A method and system for rapid identification of short-circuit faults in ultra-high voltage power supplies
By constructing a non-integer-cycle prediction window and using two-dimensional phase space analysis, and utilizing the second-order differential characteristics of current, short-circuit faults in ultra-high voltage power supply lines can be quickly identified. This solves the problems of long identification cycles and misjudgments/missed judgments in existing technologies, and achieves efficient and accurate fault identification.
Patent Information
- Application Number
- CN202611363771.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2026-09-04
- Publication Date
- 2026-11-10
- Estimated Expiration
- 2046-09-04
AI Technical Summary
Existing short-circuit fault identification technologies for ultra-high voltage power lines suffer from long identification cycles, slow response speeds, difficulty in capturing transient fault characteristics, susceptibility to interference leading to misjudgments and missed judgments, and failure to effectively utilize hierarchical prediction and phase space feature analysis, resulting in insufficient identification efficiency and accuracy.
By constructing a non-integer-cycle prediction window, the direction of sign change of the second-order difference sequence of current is extracted, mapped to a two-dimensional phase space, and the eccentricity of ellipse fitting and the characteristic value of phase trajectory jump are calculated. Combined with the threshold, the short-circuit fault of the ultra-high voltage power supply line is determined.
It significantly shortens the short-circuit fault identification cycle, improves response speed, reduces the probability of false positives and false negatives, improves identification efficiency and judgment accuracy, and ensures the stability and safety of ultra-high voltage power systems.
Smart Images

Figure CN122860202B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of fault identification technology, and in particular to a method and system for rapid identification of short-circuit faults in ultra-high voltage power supplies. Background Technology
[0002] Existing short-circuit fault identification technologies for ultra-high voltage power lines mostly rely on whole-cycle sampling and traditional time-domain feature judgment. The fault triggering and judgment cycle is relatively long, making it difficult to meet the stringent requirements of ultra-high voltage systems for fault speed and failing to achieve rapid response in the early stages of a fault. Conventional identification methods are not accurate enough in capturing the distortion and abrupt changes of current signals and are easily affected by factors such as load fluctuations and harmonic interference. The sensitivity and reliability of fault prediction are low, and misjudgments or missed judgments are prone to occur.
[0003] Traditional fault identification schemes lack a hierarchical prediction mechanism to address the transient and abrupt characteristics of short-circuit faults. A single threshold judgment logic struggles to distinguish between normal disturbances and genuine short-circuit faults, making it difficult to balance identification efficiency and accuracy. Furthermore, existing methods do not utilize advanced features such as phase space trajectories for in-depth analysis of fault signals, failing to quickly pinpoint the typical characteristics of short-circuit faults. This results in significant shortcomings in the identification speed and accuracy of short-circuit faults in ultra-high voltage power lines, hindering the safe and stable operation of power systems. Summary of the Invention
[0004] In view of the above problems, the purpose of this invention is to provide a method and system for rapid identification of short-circuit faults in ultra-high voltage power supplies, so as to solve the problems of long identification cycle, slow response speed, inaccurate capture of transient fault features, susceptibility to interference leading to false or missed judgments, and failure to adopt layered prediction and phase space feature depth analysis, which cannot balance identification efficiency and judgment accuracy.
[0005] This invention provides a method for rapid identification of short-circuit faults in ultra-high voltage power supplies, comprising: Step 1: Within the current power frequency cycle, extract the instantaneous current value of the ultra-high voltage power supply line and obtain the second-order difference sequence of the current after second-order difference operation, and determine the sign change direction at the zero crossing point. Step 2: Using the sampling moment when the direction of the sign change reverses as the starting point, construct a non-integer cycle prediction window for the ultra-high voltage power supply line. Within the non-integer cycle prediction window, when the number of alternations of the positive and negative signs of the second-order difference of the current exceeds a preset alternation threshold, generate a first-level fault prediction signal. Step 3: In response to the first-level fault prediction signal, the instantaneous current value within the non-integer cycle prediction window is mapped to the two-dimensional phase space, and the elliptic fitting eccentricity of the formed current-current differential phase trajectory and the phase trajectory jump characteristic value between two adjacent sampling times are calculated. Step 4: When the eccentricity of the ellipse fitting is greater than the preset eccentricity threshold and the cumulative sum of the phase trajectory jump characteristic values exceeds the preset characteristic threshold, the ultra-high voltage power line is determined to be a short circuit fault.
[0006] Preferably, the sign change direction of the second-order difference sequence of the current obtained after extracting the instantaneous current value of the ultra-high voltage power line through second-order difference operation is as follows: Based on the instantaneous current values continuously sampled by the ultra-high voltage power line within the current power frequency cycle, the instantaneous current difference between two adjacent sampling times is calculated sequentially to obtain a first-order current difference sequence. The difference between two adjacent first-order current differences in the first-order current difference sequence is calculated sequentially to obtain the second-order current difference sequence, and the second-order current difference sequence is used as the second-order current difference sequence. The second-order difference sequence of current is scanned point by point along the time-increasing direction, and the sign of the current sampling time is compared with the sign of the adjacent previous sampling time. When the signs of the values at two adjacent sampling times are different, the position interval formed by the two adjacent sampling times is recorded, and the direction of the jump from the sign of the previous point to the sign of the next point within the position interval is taken as the sign change direction of the position interval.
[0007] Preferably, the process of constructing a non-integer cycle prediction window for the ultra-high voltage power line, starting from the sampling moment when the direction of the sign change reverses, is as follows: The current sampling moment, in which the direction of the sign change changes from positive to negative or from negative to positive from the previous sampling moment to the current sampling moment, is determined as the sign reversal moment. Using the sign reversal moment as the starting point of the window, a preset number of sampling moments are continuously intercepted along the time increment direction, so that the length of the intercepted sampling interval is less than one-quarter of the power frequency cycle. The time interval between the sign reversal time and the last truncation sampling time is defined as the non-integer cycle prediction window of the ultra-high voltage power supply line.
[0008] Preferably, within the non-integer cycle prediction window, when the number of alternations of the positive and negative signs of the second-order current difference exceeds a preset alternation threshold, a first-level fault prediction signal is generated, the process of which is as follows: Following the increasing time sequence within the non-integer period prediction window, the numerical sign corresponding to each sampling moment in the second-order difference sequence of the current is extracted one by one, and the numerical sign is either positive or negative. The sign of the current sampling time is compared with that of the previous adjacent sampling time. If they are different, it is determined that a sign alternation has occurred, and the current value of the alternation counter is incremented by one until all sampling times within the non-integer period prediction window have been traversed. The final value of the alternation counter is read. If the final value is greater than the preset alternation threshold, a first-level fault prediction signal is output as a trigger pulse to start the subsequent fault identification process.
[0009] Preferably, the process of mapping the instantaneous current value within the non-integer cycle prediction window to a two-dimensional phase space in response to the first-level fault prediction signal is as follows: In response to the first-level fault prediction signal, the instantaneous current values arranged in chronological order within the non-integer cycle prediction window are locked. The difference between the instantaneous current values of two adjacent sampling times within the non-integer cycle prediction window is taken as the current difference component at each sampling time. Using the instantaneous current value at each sampling moment as the first dimension coordinate of the two-dimensional phase space, and the current differential component corresponding to each sampling moment as the second dimension coordinate of the two-dimensional phase space, a two-dimensional phase point is generated for each sampling moment within the non-integer period prediction window. Based on the two-dimensional phase points, the current-current differential phase trajectory of the ultra-high voltage power supply line is generated.
[0010] Preferably, the process of generating the current-current differential phase trajectory of the ultra-high voltage power supply line based on the two-dimensional phase points is as follows: The two-dimensional phase points are arranged into a phase point sequence according to the increasing order of the sampling time. Starting from the first phase point in the phase point sequence, adjacent phase points are connected sequentially with straight line segments to obtain the phase trajectory curve of the ultra-high voltage power supply line; The coordinates of all phase points in the phase trajectory curve and the connection relationship between adjacent phase points are collectively defined as the current-current differential phase trajectory of the ultra-high voltage power supply line.
[0011] Preferably, the process of calculating the eccentricity of the ellipse fitting of the formed current-current differential phase trajectory is as follows: All two-dimensional phase points in the current-current differential phase trajectory are obtained. Each two-dimensional phase point is composed of the instantaneous current value as the first dimension coordinate and the current differential component as the second dimension coordinate. Substitute all the two-dimensional phase points into the general equation of an ellipse, and use the least squares fitting method to obtain the fitted ellipse of the current-current differential phase trajectory. The lengths of the major and minor axes of the ellipse are analyzed from the coefficients of the fitted ellipse equation. The square root of the difference between the squares of the major and minor semi-axis lengths is divided by the major semi-axis length to obtain the elliptic fitting eccentricity of the current-current difference phase trajectory.
[0012] Preferably, the phase trajectory jump characteristic value between two adjacent sampling times of the calculated current-current differential phase trajectory is calculated using the following formula: ; In the formula, This indicates the first current-current differential phase trajectory. The sampling time and the first Phase trajectory jump characteristic values between sampling times Indicates the first The instantaneous current value at each sampling time. Indicates the first The instantaneous current value at each sampling time. Indicates the first The current difference components at each sampling time, Indicates the first The current difference components at each sampling time. This represents the preset weighting coefficients. This indicates the preset reference current value.
[0013] Preferably, when the eccentricity of the ellipse fitting is greater than a preset eccentricity threshold and the cumulative sum of the phase trajectory jump characteristic values exceeds a preset characteristic threshold, the ultra-high voltage power supply line is determined to have a short-circuit fault. The process is as follows: The eccentricity of the ellipse fitting is compared with a preset eccentricity threshold. If the eccentricity of the ellipse fitting is less than or equal to the eccentricity threshold, the current determination is terminated and the first-level fault prediction signal is cleared. If the eccentricity of the ellipse fitting is greater than the eccentricity threshold, then starting from the phase trajectory jump feature value corresponding to the first sampling time in the non-integer period prediction window, the phase trajectory jump feature values are accumulated sequentially to obtain the feature accumulation value of the phase trajectory jump feature values. The feature accumulation value is compared with a preset feature threshold. If the feature accumulation value is less than or equal to the feature threshold, the determination is terminated and the first-level fault prediction signal is cleared. If the cumulative sum is greater than the characteristic threshold, a short-circuit fault confirmation signal is output, and the ultra-high voltage power line is determined to be short-circuit faulted.
[0014] This invention also provides a rapid identification system for short-circuit faults in ultra-high voltage power supplies, the system comprising: The differential feature extraction module is used to extract the sign change direction of the second-order differential sequence of the current obtained by second-order differential operation of the instantaneous current value of the ultra-high voltage power line within the current power frequency cycle. The prediction window construction module is used to construct a non-integer cycle prediction window for the ultra-high voltage power line, starting from the sampling time when the direction of the sign change reverses. Within the non-integer cycle prediction window, when the number of alternations of the positive and negative signs of the second-order difference of the current exceeds a preset alternation threshold, a first-level fault prediction signal is generated. The phase trajectory feature calculation module is used to respond to the first-level fault prediction signal, map the instantaneous current value in the non-integer cycle prediction window to a two-dimensional phase space, and calculate the elliptic fitting eccentricity of the current-current differential phase trajectory and the phase trajectory jump feature value between two adjacent sampling times. The short-circuit fault determination module is used to determine that the ultra-high voltage power line is short-circuit faulted when the eccentricity of the ellipse fitting is greater than a preset eccentricity threshold and the cumulative sum of the phase trajectory jump characteristic values exceeds a preset characteristic threshold.
[0015] As can be seen from the above technical solution, the present invention provides a method and system for rapid identification of short-circuit faults in ultra-high voltage power supplies. By constructing a non-integer cycle prediction window and combining it with the alternating characteristics of the second-order differential sign of the current, a rapid initial prediction of the fault is achieved. Then, precise judgment is made by utilizing two-dimensional phase space mapping, elliptical fitting eccentricity, and phase trajectory jump characteristic values. This significantly shortens the identification cycle of short-circuit faults in ultra-high voltage power supplies, greatly improves the fault response speed, and effectively reduces the probability of misjudgment and missed judgment caused by external interference. It balances the efficiency and accuracy of fault identification and can reliably and quickly achieve accurate identification of short-circuit faults in ultra-high voltage power supply lines, ensuring the stability and safety of ultra-high voltage power system operation. Attached Figure Description
[0016] Other objects and results of the invention will become more apparent and readily understood by referring to the following description taken in conjunction with the accompanying drawings, and with a more complete understanding of the invention. In the drawings: Figure 1 This is a flowchart illustrating a method for rapid identification of short-circuit faults in ultra-high voltage power supplies according to an embodiment of the present invention. Figure 2 This is a functional block diagram of a rapid short-circuit fault identification system for ultra-high voltage power supplies according to an embodiment of the present invention; Figure 3 This is a graph showing the cumulative number of symbol alternations within a non-integer cycle prediction window under a three-phase short-circuit fault condition according to an embodiment of the present invention. Detailed Implementation
[0017] Existing ultra-high voltage power supply short circuit fault identification technologies suffer from long identification cycles, slow response speeds, inaccurate capture of transient fault characteristics, susceptibility to interference leading to misjudgments and missed judgments, and the lack of hierarchical prediction and phase space feature in-depth analysis, making it impossible to balance identification efficiency and judgment accuracy.
[0018] To address the aforementioned problems, this invention provides a method and system for rapid identification of short-circuit faults in ultra-high voltage power supplies. The specific embodiments of this invention will be described in detail below with reference to the accompanying drawings.
[0019] To illustrate the rapid identification method and system for ultra-high voltage power supply short-circuit faults provided by this invention... Figure 1 An exemplary illustration is provided for a method for rapid identification of short-circuit faults in ultra-high voltage power supplies according to an embodiment of the present invention; Figure 2 An exemplary illustration is provided for a rapid identification system for short-circuit faults in ultra-high voltage power supplies according to an embodiment of the present invention.
[0020] The following description of exemplary embodiments is merely illustrative and is in no way intended to limit the invention or its application or use. Techniques and equipment known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques and equipment should be considered part of the specification.
[0021] Reference Figure 1 The diagram shown is a flowchart illustrating a method for rapid identification of short-circuit faults in ultra-high voltage power supplies according to an embodiment of the present invention. In this embodiment, the method for rapid identification of short-circuit faults in ultra-high voltage power supplies includes: Step 1: Within the current power frequency cycle, extract the instantaneous current value of the ultra-high voltage power supply line and obtain the second-order difference sequence of the current after second-order difference operation, and determine the sign change direction at the zero crossing point. In this embodiment of the invention, the sign change direction of the second-order difference sequence of the current obtained after extracting the instantaneous current value of the ultra-high voltage power line through second-order difference operation is as follows: Based on the instantaneous current values continuously sampled by the ultra-high voltage power line within the current power frequency cycle, the instantaneous current difference between two adjacent sampling times is calculated sequentially to obtain a first-order current difference sequence. The difference between two adjacent first-order current differences in the first-order current difference sequence is calculated sequentially to obtain the second-order current difference sequence, and the second-order current difference sequence is used as the second-order current difference sequence. The second-order difference sequence of current is scanned point by point along the time-increasing direction, and the sign of the current sampling time is compared with the sign of the adjacent previous sampling time. When the signs of the values at two adjacent sampling times are different, the position interval formed by the two adjacent sampling times is recorded, and the direction of the jump from the sign of the previous point to the sign of the next point within the position interval is taken as the sign change direction of the position interval.
[0022] Continuous sampling is performed on the ultra-high voltage power line to obtain the instantaneous current values corresponding to all sampling times within the current power frequency cycle. These instantaneous current values are arranged in complete order according to the time of sampling. Starting from the first instantaneous current value after arrangement, each pair of adjacent instantaneous current values is selected in turn. The difference between the instantaneous current values is calculated by subtracting the instantaneous current value from the one in front of the one in the back. All the calculated differences are then arranged in the original order to form a complete first-order current difference sequence.
[0023] Following the original arrangement of the first-order current difference sequence, starting from the first value in the sequence, select each pair of adjacent first-order current differences in turn, subtract the first-order current difference from the second-order current difference from the first-order current difference, and complete the difference calculation for adjacent first-order current differences in turn. Arrange all the calculated differences in the original order to form a complete second-order current difference sequence, and directly determine this second-order current difference sequence as the second-order current difference sequence.
[0024] Starting from the first sampling time position of the second-order difference current sequence, each sampling time in the sequence is traversed one by one along the direction of gradually increasing time value. For each current sampling time, the corresponding numerical symbol is extracted, and the numerical symbol corresponding to the previous sampling time that is immediately adjacent to the current sampling time is extracted. The two numerical symbols are directly compared to determine whether the two numerical symbols are different symbols.
[0025] When the numerical symbols of two adjacent sampling times are determined to be different after point-by-point comparison, the complete position interval formed by these two adjacent sampling times is accurately recorded. Within this position interval, the symbol of the previous sampling time is taken as the starting point of the change, and the symbol of the next sampling time is taken as the ending point of the change. The specific direction of the change from the starting point symbol to the ending point symbol is clearly determined, and this change direction is directly determined as the symbol change direction corresponding to this position interval.
[0026] Step 2: Using the sampling moment when the direction of the sign change reverses as the starting point, construct a non-integer cycle prediction window for the ultra-high voltage power supply line. Within the non-integer cycle prediction window, when the number of alternations of the positive and negative signs of the second-order difference of the current exceeds a preset alternation threshold, generate a first-level fault prediction signal. In this embodiment of the invention, the process of constructing a non-integer cycle prediction window for the ultra-high voltage power line, starting from the sampling moment when the direction of the sign change reverses, is as follows: The current sampling moment, in which the direction of the sign change changes from positive to negative or from negative to positive from the previous sampling moment to the current sampling moment, is determined as the sign reversal moment. Using the sign reversal moment as the starting point of the window, a preset number of sampling moments are continuously intercepted along the time increment direction, so that the length of the intercepted sampling interval is less than one-quarter of the power frequency cycle. The time interval between the sign reversal time and the last truncation sampling time is defined as the non-integer cycle prediction window of the ultra-high voltage power supply line.
[0027] Within the non-integer cycle prediction window, when the number of alternations of the positive and negative signs of the second-order current difference exceeds a preset alternation threshold, a first-level fault prediction signal is generated, and the process is as follows: Following the increasing time sequence within the non-integer period prediction window, the numerical sign corresponding to each sampling moment in the second-order difference sequence of the current is extracted one by one, and the numerical sign is either positive or negative. The sign of the current sampling time is compared with that of the previous adjacent sampling time. If they are different, it is determined that a sign alternation has occurred, and the current value of the alternation counter is incremented by one until all sampling times within the non-integer period prediction window have been traversed. The final value of the alternation counter is read. If the final value is greater than the preset alternation threshold, a first-level fault prediction signal is output as a trigger pulse to start the subsequent fault identification process.
[0028] Check the sampling time corresponding to each symbol change direction one by one. When the symbol change direction from the previous sampling time to the current sampling time changes from positive to negative or from negative to positive, mark the current sampling time directly and determine it as the symbol reversal time.
[0029] Using the determined sign reversal time as the starting position of the window, a fixed number of sampling times are selected sequentially from this starting position in the direction of increasing time, ensuring that the total time length covered by all selected sampling times is less than one-quarter of the power frequency cycle.
[0030] The complete time range covered from the sign inversion moment to the last selected sampling moment is formally defined as the non-integer cycle prediction window corresponding to the ultra-high voltage power supply line.
[0031] Following the order in which the time within the non-integer period prediction window gradually increases, the numerical symbols corresponding to each sampling moment in the second-order difference sequence of the current are extracted one by one. The extracted numerical symbols only include two types: positive and negative.
[0032] For each sampling moment within the non-integer period prediction window, the numerical sign of the current sampling moment is directly compared with the numerical sign of the immediately preceding sampling moment. When the two numerical signs are inconsistent, it is determined that a sign alternation has been completed. At the same time, the value recorded in the alternation counter is increased by one unit. The comparison and counting operations are repeated in chronological order until all sampling moments within the non-integer period prediction window have been compared and counted.
[0033] Read the final value recorded by the alternation counter, compare the final value with the preset alternation threshold, and when the final value is greater than the preset alternation threshold, directly output the first-level fault prediction signal, which serves as the start signal to trigger the subsequent fault identification process.
[0034] Regarding the setting of the aforementioned alternation threshold, in this embodiment, it is determined based on the following principle: Under normal steady-state operating conditions, the sign change of the second-order differential current sequence is mainly caused by sampling noise and minor load fluctuations. By collecting current data from the line under typical load conditions for at least 100 power frequency cycles, the maximum value of the number of second-order differential sign alternations within the non-integer cycle prediction window is statistically analyzed. Take alternating thresholds ,in For safety margin, the prediction window length in this embodiment is taken from 20 sampling points, based on statistical analysis under normal operating conditions. =3, take =2, therefore, an alternating threshold is set. =5.
[0035] Step 3: In response to the first-level fault prediction signal, the instantaneous current value within the non-integer cycle prediction window is mapped to the two-dimensional phase space, and the elliptic fitting eccentricity of the formed current-current differential phase trajectory and the phase trajectory jump characteristic value between two adjacent sampling times are calculated. In this embodiment of the invention, the process of mapping the instantaneous current value within the non-integer cycle prediction window to a two-dimensional phase space in response to the first-level fault prediction signal is as follows: In response to the first-level fault prediction signal, the instantaneous current values arranged in chronological order within the non-integer cycle prediction window are locked. The difference between the instantaneous current values of two adjacent sampling times within the non-integer cycle prediction window is taken as the current difference component at each sampling time. Using the instantaneous current value at each sampling moment as the first dimension coordinate of the two-dimensional phase space, and the current differential component corresponding to each sampling moment as the second dimension coordinate of the two-dimensional phase space, a two-dimensional phase point is generated for each sampling moment within the non-integer period prediction window. Based on the two-dimensional phase points, the current-current differential phase trajectory of the ultra-high voltage power supply line is generated.
[0036] The process of generating the current-current differential phase trajectory of the ultra-high voltage power supply line based on the two-dimensional phase points is as follows: The two-dimensional phase points are arranged into a phase point sequence according to the increasing order of the sampling time. Starting from the first phase point in the phase point sequence, adjacent phase points are connected sequentially with straight line segments to obtain the phase trajectory curve of the ultra-high voltage power supply line; The coordinates of all phase points in the phase trajectory curve and the connection relationship between adjacent phase points are collectively defined as the current-current differential phase trajectory of the ultra-high voltage power supply line.
[0037] The process for calculating the eccentricity of the elliptic fitting of the formed current-current differential phase trajectory is as follows: All two-dimensional phase points in the current-current differential phase trajectory are obtained. Each two-dimensional phase point is composed of the instantaneous current value as the first dimension coordinate and the current differential component as the second dimension coordinate. Substitute all the two-dimensional phase points into the general equation of an ellipse, and use the least squares fitting method to obtain the fitted ellipse of the current-current differential phase trajectory. The lengths of the major and minor axes of the ellipse are analyzed from the coefficients of the fitted ellipse equation. The square root of the difference between the squares of the major and minor semi-axis lengths is divided by the major semi-axis length to obtain the elliptic fitting eccentricity of the current-current difference phase trajectory.
[0038] Regarding the setting of the aforementioned eccentricity threshold, in this embodiment, it is determined based on the following principles: Under normal operating conditions, the current-current differential phase trajectory is approximately a regular ellipse with a small eccentricity; during a short-circuit fault, the phase trajectory is distorted, and the eccentricity increases significantly. Phase trajectory data of the line under normal operating conditions and typical short-circuit fault conditions are obtained through simulation, and the ellipse fitting eccentricity is calculated respectively. Through extensive simulation statistics, the eccentricity distribution range under normal operating conditions is [0.05, 0.25], and the eccentricity distribution range under short-circuit fault conditions is [0.45, 0.85]. Taking the critical value and leaving a certain margin, the eccentricity threshold is set to 0.35.
[0039] The formula for calculating the phase trajectory jump characteristic value between two adjacent sampling times of the current-current differential phase trajectory is as follows: ; In the formula, This indicates the first current-current differential phase trajectory. The sampling time and the first Phase trajectory jump characteristic values between sampling times Indicates the first The instantaneous current value at each sampling time. Indicates the first The instantaneous current value at each sampling time. Indicates the first The current difference components at each sampling time. Indicates the first The current difference components at each sampling time. This represents the preset weighting coefficients. This indicates the preset reference current value.
[0040] Upon receiving the first-level fault prediction signal, all instantaneous current values that have been collected within the non-integer cycle prediction window are immediately locked and saved. These instantaneous current values are arranged in complete order according to the actual sampling time sequence. After locking, there will be no numerical modification, order change, data loss, or additional addition, ensuring that the data used for subsequent calculations is completely authentic and stable.
[0041] Within the entire sampling time range covered by the non-integer cycle prediction window, two adjacent sampling times are selected one by one in strict order from first to last, and the instantaneous current value corresponding to the earlier sampling time is subtracted from the instantaneous current value corresponding to the later sampling time to accurately obtain the numerical difference between the two. This calculated difference is directly determined as the current difference component corresponding to the sampling time currently involved in the calculation. This calculation is performed for each sampling time one by one and the results are recorded.
[0042] For each independent sampling moment within the non-integer period prediction window, a unique two-dimensional coordinate is constructed. The instantaneous current value obtained in the actual sampling at that sampling moment is used as the specific value of the first dimension coordinate in the two-dimensional phase space. The current difference component calculated by the adjacent current difference at the same sampling moment is used as the specific value of the second dimension coordinate in the two-dimensional phase space. Based on the two sets of accurate coordinate values, a unique two-dimensional phase point corresponding to that sampling moment is generated.
[0043] All two-dimensional phase points generated corresponding to all sampling times within the non-integer cycle prediction window are arranged sequentially in strict accordance with the increasing order of sampling time, ensuring that the order of phase points is completely consistent with the actual current sampling time order, so that all phase points form a continuous and orderly phase point sequence.
[0044] Starting from the first phase point in the phase point sequence, strictly following the established order of the phase point sequence, each adjacent phase point is steadily connected with a straight and uninterrupted line segment. This process continues until the last phase point is connected, ultimately forming the phase trajectory curve of the ultra-high voltage power line that fully reflects the dynamic trend of current changes.
[0045] The specific coordinate information of all two-dimensional phase points contained in the phase trajectory curve, as well as all the corresponding relationships between adjacent phase points connected by straight line segments, are fully integrated, and all the integrated information is uniformly defined as the current-current differential phase trajectory of the ultra-high voltage power supply line.
[0046] All two-dimensional phase points contained in the current-current differential phase trajectory are fully extracted. Each extracted two-dimensional phase point retains its original structure, with the instantaneous current value forming the first dimension and the current differential component forming the second dimension, without any coordinate adjustment or data modification.
[0047] The coordinate data of all extracted two-dimensional phase points are substituted into the conventional expression equation of the ellipse one by one. The least squares fitting method is used to perform overall matching calculations on all phase points. The parameters of the equation are continuously adjusted to make the fitted graph as close as possible to the actual distribution of all phase points, reduce the overall deviation, and finally generate a fitted ellipse that highly matches the current-current differential phase trajectory.
[0048] The coefficients of each term in the equation corresponding to the fitted ellipse are analyzed one by one. Key parameters that can accurately characterize the size and shape of the ellipse are separated from the coefficient information. Based on these key parameters, the specific lengths of the major and minor axes of the fitted ellipse are further accurately determined.
[0049] First, calculate the square values of the lengths of the major and minor semi-axis respectively. Then, subtract the square value of the minor semi-axis from the square value of the major semi-axis to obtain the accurate square difference. Perform standard square root processing on this square difference to obtain the corresponding square root value. Divide this square root value by the length of the major semi-axis. After calculation, the eccentricity of the ellipse fitting corresponding to the current-current difference phase trajectory is finally obtained.
[0050] Regarding the setting of the aforementioned weighting coefficient and reference current, in this embodiment: the weighting coefficient is used to adjust the weight of the area variation component in the phase trajectory jump characteristic value, and its value ranges from 0.5 to 2.0. In this embodiment, the weighting coefficient is set to 1.0. The reference current is taken as the rated current value of the line. In this embodiment, the rated current of the ultra-high voltage line is 2000A.
[0051] On the current-current differential phase trajectory, the phase point positions corresponding to two adjacent sampling times are selected sequentially, and the spatial distance and positional change amplitude between the two points are calculated respectively. Combined with the coordinate changes and displacement characteristics of the phase points, a comprehensive calculation is completed to obtain the unique phase trajectory jump characteristic value corresponding to two adjacent sampling times. The radical operation is used to measure the linear interval between two adjacent phase points in the two-dimensional phase plane. The interval is calculated step-by-step using the data differences between the instantaneous current dimension and the current differential component dimension, accurately reflecting the positional movement distance of adjacent phase points in phase space. The fractional operation is used to characterize the area change amplitude of the closed region formed by the combination of two sets of coordinates. The numerical calculation is completed by cross-multiplying two sets of instantaneous currents and two sets of current differential components and then obtaining the difference, supplementing the information on trajectory shape changes caused by phase point deflection. Weighting coefficients are used to limit the proportion of the area change component in the overall characteristic value, and, together with the reference current, complete the standardization and conversion of the area component, eliminating the interference effect of reference current value differences on the component calculation results. The phase trajectory jump characteristic value obtained after integrating all the calculations can simultaneously include two types of data information: the magnitude of phase point spatial displacement and the phase point coordinate deflection change, accurately quantifying the magnitude of morphological changes in the current-current difference phase trajectory within adjacent sampling intervals.
[0052] Among them, the The instantaneous current value corresponding to each sampling moment is taken from the actual current measurement data collected on-site at the corresponding sampling point within the non-integer cycle prediction window. This data originates from the original sampling results retained from previous continuous sampling of ultra-high voltage power lines within the power frequency cycle. The instantaneous current value corresponding to each sampling moment is taken from the actual measured current data collected from the subsequent sampling points immediately adjacent to the previous sampling point within the non-integer cycle prediction window, and the data is collected and saved synchronously with the synchronous sampling operation. The current differential component corresponding to each sampling moment is generated by the difference calculation between the instantaneous current of the corresponding sampling point and the preceding adjacent point, and originates from the predetermined calculation results obtained by batch solving the current differential components of each sampling point. The current differential component corresponding to each sampling moment is generated by calculating the difference between the instantaneous current at the corresponding sampling point and the preceding adjacent point, and it originates from the same source as the data obtained from the unified calculation process of the current differential component of the full prediction window. The weighting coefficient is fixedly assigned based on the inherent operating parameters of the line before the entire fault identification process is officially started, and it will not be changed throughout the process after it is set. The reference current value is fixed in advance based on the factory rated operating indicators of the ultra-high voltage power supply line, and the value remains constant throughout the entire fault determination cycle.
[0053] When the line is operating normally, the instantaneous current and corresponding current differential components at adjacent sampling points change only slightly. The phase interval calculation and area variation calculation results remain at low levels, and the final phase trajectory jump characteristic value is consistently within a small range. When a short-circuit fault occurs, the instantaneous current and corresponding current differential components obtained from the instantaneous sampling change abruptly. The coordinate difference and cross-product difference between adjacent points increase significantly, and the phase interval calculation and area variation calculation results increase synchronously. The final phase trajectory jump characteristic value shows a significant upward trend. The weighting coefficient and reference current remain fixed during the full sampling calculation phase and do not change with the sampling data. They only constrain the fluctuation range of the area component and cannot alter the inherent trend of a significant increase in characteristic values after a fault occurs.
[0054] Step 4: When the eccentricity of the ellipse fitting is greater than the preset eccentricity threshold and the cumulative sum of the phase trajectory jump characteristic values exceeds the preset characteristic threshold, the ultra-high voltage power line is determined to be a short circuit fault.
[0055] In this embodiment of the invention, when the eccentricity of the ellipse fitting is greater than a preset eccentricity threshold and the cumulative sum of the phase trajectory jump feature values exceeds a preset feature threshold, the ultra-high voltage power line is determined to have a short-circuit fault. The process is as follows: The eccentricity of the ellipse fitting is compared with a preset eccentricity threshold. If the eccentricity of the ellipse fitting is less than or equal to the eccentricity threshold, the current determination is terminated and the first-level fault prediction signal is cleared. If the eccentricity of the ellipse fitting is greater than the eccentricity threshold, then starting from the phase trajectory jump feature value corresponding to the first sampling time in the non-integer period prediction window, the phase trajectory jump feature values are accumulated sequentially to obtain the feature accumulation value of the phase trajectory jump feature values. The feature accumulation value is compared with a preset feature threshold. If the feature accumulation value is less than or equal to the feature threshold, the determination is terminated and the first-level fault prediction signal is cleared. If the cumulative sum is greater than the characteristic threshold, a short-circuit fault confirmation signal is output, and the ultra-high voltage power line is determined to be short-circuit faulted.
[0056] Regarding the setting of the aforementioned feature threshold, in this embodiment, it is determined based on the following principles: Under normal operating conditions, the changes in current and current difference between adjacent sampling times are gradual, and the cumulative sum of phase trajectory jump feature values is at a low level; during short-circuit faults, sudden current changes lead to a significant increase in the cumulative sum. The cumulative sum and distribution of phase trajectory jump feature values within the prediction window under normal and short-circuit fault conditions are obtained through simulation, and the lower boundary of the non-overlapping region of the two distributions is used as the feature threshold. Simulation statistics show that the upper limit of the feature cumulative value distribution under normal operating conditions is approximately 0.15, and the lower limit of the feature cumulative value distribution under short-circuit fault conditions is approximately 0.45; therefore, the feature threshold is set to 0.30.
[0057] Extract the eccentricity of the ellipse fitting determined by the fitting calculation, and compare this fixed value with the eccentricity threshold set in advance according to the line operation standard. If the actual value of the ellipse fitting eccentricity is less than the eccentricity threshold or the two values are completely consistent after comparison, immediately shut down the fault judgment process for the ultra-high voltage power line, and simultaneously cancel the identification process started after receiving the first-level fault prediction signal, and completely clear the first-level fault prediction signal that is already in an effective trigger state.
[0058] After comparing the complete values of the ellipse fitting eccentricity and the eccentricity threshold, and determining that the actual value of the ellipse fitting eccentricity is higher than the eccentricity threshold, the phase trajectory jump feature value calculated at the earliest sampling time within the non-integer period prediction window is selected as the initial base value for the cumulative calculation. According to the order of the sampling times within the non-integer period prediction window, the phase trajectory jump feature values corresponding to each subsequent sampling time are extracted one by one, and the extracted feature values are successively superimposed on the current cumulative result. After traversing all the feature data corresponding to all sampling points within the prediction window, all superimposed results are integrated to generate the feature accumulation value of the complete phase trajectory jump feature value.
[0059] After retrieving all feature values and accumulating them, the accumulated feature value is compared with the feature threshold set in advance based on the line fault characteristics. If the comparison result shows that the actual value of the accumulated feature value is less than the feature threshold or the two values are equal, all fault determination related operations in this round are immediately terminated, the process lock state caused by the first-level fault prediction signal is released, and the prediction signal is completely cleared.
[0060] After comparing the accumulated feature value with the preset feature threshold, if the actual value of the accumulated feature value exceeds the feature threshold, a dedicated short-circuit fault confirmation signal is sent out. Based on the judgment criteria obtained from both threshold comparisons meeting the standards, it is finally determined that a short-circuit fault has occurred in the currently monitored ultra-high voltage power line.
[0061] As can be seen from the above embodiments, the rapid identification method for ultra-high voltage power supply short-circuit faults provided by the present invention achieves rapid initial fault prediction by constructing a non-integer cycle prediction window and combining it with the alternating characteristics of the second-order differential sign of the current. Then, it uses two-dimensional phase space mapping, ellipse fitting eccentricity and phase trajectory jump characteristic values for accurate judgment. This significantly shortens the identification cycle of ultra-high voltage power supply short-circuit faults, greatly improves the fault response speed, and effectively reduces the probability of misjudgment and missed judgment caused by external interference. It balances the efficiency and accuracy of fault identification and can reliably and quickly achieve accurate identification of ultra-high voltage power supply line short-circuit faults, ensuring the stability and safety of ultra-high voltage power system operation.
[0062] To verify the effectiveness of the technical solution in this application, a 500kV ultra-high voltage transmission line model was built in the PSCAD / EMTDC simulation platform. The line parameters are as follows: positive sequence resistance is 0.02Ω / km, positive sequence reactance is 0.28Ω / km, line length is 300km, rated current is 2000A, and sampling frequency is 10kHz. The non-integer cycle prediction window length is 20 sampling points. The alternation threshold is 5, the eccentricity threshold is 0.35, and the characteristic threshold is 0.30. The weighting coefficient α = 1.0, and the reference current... The rated current of the line is 2000A.
[0063] To maintain dimensional consistency with the feature threshold, this embodiment writes the phase trajectory jump feature value in the following per-unit form: ; The feature cumulative sum is the sum of all adjacent sampling segments within the prediction window. The sum of When N=20, 19 segments should be accumulated; It is a dimensionless quantity, and is in the same dimensional system as the feature threshold of 0.30.
[0064] Simulate the following typical working conditions respectively: Operating Condition 1: Normal Operation The line operated in steady state under rated load, continuously collecting current data for 100 power frequency cycles. Statistical results showed that the maximum number of second-order difference sign alternations within the non-integer cycle prediction window was 3, which did not exceed the alternation threshold of 5; the average elliptic fitting eccentricity was 0.12, lower than the eccentricity threshold of 0.35; and the average cumulative sum of phase trajectory jump characteristic values calculated using the per-unit formula on the continuous prediction window was 0.08, lower than the characteristic threshold of 0.30. None of the indicators triggered any thresholds, and the system was determined to be operating normally, with no misjudgments occurring within the 100 power frequency cycles.
[0065] Operating Condition 2: Three-phase short circuit fault A three-phase metallic short-circuit fault occurred at t=0.205s. After the fault occurred, the system detected second-order differential sign reversal at t=0.206s and constructed a prediction window. The number of second-order differential sign alternations within the prediction window was 8, triggering the first-level fault prediction signal. Through phase space mapping and elliptic fitting calculations, the eccentricity was found to be 0.62, and the cumulative sum of phase trajectory jump characteristic values was 0.88. The system output a short-circuit fault confirmation signal at t=0.208s. The total time from fault occurrence to fault confirmation was approximately 3ms.
[0066] Operating Condition 3: Sudden Load Change The simulation showed a 50% sudden increase in load. The results showed that the number of second-order differential symbol alternations within the prediction window was 4, and the first-level fault prediction signal was not triggered. The system judged it as a normal disturbance, with no misjudgment.
[0067] Operating Condition 4: Harmonic Interference Simulated harmonic interference conditions. The results show that the number of second-order difference symbol alternations within the prediction window is 4, and the first-level fault prediction signal is not triggered. Even if the alternation threshold is manually lowered to trigger the prediction signal, the ellipse fitting eccentricity is 0.18, and the cumulative sum of the jump characteristics obtained by accumulating the per-unit values of 19 consecutive prediction window segments is 0.11, which is judged as a normal operating condition, effectively avoiding misjudgments caused by harmonic interference.
[0068] Comparison with existing technologies A comparison was made between the proposed method and a traditional overcurrent protection method based on whole-cycle sampling: Under the same three-phase short-circuit fault conditions, the traditional method requires at least one complete power frequency cycle to make a judgment, while the proposed method can complete the judgment in about 3ms after the fault occurs, shortening the identification cycle by about 85%. In 100 Monte Carlo simulation experiments, the proposed method had an average identification time of 2.8ms, a false positive rate of 0%, and a false negative rate of 0%; the traditional method had an average identification time of 21.5ms and a false negative rate of 3% when dealing with high-resistance grounding faults.
[0069] like Figure 2 The diagram shown is a functional block diagram of a rapid short-circuit fault identification system 100 for ultra-high voltage power supplies provided in an embodiment of the present invention, including a differential feature extraction module 101, a prediction window construction module 102, a phase trajectory feature calculation module 103, and a short-circuit fault determination module 104.
[0070] In this embodiment, the functions of each module are as follows: The differential feature extraction module 101 is used to extract the sign change direction of the second-order difference sequence of the current obtained by the second-order difference operation of the instantaneous current value of the ultra-high voltage power line within the current power frequency cycle, at the zero crossing point. The prediction window construction module 102 is used to construct a non-integer cycle prediction window for the ultra-high voltage power line, starting from the sampling time when the direction of the sign change reverses. Within the non-integer cycle prediction window, when the number of alternations of the positive and negative signs of the second-order difference of the current exceeds a preset alternation threshold, a first-level fault prediction signal is generated. The phase trajectory feature calculation module 103 is used to respond to the first-level fault prediction signal, map the instantaneous current value in the non-integer cycle prediction window to the two-dimensional phase space, and calculate the elliptic fitting eccentricity of the current-current difference phase trajectory and the phase trajectory jump feature value between two adjacent sampling times. The short-circuit fault determination module 104 is used to determine that the ultra-high voltage power line is short-circuit fault when the eccentricity of the ellipse fitting is greater than a preset eccentricity threshold and the cumulative sum of the phase trajectory jump characteristic values exceeds a preset characteristic threshold.
[0071] As can be seen from the above embodiments, the ultra-high voltage power supply short-circuit fault rapid identification system provided by the present invention achieves rapid initial fault prediction by constructing a non-integer cycle prediction window and combining it with the alternating characteristics of the second-order differential sign of the current. Then, it uses two-dimensional phase space mapping, ellipse fitting eccentricity and phase trajectory jump characteristic values for accurate judgment. This significantly shortens the identification cycle of ultra-high voltage power supply short-circuit faults, greatly improves the fault response speed, and effectively reduces the probability of misjudgment and missed judgment caused by external interference. It balances the efficiency and accuracy of fault identification and can reliably and quickly achieve accurate identification of ultra-high voltage power supply line short-circuit faults, ensuring the stability and safety of ultra-high voltage power system operation.
[0072] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.
[0073] Figure 3 This is a graph showing the cumulative number of sign alternations within a non-integer cycle prediction window under a three-phase short-circuit fault condition. The horizontal axis represents the prediction window sampling number k, and the vertical axis represents the cumulative number of alternations. Light-colored backgrounds indicate a 20-point prediction window; solid lines represent the stepped curve of the cumulative number of sign alternations; horizontal dashed lines represent the alternation threshold; and solid dots indicate the position where the threshold is first exceeded. Figure 3 As can be seen, the cumulative number of signs increases gradually from 0 at the beginning of the window, reaching 6 for the first time at k=6 and exceeding the alternation threshold of 5. It then continues to increase, stabilizing at 8 after k=7.5 until k=20. This figure is used to fully illustrate the dynamic change process of the cumulative number of signs alternating throughout the prediction window under this fault condition and the final statistical result at the end of the window.
[0074] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.
Claims
1. A method for rapid identification of short-circuit faults in ultra-high voltage power supplies, characterized in that, The method includes: Step 1: Within the current power frequency cycle, extract the instantaneous current value of the ultra-high voltage power supply line and obtain the second-order difference sequence of the current after second-order difference operation, and determine the sign change direction at the zero crossing point. Step 2: Using the sampling moment when the direction of the sign change reverses as the starting point, construct a non-integer cycle prediction window for the ultra-high voltage power supply line. Within the non-integer cycle prediction window, when the number of alternations of the positive and negative signs of the second-order difference of the current exceeds a preset alternation threshold, generate a first-level fault prediction signal. Step 3: In response to the first-level fault prediction signal, the instantaneous current value within the non-integer cycle prediction window is mapped to the two-dimensional phase space, and the elliptic fitting eccentricity of the formed current-current differential phase trajectory and the phase trajectory jump characteristic value between two adjacent sampling times are calculated. Step 4: When the eccentricity of the ellipse fitting is greater than the preset eccentricity threshold and the cumulative sum of the phase trajectory jump characteristic values exceeds the preset characteristic threshold, the ultra-high voltage power line is determined to be a short circuit fault.
2. The method for rapid identification of short-circuit faults in ultra-high voltage power supplies as described in claim 1, characterized in that, The sign change direction of the second-order difference sequence of the current obtained by extracting the instantaneous current value of the ultra-high voltage power line and performing a second-order difference operation is as follows: Based on the instantaneous current values continuously sampled by the ultra-high voltage power line within the current power frequency cycle, the instantaneous current difference between two adjacent sampling times is calculated sequentially to obtain a first-order current difference sequence. The difference between two adjacent first-order current differences in the first-order current difference sequence is calculated sequentially to obtain the second-order current difference sequence, and the second-order current difference sequence is used as the second-order current difference sequence. The second-order difference sequence of current is scanned point by point along the time-increasing direction, and the sign of the current sampling time is compared with the sign of the adjacent previous sampling time. When the signs of the values at two adjacent sampling times are different, the position interval formed by the two adjacent sampling times is recorded, and the direction of the jump from the sign of the previous point to the sign of the next point within the position interval is taken as the sign change direction of the position interval.
3. The method for rapid identification of short-circuit faults in ultra-high voltage power supplies as described in claim 1, characterized in that, The process of constructing a non-integer cycle prediction window for the ultra-high voltage power line, starting from the sampling moment when the direction of the sign change reverses, is as follows: The current sampling moment, in which the direction of the sign change changes from positive to negative or from negative to positive from the previous sampling moment to the current sampling moment, is determined as the sign reversal moment. Using the sign reversal moment as the starting point of the window, a preset number of sampling moments are continuously intercepted along the time increment direction, so that the length of the intercepted sampling interval is less than one-quarter of the power frequency cycle. The time interval between the sign reversal time and the last truncation sampling time is defined as the non-integer cycle prediction window of the ultra-high voltage power supply line.
4. The method for rapid identification of short-circuit faults in ultra-high voltage power supplies as described in claim 3, characterized in that, Within the non-integer cycle prediction window, when the number of alternations of the positive and negative signs of the second-order current difference exceeds a preset alternation threshold, a first-level fault prediction signal is generated, and the process is as follows: Following the increasing time sequence within the non-integer period prediction window, the numerical sign corresponding to each sampling moment in the second-order difference sequence of the current is extracted one by one, and the numerical sign is either positive or negative. The sign of the current sampling time is compared with that of the previous adjacent sampling time. If they are different, it is determined that a sign alternation has occurred, and the current value of the alternation counter is incremented by one until all sampling times within the non-integer period prediction window have been traversed. The final value of the alternation counter is read. If the final value is greater than the preset alternation threshold, a first-level fault prediction signal is output as a trigger pulse to start the subsequent fault identification process.
5. The method for rapid identification of short-circuit faults in ultra-high voltage power supplies as described in claim 1, characterized in that, In response to the first-level fault prediction signal, the instantaneous current value within the non-integer cycle prediction window is mapped to a two-dimensional phase space, and the process is as follows: In response to the first-level fault prediction signal, the instantaneous current values arranged in chronological order within the non-integer cycle prediction window are locked. The difference between the instantaneous current values of two adjacent sampling times within the non-integer cycle prediction window is taken as the current difference component at each sampling time. Using the instantaneous current value at each sampling moment as the first dimension coordinate of the two-dimensional phase space, and the current differential component corresponding to each sampling moment as the second dimension coordinate of the two-dimensional phase space, a two-dimensional phase point is generated for each sampling moment within the non-integer period prediction window. Based on the two-dimensional phase points, the current-current differential phase trajectory of the ultra-high voltage power supply line is generated.
6. The method for rapid identification of short-circuit faults in ultra-high voltage power supplies as described in claim 5, characterized in that, The process of generating the current-current differential phase trajectory of the ultra-high voltage power supply line based on the two-dimensional phase points is as follows: The two-dimensional phase points are arranged into a phase point sequence according to the increasing order of the sampling time. Starting from the first phase point in the phase point sequence, adjacent phase points are connected sequentially with straight line segments to obtain the phase trajectory curve of the ultra-high voltage power supply line; The coordinates of all phase points in the phase trajectory curve and the connection relationship between adjacent phase points are collectively defined as the current-current differential phase trajectory of the ultra-high voltage power supply line.
7. The method for rapid identification of short-circuit faults in ultra-high voltage power supplies as described in claim 6, characterized in that, The process for calculating the eccentricity of the elliptic fitting of the formed current-current differential phase trajectory is as follows: All two-dimensional phase points in the current-current differential phase trajectory are obtained. Each two-dimensional phase point is composed of the instantaneous current value as the first dimension coordinate and the current differential component as the second dimension coordinate. Substitute all the two-dimensional phase points into the general equation of an ellipse, and use the least squares fitting method to obtain the fitted ellipse of the current-current differential phase trajectory. The lengths of the major and minor axes of the ellipse are analyzed from the coefficients of the fitted ellipse equation. The square root of the difference between the squares of the major and minor semi-axis lengths is divided by the major semi-axis length to obtain the elliptic fitting eccentricity of the current-current difference phase trajectory.
8. The method for rapid identification of short-circuit faults in ultra-high voltage power supplies as described in claim 7, characterized in that, The formula for calculating the phase trajectory jump characteristic value between two adjacent sampling times of the current-current differential phase trajectory is as follows: ; In the formula, This indicates the first current-current differential phase trajectory. The sampling time and the first Phase trajectory jump characteristic values between sampling times Indicates the first The instantaneous current value at each sampling time. Indicates the first The instantaneous current value at each sampling time. Indicates the first The current difference components at each sampling time. Indicates the first The current difference components at each sampling time. This represents the preset weighting coefficients. This indicates the preset reference current value.
9. The method for rapid identification of short-circuit faults in ultra-high voltage power supplies as described in claim 1, characterized in that, When the eccentricity of the ellipse fitting is greater than a preset eccentricity threshold and the cumulative sum of the phase trajectory jump characteristic values exceeds a preset characteristic threshold, the ultra-high voltage power supply line is determined to have a short circuit fault. The process is as follows: The eccentricity of the ellipse fitting is compared with a preset eccentricity threshold. If the eccentricity of the ellipse fitting is less than or equal to the eccentricity threshold, the current determination is terminated and the first-level fault prediction signal is cleared. If the eccentricity of the ellipse fitting is greater than the eccentricity threshold, then starting from the phase trajectory jump feature value corresponding to the first sampling time in the non-integer period prediction window, the phase trajectory jump feature values are accumulated sequentially to obtain the feature accumulation value of the phase trajectory jump feature values. The feature accumulation value is compared with a preset feature threshold. If the feature accumulation value is less than or equal to the feature threshold, the determination is terminated and the first-level fault prediction signal is cleared. If the cumulative sum is greater than the characteristic threshold, a short-circuit fault confirmation signal is output, and the ultra-high voltage power line is determined to be short-circuit faulted.
10. A rapid identification system for short-circuit faults in ultra-high voltage power supplies, characterized in that, The system is used to implement the rapid identification method for short-circuit faults in ultra-high voltage power supplies according to any one of claims 1-9, the system comprising: The differential feature extraction module is used to extract the sign change direction of the second-order differential sequence of the current obtained by second-order differential operation of the instantaneous current value of the ultra-high voltage power line within the current power frequency cycle. The prediction window construction module is used to construct a non-integer cycle prediction window for the ultra-high voltage power line, starting from the sampling time when the direction of the sign change reverses. Within the non-integer cycle prediction window, when the number of alternations of the positive and negative signs of the second-order difference of the current exceeds a preset alternation threshold, a first-level fault prediction signal is generated. The phase trajectory feature calculation module is used to respond to the first-level fault prediction signal, map the instantaneous current value in the non-integer cycle prediction window to a two-dimensional phase space, and calculate the elliptic fitting eccentricity of the current-current differential phase trajectory and the phase trajectory jump feature value between two adjacent sampling times. The short-circuit fault determination module is used to determine that the ultra-high voltage power line is short-circuit faulted when the eccentricity of the ellipse fitting is greater than a preset eccentricity threshold and the cumulative sum of the phase trajectory jump characteristic values exceeds a preset characteristic threshold.
Citation Information
Patent Citations
Power distribution network successive grounding fault identification method based on phase current break variable characteristics
CN122136750A
Ground fault intelligent early warning method and system for high-voltage cable
CN122525295A
