium having stored thereon computer program instructions which, when executed by a processor, implement the method as described above.
According to the embodiment of the invention, the multi-physical-field time domain signals of the cable sheath at a plurality of grounding points are obtained, various characteristics are extracted, a sheath electrical connectivity topological graph is constructed by combining the circulation phase characteristics and the amplitude characteristics, the accurate division of the sheath electrical homogeneity sections is realized, the accuracy of fault area identification is improved, the energy dissipation space-time field distribution is constructed based on the product accumulation integration of the temperature gradient evolution characteristics and the circulation amplitude characteristics, the suspected fault source area can be effectively calibrated through the entropy analysis of the energy concentration degree, the problems of misjudgment and missed judgment in the traditional method are avoided, the Lyapunov index is calculated by analyzing the association response relation between the voltage disturbance characteristics and the circulation amplitude characteristics, the fault type judgment is realized, the fault source accurate positioning is carried out by combining the constraint of the electromagnetic field boundary conditions, the efficiency and the accuracy of the cable sheath fault diagnosis are improved, and the reliable guarantee is provided for the safe operation of the power system.
Drawings
FIG. 1 is a schematic flow chart of a method for online monitoring and fault source positioning diagnosis of a sheath grounding loop according to an embodiment of the present invention;
FIG. 2 is a flow chart for fault diagnosis and localization based on nonlinear dynamics.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are only some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The technical scheme of the invention is described in detail below by specific examples. The following embodiments may be combined with each other, and some embodiments may not be repeated for the same or similar concepts or processes.
Fig. 1 is a flow chart of a method for online monitoring and fault source positioning diagnosis of a sheath grounding loop according to an embodiment of the present invention, as shown in fig. 1, the method includes:
Acquiring multi-physical field time domain signals of the cable sheath at a plurality of grounding points, and extracting a circulation amplitude characteristic, a circulation phase characteristic, a temperature gradient evolution characteristic and a voltage disturbance characteristic;
constructing a sheath electrical connectivity topological graph based on a phase gradient tensor of a loop phase characteristic and a transfer attenuation factor of a loop amplitude characteristic, performing spectrum decomposition on the sheath electrical connectivity topological graph, screening effective eigenmodes, and dividing the sheath into a plurality of electrical homogeneous sections based on a space gradient abrupt change characteristic of the effective eigenmodes;
Performing product accumulation integration on the temperature gradient evolution characteristic sequence and the circulating amplitude characteristic sequence to construct energy dissipation space-time field distribution, calculating an energy concentration degree through an energy duty entropy value, and calibrating a space-time region in which the energy concentration degree exceeds a preset entropy value threshold and the spatial gradient is in a converging mode as a suspected fault source region;
Aiming at the suspected fault source region, analyzing the association response relation between the voltage disturbance characteristic and the loop amplitude characteristic to obtain a voltage-loop dynamic response curved surface;
and carrying out phase space reconstruction on the voltage-loop dynamic response curved surface to calculate the Lyapunov index, judging the fault type, and binding and positioning the fault source position by combining the electromagnetic field boundary condition of the fault type.
In an alternative embodiment, constructing the sheath electrical connectivity topology map based on the phase gradient tensor of the loop phase characteristic and the transfer attenuation factor of the loop amplitude characteristic includes:
Performing space-time coupled differential operation on the circulation phase characteristics, calculating phase increment vectors between adjacent grounding points, performing tensor decomposition according to the axial component, the circumferential component and the time evolution component of the sheath, and constructing a phase gradient tensor field;
Eigenvalue decomposition is carried out on the phase gradient tensor field, a main eigenvector and a secondary eigenvector of each grounding point are extracted, and the ratio is calculated to determine the phase propagation anisotropism degree of each grounding point;
carrying out multipath propagation tracking on the circulation amplitude characteristics, calculating the accumulation attenuation of the circulation amplitude, and carrying out weighted fusion normalization on the accumulation attenuation of different propagation paths to obtain a transmission attenuation factor;
Establishing a connectivity graph structure taking grounding points as nodes, marking the grounding points with the phase propagation anisotropy degree larger than a preset judging threshold value as anisotropic nodes, respectively establishing strong connecting edges and weak connecting edges based on a main eigenvector and a secondary eigenvector, establishing conventional connecting edges for other grounding points according to the gradient direction of a phase gradient tensor field, and assigning transfer attenuation factors to the corresponding connecting edges to form differential edge weights to obtain a weighted directed graph;
and calculating the relative error of the sum of the inflow edge weights and the sum of the outflow edge weights of each node in the weighted directed graph, and when the relative error exceeds a preset flow conservation threshold value, reversely tracing and identifying an abnormal propagation path along a strong connecting edge, removing the abnormal edge and the associated node, and generating the sheath electric connectivity topological graph.
In one embodiment, phase increment vectors between adjacent ground points are calculated when performing a space-time coupled differential operation on the loop phase characteristics. Illustratively, the axial phase increment between ground point 1 (0, 0) and ground point 2 (1, 0) is 7.5 degrees, the circumferential phase increment between ground point 1 (0, 0) and ground point 3 (0, 1) is 4.2 degrees, and the phase change rate at adjacent time points is 0.6 degrees/hour. And carrying out tensor decomposition on the increments according to the axial component, the circumferential component and the time evolution component of the protective layer to construct a phase gradient tensor field. For example, the phase gradient tensor at the ground point 1 is [ [7.5,1.2], [1.2,4.2], [0.6,0.3] ], which indicates the rate of change of the phase of the point in each direction.
When eigenvalue decomposition is performed on the phase gradient tensor field, the principal eigenvector and the secondary eigenvector of each grounding point are extracted. At ground point 1, the calculated principal eigenvalue is 8.2, the corresponding principal eigenvector is [0.82,0.57] representing the direction in which the phase change is most pronounced, the minor eigenvalue is 3.5, and the corresponding minor eigenvector is [ -0.57,0.82]. Calculating the ratio of the principal eigenvalue to the secondary eigenvalue gives a degree of phase propagation anisotropy of 2.34 for ground point 1. And when the anisotropy is greater than a preset discrimination threshold value of 1.8, marking the grounding point as an anisotropic node.
When the circulating current amplitude characteristics are subjected to multipath propagation tracking, a propagation model is built based on the sheath structure. On the propagation path from the ground point 1 to the ground point 2, the circulation amplitude was attenuated from 100 amperes to 85 amperes, and the propagation attenuation was calculated to be 0.15. While looking at the other paths from ground point 1 to ground point 2, such as paths through ground point 4to ground point 2, the cumulative attenuation is 0.22. The cumulative attenuation of these paths is weighted according to the physical feasibility of the paths, wherein the direct path weight is 0.7, the indirect path weight is 0.3, and the attenuation after weighted fusion is 0.17. By the normalization processing, the transmission attenuation factor from the ground point 1 to the ground point 2 is calculated to be 0.83, which represents the proportion of the amplitude reserved when the signal propagates from the ground point 1 to the ground point 2.
When the connectivity graph structure taking the grounding points as nodes is established, each grounding point is taken as one node in the graph. For ground point 1 marked as an anisotropic node, a strong connection edge is established with ground point 2 along the direction of the principal eigenvector, the weight of the edge is 0.83, and a weak connection edge is established with ground point 3 along the direction of the secondary eigenvector, the weight of the edge is 0.76. For a normal ground point 5, which is not marked anisotropic, a normal connection edge with the ground point 6 is established, with an edge weight of 0.79, depending on the gradient direction of the phase gradient tensor field. In this way, a weighted directed graph is constructed that includes strong, weak, and conventional connecting edges.
When each node in the weighted directed graph is subjected to flow conservation test, the sum of the inflow edge weight and the sum of the outflow edge weight of each node are calculated. Taking the grounding point 7 as an example, the sum of the inflow edge weights is 2.35, the sum of the outflow edge weights is 1.98, and the relative error is [ (2.35-1.98)/2.35 ] ×100% =15.7%. When the relative error exceeds the preset flow conservation threshold by 12%, this point is interpreted as having an abnormal propagation. The abnormal propagation path from the ground point 8 to the ground point 7 was identified by tracing back along the strong connecting edge flowing into the ground point 7, and the path edge weight was 0.92. After temporarily removing the abnormal edge and the possibly affected associated node, the flow conservation of the grounding point 7 is recalculated, the relative error is reduced to 8.5%, the threshold requirement is met, and the abnormality is confirmed to originate from the grounding point 8-grounding point 7 path.
The finally generated sheath electrical connectivity topological graph intuitively shows the electrical connection relation among all the grounding points, and comprises 42 nodes and 78 edges, wherein the strong connection edge is 24, the weak connection edge is 31 and the conventional connection edge is 23. The main current propagation path and the abnormal region can be clearly identified through the topological graph, and important references are provided for maintenance of the sheath system.
In an alternative embodiment, spectrally decomposing the sheath electrical connectivity topology map, screening for effective eigenmodes, and dividing the sheath into a plurality of electrical homogeneity segments based on spatial gradient abrupt change characteristics of the effective eigenmodes comprises:
expressing the sheath electric connectivity topological graph as a graph Laplace matrix, and performing spectrum decomposition operation to obtain a plurality of eigenvectors ordered according to eigenvalue sizes;
screening eigenvectors with eigenvalues in a preset spectrum energy concentration interval from the eigenvectors as effective eigenvalues, calculating the amplitude distribution of each effective eigenvalue at each grounding point position of the protective layer, and carrying out weighted superposition on the effective eigenvalue amplitude of each grounding point position to construct the response space field of the electrical modal of the protective layer;
Carrying out spatial gradient calculation on the sheath electric mode response space field, extracting gradient change rate of the sheath electric mode response space field along the extension direction of the sheath, and marking the extreme point position of the gradient change rate as a section boundary candidate point;
And carrying out electric continuity verification on each section boundary candidate point, calculating the amplitude difference degree of effective eigenmodes at two sides of the section boundary candidate points, and when the amplitude difference degree is larger than a preset mode mutation threshold value, confirming that the section boundary candidate points are section boundary points, and dividing the protective layer into a plurality of electric homogeneous sections according to the protective layer range between adjacent section boundary points.
In one embodiment, the electrical connectivity of the cable sheath may be represented as a topology. The topological graph comprises a plurality of nodes which represent grounding points or measuring points on the protective layer, the nodes are connected through edges, and the weight of the edges represents the electric conductivity intensity between the corresponding grounding points. Converting this topology to a graph laplace matrix is the first step of analysis. Specifically, for the sheath topology map of N nodes, an n×n laplace matrix L is constructed, where the diagonal element L (i, i) represents the sum of the connection strengths of the node i and all neighboring nodes, and the non-diagonal element L (i, j) represents the negative weight value of the connection between the node i and the node j.
And carrying out spectrum decomposition on the constructed Laplace matrix L to obtain N eigenvectors and corresponding eigenvalues which are ordered according to the eigenvalue size. In an actual 500 meter long cable sheath case, 25 measurement points are total, and 25 eigenvectors are obtained. The spectral decomposition results show that the first 10 eigenvalues are respectively 0, 0.0326, 0.0871, 0.1574, 0.2408, 0.3329, 0.4312, 0.5325, 0.6339, 0.7332 and the like, and the corresponding eigenvectors respectively represent modal components of different spatial frequencies.
The effective eigenvalues are screened from all eigenvectors, and the energy is mainly concentrated in the range of 0.05 to 0.45 eigenvalues and accounts for 87.6% of the total energy by calculating the spectrum energy distribution. Thus, eigenvectors with corresponding eigenvalues in this interval are chosen as the effective eigenvalues, which in the above case include five eigenvectors with eigenvalues of 0.0871, 0.1574, 0.2408, 0.3329, 0.4312.
For each effective eigenmode, the amplitude distribution at each ground point location of the sheath is calculated. For example, in the 3 rd eigenvector (eigenvalue 0.0871), the magnitudes of the 25 measurement points are respectively -0.0534、-0.0912、-0.1254、-0.1496、-0.1624、-0.1615、-0.1462、-0.1172、-0.0758、-0.0249、0.0312、0.0876、0.1398、0.1835、0.2153、0.2328、0.2347、0.2204、0.1904、0.1464、0.0906、0.0259、-0.0437、-0.1143、-0.1815.
And carrying out weighted superposition on the amplitude values of the effective eigenmodes to construct a sheath electric mode response space field. The weight coefficients are determined from the relative energy contributions of the eigenvalues, in this embodiment the weights of the five effective eigenmodes are 0.2136, 0.3249, 0.2541, 0.1426, 0.0648, respectively. The value of the modal response space field obtained after weighted superposition at 25 measuring points is 0.0124、0.0287、0.0398、0.0456、0.0461、0.0412、0.0317、0.0182、0.0024、-0.0132、-0.0268、-0.0368、-0.0419、-0.0416、-0.0359、-0.0256、-0.0123、0.0028、0.0175、0.0301、0.0388、0.0427、0.0413、0.0348、0.0243.
And carrying out spatial gradient calculation on the obtained sheath electric modal response space field, namely calculating the difference of modal response values between adjacent measuring points. The calculation result of the spatial gradient value in 24 adjacent point pairs is that 0.0163、0.0111、0.0058、0.0005、-0.0049、-0.0095、-0.0135、-0.0158、-0.0156、-0.0136、-0.0100、-0.0051、0.0003、0.0057、0.0103、0.0133、0.0151、0.0147、0.0126、0.0087、0.0039、-0.0014、-0.0065、-0.0105.
The gradient change rate, i.e. the first order difference of the gradient, is further calculated. 23 values of gradient change rate -0.0052、-0.0053、-0.0053、-0.0054、-0.0046、-0.0040、-0.0023、0.0002、0.0020、0.0036、0.0049、0.0054、0.0054、0.0046、0.0030、0.0018、-0.0004、-0.0021、-0.0039、-0.0048、-0.0053、-0.0051、-0.0040.
And determining a section boundary candidate point according to the extreme points of the gradient change rate, wherein the extreme points of the gradient change rate appear between 7 th to 8 th, 13 th to 14 th and 18 th to 19 th measuring points, and the corresponding gradient change rate values are respectively 0.0025, 0.0054 and-0.0039. These locations are marked as sector boundary candidate points.
And carrying out electric continuity verification on the section boundary candidate points. And calculating the amplitude difference degree of the effective eigenmodes at two sides of the candidate point, and taking Euclidean distance as a measure. The amplitude differences at the 7 th to 8 th, 13 th to 14 th and 18 th to 19 th measurement points were 0.0826, 0.1243, 0.0735, respectively. If the mode mutation threshold value is set to be 0.08, the degree of difference between the 7 th to 8 th and 13 th to 14 th measurement points exceeds the threshold value, and the section boundary point is confirmed.
According to the confirmed section boundary points, the 500 m long cable protection layer is divided into three electric homogeneity sections, namely a 1 st section (0-140 m and containing measuring points 1-7), a 2 nd section (140-280 m and containing measuring points 8-13) and a 3 rd section (280-500 m and containing measuring points 14-25). The division result is consistent with the grounding structure of the actual cable sheath and the distribution of external environment interference.
In an alternative embodiment, integrating the temperature gradient evolution feature sequence with the circulating current amplitude feature sequence to build up the energy dissipation spatiotemporal field distribution comprises:
Temperature acquisition is carried out on the grounding points in each electrical homogeneous section, the temperature change rate of each grounding point in a continuous time window is calculated, and the temperature change rate is subjected to space difference operation along the axial direction of the protective layer to obtain a temperature gradient evolution characteristic sequence of each electrical homogeneous section;
Extracting circulation amplitude measurement data of each grounding point in each electrical homogeneous section, and performing time sequence arrangement according to a time window to obtain a circulation amplitude characteristic sequence of each electrical homogeneous section;
performing time alignment on the temperature gradient evolution characteristic sequence and the circulating amplitude characteristic sequence, calculating the product of the temperature gradient evolution characteristic sequence and the circulating amplitude characteristic sequence in each time window, and performing accumulated integration along the time dimension to obtain the local energy dissipation density of each electric homogeneity section in each time window;
And carrying out three-dimensional mapping on the local energy dissipation density of each electrical homogeneity section in each time window according to the space coordinates of the grounding point and the time sequence coordinates of the time window, and constructing a three-dimensional tensor structure on the axial space dimension of the sheath, the circumferential space dimension of the sheath and the time evolution dimension to form energy dissipation space-time field distribution.
In one embodiment, a ground point is selected for temperature acquisition within each electrical homogeneity section of the cable sheath system. The ground points are typically disposed at the cable joint or within a sheath grounding box, with at least 2 ground points disposed within each electrical homogeneity section. The temperature sampling frequency was set to once every 10 minutes.
And after the temperature data acquisition is completed, calculating the temperature change rate of each grounding point in a continuous time window. And selecting 30 minutes as a basic time window, calculating the temperature difference of each grounding point at two adjacent sampling moments, and dividing the temperature difference by the sampling time interval to obtain a temperature change rate value. For example, when the temperature of a certain grounding point is 25.6 ℃ at the time t1, the temperature is 26.4 ℃ at the time t2, and the sampling time interval is 10 minutes, the temperature change rate in the time window is 0.08 ℃ per minute.
And performing space difference operation on the temperature change rate along the axial direction of the protective layer, and calculating the temperature change rate difference between adjacent places. For each electric homogeneity section, arranging all grounding points according to the axial coordinates of the protective layer from small to large, calculating the difference of the temperature change rates between adjacent points, and dividing the difference by the space distance between the two points to obtain a temperature gradient value. For example, when the spatial distance between two adjacent grounding points in a certain section is 50m, the temperature change rate of the grounding point 1 is 0.08 ℃ per minute and the temperature change rate of the grounding point 2 is 0.05 ℃ per minute in the same time window, the temperature gradient of the section in the time window is 0.0006 ℃/(min·m). And obtaining the temperature gradient evolution characteristic sequence of each electric homogeneity section by carrying out the calculation in a plurality of continuous time windows.
And extracting circulation amplitude measurement data of each grounding point in each electrical homogeneity section from the cable sheath monitoring system. The circulation data acquisition equipment can adopt a rogowski coil or a Hall current sensor, the measurement range is 0-100A, the accuracy is not lower than +/-2% of the reading, the sampling frequency is consistent with that of the temperature sensor, and the sampling frequency is set to be once every 10 minutes. And (3) carrying out time sequence arrangement on the acquired circulating current amplitude data according to a time window which is the same as the temperature data to form a circulating current amplitude characteristic sequence of each electric homogeneity section.
In order to ensure the time consistency of the data, the temperature gradient evolution characteristic sequence and the circulation amplitude characteristic sequence are subjected to time alignment treatment. If the sampling time of the two data acquisition devices has deviation, a linear interpolation method is adopted for time alignment, so that the data points used in the calculation process correspond to the same time.
After time alignment is completed, calculating the product of the temperature gradient evolution characteristic sequence and the circulation amplitude characteristic sequence in each time window. For example, if the temperature gradient value of a certain section in a specific time window is 0.0006 ℃/(min·m) and the average value of the circulation amplitude of the section in the same time window is 32.5A, the product of the two is 0.0195 ℃ ·a/(min·m).
And accumulating integration along the time dimension, and accumulating the product value of each time window. For each electrical homogeneity section, starting from the monitoring start time t0, the product value of all time windows up to the current time t is accumulated to obtain the local energy dissipation density of the section at the time t. The integration calculation can adopt a trapezoid integration method, so that the integration precision is improved. For example, the product value of a segment over consecutive 5 time windows is 0.0195, 0.0210, 0.0225, 0.0240, 0.0255 ℃ a/(min·m), respectively, and the time window length is 30 minutes, then the cumulative energy dissipation density of the segment after 150 minutes is 3.375 ℃ a·min/m.
And carrying out three-dimensional mapping on the local energy dissipation density of each electrical homogeneity section in each time window according to the space coordinates of the grounding point and the time sequence coordinates of the time window. The method comprises the steps of establishing a coordinate axis along a cable laying direction by taking a cable starting end as an origin in a sheath axial space dimension, establishing a polar coordinate system by taking the center of a cable section as the origin in a sheath circumferential space dimension, and establishing a time coordinate axis by taking a monitoring starting moment as the origin in a time evolution dimension. And mapping the calculated local energy dissipation density value to a corresponding three-dimensional space point to construct a three-dimensional tensor structure.
To achieve visual representation of the data, thermodynamic diagrams may be used to demonstrate the energy dissipation spatiotemporal field distribution. For example, in a 110kV crosslinked polyethylene cable system, 3 electrical homogeneity sections are monitored, each section being provided with 4 grounding points, for a period of 48 hours. The energy dissipation space-time field distribution obtained by calculation through the method shows that the phenomenon that the energy dissipation density value is increased sharply occurs in the middle position of the 2 nd section after the 36 th hour of monitoring, the highest value reaches 12.56 ℃ A.hour/meter, the highest value is higher than that of the normal area by more than 3 times, and the situation that the protective layer loop is abnormal in the position is indicated. After field inspection, the ground lead at the position is found to have a poor contact problem, and the energy dissipation density value is recovered to a normal level after repair.
By constructing energy dissipation space-time field distribution, the space-time evolution rule of energy conversion and dissipation in the cable sheath system can be intuitively reflected, and an effective basis is provided for early identification and positioning of cable sheath loop faults.
In an alternative embodiment, calculating the energy concentration through the energy duty entropy value, and calibrating the space-time area with the energy concentration exceeding the preset entropy threshold and the spatial gradient in the convergence mode as the suspected fault source area includes:
extracting local energy dissipation density of each electric homogeneous section in each time window, and calculating the energy ratio of the local energy dissipation density to the total energy dissipation density of all electric homogeneous sections;
the energy duty ratio is used as probability weight to carry out entropy calculation, global energy dispersity of each time window is obtained, and the energy concentration degree is determined through the difference value between the maximum dispersity and the global energy dispersity;
extracting a time evolution sequence of the energy concentration degree, and identifying a time window in which the energy concentration degree is mutated;
Calculating the spatial gradient of the energy concentration degree between each electrical homogeneity section and the adjacent electrical homogeneity section, and calibrating the electrical homogeneity section of the spatial gradient presenting aggregation increasing mode and the corresponding time window as a suspected fault source region, wherein the energy concentration degree exceeds a preset entropy threshold value.
In a specific embodiment, the sheath grounding circulation on-line monitoring device collects current and voltage data of each measuring point, after data preprocessing, the continuous time sequence is divided into a plurality of time windows, and each window comprises a fixed number of sampling points. For each time window, the local energy dissipation density of the respective electrical homogeneity section is calculated. The local energy dissipation density represents the energy consumption per unit volume of area per unit time, and is obtained by dividing the product of the section current and the section resistance by the section volume. Assuming that a section measures 3.5 amps, a section resistance of 0.25 ohms, and a section volume of 0.5 cubic meters for a time window, the local energy dissipation density for that section is (3.5 x 0.25)/0.5 = 6.125 watts/cubic meter.
The total energy dissipation density was counted for all electrical homogeneity zones throughout the monitoring system. If there are 10 electrically homogeneous segments, the total energy dissipation density is 125 watts/cubic meter, and the local energy dissipation density for a particular segment is 6.125 watts/cubic meter, then the energy ratio for that segment is 6.125/125=0.049 or 4.9%. This energy duty cycle is used as a probability weight for subsequent entropy calculation.
For each time window, a global energy dispersion is calculated based on the energy duty cycle. The energy dispersion degree reflects the uniformity degree of the energy distribution of the system based on the entropy principle. Maximum dispersion is achieved when the energy duty cycle of all the electrical homogeneity sections is completely equal. For example, if there are 10 segments, the maximum dispersion corresponds to a case where the energy ratio per segment is 10%. And obtaining an energy concentration index by calculating the difference between the maximum dispersity and the actual global energy dispersity. The larger the energy concentration value, the more uneven the energy distribution, with some sections consuming significantly more energy than others.
And extracting the energy concentration value of the continuous time window to form a time evolution sequence of the energy concentration. The sequence is analyzed to identify a time window in which the energy concentration is mutated. The abrupt change judgment is based on the change rate of the energy concentration, and when the change rate of the energy concentration of a certain time window exceeds a preset threshold value, such as 30%, compared with the change rate of the previous window, the abrupt change judgment is marked as an abrupt change point. For example, if the energy concentration in the previous time window is 0.35, the current window is 0.49, the rate of change is (0.49-0.35)/0.35=40%, and the threshold value of 30% is exceeded, the mutation point is determined.
For each time window, an energy concentration spatial gradient between each electrical homogeneity section and the adjacent section is calculated. The spatial gradient represents the rate of change in the spatial concentration of energy reflecting the spatial non-uniformity of the energy distribution. And calculating the energy concentration difference value of the adjacent sections, and dividing the energy concentration difference value by the physical distance between the sections to obtain a spatial gradient value. For example, if the energy concentration of a certain segment is 0.45, the adjacent segment is 0.32, and the distance between two segments is 10 meters, the spatial gradient is (0.45-0.32)/10=0.013/meter.
Analyzing the distribution pattern of the spatial gradient, and identifying an incremental convergence pattern. The phenomenon that the spatial gradient value continuously increases as the incremental convergence pattern moves toward a segment indicates that energy is concentrated toward that segment. For example, moving from the periphery to the center section, the spatial gradient increases from 0.005/meter to 0.008/meter, and then to 0.013/meter, creating an incremental convergence pattern.
A preset entropy threshold for the energy concentration is set, such as 0.4. The energy concentration exceeds the threshold and the electrical homogeneity sections of the spatial gradient rendering convergence increasing pattern are marked as suspected fault sources. And recording time windows corresponding to the sections to form a space-time two-dimensional fault source positioning result.
The fault source positioning result can be presented through a visual interface, and the fault source positioning result comprises various forms such as an energy concentration thermodynamic diagram, a spatial gradient vector diagram and the like, so that operation and maintenance personnel can be helped to quickly position the fault source. In the embodiment, the fault source determined through the energy dissipation analysis is located in the 7 th section, and after the actual overhaul, the cable sheath insulating layer of the section is found to have aging damage, so that the sheath grounding current is increased, and the accuracy of the positioning result is verified.
Compared with the traditional fault positioning method, the method of the embodiment does not need an external excitation signal, and can accurately position the sheath grounding annular fault source by analyzing the inherent energy dissipation characteristic, so that the operation reliability and maintenance efficiency of the power system are improved.
In an alternative embodiment, performing phase space reconstruction calculation on the voltage-loop dynamic response curved surface to obtain Lyapunov exponent, judging the fault type, and locating the fault source position in combination with the electromagnetic field boundary condition constraint of the fault type comprises:
extracting sampling point data of the voltage-loop dynamic response curved surface on a time sequence, and selecting an embedding dimension and a time delay parameter to reconstruct the sampling point data in a phase space so as to obtain a phase space track;
Calculating the maximum Lyapunov index of the phase space track to obtain a numerical variation curve, determining a stability characterization vector through zero residence time statistics and piecewise fitting slope analysis, and judging the fault type according to the directed distance from the stability characterization vector to the feature space judgment hyperplane;
And searching a preset electromagnetic field boundary condition constraint library according to the fault type, acquiring corresponding electromagnetic field boundary condition constraints, screening energy density extreme points meeting the electromagnetic field boundary condition constraints in energy dissipation space-time field distribution, and calibrating space position coordinates corresponding to the energy density extreme points as fault source positions.
In one embodiment, time series sampling point data is extracted from a voltage-loop dynamic response surface. The voltage data collected may be {220.5v,219.8v,221.3v,218.7v, }, the circulation data may be {15.2a,15.5a,14.9a,15.8a, }, the sampling frequency is 10kHz, i.e. one data point is collected every 0.1 ms. And carrying out phase space reconstruction on the sampling point data by selecting proper embedding dimension and time delay parameters. The embedding dimension may be determined by assuming a correlation method, in this embodiment, the embedding dimension is selected to be 3, and the time delay parameter is determined to be 8 sampling point intervals, i.e., 0.8ms, by a mutual information method. For the voltage sequences { V1, V2, V3,.. } a set of phase space points is constructed { (V1, V9, V17), (V2, V10, V18),... The point sets form a phase space track in a three-dimensional space, and reflect the dynamic behavior characteristics of the system.
And calculating the maximum Lyapunov exponent of the phase space track to obtain a numerical variation curve. Specifically, a reference point in a phase space is selected, a nearby point is found in a neighborhood of the reference point, and the divergence condition of the two-point track along with time is calculated. And recording the divergence rate of each time step, and performing linear regression on the data, wherein the slope of the regression line is the maximum Lyapunov exponent. In an example, the index curve in normal operation state is { -0.02, -0.01, -0.03, & gt}, fluctuation is small and mean value is close to zero, the index curve in short-circuit fault state is {1.23,1.45,1.37,1.51, & gt}, the value is large and positive, and the index curve in open-circuit fault state is { -1.05, -1.21, -0.98, -1.15, & gt, the value is large and negative. Through the statistical analysis of the zero residence time, the zero residence proportion in the normal state reaches 85%, the zero residence proportion in the short-circuit fault state is less than 10%, and the zero residence proportion in the open-circuit fault state is less than 15%. The curves are fitted piecewise to obtain a slope sequence for each time window, such as normal state {0.003, -0.005,0.002,. }, short-circuit fault state {0.217,0.198,0.225,. }, open-circuit fault state { -0.175, -0.183, -0.166,. }. Based on the zero residence ratio and the slope distribution characteristics, a stability characterization vector t= [ residence ratio, positive slope mean, negative slope mean ] is constructed. For example, the normal state t1= [0.85,0.004, -0.003], the short-circuit fault t2= [0.08,0.205, -0.025], and the open-circuit fault t3= [0.12,0.018, -0.175].
In the feature space, the hyper plane is trained and judged in advance, and a fault type classifier is constructed by using a support vector machine algorithm. And calculating the directed distance from the newly input stability characterization vector to each discrimination hyperplane, and determining the fault type. For example, a certain measurement results in a stability characterization vector t= [0.09,0.211, -0.022], the distance to the hyperplane of the short-circuit fault type is 0.87, the distance to the hyperplane of the open-circuit fault type is-2.15, the distance to the hyperplane of the normal state is-5.23, and the short-circuit fault is determined.
And searching a preset constraint library of the electromagnetic field boundary conditions according to the judged fault type. For short circuit faults, the boundary condition constraint is that the electric field strength is suddenly changed and the tangential component is continuous, and the magnetic field strength is suddenly changed and the normal component is continuous. And acquiring the distribution data of the energy dissipation space-time field of the whole network, wherein the distribution data comprise the electric field intensity, the magnetic field intensity and the energy density value of each area node. For example, the data of node 1 is { co-ordinate (12.5 m,8.3m,1.2 m), the electric field strength is 500V/m, the magnetic field strength is 1.5A/m, the energy density is 2750J/m 3 }, the data of node 2 is { co-ordinate (12.7 m,8.3m,1.2 m), the electric field strength is 85V/m, the magnetic field strength is 0.8A/m, the energy density is 320J/m 3 }. Between the adjacent nodes 1 and 2, the tangential components of the electric field intensity are 480V/m and 82V/m respectively, the mutation ratio is 5.85, the normal components of the magnetic field intensity are 0.9A/m and 0.85A/m respectively, the ratio is 1.06 and is nearly continuous, and the energy density ratio is 8.59.
Screening the area meeting the constraint of the boundary condition of the electromagnetic field, namely, the tangential component ratio of the electric field intensity is more than 3, the normal component ratio of the magnetic field intensity is close to 1 (in the range of 0.9-1.1), and searching the extreme point of the energy density in the screened area. In the example, 3 extreme points of energy density meeting the constraint of boundary conditions are found, namely { coordinates (12.6 m,8.3m,1.2 m), energy density 4120J/m 3 }, { coordinates (25.3 m,12.7m,0.8 m), energy density 3580J/m 3 }, { coordinates (18.9 m,15.5m, 2.1), and energy density 2980J/m 3 }. The three points are subjected to energy density sorting, and the highest value point (12.6 m,8.3m and 1.2 m) is selected as a fault source position.
In the embodiment, the fault type of the power system is accurately judged, the fault source position is accurately positioned, the automation and the accuracy of fault diagnosis are realized, the fault positioning time is shortened from the hour level of traditional manual inspection to the second level, the positioning accuracy is improved from the meter level to the centimeter level, and the operation and maintenance efficiency and the reliability of the power system are greatly improved.
In an alternative embodiment, calculating the maximum lyapunov exponent of the phase space trajectory to obtain a numerical variation curve, determining a stability characterization vector by zero residence time statistics and piecewise fitting slope analysis, and determining the fault type according to the directed distance from the stability characterization vector to the feature space discrimination hyperplane includes:
calculating the maximum Lyapunov exponent of the phase space track to obtain a numerical variation curve of the maximum Lyapunov exponent on a time sequence;
Performing zero detection on the numerical variation curve, counting residence time distribution on two sides of the zero point, extracting statistical moments of the residence time distribution, determining a symbol stability metric, performing piecewise linear fitting on the numerical variation curve, extracting slope symbol sequences and slope amplitude sequences of all fitting segments, calculating the conversion entropy of the slope symbol sequences and the energy concentration degree of the slope amplitude sequences to determine a time domain evolution metric, and determining a stability characterization vector based on the symbol stability metric and the time domain evolution metric;
Projecting the stability characterization vector in a feature space, presetting a discrimination hyperplane corresponding to each fault type in the feature space, calculating the directional distance from the stability characterization vector to each discrimination hyperplane, determining a fault type area corresponding to the stability characterization vector according to the sign and the amplitude of the directional distance, and determining the fault type.
In a specific embodiment, in the method for online monitoring of sheath grounding loop and fault source positioning diagnosis, the phase space track is processed, the maximum Lyapunov exponent is calculated, and a small data volume method is adopted for calculation, so that the method is suitable for time series with limited length. And selecting the acquired sheath voltage signal as an analysis object, and reconstructing the acquired sheath voltage signal into a phase space track. When reconstructing the phase space, the embedding dimension can be set to 3, the time delay parameter can be determined by a mutual information function method, and the time delay value corresponding to the first local minimum point is generally taken, and in practical application, the time delay value is generally between 5 and 15.
In the calculation process, selecting reference points in a phase space, searching neighbor points closest to each reference point, tracking the evolution of the two points in the phase space, and calculating the change of the distance between the two points along with time. When the initial distance is d 0 and becomes d after the time t, the maximum Lyapunov exponent lambda can be obtained by calculating the average of ln (d/d 0)/t. In the actual calculation, a plurality of reference points can be selected for calculation and then an average value can be obtained so as to increase the reliability of the result.
And after obtaining a numerical curve of the maximum Lyapunov exponent changing along with time, performing zero point detection on the curve. The zero point detection adopts a linear interpolation method, namely, at the curve variation position, the zero point position is determined by carrying out linear interpolation through the values of two adjacent points. And recording the time position of each detected zero point, and counting the time interval between adjacent zero points, namely the residence time. For example, in a test, the detected zero time positions are 52 seconds, 78 seconds, 103 seconds, 129 seconds, respectively, and the corresponding dwell time sequences are 26 seconds, 25 seconds, 26 seconds.
And extracting statistical moment characteristics aiming at the obtained residence time distribution. The mean, variance, skewness, and kurtosis of the dwell times were calculated. The average value reflects the average period of the system state change, the variance reflects the period stability, the skewness reflects the asymmetry of the distribution, and the kurtosis reflects the kurtosis degree of the distribution. In practical application, the residence time average value in the normal running state may be in the range of 25-30 seconds, the variance is smaller, which indicates that the system state transition is more regular, while in the fault state, the residence time average value may be obviously shortened to 10-15 seconds, and the variance is increased, which indicates that the system state transition is accelerated and unstable.
A symbol stability metric is determined based on the statistical characteristics of the dwell time. The symbol stability metric includes a dwell time ratio and a dwell number ratio of the positive and negative regions. The positive region refers to the period when the lyapunov exponent is greater than zero, and the negative region refers to the period when the lyapunov exponent is less than zero. The dwell time ratio is calculated as the total dwell time of the positive region divided by the total dwell time of the negative region, and the dwell number ratio is calculated as the dwell number of the positive region divided by the dwell number of the negative region. Under normal operating conditions, these two ratios are typically close to 1, indicating that the system is balanced between stable and unstable conditions, while under certain fault conditions the ratio may deviate significantly from 1, e.g., an insulation ageing fault may cause the positive and negative area residence time ratio to increase above 2.
And performing piecewise linear fitting on a numerical variation curve of the maximum Lyapunov exponent. The selection of the segmentation points can be based on turning points of the curves, and points with absolute values exceeding a preset threshold value of the second-order differences can be used as potential segmentation points by calculating the second-order differences of the curves. In practical application, the threshold may be set to 3 times the mean value of the second order difference absolute value. For example, for data of 300 seconds in length, 8-12 segmentation points are obtained, dividing the entire curve into 9-13 linear segments.
For each linear fit segment, slope values are extracted to form a slope sequence. The slope sequence is divided into a positive slope sequence and a negative slope sequence according to the sign, and the length distribution and the conversion frequency of the positive slope sequence and the negative slope sequence are counted. Positive slope indicates an increase in system instability and negative slope indicates a decrease in system instability. By calculating the transition entropy of the slope symbol sequence, the complexity of the system state transition can be quantified. The higher the transition entropy value, the more complex the system state change. In some test, the transition entropy of the normal operation state is about 0.8, and the transition entropy of the ground fault state may drop below 0.4, indicating that the system enters a single dominant state.
The energy concentration of the slope magnitude sequence is determined by calculating the power spectral density of the sequence and counting the proportion of energy occupied by the dominant frequency component. A high energy concentration indicates a significant periodicity of system state changes, and a low energy concentration indicates a more random system state change. Under normal operating conditions, the energy concentration is typically between 0.4 and 0.6, while certain fault conditions, such as a sheath shield short circuit fault, may increase to above 0.8.
Based on the symbol stability measurement and the time domain evolution measurement obtained by the calculation, a stability characterization vector is constructed, wherein the stability characterization vector comprises four components of a residence time ratio, a residence times ratio, a conversion entropy and an energy concentration degree. In the feature space, each fault type corresponds to a specific region, and the region boundary can be defined by discriminating the hyperplane. The discrimination hyperplane can be obtained by training the data set by using a support vector machine method.
The method comprises the steps of presetting a plurality of fault types of discrimination hyperplanes, including ground faults, insulation aging faults, short-circuit faults of a protective layer shielding layer and the like. And projecting the calculated stability characterization vector in a feature space, and calculating the directed distance from the vector to each discrimination hyperplane. The directional distance is calculated by subtracting the closest point on the hyperplane from the characterization vector and then performing a dot product with the normal vector of the hyperplane. The sign of the distance indicates on which side of the hyperplane the vector is located and the absolute value of the distance indicates how close the vector is to the hyperplane.
Based on the calculated directed distance, it can be determined which fault type the current state belongs to. The decision rule may be set such that if the directed distance of the vector to the fault hyperplane of a certain type is negative and the absolute value is maximum, then the fault is decided as such. In a certain diagnosis, the calculated directional distance from the stability characterization vector to the ground fault hyperplane is-0.32, the directional distance from the stability characterization vector to the insulation aging fault hyperplane is 0.15, the directional distance from the stability characterization vector to the short circuit fault hyperplane of the protective layer shielding layer is 0.08, and the current fault type is determined to be the ground fault according to a judgment rule.
In the embodiment, the real-time monitoring of the sheath grounding circulation state is realized, the fault source type is accurately positioned, a decision basis is provided for cable system maintenance, and the nonlinear characteristic of the sheath grounding circulation system can be effectively processed by utilizing the chaos dynamics theory, so that the method has higher fault recognition rate and stronger anti-interference capability compared with the traditional method.
The online monitoring and fault source positioning and diagnosing system for the sheath grounding loop comprises the following components:
the first unit is used for acquiring multi-physical-field time domain signals of the cable sheath at a plurality of grounding points and extracting a circulation amplitude characteristic, a circulation phase characteristic, a temperature gradient evolution characteristic and a voltage disturbance characteristic;
The second unit is used for constructing a sheath electrical connectivity topological graph based on a phase gradient tensor of a loop phase characteristic and a transfer attenuation factor of a loop amplitude characteristic, carrying out spectrum decomposition on the sheath electrical connectivity topological graph, screening effective eigenmodes, and dividing the sheath into a plurality of electrical homogeneity sections based on the space gradient abrupt change characteristics of the effective eigenmodes;
the third unit is used for carrying out product accumulation integration on the temperature gradient evolution characteristic sequence and the circulation amplitude characteristic sequence to construct energy dissipation space-time field distribution, calculating an energy concentration degree through an energy duty entropy value, and calibrating a space-time region which exceeds a preset entropy value threshold and has a space gradient in a convergence mode as a suspected fault source region;
A fourth unit, configured to analyze, for the suspected fault source region, an association response relationship between the voltage disturbance characteristic and the circulation amplitude characteristic, to obtain a voltage-circulation dynamic response curved surface;
And the fifth unit is used for carrying out phase space reconstruction calculation on the voltage-loop dynamic response curved surface, judging the fault type, and positioning the fault source position by combining the electromagnetic field boundary condition constraint of the fault type.
In a third aspect of an embodiment of the present invention, there is provided an electronic device including:
a processor;
A memory for storing processor-executable instructions;
wherein the processor is configured to invoke the instructions stored in the memory to perform the method described previously.
In a fourth aspect of embodiments of the present invention, there is provided a computer readable storage medium having stored thereon computer program instructions which, when executed by a processor, implement the method as described above.
The present invention may be a method, apparatus, system, and/or computer program product. The computer program product may include a computer readable storage medium having computer readable program instructions embodied thereon for performing various aspects of the present invention.
It should be noted that the above embodiments are merely for illustrating the technical solution of the present invention and not for limiting the same, and although the present invention has been described in detail with reference to the above embodiments, it should be understood by those skilled in the art that the technical solution described in the above embodiments may be modified or some or all of the technical features may be equivalently replaced, and these modifications or substitutions do not make the essence of the corresponding technical solution deviate from the scope of the technical solution of the embodiments of the present invention.
Claims (10)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202511706181.4A CN121164829B (en) | 2025-11-20 | 2025-11-20 | Method and System for Online Monitoring and Fault Source Location Diagnosis of Sheath Grounding Circulation Current |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202511706181.4A CN121164829B (en) | 2025-11-20 | 2025-11-20 | Method and System for Online Monitoring and Fault Source Location Diagnosis of Sheath Grounding Circulation Current |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN121164829A true CN121164829A (en) | 2025-12-19 |
| CN121164829B CN121164829B (en) | 2026-01-27 |
Family
ID=98047740
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202511706181.4A Active CN121164829B (en) | 2025-11-20 | 2025-11-20 | Method and System for Online Monitoring and Fault Source Location Diagnosis of Sheath Grounding Circulation Current |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN121164829B (en) |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110006761A1 (en) * | 2005-11-07 | 2011-01-13 | Enerize Corporation | Method and eddy current system for non-contact determination of interface resistance |
| US11150291B1 (en) * | 2021-05-14 | 2021-10-19 | Underground Systems, Inc. | Functional reliability assessment for insulated power cable systems |
| CN114035118A (en) * | 2021-11-30 | 2022-02-11 | 徐忠林 | Covering ground fault detection method, location method, detection system and location system |
| EP4257988A1 (en) * | 2022-04-06 | 2023-10-11 | Lumiker Aplicaciones Tecnologicas S.L. | Method for locating a fault point on a high-voltage three-phase ac cable, and system for locating a fault point |
| CN119881542A (en) * | 2025-03-25 | 2025-04-25 | 内蒙古电力(集团)有限责任公司包头供电分公司 | Intelligent positioning and diagnosing method for optical fiber composite faults of power distribution network based on hybrid sensing |
| WO2025087985A1 (en) * | 2023-10-23 | 2025-05-01 | Puls8 Ltd | Method and system for identifying a cable |
| CN120233186A (en) * | 2025-05-19 | 2025-07-01 | 曲阜市虹飞电缆有限公司 | Intelligent cable fault monitoring and early warning method and system |
| CN120632543A (en) * | 2025-08-13 | 2025-09-12 | 国网上海市电力公司 | In-loop simulation method and system for cable sheath circulating current fault location based on LABVIEW |
| WO2025213778A1 (en) * | 2024-04-08 | 2025-10-16 | 国网四川省电力公司成都供电公司 | Method and system for testing sheath impedance in high-voltage cable cross-bonding system |
| CN120910667A (en) * | 2025-10-09 | 2025-11-07 | 艾斯安热能科技(太仓)有限公司 | Fault diagnosis and early warning method and system for double bypass valve of industrial cooler |
-
2025
- 2025-11-20 CN CN202511706181.4A patent/CN121164829B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110006761A1 (en) * | 2005-11-07 | 2011-01-13 | Enerize Corporation | Method and eddy current system for non-contact determination of interface resistance |
| US11150291B1 (en) * | 2021-05-14 | 2021-10-19 | Underground Systems, Inc. | Functional reliability assessment for insulated power cable systems |
| CN114035118A (en) * | 2021-11-30 | 2022-02-11 | 徐忠林 | Covering ground fault detection method, location method, detection system and location system |
| EP4257988A1 (en) * | 2022-04-06 | 2023-10-11 | Lumiker Aplicaciones Tecnologicas S.L. | Method for locating a fault point on a high-voltage three-phase ac cable, and system for locating a fault point |
| WO2025087985A1 (en) * | 2023-10-23 | 2025-05-01 | Puls8 Ltd | Method and system for identifying a cable |
| WO2025213778A1 (en) * | 2024-04-08 | 2025-10-16 | 国网四川省电力公司成都供电公司 | Method and system for testing sheath impedance in high-voltage cable cross-bonding system |
| CN119881542A (en) * | 2025-03-25 | 2025-04-25 | 内蒙古电力(集团)有限责任公司包头供电分公司 | Intelligent positioning and diagnosing method for optical fiber composite faults of power distribution network based on hybrid sensing |
| CN120233186A (en) * | 2025-05-19 | 2025-07-01 | 曲阜市虹飞电缆有限公司 | Intelligent cable fault monitoring and early warning method and system |
| CN120632543A (en) * | 2025-08-13 | 2025-09-12 | 国网上海市电力公司 | In-loop simulation method and system for cable sheath circulating current fault location based on LABVIEW |
| CN120910667A (en) * | 2025-10-09 | 2025-11-07 | 艾斯安热能科技(太仓)有限公司 | Fault diagnosis and early warning method and system for double bypass valve of industrial cooler |
Also Published As
| Publication number | Publication date |
|---|---|
| CN121164829B (en) | 2026-01-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Jamali et al. | A new fault location method for distribution networks using sparse measurements | |
| CN102221651B (en) | Fault on-line diagnosis and early warning method of flameproof dry-type transformer for mine | |
| Li et al. | Damage localization of stacker’s track based on EEMD-EMD and DBSCAN cluster algorithms | |
| CN106249101A (en) | A kind of intelligent distribution network fault identification method | |
| CN111856209A (en) | Power transmission line fault classification method and device | |
| CN120103062B (en) | Abnormal location method and system for distribution station area based on multi-modal dynamic feature analysis | |
| CN110414412A (en) | Method and device for accurate identification of multiple disturbances in wide area power grid based on big data analysis | |
| CN117849691B (en) | A multi-dimensional coordinated operation monitoring and early warning system and method for capacitive voltage transformer | |
| Jamali et al. | A fast and accurate fault location method for distribution networks with dg using genetic algorithms | |
| Wu et al. | Quantifying harmonic responsibilities based on kurtosis detection principle of amplitude fluctuations | |
| CN120929891B (en) | A photovoltaic power grid fault identification method and system based on circuit analysis | |
| CN111352365B (en) | A dust-proof and ventilated electrical equipment cabinet and its control method | |
| Mahmoud et al. | Review of fuzzy and ANN fault location methods for distribution power system in oil and gas sectors | |
| CN121164829B (en) | Method and System for Online Monitoring and Fault Source Location Diagnosis of Sheath Grounding Circulation Current | |
| CN120222635B (en) | Self-adaptive control system based on optical measurement 10KV outdoor transformer | |
| CN106646106B (en) | Grid fault detection method based on change point detection technology | |
| CN105552887A (en) | Branch active power flow-based power network topology error identification method | |
| Tian et al. | Intelligent Detection Method of Power Grid Fault Big Data Based on Deep Learning | |
| Yan et al. | Application of random matrix model in multiple abnormal sources detection and location based on PMU monitoring data in distribution network | |
| CN121186528B (en) | Power transmission line sheath grounding loop fault diagnosis method and system | |
| CN115343579A (en) | Power grid fault analysis method and device and electronic equipment | |
| Wang et al. | A Fast Fault Location Method for Distribution Network Based on Line Topology and Beat Frequency Attenuation Discrimination | |
| Han et al. | Data repair and anomaly detection of power meter based on deep reinforcement learning | |
| CN117289072B (en) | A method for identifying bad power grid data based on voltage coding | |
| Wu et al. | Online State Monitoring of Distribution Cable Faults Based Empirical Mode Decomposition with Convolutional Neural Network and Spiking Neural Network |
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 |