Voltage sag source positioning method based on measured data

Abstract

The invention relates to a voltage sag source positioning method based on measured data. The method comprises the following steps: S1, acquiring monitoring data; s2, performing sag-fault matching based on time sequence matching; the voltage sag of one monitoring point is matched with the same bus fault and the external fault according to the time sequence of events, and the time difference between each pair of voltage sag and fault in the matching result is calculated; s3, dynamically adjusting a window to calculate a quantile detection abnormal value; s4, performing secondary matching and iteration termination based on the reduced data set; and S5, repeating the steps S2-S4 for other monitoring points to obtain all matching data in the region, and realizing accurate positioning of the voltage sag source. According to the method, accurate and efficient fault positioning and matching can be carried out on the voltage sag event.

Description

Technical Field

[0001] The present invention relates to the technical field of voltage sag source location, and particularly relates to a voltage sag source location method based on measured data. Background Art

[0002] With the development of smart grid and industrial automation, voltage sag has become a key issue affecting the reliable operation of power systems. Although the amplitude of its millisecond-level sudden drop lasts for a short time, it may cause malfunction of sensitive electronic devices, breakdown of semiconductor components, and attenuation of the lifespan of precision instruments. To address these problems, the industry has taken a series of measures, such as using compensation devices like dynamic voltage restorers and energy storage devices. However, there are limitations such as lagging response speed, limited regulation performance, high investment cost, and fast lifespan attenuation. Therefore, even with countermeasures taken, it is still difficult to completely avoid the adverse effects brought by voltage sag in industrial production. Precise location of the sag source can narrow the scope of fault troubleshooting from the entire network to several fault points, improve the operation and maintenance efficiency, and at the same time provide data support for the self-healing control of smart grids.

[0003] Due to the continuous expansion of the scale of power systems and the increasing complexity of grid structures, the factors that may cause voltage sags in power systems have increased. Coupled with the access of distributed power sources, the power grid has become a multi-source loop structure, and fault currents flow bidirectionally. Voltage sags may be caused by faults in distributed power sources themselves or disturbances at the connection points, significantly increasing the difficulty of sag source location. With the increasing attention of power systems to power quality issues, various types of monitoring devices are widely deployed in power systems, which can obtain the waveforms and occurrence times of sags at different monitoring points, and further calculate characteristic quantities such as sag duration and sag amplitude. At the same time, when a fault occurs in the power system, the line and time of the fault occurrence will also be recorded.

[0004] In the current voltage sag source location method, traditional techniques generally locate by analyzing the flow of energy or changes in power grid parameters. For example, methods such as analyzing the change in disturbance power and energy flow direction, the change in voltage and current measurement values, and the polarity of the real part of the current are used. In recent years, some studies have improved traditional location methods through algorithms, such as location methods that combine wavelet singular entropy, positive sequence component phase difference, etc. with waveforms. However, such methods can only determine the relative position of the sag source, that is, whether it is upstream or downstream of the monitoring point, and cannot accurately locate the position of the sag source. The method based on artificial intelligence is trained through a large amount of sample data. Although it can accurately locate the position of the sag source, its calculation is complex, requiring high computing power, and has high requirements for data. Practical use depends on the large-scale deployment of monitoring equipment, while simulation analysis cannot reflect the changes in power grid parameters caused by the actual system operation, and the interpretability of the method based on artificial intelligence is poor. The traditional time-based location method has high requirements for the time synchronization of different monitoring points, high dependence on high-precision equipment, high equipment costs, installation and maintenance costs, and large data transmission and processing pressure, and is vulnerable to electromagnetic interference, resulting in measurement deviation. Moreover, in practical applications, it is difficult to ensure the time synchronization of each monitoring point, and with the aging of monitoring equipment and other reasons, the time offset will become larger and larger. Therefore, a method for accurately locating the voltage sag source in the case of time offset and continuous increase in the offset is needed. Summary of the Invention

[0005] The purpose of the present invention is to provide a voltage sag source location method based on measured data, which can accurately and efficiently perform fault location and matching for voltage sag events.

[0006] To achieve the above purpose, the technical solution adopted by the present invention is: a voltage sag source location method based on measured data, including the following steps:

[0007] Step S1: Obtain monitoring data; obtain the time data of the same bus fault, external bus fault, and voltage sag event of each monitoring point on the bus within a certain period of time in a region as the data source;

[0008] Step S2: Sag-fault matching based on time sequence matching; match the voltage sag of a monitoring point with the same bus fault and external fault according to the chronological order of event occurrence, and calculate the time difference between each pair of voltage sag and fault in the matching result;

[0009] Step S3: Dynamically adjust the window to calculate the quantile to detect outliers; first set the initial window size and density threshold, calculate the quantile in the window and generate a dynamic threshold, and then mark the time difference outliers based on the dynamic threshold; move the window, calculate the event density and dynamically adjust the window size, calculate the window quantile and generate a dynamic threshold to detect abnormal data; repeat the dynamic window adjustment until all data is traversed, and output the set of abnormal data points;

[0010] Step S4: Remove the abnormal data S' in the matching result from the original voltage sag dataset, and use the mis-matched fault dataset F' as a new fault dataset for secondary matching; after secondary matching, iteratively perform dynamic window adjustment, add the successfully matched results to the preliminary matching result dataset, and then perform detection of the time difference constant value. Keep iterating until no abnormal time difference data is detected, and obtain the matching result of this monitoring point;

[0011] Step S5: For the remaining monitoring points, repeat Steps S2 - S4 to obtain all the matching data within the region and achieve precise location of the voltage sag source.

[0012] Furthermore, in Step S1, for a region, different groups of data are matched separately according to the different positions of the voltage sag monitoring points. The same - bus fault is the fault corresponding to the 10kV bus under the substation where this monitoring point is located, that is, the same - bus fault will only cause the voltage sag event to be detected at this monitoring point. The off - bus fault is the fault corresponding to the 110kV bus, 220kV substation, and 500kV substation within this region, that is, the off - bus fault will cause the voltage sag event to be detected at all the monitoring points within this region.

[0013] Furthermore, Step S2 specifically includes the following steps:

[0014] Step S201: Unify the timestamp formats of different data sources;

[0015] Filter out short - time and intensive sags according to the sag amplitude and detection frequency; read the time of the voltage sag at a monitoring point within the region and sort it in ascending order S = {S1, S2, …, S n}; read the times of the same - bus faults and off - bus faults within the region, synthesize a dataset and sort it in ascending order to obtain the fault dataset F = {F1, F2, …, F m};

[0016] Step S202: Starting from the first fault, traverse the voltage sag data. If the first voltage sag is found to satisfy:

[0017] S j ∈S, |F1 - S j | ≤ T0

[0018] where F1 is the timestamp of the first fault, S j is the timestamp of the j - th voltage sag data, and T0 is the matching window size;

[0019] then the fault timestamp F1 and the voltage sag timestamp S jIf a match is found, stop further searching; otherwise, if no match is found, it is considered that the fault does not match, and proceed to match the next fault timestamp F2; if S j Fault data has been matched, and there are other voltage sag timestamps at this monitoring point that satisfy the following relationship:

[0020]

[0021] Then merge the two voltage sags into one voltage sag event, and take the voltage sag timestamp S j as the merged voltage sag timestamp; for the already matched fault and voltage sag data, they need to be removed from the original set and no longer participate in subsequent matching;

[0022] Step S203: After the preliminary matching is completed, combine each pair of matched fault and sag data in the matched results into a new data. If one fault matches multiple sags, there will be multiple data; for each matching result, calculate the time difference between the fault and the sag respectively. The formula is as follows:

[0023]

[0024] where, represents the timestamp of the fault in the i-th matching result, represents the timestamp of the voltage sag in the i-th matching result, d i represents the time difference between the fault and the voltage sag in the i-th matching result.

[0025] Furthermore, the formula for unifying the timestamp formats of different data sources is:

[0026]

[0027] where, T is the timestamp, Y is the year offset, y i is the number of days in the i-th year, M is the month offset, m j is the number of days in the j-th month, D is the number of days in the current month, and h, m, s are hours, minutes, and seconds respectively.

[0028] Furthermore, step S3 specifically includes the following steps:

[0029] Step S301: Input the time difference sequence data set D = {d1, d2,..., d i ,..., d n}, and initialize the window parameters, including the initial window size W0, the density threshold σ0, the maximum window size W max , and the initial window

[0030] Calculate the quantile threshold of the initial window. The threshold is calculated using the interquartile range: IQR = Q3 - Q1, where Q1 = P 25 (W1), Q3 = P 75 (W1), P 25 () represents the 25th percentile, and P 75 () represents the 75th percentile; Define the abnormal threshold range:

[0031]

[0032] where α is a preset coefficient;

[0033] Detect whether there are outliers in the initial window. If then mark d i as abnormal data;

[0034] Step S302: Slide the window backward, and add d i to the current window W i ={d i-W+1 , d i-W+2 , …, d i}, where W is the size of the current window; Define the event density as the standard deviation of the differences between adjacent data in the window, and adjust the window size according to the event density;

[0035] Adjust the window size according to the event density σ. When the time interval fluctuates greatly, expand the window to adapt to the data fluctuation. Conversely, shrink the window to improve the sensitivity; The window size adjustment rule is:

[0036]

[0037] Step S303: Calculate the quantile threshold based on the current window: IQR = Q3 - Q1, where Q1 = P 25 (W i ), Q3 = P 75 (W i ); Define the abnormal threshold range:

[0038]

[0039] where α is a preset coefficient;

[0040] Detect whether the current data point d i is an outlier. If then mark d i as abnormal data; Otherwise, retain it as normal data;

[0041] Step S304: Repeat steps S302 - S303 until all data is traversed, and output the index set A of abnormal data points = {i | d iMarked as abnormal}, output the faulty dataset F' = {F A1 , F A2 , …, F An} and the sag dataset S' = {S A1 , S A2 , …, S An}, where n is the size of the abnormal data index A.

[0042] Furthermore, adjust the window size according to the event density, and its implementation method is as follows:

[0043] Calculate the time interval of adjacent time difference data within the window W, and the formula is as follows:

[0044] ΔT = {Δt1, Δt2,..., Δt W-1}

[0045] where, Δt k = d i-W+k+1 - d i-W+k ;

[0046] Calculate the standard deviation σ of the time interval, and the formula is as follows:

[0047]

[0048] where, μ is the mean value of the time interval,

[0049] The present invention also provides a computer device, including: at least one processor, at least one memory, and computer program instructions stored in the memory, and when the computer program instructions are executed by the processor, the above method is implemented.

[0050] The present invention also provides a computer-readable storage medium, on which computer program instructions are stored, and when the computer program instructions are executed by a processor, the above method is implemented.

[0051] Compared with the prior art, the present invention has the following beneficial effects: The present invention proposes a method for locating voltage sag sources based on measured data. Considering that there are time deviations in the records of each monitoring point and the deviations will gradually increase, the method adopts the way of approximate time correlation, takes the time difference between the sag time and the fault time as the main criterion, and dynamically expands or contracts the analysis window by analyzing the time distribution density of events to adapt to data fluctuations, accurately detect abnormal data, and achieve accurate location of voltage sag sources. Different from traditional location methods and artificial intelligence-based location methods, the method for locating voltage sag sources based on measured data proposed by the present invention is simple in calculation, has simple requirements for data and does not require modification of existing monitoring equipment, supports asynchronous distributed node clocks and gradually increasing clock offsets, and can accurately match voltage sags and faults. Therefore, the present invention has strong practicability and broad application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 is a flowchart of the method implementation of an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0053] The present invention will be further described below in conjunction with the drawings and embodiments.

[0054] It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the present application. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which the present application belongs.

[0055] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0056] This embodiment provides a method for locating voltage sag sources based on measured data, which realizes the matching of voltage sags with faults on the same bus and faults outside the bus based on timestamp sorting, a sliding window for adaptive data fluctuations that dynamically expands / contracts the window based on event density, the definition of thresholds with dynamic quantiles combined with local features, and the voltage sag-fault matching that fuses time series matching and dynamically adjusts the window based on event density. As Figure 1 shown, the method specifically includes the following implementation steps.

[0057] Step S1: Obtain monitoring data. Obtain the time data of faults on the same bus, faults outside the bus, and voltage sag events of each monitoring point on the bus within a certain period of time in a region as the data source.

[0058] Specifically, for a region, different groups of data are matched separately according to the different positions of voltage sag monitoring points. The same-bus fault is the fault corresponding to the 10kV bus under the substation where the monitoring point is located. That is, the same-bus fault will only cause the voltage sag event to be detected at this monitoring point. The off-bus fault is the fault corresponding to the 110kV bus, 220kV substation, and 500kV substation within this region. That is, the off-bus fault will cause all the monitoring points within this region to detect voltage sag events.

[0059] Step S2: Sag-fault matching based on time sequence matching. Match the voltage sags of a monitoring point with the same-bus faults and external faults according to the time sequence of event occurrence. The sags occurring within one minute need to be aggregated. The successfully matched data needs to be removed from the original dataset, and each pair of voltage sag and fault in the matching result forms a new piece of data. Calculate the time difference between each pair of voltage sag and fault in the matching result. The specific implementation steps of step S2 are as follows.

[0060] Step S201: Unify the timestamp formats of different data sources with the unit of seconds to ensure effective time comparison; the formula is:

[0061]

[0062] where T is the timestamp, Y is the year offset, y i is the number of days in the i-th year, M is the month offset, m j is the number of days in the j-th month, D is the number of days in the current month, and h, m, s are hours, minutes, and seconds respectively.

[0063] Filter out the short-term intensive sags detected due to low voltage and other reasons according to the sag amplitude and detection frequency; read the time of the voltage sag of a monitoring point within the region and sort it in ascending order S = {S1, S2, …, S n}; read the times of the same-bus faults within the region and off-bus faults of the region, synthesize a dataset and sort it in ascending order to obtain the fault dataset F = {F1, F2, …, F m}.

[0064] Step S202: Starting from the first fault, traverse the voltage sag data. If the first voltage sag that meets the following conditions is found:

[0065] S j ∈S, |F1 - S j | ≤ T0

[0066] where F1 is the timestamp of the first fault, S jis the timestamp of the j-th voltage sag data, and T0 is the size of the matching window. The value of T0 is selected according to the actual data situation.

[0067] Then it is considered that the fault timestamp F1 and the voltage sag timestamp S j are matched and no further search is needed; if not, it is considered that the fault is not matched, and the next fault timestamp F2 is matched; if S j has matched the fault data, and there are other voltage sag timestamps at this monitoring point that satisfy the following relationship:

[0068]

[0069] Then the two voltage sags are combined into one voltage sag event, and the voltage sag timestamp S j is taken as the combined voltage sag timestamp; for the matched fault and voltage sag data, they need to be removed from the original set and no longer participate in subsequent matching.

[0070] Step S203: After the initial matching is completed, each pair of matched fault and sag data in the matched results is combined into a new data. If one fault matches multiple sags, there will be multiple data; for each matching result, the time difference between the fault and the sag is calculated respectively, and the formula is as follows:

[0071]

[0072] Among them, represents the timestamp of the fault in the i-th matching result, represents the timestamp of the voltage sag in the i-th matching result, and d i represents the time difference between the fault and the voltage sag in the i-th matching result, with the unit of seconds.

[0073] Step S3: Dynamically adjust the window to calculate the quantile to detect outliers. First, set the initial window size and density threshold, calculate the quantile in the window and generate a dynamic threshold, and then mark the time difference outliers based on the dynamic threshold; move the window, calculate the event density and dynamically adjust the window size, calculate the window quantile and generate a dynamic threshold to detect abnormal data. Repeat the dynamic window adjustment until all data is traversed, and output the set of abnormal data points. The specific implementation steps of Step S3 are as follows.

[0074] Step S301: Input the time difference sequence data set D = {d1, d2,..., d i ,..., d n}, and initialize the window parameters, including the initial window size W0, the density threshold σ0, the maximum window size W max , and the initial window

[0075] W0: The selection of the initial window size usually takes 5% - 10% of the total dataset volume. The window sensitivity can be observed by running the algorithm on a small amount of data, and the W0 with the smallest error is selected.

[0076] σ0: Calculate the average density of the data points within the initial window and set it as the density threshold.

[0077] W max : The size of the maximum window usually does not exceed 30% - 50% of the total dataset volume to avoid a decrease in sensitivity caused by an overly large window.

[0078] Calculate the quantile threshold of the initial window. The threshold is calculated using the interquartile range: IQR = Q3 - Q1, where Q1 = P 25 (W1), Q3 = P 75 (W1), and P() is the percentile function, indicating at least what percentage of the data is less than or equal to it. P 25 () represents the 25th percentile, and P 75 () represents the 75th percentile.

[0079] Define the outlier threshold range:

[0080]

[0081] Among them, α is a preset coefficient, and based on statistical experience, α = 1.5 is usually taken.

[0082] Detect whether there are outliers in the initial window. If then mark d i as outlier data.

[0083] Step S302: Slide the window backward and add d i to the current window W i ={d i-W+1 , d i-W+2 ,..., d i}, where W is the current window size; Define the event density as the standard deviation of the differences between adjacent data within the window, and adjust the window size according to the event density. The specific method is as follows:

[0084] Calculate the time intervals of the adjacent time difference data within the window W. The formula is as follows:

[0085] ΔT = {Δt1, Δt2,..., Δt W-1}

[0086] Among them, Δt k = d i-W+k+1 - d i-W+k .

[0087] Calculate the standard deviation σ of the time interval, and the formula is as follows:

[0088]

[0089] where μ is the mean value of the time interval,

[0090] Adjust the size of the window according to the event density σ. When the time interval fluctuates greatly, expand the window to adapt to the data fluctuation. Conversely, shrink the window to improve the sensitivity. The window size adjustment rule is as follows:

[0091]

[0092] Step S303: Calculate the quantile threshold based on the current window: IQR = Q3 - Q1, where Q1 = P 25 (W i ), Q3 = P 75 (W i ); Define the abnormal threshold range:

[0093]

[0094] where α is a preset coefficient, and usually α = 1.5 based on statistical experience.

[0095] Detect whether the current data point d i is an outlier. If then d i is marked as abnormal data; otherwise, it is retained as normal data.

[0096] Step S304: Repeat steps S302 - S303 until all data is traversed, and output the index set A of abnormal data points A = {i | d i is marked as abnormal}. Output the faulty data set F' = {F A1 , F A2 , …, F An} and the sag data set S' = {S A1 , S A2 , …, S An} that matches incorrectly, where n is the size of the abnormal data index A.

[0097] Step S4: Remove the abnormal data S' in the matching result from the original voltage sag data set, and use the faulty data set F' that matches incorrectly as a new faulty data set for secondary matching; after secondary matching, perform dynamic window adjustment iteratively. The specific steps are the same as step S2 above. Add the successfully matched results to the preliminary matching result data set, and then perform the detection of time difference outliers as in step S3. Keep iterating until no abnormal time difference data is detected, and obtain the matching result of this monitoring point.

[0098] Step S5: For the remaining monitoring points, repeat Steps S2 - S4 to obtain all the matching data within the region, achieving precise positioning of the voltage sag source.

[0099] The voltage sag source positioning method based on measured data provided by the present invention performs a temporal sorting of the sag time and the fault time, and uses the characteristic that events occur in sequence for preliminary matching; by analyzing the pattern of the time difference in the preliminary matching results, a method of adaptively adjusting the sliding window in combination with the window dynamic quantile to define the threshold is used to precisely identify abnormal data by adapting to the data fluctuations.

[0100] This embodiment also provides a computer device, including: at least one processor, at least one memory, and computer program instructions stored in the memory, which implement the above - mentioned method when executed by the processor.

[0101] This embodiment also provides a computer - readable storage medium, on which computer program instructions are stored, which implement the above - mentioned method when executed by a processor.

[0102] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer - usable storage media (including but not limited to disk memories, CD - ROMs, optical memories, etc.) containing computer - usable program codes.

[0103] The present application is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, and the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general - purpose computer, a special - purpose computer, an embedded processor, or other programmable data - processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data - processing devices generate a device for implementing the specified functions in Figure 1 one or more flows or multiple flows and / or blocks Figure 1 one or more blocks or multiple blocks.

[0104] These computer program instructions can also be stored in a computer - readable memory that can direct a computer or other programmable data - processing device to work in a specific manner, such that the instructions stored in the computer - readable memory generate a manufactured article including an instruction device that implements the specified functions in the flowFigure 1 one process or multiple processes and / or blocks Figure 1 the functions specified in one block or multiple blocks.

[0105] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process. Thus, the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0106] As described above, it is only a preferred embodiment of the present invention, and it is not a limitation of the present invention in other forms. Any person skilled in the art may use the disclosed technical content to make changes or modifications into equivalent embodiments with equivalent changes. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the technical solution content of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

1. A method for locating voltage sag sources based on measured data, characterized in that Including the following steps: Step S1: Obtain monitoring data; obtain the time data of the same-bus fault, external-bus fault, and voltage sag events of the bus where each monitoring point is located within a certain period of time in a region as the data source; Step S2: Sag-fault matching based on time series matching; match the voltage sags of a monitoring point with the same-bus faults and external faults according to the chronological order of event occurrence, and calculate the time difference between each pair of voltage sags and faults in the matching result; Step S3: Dynamically adjust the window to calculate the quantile to detect outliers; first set the initial window size and density threshold, calculate the quantile in the window and generate a dynamic threshold, and then mark the time difference outliers based on the dynamic threshold; move the window, calculate the event density and dynamically adjust the window size, calculate the window quantile and generate a dynamic threshold to detect abnormal data; repeat the dynamic window adjustment until all data is traversed, and output the set of abnormal data points; Step S4: Remove the abnormal data S' in the matching result from the original voltage sag dataset, and use the mis-matched fault dataset F' as the new fault dataset for secondary matching; after secondary matching, iteratively perform dynamic window adjustment, add the successfully matched results to the preliminary matching result dataset, and then perform the detection of time difference outliers, and continuously iterate until no abnormal time difference data is detected to obtain the matching result of this monitoring point; Step S5: For the remaining monitoring points, repeat steps S2 - S4 to obtain all the matching data in the region and achieve the accurate positioning of the voltage sag source.

2. The voltage sag source location method based on measured data according to claim 1, characterized in that, In step S1, for a region, different groups of data are separately matched according to the different positions of the voltage sag monitoring points. The same-bus fault is the fault corresponding to the 10kV bus under the substation where this monitoring point is located, that is, the same-bus fault will only cause the voltage sag event monitored by this monitoring point. The external-bus fault is the fault corresponding to the 110kV bus, 220kV substation, and 500kV substation in this region, that is, the external-bus fault will cause all the monitoring points in this region to monitor the voltage sag event.

3. A voltage sag source location method based on measured data according to claim 1, characterized in that Step S2 specifically includes the following steps: Step S201: Unify the timestamp formats of different data sources; Filter out short-term and intensive voltage sags according to the temporary voltage sag value and the detection frequency; read the time of voltage sags at a monitoring point in the area and sort them in ascending order S = {S1, S2, …, S n}; read the time of faults on the same bus in the area and faults outside the area bus, synthesize a data set and sort it in ascending order to obtain the fault data set F = {F1, F2, …, F m}; Step S202: Starting from the first fault, traverse the voltage sag data. If the first voltage sag is found to satisfy: S j ∈ S, |F1 - S j | ≤ T0 Among them, F1 is the timestamp of the first fault, and S j is the timestamp of the j-th voltage sag data, and T0 is the matching window size; Then it is considered that the fault timestamp F1 and the voltage sag timestamp S j match, and no further search is continued; if not, it is considered that the fault does not match, and the next fault timestamp F2 is matched; if S j has matched the fault data, and there are other voltage sag timestamps at this monitoring point that satisfy the following relationship: Then the two voltage sags are combined into one voltage sag event, and the voltage sag timestamp S is taken j as the voltage sag timestamp after combination; for the already matched fault and voltage sag data, they need to be removed from the original set and no longer participate in subsequent matching; After the preliminary matching is completed, combine each pair of matched faults and sag data in the matching result into a new data. If one fault matches multiple sags, there will be multiple data; for each matching result, calculate the time difference between the fault and the sag respectively, and the formula is as follows: Among them, represents the timestamp of the fault in the i-th matching result, represents the timestamp of the voltage sag in the i-th matching result, d i represents the time difference between the fault and the voltage sag in the i-th matching result.

4. A method for locating voltage sag sources based on measured data according to claim 3, characterized in that The formula for unifying the timestamp formats of different data sources is: where T is the timestamp, Y is the year offset, y i is the number of days in the i-th year, M is the month offset, m j is the number of days in the j-th month, D is the number of days in the current month, and h, m, and s are the hours, minutes, and seconds respectively.

5. A voltage sag source location method based on measured data according to claim 1, characterized in that Step S3 specifically includes the following steps: Step S301: Input the time difference sequence dataset D = {d1, d2,..., d i ,..., d n}, initialize the window parameters, including the initial window size W0, the density threshold σ0, the maximum window size W max , and the initial window W1 = {d1, d2,..., d W0}; Calculate the quantile threshold of the initial window, and use the interquartile range to calculate the threshold: IQR = Q3 - Q1, where Q1 = P 25 (W1), Q3 = P 75 (W1), P 25 () represents the 25th percentile, P 75 () represents the 75th percentile; Define the abnormal threshold range: where α is a preset coefficient; Check whether there are outliers in the initial window. If then mark d i as abnormal data; Step S302: Slide the window backward and add d i to the current window W i ={d i-W+1 , d i-W+2 ,..., d i}, where W is the size of the current window; define the event density as the standard deviation of the differences between adjacent data within the window, and adjust the window size according to the event density; Adjust the window size according to the event density σ. When the time interval fluctuates greatly, expand the window to adapt to the data fluctuation, and vice versa, shrink the window to improve the sensitivity; the window size adjustment rule is: Step S303: Calculate the quantile threshold based on the current window: IQR = Q3 - Q1, where Q1 = P 25 (W i ), Q3 = P 75 (W i ); Define the abnormal threshold range: where α is a preset coefficient; Detect the current data point d i Whether it is an outlier. If then mark d i as abnormal data; otherwise, retain it as normal data; Step S304: Repeat steps S302 - S303 until all data is traversed, and output the index set A of abnormal data points A = {i | d i is marked as abnormal}, and output the faulty data set F' = {F A1 , F A2 ,..., F An} and the sag data set S' = {S A1 , S A2 ,..., S An} that match the error according to the abnormal data index, where n is the size of the abnormal data index A.

6. The voltage sag source location method based on measured data according to claim 5, characterized in that Adjust the window size according to the event density, and its implementation method is: Calculate the time interval of adjacent time difference data within the window W, and the formula is as follows: ΔT = {Δt1, Δt2,..., Δt W-1} where, Δt k = d i-W+k+1 - d i-W+k ; Calculate the standard deviation σ of the time interval, and the formula is as follows: where μ is the mean of the time intervals, 7. A computer device, characterized in that, Comprising: At least one processor, at least one memory, and computer program instructions stored in the memory, which, when executed by the processor, implement the method according to any one of claims 1-6.

8. A computer-readable storage medium, on which computer program instructions are stored, characterized in that, When the computer program instructions are executed by a processor, the method according to any one of claims 1-6 is implemented.

Images