Fault line selection method based on transient zero-sequence current state trajectory analysis

Abstract

The embodiment of the invention discloses a fault line selection method based on transient zero-sequence current state trajectory analysis, relates to the technical field of power system protection, and aims to solve the problem that high-resistance grounding fault line selection of a resonant grounding system is low in reliability due to dependence on signal values. The method comprises the following steps: acquiring a zero-sequence current time sequence of a plurality of feeder lines; performing state space reconstruction on the sequence of each line to generate a multi-dimensional state track; calculating a normalized time-varying form descriptor representing the geometrical form evolution trend of the trajectory; randomly selecting a reference line, and calculating a cumulative difference value between the descriptors of the non-reference line and the reference line; and a decision rule is applied to identify a fault line, and the rule can distinguish a non-reference line fault, a reference line fault and a bus fault. According to the invention, line selection is converted from numerical value comparison to geometrical morphology recognition, and the criterion does not depend on the absolute value of a signal, so that immunity to noise and measurement errors is realized, and the sensitivity and reliability of line selection are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of power system protection technology, and in particular to a fault location method based on transient zero-sequence current state trajectory analysis. Background Technology

[0002] 35kV systems often employ resonant grounding, which involves compensation via arc-suppression coils. While this compensation mechanism facilitates self-extinguishing of transient faults, its side effect is that it minimizes the residual current at the fault point, significantly weakening the electrical characteristics of the fault. When the fault transition resistance reaches 1000Ω or even higher, the weak transient or steady-state signals generated by the fault are easily drowned out by the system's inherent unbalanced current, measurement noise, and interference introduced by control switch actions.

[0003] For the fault location problem in resonant grounding systems, existing traditional methods, whether based on steady-state or transient quantities, are all based on the analysis of one-dimensional time-domain signals, and therefore generally face insurmountable limitations.

[0004] Line selection methods based on steady-state quantities include the zero-sequence admittance method and the power direction method, which mainly utilize the amplitude and phase of the zero-sequence current after the fault enters steady state. Under high-resistance faults, these methods suffer from extremely low signal-to-noise ratios due to the arc-suppression coil compensating for the minimum steady-state zero-sequence current, making effective extraction difficult. Even minor system imbalances or measurement errors can lead to incorrect power direction determination or inaccurate admittance calculations, resulting in extremely poor reliability.

[0005] Line selection methods based on transient quantities include the first half-wave method and the transient energy method, which attempt to utilize the transient process at the moment of fault occurrence. However, these methods often fail to fully utilize the transient process. High-resistivity faults typically exhibit underdamped oscillatory characteristics, and their transient processes last for a relatively long time, containing rich dynamic information about the system. However, traditional transient methods often only utilize local information such as polarity or energy in the initial stage of the transient process, ignoring the global characteristics inherent in the entire dynamic evolution process.

[0006] In summary, traditional line selection methods generally suffer from limitations such as low signal-to-noise ratio, insufficient utilization of transient processes, and extremely high requirements for measurement errors (e.g., transmission error of zero-sequence CTs and sampling synchronization). In the important and complex 35kV power grid environment, these limitations have become major technical challenges restricting power supply reliability and protection accuracy. Summary of the Invention

[0007] The main objective of this invention is to provide a fault location method based on transient zero-sequence current state trajectory analysis. This method addresses the problem that current fault location methods for high-resistance grounding systems rely on numerical comparisons of weak signals, which are susceptible to interference from factors such as signal-to-noise ratio, measurement errors, and differences in line parameters, resulting in low sensitivity and reliability. The invention provides a fault location method and system based on transient zero-sequence current state trajectory analysis, which is fundamentally immune to signal amplitude distortion and significantly improves fault location reliability.

[0008] To achieve the above objectives, the first aspect of this application provides a fault location method based on transient zero-sequence current state trajectory analysis, the method comprising: Obtain the transient zero-sequence current time series of each of the multiple feeder lines in the resonant grounding system; For each of the multiple feeder lines, state space reconstruction is performed based on its transient zero-sequence current time series to generate a multidimensional state trajectory characterizing the dynamic evolution of the current signal. For each generated state trajectory, a normalized time-varying morphological descriptor is calculated to quantitatively characterize the evolution trend of its geometric shape over time; A feeder line is randomly selected as the reference line, and the time-varying morphology descriptor of any non-reference line is compared with the time-varying morphology descriptor of the reference line hour by hour. The comparison results are accumulated within a preset time window to calculate the cumulative difference score that characterizes the difference in dynamic evolution trend between the non-reference line and the reference line. The faulty line is identified by applying a preset decision rule to the cumulative difference score of each of the non-baseline lines.

[0009] A second aspect of this application provides a fault location device based on transient zero-sequence current state trajectory analysis, comprising: The data acquisition module is used to acquire the transient zero-sequence current time series of each of the multiple feeder lines in the resonant grounding system; The trajectory generation module is used to reconstruct the state space of each of the multiple feeder lines based on its transient zero-sequence current time series, so as to generate a multi-dimensional state trajectory characterizing the dynamic evolution process of the current signal. The morphology analysis module is used to calculate a normalized time-varying morphological descriptor for each generated state trajectory, which quantitatively characterizes the evolution trend of its geometric shape over time. The difference calculation module is used to randomly select a feeder line as the reference line, and calculate the cumulative difference score that characterizes the difference in dynamic evolution trend between the non-reference line and the reference line by comparing the time-varying morphology descriptor of any non-reference line with the time-varying morphology descriptor of the reference line hourly, and accumulating the comparison results within a preset time window. The decision output module is used to apply preset decision rules to the cumulative difference scores of each of the non-baseline lines to identify faulty lines.

[0010] A third aspect of this application provides an electronic device including a memory and a processor, the memory storing a computer program that, when executed by the processor, causes the processor to perform steps as described in the first aspect and any possible implementation thereof.

[0011] A fourth aspect of this application provides a computer-readable storage medium, wherein when a computer program is executed by a processor, the processor performs the steps of the first aspect and any possible implementation thereof.

[0012] The technical solution provided in this application has the following beneficial effects: 1. High sensitivity and high reliability: This application transforms the fault location problem from the traditional "numerical comparison" paradigm to "geometric shape recognition" of the dynamic evolution process of the signal. Its criterion does not depend on the absolute value of the signal, but rather on the normalized dynamic evolution pattern, thus achieving natural immunity to practical engineering interference such as system noise and zero-sequence current transformer transmission errors, greatly improving the sensitivity and reliability of fault location. 2. Strong high-resistance capability: This application can reliably identify extreme high-resistance grounding faults with transition resistances as high as several thousand ohms or even higher, effectively solving the protection blind zone problem of traditional methods. 3. Strong robustness and universality: The fault location results of this application are not affected by the fault location, the initial phase angle of the fault, or the line type (such as overhead lines, cables) and its differences in ground capacitance parameters. It performs stably in complex power grid environments and has good engineering applicability. 4. Innovative Principle: This application introduces a completely new analytical dimension and technical approach to the field of power system fault diagnosis. By analyzing the geometric evolution trend of the state trajectory, it fundamentally solves the limitations of traditional methods in addressing the core pain point of "low signal-to-noise ratio and large measurement error". Attached Figure Description

[0013] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0014] in: Figure 1 This application provides a schematic diagram of the phase space trajectory of a normal line during a high-resistance grounding fault in a power distribution network. Figure 2This application provides a schematic diagram of the phase space trajectory of different faulted lines during a high-resistance grounding fault in a distribution network. Figure 3 This is a schematic diagram of a typical resonant grounding power distribution system provided in an embodiment of this application; Figure 4 A flowchart illustrating a fault location method based on transient zero-sequence current state trajectory analysis provided in this application embodiment; Figure 5 This is a schematic diagram of a simulation model architecture for a 35kV arc-suppression coil power distribution system provided in an embodiment of this application. Figure 6 This application provides a schematic diagram of the phase space trajectory of zero-sequence current in each line during an F1 fault. Figure 7 This is a schematic diagram illustrating the change in the distance between the origin points of each line in an F1 fault, provided as an embodiment of this application. Figure 8 This is a schematic diagram of a fault location device based on transient zero-sequence current state trajectory analysis provided in an embodiment of this application. Detailed Implementation

[0015] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present application.

[0016] The terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or apparatuses.

[0017] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0018] The embodiments of this application are described below with reference to the accompanying drawings.

[0019] To optimize the operating efficiency and stability of power systems, this application proposes a fundamental shift in technical paradigm: instead of analyzing the instantaneous value of transient current in isolation, it treats the zero-sequence current evolving after a fault in each line as a whole and analyzes its geometric trajectory in state space. This introduces state space reconstruction theory, cleverly transforming the fault line selection problem from a traditional numerical comparison problem into a problem of pattern recognition of the geometric shape of global electrical parameters.

[0020] For all healthy lines, they experience the same system-level disturbance caused by a distant fault point; therefore, their dynamic response processes are essentially similar, and their corresponding state trajectories will be highly consistent in geometric topology. For faulty lines, however, they not only experience system disturbances but are also the source and channel of fault energy; their dynamic response processes will inevitably differ fundamentally from those of healthy lines. Therefore, the topological form of their state trajectories will exhibit significant differences. Accordingly, the fault location problem is cleverly transformed from a comparison of the magnitudes of weak, noisy signals into a pattern recognition problem of the complex planar geometry of global electrical parameters.

[0021] First, the power system environment in which the application is applied will be described.

[0022] Please see Figure 3 , Figure 3 This application provides a schematic diagram of a typical resonant grounding power distribution system. The resonant grounding system 300 includes a power supply 350 that supplies power to the system, connected to a busbar 310. The busbar 310, as the core node for power collection and distribution, leads out multiple feeder lines 320 (as shown in the diagram, feeder lines L1, L2 to Ln) to supply power to different load areas. To compensate for the capacitive current generated during a single-phase ground fault, the neutral point of the resonant grounding system 300 is grounded through an arc suppression coil 330. When a single-phase ground fault occurs on one of the feeder lines (e.g., feeder line L2), a fault point 340 will appear. The technical solution provided in this application aims to accurately determine the specific line where the fault point 340 is located by analyzing the electrical quantities of all feeder lines 320.

[0023] Figure 4 This is a flowchart illustrating a fault location method based on transient zero-sequence current state trajectory analysis provided in an embodiment of this application. The method includes the steps of obtaining a zero-sequence current time series, 401; reconstructing the state space, 402; calculating a time-varying morphological descriptor, 403; calculating a cumulative difference score, 404; and performing a decision discrimination, 405.

[0024] Reference Figure 3 and Figure 4 The specific steps of the method in this embodiment are described below: 401. Obtain the zero-sequence current time series of multiple feeder lines. The zero-sequence current time series in this embodiment is a transient zero-sequence current time series.

[0025] When the system detects a suspected single-phase ground fault (e.g., zero-sequence voltage exceeds a preset trigger value), the fault selection procedure is triggered. A data acquisition device (such as a microprocessor-based protection device or a dedicated fault recorder) deployed within the substation and connected to the zero-sequence current transformers of each feeder line 320 begins data acquisition. In this embodiment, the data acquisition device synchronously acquires the transient zero-sequence current of all feeder lines in the system (taking a system containing five lines L1 to L5 as an example) at a relatively high sampling frequency (e.g., 4kHz). The acquired data length covers a preset time window after the fault occurs, for example, three power frequency cycles (60 milliseconds for a 50Hz power grid). Thus, for each line k (k=1, 2, ..., 5), a discrete zero-sequence current time series can be obtained, which contains rich dynamic information during the fault transient process, forming the basis for subsequent analysis.

[0026] 402. Reconstruct the state space for each line to generate a multi-dimensional state trajectory.

[0027] In order to reveal the inherent dynamic structure and geometric shape of a one-dimensional time-domain signal and thus get rid of the direct dependence on the signal amplitude, this step maps it to a higher-dimensional space.

[0028] Specifically, a two-dimensional phase space (d=2) can be selected to visually display the transient process trajectory changes of each line after a fault, with less computation, resulting in high analysis efficiency and fast line selection speed. Through the two-dimensional phase space, the two-dimensional phase space vector expression P of each sampling point of the zero-sequence current of line k, with the origin as the starting point, can be obtained. km The vector ends of all sampling points along this line are connected sequentially to form the phase space trajectory in two-dimensional space.

[0029] (1) (2) As a specific implementation method, this embodiment can use the time delay method to construct a two-dimensional state trajectory.

[0030] Considering that the components of the zero-sequence current transient process are all power frequency or near-power frequency free oscillation components, based on the mutual information method in chaos theory, for the time series of each line k, its time delay sequence can be constructed, in which a fixed time delay parameter is set.

[0031] Optionally, the delay time τ can be selected as 1 / 4 of a power frequency cycle. For a 50Hz power grid, one power frequency cycle is 20 milliseconds, so τ is set to 5 milliseconds accordingly. The reason for choosing this delay time is to ensure that the two-dimensional phase space trajectory presents a clear, geometrically recognizable spiral curve during the transient process.

[0032] Using the zero-sequence current of the 35kV resonant grounding system as a reference, the phase space trajectories of the normal line and the fault line are constructed respectively. The horizontal and vertical coordinates of their phase space trajectories are shown in Equation (3) and Equation (4) respectively. The phase space trajectories are shown in Equation (7).

[0033] (3) (4) (5) In the formula: x k ( t ), y k ( t ) represents the abscissa and ordinate expressions of the phase space trajectories of each line; the abscissa and ordinate of the center of the circle are respectively expressed as: a k ( t ), a k ( t + τ ); b k = D k Its absolute value represents the radius of the phase space trajectory of each line. The x and y coordinates of the center are respectively represented as... a k ( t )=- D k e -δt sin( ωt + φ 2), a k ( t + τ )=- D k e -δ(t+τ) sin( ωt + φ2+π / 2), both are periodic oscillation components whose amplitude decreases and tends to 0 as time increases; b k = D k Its absolute value represents the phase space trajectory radius of a normal line, and it is defined as: (6) In the above formula, L p This represents the inductance value of the arc suppression coil in a 35kV resonant grounding system. C k It is the first k The capacitance parameters to ground of the feeder.

[0034] By using the original time series as the horizontal axis and its time delay sequence as the vertical axis, a series of trajectory points can be constructed on a two-dimensional plane. Connecting these discrete points that evolve over time in sequence forms a two-dimensional state trajectory that can characterize the dynamic evolution of the zero-sequence current of the line.

[0035] Figure 1 and Figure 2 The diagram illustrates the trajectory patterns under different line conditions.

[0036] in, Figure 1 This application provides a schematic diagram of the phase space trajectory of a normal line during a high-resistance grounding fault in a distribution network, specifically illustrating the occurrence of such a fault in the distribution network. R f =4kΩ φ The phase space trajectory of the normal line within 5 cycles after a high-resistance ground fault at 0° forms a spiral curve that develops from the inside out.

[0037] Since the transient zero-sequence current of a sound line is mainly composed of the transient oscillation component caused by the line-to-ground capacitance, its energy decays rapidly over time. Therefore, its state trajectory is a curve that starts from a certain initial state point and rapidly converges to the origin of the coordinate system in a spiral manner.

[0038] For any healthy circuit, its transient zero-sequence current is mainly determined by system-level disturbances caused by distant fault points. Mathematical analysis proves that there is no significant phase difference (angle difference) between its transient oscillatory component and its final steady-state component. θ (Always 0 or 2π). Analysis of the phase space trajectory characteristics of the zero-sequence current in a normal line reveals that during the transient process, the horizontal and vertical coordinates of the phase space trajectory continuously change due to the attenuation factor, causing the trajectory circle to undergo some deformation and exhibit a spiral curve trend. As time increases, the distance dk between each vector point on the trajectory and the spatial origin continuously increases until a steady state is reached.

[0039] In addition to transient oscillation components, faulty circuits also contain continuous fault current components injected from the fault point, which causes the energy decay of the transient process to be slow or even divergent. Therefore, its state trajectory shows a divergence outward from the vicinity of the origin, or exhibits a continuous, large-amplitude irregular oscillation pattern.

[0040] Accordingly, Figure 2 This is a schematic diagram of the phase space trajectory of different fault lines during a high-resistance grounding fault in a distribution network, provided as an embodiment of this application.

[0041] Unlike healthy lines, faulty lines not only bear the brunt of system disturbances, but also serve as conduits for fault energy. This fundamental difference in role results in a significant, fault-parameter-dependent phase difference between the transient and steady-state components of their transient zero-sequence current. θ n The range of values ​​is approximately [ π / 2, π], the value of which is affected by the transition resistance R f and the percentage of capacitance to ground of the faulty line C n / C The impact.

[0042] Set up distribution network generation R f =3kΩ φ A high-resistance grounding fault with a radius of 90°, such as Figure 2 The diagram shows the phase-space trajectory of the faulted line within five cycles after a fault occurs in the distribution network under two different fault conditions. Figure 2 (a) represents the percentage of capacitance to ground, C. n The phase space trajectory of the faulty overhead line with a fault rate of / C=4% shows a spiral line that continuously develops from the outside in and tends to stabilize. Figure 2 (b) is C n The phase space trajectory of a faulty cable line with a fault rate of / C=70% exhibits a spiral shape that develops from the inside out. At this time, the trajectory shape of the faulty line is similar to that of the normal line.

[0043] when R f and C n / C When it is large, θ n Approaching π. The initial point of the phase space trajectory of the faulty line is inside the steady-state circle, and the whole line presents a spiral curve that develops from the inside out (similar to the healthy line).

[0044] And when R f and Cn / C When it is extremely small, θ n close to π / 2. The initial point of the phase space trajectory is outside the steady-state circle, and the overall trajectory will exhibit a spiral curve that decays from the outside to the inside.

[0045] In an alternative implementation, instead of using the time delay method, the signal and its first derivative can be used to construct the two-dimensional state trajectory.

[0046] Optionally, those skilled in the art will understand that, in addition to the two-dimensional state trajectory described above, three-dimensional or higher-dimensional state trajectories can also be constructed by introducing components such as the second derivative of the signal, which also fall within the scope of the protection concept of this invention.

[0047] 403. Calculate the normalized time-varying morphological descriptor for each trajectory.

[0048] To quantitatively describe the trajectory evolution trend (contraction or expansion), this embodiment introduces a normalized quantization index that is insensitive to signal amplitude, namely, the morphological descriptor. In this embodiment, this descriptor is specifically implemented as the "origin distance ratio coefficient". μ k ".

[0049] The ratio of the origin distance to the normal route μ k ( t The expression for ) is: (7) Specifically, the ratio coefficient of the origin distance on a normal route. μ k The values ​​of the line and their respective capacitance to ground C k Regardless of the magnitude, the change trend of the distance between the origin and the phase space trajectory of normal lines is consistent. Therefore, after a high-resistance ground fault occurs in the distribution network, the ratio coefficient of the distance between the origin and the origin of each normal line is the same. The faulty line was due to and There is an angle difference between them, and the ratio coefficient of their origin distance during the fault transient process. μ n ( t The ratio coefficient is definitely different from that of a normal line, and the faulty lines in the distribution network can be distinguished based on the above characteristics.

[0050] 404. Randomly select a baseline route and calculate the cumulative difference score for each non-baseline route.

[0051] After calculating the time-varying morphological descriptor sequences for all five lines (e.g., L1-L5), this step aims to quantify the differences between them.

[0052] This application employs a relative comparison strategy: randomly selecting a feeder as the reference line. For example, in this discrimination, line L2 is randomly selected as the reference line h. This random selection strategy is feasible because the curve shapes of all healthy lines are highly similar, and arbitrarily selecting a healthy line as the reference yields equivalent results.

[0053] In the determination process of the method proposed in this application, after a single-phase ground fault is detected, a feeder can be randomly selected as a reference line. By comparing the response differences between the reference line and other feeders one by one, the faulty feeder is selected. The implementation process will be described below.

[0054] The ratio coefficient of the origin distance of other lines in the distribution network and μ h ( t The difference is calculated at each point in the time series to analyze the degree of difference between the current line and the reference line. This difference in the ratio coefficient is defined as the ratio difference. μ k ( t Its expression is: (8) In the formula: , k ≠ h .

[0055] 405. Make decisions using pre-defined decision rules.

[0056] Finally, based on the calculated cumulative difference scores of each non-benchmark line, a pre-defined decision rule is applied to identify faulty lines. The core of this decision rule is to determine which lines' cumulative difference scores meet a pre-defined anomaly condition.

[0057] Specifically, to analyze the difference between the line and the reference line during the entire transient process after the fault occurs, and to amplify the differences of the line, the sum of the ratios at each time point in the time series is obtained, thus determining the line's performance. k The ratio difference and S k for: (9) If the base line h No fault occurred on the line. k When it is a normal line, μ k (t If )≡0, then the line k The ratio difference and S k =0, theoretically a normal line k The trend of change is completely consistent with that of the baseline line; when the line k The faulty line, i.e. k = n ,at this time μ n ( t )=| μ n ( t )- μ h ( t )|≠0, S k >0. Based on this characteristic, once the sum of the differences between a certain line and the reference line is not zero, it can be determined that the line is faulty.

[0058] If the base line h A malfunction occurred, that is h = n hour, μ k ( t )=| μ k ( t )- μ n ( t )|≠0, From this, we can know the ratio difference of other lines and S k If both are greater than 0, the reference line can be determined to be a faulty line; conversely, if the line ratio difference and sum are both greater than 0, the reference line can be determined to be a faulty line. S k If all values ​​are 0, it can be concluded that no faults have occurred on any of the lines in the distribution network, and the fault can be determined to be on the distribution network bus.

[0059] In this embodiment, the abnormal condition can be set as the accumulated difference score exceeding a preset fixed threshold. For example, the threshold can be set to 3, which provides a clear boundary between normal fluctuations in a healthy circuit and significant differences in a faulty circuit.

[0060] The threshold setting is intended to enhance the robustness of the method. Through extensive simulation and experimental research, the recommended criterion for fault selection in distribution networks under high-resistance grounding fault conditions is as follows: (10) If equation (10) is satisfied, then the line can be determined. kThe faulty line is identified; if the sum of the differences between all lines satisfies the above formula, then the reference line can be determined. h If the above formula is not met, the faulty line can be identified; conversely, if none of the above formulas are met, the faulty busbar of the distribution network can be identified, thus achieving accurate fault line selection.

[0061] For example, decision-making logic specifically includes the following three situations: Scenario 1: When only one non-baseline line's cumulative difference score meets the abnormal condition, that non-baseline line is determined to be a faulty line. For example, suppose line L1 is faulty, while the baseline line L2 is intact. Because L1's... u 1( t ) curve and L2 u 2( t The curves differ greatly, and the calculated... S 1 will be much greater than 3 (for example, the simulation result is 129.877). And other healthy lines L3, L4, L5... u ( t The curve is very similar in height to the curve of L2, therefore the calculated... S 3. S 4. S 5 will all be much smaller than 3 (e.g., 0.007, 0.186, 0.218 respectively). At this point, the decision logic finds that only... S Since the condition of "greater than 3" is met, line L1 is ultimately determined to be a faulty line. Scenario 2: When the cumulative difference scores of all non-baseline lines meet the preset abnormality conditions, the selected baseline line is determined to be a faulty line. For example, suppose the randomly selected baseline line L2 is itself a faulty line. Then, all non-baseline lines (L1, L3, L4, L5) are healthy lines. u ( t The curves are similar to each other, but both are related to the faulty line L2. u 2( t The curves show significant differences. Therefore, all cumulative difference scores were calculated. S 1. S 3. S 4. S The values ​​of 5 are all significantly greater than 3 (e.g., 13.359, 13.357, 13.359, 13.295 respectively). At this point, the decision logic finds that the difference scores of all non-baseline lines are abnormal, which in turn indicates that line L2, which serves as the comparison benchmark, is the real faulty line.

[0062] Scenario 3: When the cumulative difference scores of all non-baseline lines do not meet the preset abnormal conditions, it is determined to be a bus fault or an external fault. If the fault occurs on bus 310 or outside the system, then all feeder lines 320 within the substation are in good condition, and the transient zero-sequence current flowing through them is determined by their respective line-to-ground capacitance, with highly consistent dynamic evolution trends. Therefore, regardless of which line is selected as the base, all other lines... u ( t The curves are similar, and all cumulative difference scores are calculated. S k All scores will be much less than 3. At this point, the decision logic finds that no score meets the abnormal condition, so it determines that there is a bus fault (or no fault).

[0063] The use of a relatively sparse sampling frequency in power grid substations does not affect the line selection accuracy of the phase space trajectory identification method and improves the noise resistance of the method. At the same time, the criteria used in the method of this application embodiment are not affected by the zero-sequence CT transmission error under high-resistance grounding fault and the phase angle and amplitude sampling error caused by operational errors in actual engineering, and always maintain good line selection accuracy and reliability.

[0064] In an optional implementation, in the step of calculating the time-varying morphological descriptor (corresponding to step S603), instead of using the origin distance ratio coefficient, a descriptor based on the trajectory geometric coverage range can be used, such as the "normalized convex hull area change rate". The specific operation is as follows: At every moment t n Consider all trajectory points within a short time window (e.g., half a power frequency cycle) since that moment. Calculate the convex hull of this set of points, which is the smallest convex polygon that can contain all these points. Then, calculate the area of ​​this convex hull. A k ( t n ). This area A k ( t n This intuitively reflects the size of the trajectory's activity range in the recent period. Subsequently, the "normalized convex hull area change rate" is calculated as a time-varying morphological descriptor, and its calculation formula can be defined as: (11) in This is the time step used to calculate the rate of change, which can be, for example, several sampling periods. This descriptor is also normalized; it quantifies whether the trajectory's range of activity is expanding (positive descriptor) or shrinking (negative descriptor). For a healthy line, its trajectory converges, and the convex hull area decreases over time; the descriptor will primarily be negative or close to zero. For a faulty line, its trajectory diverges or oscillates continuously; the convex hull area increases or fluctuates violently; the descriptor will exhibit large positive values ​​or violent oscillations.

[0065] The subsequent steps are similar to those in the aforementioned embodiments. A reference line is randomly selected, and then the cumulative difference score between the morphological descriptor sequences of any non-reference line k and the reference line h is calculated. For example, the following calculation is performed: (12) Finally, the same decision rule (comparison) is applied. S k The faulty circuit can be identified by comparing it with the preset threshold.

[0066] The simulation verification based on the aforementioned method is given below.

[0067] To comprehensively and rigorously evaluate the performance of the method proposed in this application, a simulation environment such as PSCAD / EMTDC was built. Figure 5 The simulation model shown is a typical simulation model of a 35kV resonant grounding distribution network. This simulation covers different transition resistances, different fault locations, and different line types, and specifically verifies the high resistance withstand capability and anti-interference capability of the method proposed in this application.

[0068] The system's main transformer has a rated voltage of 110 / 35kV and a rated capacity of 100MVA. The system's ground capacitance current is 84.07A, the arc suppression coil detuning is -5%, and the inductance LP is 0.217H. Ground fault points F1~F4 are set at different locations in the system, with line L2 designated as the reference line. The simulation uses a sampling frequency of 4kHz to collect zero-sequence current data within three cycles after the fault for line selection analysis.

[0069] A single-phase ground fault with a transition resistance of 4kΩ and an initial phase angle of 0° occurred at ground fault point F1. Zero-sequence current data for three cycles after the fault occurred on each line were collected, and the phase space trajectories of the zero-sequence currents of the normal line L4 and the faulted line L1 were constructed as follows: Figure 6 As shown in (a) and (b), it can be seen that during the fault transient process, the phase space trajectory of the zero-sequence current of the normal line is a spiral curve that develops from the inside to the outside, while the phase space trajectory of the fault line develops from the outside to the inside. After reaching steady state, each line forms a circle with the origin as the center.

[0070] The changes in the distance dk from the origin of each line's phase space trajectory are obtained as follows: Figure 7 As shown in (a) and (b), (a) corresponds to the normal line L4 and (b) corresponds to the faulty line L1. When a high-resistance grounding fault occurs on an overhead line, the development trends of the phase space trajectories of the normal line and the faulty line are opposite, and the difference in phase space trajectories is very obvious.

[0071] To verify the universality and accuracy of the criterion under high-resistance grounding faults in the distribution network, line L2 was still set as the reference line. Multiple fault occurrence states with different initial phase angles and transition resistances were randomly set at grounding fault points F1, F2, F3 and F4. Data from the first three cycles after the fault occurred were also analyzed.

[0072] Table 1 shows the ratio difference of each line relative to the reference line L2 under different fault conditions. S k And the route selection results.

[0073]

[0074] Table 1 Meanwhile, simulations verified that the proposed fault selection method has high noise resistance and excellent fault selection performance under noisy system conditions.

[0075] Table 2 shows the distribution network route selection results under system noise interference.

[0076]

[0077] Table 2 When the system contains noise, the current data collected after a high-resistance grounding fault will be disturbed to some extent, but the difference between the normal line and the reference line will remain unchanged. S k The value has virtually no impact, while the faulty line ratio and S k This is significantly larger than the case where there is no noise in the system. The presence of system noise not only does not affect the route selection results, but even improves the reliability of route selection.

[0078] Table 3 shows the distribution network line selection results under the condition that the zero-sequence CT has transmission error.

[0079]

[0080] Table 3 As shown in Table 3, regardless of the changes in the transformation ratio and phase angle errors of the zero-sequence CTs on different lines, the calculated ratio difference of each line remains basically unchanged, achieving accurate fault line selection. This fully demonstrates that the fault line selection criterion can effectively avoid the influence of the zero-sequence CT transmission error on the line selection result under high-resistance grounding faults in actual engineering practice. At the same time, the different ground capacitance values ​​of each line do not interfere with the criterion, ensuring the reliability of fault line selection.

[0081] In summary, the novel fault selection criterion proposed in this application for phase-space trajectory identification in distribution networks can accurately and reliably select faulty lines. The selection results are unaffected by the fault location, the initial phase angle of the fault, or the magnitude of the transition resistance at the fault point. It exhibits strong high-resistance capability, natural immunity to system noise and severe CT transmission errors, and stable and reliable performance. This characteristic fundamentally overcomes the limitations of traditional methods in complex electromagnetic environments and measurement conditions at 35kV.

[0082] This application presents a systematic study of the highly challenging technical problem of accurate fault location for high-impedance grounding faults in 35kV resonant grounding systems. Given that the 35kV network serves as the backbone of regional power supply, its higher voltage level, larger capacitive current, and more important load role place far more stringent demands on the sensitivity, reliability, and robustness of protection technologies than on conventional distribution networks. The limitations of traditional fault location methods become increasingly apparent under this complex context. Therefore, the starting point of this application is to seek a novel protection principle that can adapt to the characteristics of 35kV systems.

[0083] At the theoretical level, this study innovatively introduces state-space reconstruction theory, transforming fault route selection from the traditional "numerical comparison" paradigm into a pattern recognition problem involving the identification of geometric manifolds in high-dimensional space. For the first time, the study theoretically proves that the state trajectory manifolds of all healthy routes exhibit a high degree of consistency in geometric evolution trends, while faulty routes show fundamental differences. This core finding lays a solid physical and mathematical foundation for moving away from dependence on absolute signal values. Simultaneously, it proposes an index capable of quantitatively describing the differences in trajectory manifolds—the "ratio difference sum." S k "and based on this, a system was constructed S k The clear line selection criterion of >3 demonstrates its inherent immunity to major interference factors in practical engineering, such as system noise and CT transmission error. This characteristic of not relying on absolute measurement values ​​is the key to fundamentally solving the core pain point of "low signal-to-noise ratio and large measurement error" in 35kV systems, and is also the core advantage of this technical solution.

[0084] In summary, this application, through its unique morphological identification principle, effectively overcomes the technical challenges of weak fault characteristics and complex measurement environments in 35kV systems with high current capacity. It successfully proposes and verifies a precise fault location technology for high-resistance grounding faults in 35kV resonant grounding systems based on transient current state trajectory manifold analysis. This technical solution, through its unique morphological identification principle, effectively overcomes the technical challenges of weak fault characteristics and complex measurement environments in 35kV systems with high current capacity. The research results not only provide a forward-looking and highly reliable innovative solution for addressing protection blind spots in the industry, but also introduce new analytical dimensions and technical approaches to the field of power system fault diagnosis, possessing significant theoretical value and broad engineering application prospects.

[0085] Figure 8 This is a schematic diagram of a fault location device based on transient zero-sequence current state trajectory analysis, provided as an embodiment of this application. Figure 8 As shown, the fault location device 800 based on transient zero-sequence current state trajectory analysis includes: Data acquisition module 810 is used to acquire the transient zero-sequence current time series of each of the multiple feeder lines in the resonant grounding system; The trajectory generation module 820 is used to reconstruct the state space of each of the multiple feeder lines based on its transient zero-sequence current time series, so as to generate a multi-dimensional state trajectory characterizing the dynamic evolution process of the current signal. The morphological analysis module 830 is used to calculate a normalized time-varying morphological descriptor for each generated state trajectory, which is used to quantitatively characterize the trend of its geometric shape evolution over time. The difference calculation module 840 is used to randomly select a feeder line as a reference line, and calculate the cumulative difference score that characterizes the difference in dynamic evolution trend between the non-reference line and the reference line by comparing the time-varying morphological descriptor of any non-reference line with the time-varying morphological descriptor of the reference line hourly, and accumulating the comparison results within a preset time window. The decision output module 850 is used to apply preset decision rules to the cumulative difference scores of each of the non-baseline lines to identify faulty lines.

[0086] Understandably, this involves Figure 8 The relevant content of each module in the above method embodiments has been described in detail, and you can refer to the content of the method embodiments for details; that is... Figure 8 The provided fault selection device 800 based on transient zero-sequence current state trajectory analysis can perform tasks such as... Figure 4 Any steps in the illustrated embodiments will not be described in detail here.

[0087] In one embodiment, a computer-readable storage medium is also provided, which stores a computer program that, when executed by a processor, causes the processor to perform any of the steps in the above method embodiments.

[0088] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments described above. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM), etc.

[0089] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0090] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A fault location method based on transient zero-sequence current state trajectory analysis, characterized in that, The method includes: Obtain the transient zero-sequence current time series of each of the multiple feeder lines in the resonant grounding system; For each of the multiple feeder lines, state space reconstruction is performed based on its transient zero-sequence current time series to generate a multidimensional state trajectory characterizing the dynamic evolution of the current signal. For each generated state trajectory, a normalized time-varying morphological descriptor is calculated to quantitatively characterize the evolution trend of its geometric shape over time; A feeder line is randomly selected as the reference line, and the time-varying morphology descriptor of any non-reference line is compared with the time-varying morphology descriptor of the reference line hour by hour. The comparison results are accumulated within a preset time window to calculate the cumulative difference score that characterizes the difference in dynamic evolution trend between the non-reference line and the reference line. The faulty line is identified by applying a preset decision rule to the cumulative difference score of each of the non-baseline lines.

2. The fault location method based on transient zero-sequence current state trajectory analysis according to claim 1, characterized in that, The preset decision rules include: When only one of the non-baseline lines has a cumulative difference score that meets the preset abnormal condition, the non-baseline line is determined to be a faulty line. When the cumulative difference scores of all the non-reference lines meet the preset abnormality condition, the reference line is determined to be a faulty line; and, When the cumulative difference score of all the non-baseline lines does not meet the preset abnormal condition, it is determined to be a bus fault.

3. The fault location method based on transient zero-sequence current state trajectory analysis according to claim 1, characterized in that, The state space reconstruction specifically includes: A two-dimensional state trajectory is constructed using the transient zero-sequence current time series and its time delay series.

4. The fault location method based on transient zero-sequence current state trajectory analysis according to claim 3, characterized in that, The delay time is 1 / 4 of a power frequency cycle.

5. The fault location method based on transient zero-sequence current state trajectory analysis according to claim 1, characterized in that, The time-varying morphological descriptor is the origin distance ratio coefficient; The origin distance ratio coefficient is the ratio of the distance from the state trajectory point at the current sampling time to the origin to the distance from the state trajectory point at the previous sampling time to the origin.

6. The fault location method based on transient zero-sequence current state trajectory analysis according to claim 1, characterized in that, The method for calculating the cumulative difference score specifically includes: Within the preset time window, the absolute values ​​of the differences between the time-varying morphology descriptors of the non-reference line and the time-varying morphology descriptors of the reference line are summed.

7. The fault location method based on transient zero-sequence current state trajectory analysis according to claim 2, characterized in that, The preset abnormal condition is that the cumulative difference score is greater than a preset threshold.

8. A fault location device based on transient zero-sequence current state trajectory analysis, characterized in that, include: The data acquisition module is used to acquire the transient zero-sequence current time series of each of the multiple feeder lines in the resonant grounding system; The trajectory generation module is used to reconstruct the state space of each of the multiple feeder lines based on its transient zero-sequence current time series, so as to generate a multi-dimensional state trajectory characterizing the dynamic evolution process of the current signal. The morphology analysis module is used to calculate a normalized time-varying morphological descriptor for each generated state trajectory, which quantitatively characterizes the evolution trend of its geometric shape over time. The difference calculation module is used to randomly select a feeder line as the reference line, and calculate the cumulative difference score that characterizes the difference in dynamic evolution trend between the non-reference line and the reference line by comparing the time-varying morphology descriptor of any non-reference line with the time-varying morphology descriptor of the reference line hourly, and accumulating the comparison results within a preset time window. The decision output module is used to apply preset decision rules to the cumulative difference scores of each of the non-baseline lines to identify faulty lines.

9. An electronic device, characterized in that, It includes a memory and a processor, the memory storing a computer program that, when executed by the processor, causes the processor to perform the steps of the method as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, When the computer program is executed by a processor, the processor performs the steps of the method as described in any one of claims 1-7.

Images