Power failure anomaly recognition method and apparatus

Abstract

A power failure anomaly recognition method and apparatus. The method comprises: acquiring cluster data from a target cluster (S1); on the basis of the cluster data, acquiring distribution transformer data of each distribution transformer at each moment, the distribution transformer data comprising three-phase voltage imbalance, a three-phase voltage, the fluctuation amplitude of a phase current corresponding to voltage sag, a three-phase current, the sag amplitude of a three-phase instantaneous total active power, and the sag amplitude of each phase current (S2); on the basis of the distribution transformer data, screening out a corresponding distribution transformer that meets a first preset anomaly condition, so as to achieve preliminary anomaly recognition (S3); and, on the basis of the distribution transformer data, performing in-depth anomaly recognition on the distribution transformer that meets the first preset anomaly condition, so as to screen out a corresponding distribution transformer that meets a second preset anomaly condition and obtain a distribution transformer having power failure anomaly (S4). On the basis of the three-phase voltage imbalance, the three-phase voltage, the three-phase current and other distribution transformer data, preliminary anomaly recognition and in-depth anomaly recognition are performed; compared with manual inspection, the present method reduces manual workload and eliminates factors that affects power failure anomaly recognition such as limited human energy, thereby improving the timeliness and accuracy in power failure anomaly recognition.

Description

Power outage anomaly identification method and device TECHNICAL FIELD

[0001] The present application relates to the field of power anomaly detection, and in particular to a power outage anomaly identification method and device. BACKGROUND

[0002] In a power system, detection and identification of power outage of distribution transformers are crucial for ensuring safe and stable operation of power supply. At present, detection and identification of power outage of distribution transformers mainly rely on manual inspection or manual analysis based on monitoring data. The above detection and identification methods have high dependence on human work. Considering the limited energy of human beings, power outage identification and detection work dependent on human beings have problems of timeliness and accuracy to some extent. SUMMARY

[0003] The present application provides a power outage anomaly identification method and device to solve the technical problem of how to improve the timeliness and accuracy of power outage anomaly identification.

[0004] To solve the above technical problem, the present application provides a power outage anomaly identification method, comprising:

[0005] obtaining cluster data from a target cluster;

[0006] obtaining distribution transformer data of each distribution transformer at each time according to the cluster data; wherein the distribution transformer data includes three-phase voltage unbalance degree, three-phase voltage, fluctuation amplitude of corresponding phase current of voltage sag, three-phase current, sag amplitude of three-phase instantaneous total active power, and sag amplitude of each phase current;

[0007] screening out distribution transformers corresponding to a first preset abnormal condition according to the distribution transformer data to realize preliminary anomaly identification;

[0008] performing deep anomaly identification on the distribution transformers corresponding to the first preset abnormal condition according to the distribution transformer data, screening out distribution transformers corresponding to a second preset abnormal condition, and obtaining power outage abnormal distribution transformers.

[0009] As a preferred scheme, the screening out of distribution transformers corresponding to the first preset abnormal condition comprises:

[0010] selecting distribution transformers with three-phase voltage unbalance degree greater than or equal to a preset first unbalance degree threshold, or three-phase voltage unbalance degree of zero and fluctuation amplitude of corresponding phase current greater than or equal to a preset first fluctuation amplitude threshold;

[0011] selecting distribution transformers with three-phase current less than a preset first current threshold;

[0012] select a distribution transformer whose three-phase instantaneous total active sudden drop amplitude is greater than or equal to a first sudden drop amplitude threshold value or whose current of a certain phase in the three-phase current has a sudden drop amplitude greater than or equal to a second current threshold value;

[0013] obtain the distribution transformer that meets the second preset abnormal condition.

[0014] As a preferred solution, the distribution transformer meeting the second preset abnormal condition is screened out, comprising:

[0015] statistically screen out the distribution transformer meeting the first preset abnormal condition for distribution transformer data in the past three days, and eliminate the distribution transformer with less than 197 data;

[0016] Eliminate the distribution transformer whose data amount of three-phase current less than 1A is more than 32 in the past three days, or whose number of days of three-phase current less than 1A within 90 minutes before and after the same time point in the past three days is greater than 1;

[0017] Eliminate the distribution transformer whose data amount of three-phase voltage less than 180V is more than 32 in the past three days, or whose three-phase voltage imbalance degree is greater than or equal to 0.4 or equal to 0 in the past three days, or whose three-phase voltage imbalance degree is greater than or equal to 0.4 or equal to 0 within 90 minutes before and after the same time point in the past three days;

[0018] Eliminate the distribution transformer whose sudden drop amplitude of current of a certain phase in the three-phase current is greater than or equal to the second current threshold value and the data number is greater than or equal to 3 in the past three days;

[0019] Eliminate the distribution transformer whose current minimum value of the current of the three-phase current plus 1 is greater than the current minimum value of the three-phase current in the past three days;

[0020] Eliminate the distribution transformer whose three-phase current is less than 80A in the past three days;

[0021] Eliminate the distribution transformer whose data amount of three-phase current fluctuation amplitude less than 0.02 is greater than or equal to 48 in the past three days;

[0022] Eliminate the distribution transformer whose number of days of three-phase current less than 1A or three-phase voltage less than 157V is greater than 2 in the past month;

[0023] obtain the distribution transformer that meets the second preset abnormal condition.

[0024] As a preferred solution, the power outage abnormality identification method further comprises:

[0025] Obtain the measurement point identifier of the distribution transformer meeting the second preset abnormal condition;

[0026] The measurement points are identified, the same measurement points are merged, and the power distribution transformer whose previous time data is not empty, the number of users of the power outage area of the missing phase type is not less than 8, and the name of the distribution transformer does not contain specific words is fed back to the power grid management platform.

[0027] As a preferred solution, the obtaining of the cluster data from the target cluster comprises:

[0028] The original data is obtained from the metering system, the original data is parsed, and the parsed result is stored in the target cluster based on KAFKA;

[0029] The data stored in the target cluster is processed by stream computing through Spark Streaming;

[0030] The stream computing processing result is obtained through the dispatching center.

[0031] As a preferred solution, the stream computing processing comprises abnormal data processing, missing value processing, data type conversion, three-phase current multiplied by the corresponding CT value processing, and three-phase voltage multiplied by the corresponding PT value processing.

[0032] Correspondingly, the embodiment of the application also provides a power outage anomaly identification device, comprising a cluster data acquisition module, a distribution transformer data acquisition module, a preliminary identification module and a deep anomaly identification module; wherein,

[0033] The cluster data acquisition module is used for obtaining cluster data from a target cluster;

[0034] The distribution transformer data acquisition module is used for obtaining distribution transformer data of each time according to the cluster data; wherein, the distribution transformer data comprises three-phase voltage unbalance degree, three-phase voltage, fluctuation amplitude of corresponding phase current of voltage sudden drop, three-phase current, sudden drop amplitude of three-phase instantaneous total active power and sudden drop amplitude of each phase current;

[0035] The preliminary identification module is used for screening out distribution transformers corresponding to the first preset anomaly condition according to the distribution transformer data, and realizing preliminary anomaly identification;

[0036] The deep anomaly identification module is used for performing deep anomaly identification on the distribution transformers meeting the first preset anomaly condition according to the distribution transformer data, screening out distribution transformers corresponding to the second preset anomaly condition, and obtaining the power outage anomaly distribution transformer.

[0037] As a preferred solution, the preliminary identification module screens out the distribution transformers corresponding to the first preset anomaly condition, comprising:

[0038] The preliminary identification module selects the distribution transformer with the three-phase voltage unbalance degree greater than or equal to a preset first unbalance degree threshold, or the three-phase voltage unbalance degree is zero and the fluctuation amplitude of the corresponding phase current is greater than or equal to a preset first fluctuation amplitude threshold;

[0039] The distribution transformer with the three-phase current less than a preset first current threshold is selected;

[0040] The distribution transformer with the three-phase transient total active sudden drop amplitude greater than or equal to a first sudden drop amplitude threshold, or the sudden drop amplitude of a certain phase current in the three-phase current greater than or equal to a second current threshold is selected;

[0041] The distribution transformer corresponding to the first preset abnormal condition is obtained.

[0042] As a preferred solution, the deep abnormality identification module screens the distribution transformer corresponding to the second preset abnormal condition, comprising:

[0043] The deep abnormality identification module screens out the distribution transformer with the distribution data in the last three days corresponding to the first preset abnormal condition, and removes the distribution transformer with less than 197 data amounts;

[0044] The distribution transformer with the data amount of the three-phase current less than 1A exceeding 32 in the last three days, or the number of days with the three-phase current less than 1A within 90 minutes before and after the same time point in the last three days greater than 1 is removed;

[0045] The distribution transformer with the data amount of the three-phase voltage less than 180V exceeding 32 in the last three days, the three-phase voltage unbalance degree greater than or equal to 0.4 or equal to 0 in the last three days, or the three-phase voltage unbalance degree greater than or equal to 0.4 or equal to 0 within 90 minutes before and after the same time point in the last three days is removed;

[0046] The distribution transformer with the sudden drop amplitude of a certain phase current in the three-phase current greater than or equal to the second current threshold and the data amount greater than or equal to 3 in the last three days is removed;

[0047] The distribution transformer with the current minimum value plus 1 greater than the current minimum value in the last three days is removed;

[0048] The distribution transformer with the three-phase current less than 80A in the last three days is removed;

[0049] The distribution transformer with the data amount of the three-phase current fluctuation amplitude less than 0.02 greater than or equal to 48 in the last three days is removed;

[0050] The distribution transformer with the number of days with the three-phase current less than 1A or the three-phase voltage less than 157V greater than 2 in the last month is removed;

[0051] The power distribution transformer corresponding to the second preset abnormal condition is obtained.

[0052] As a preferred solution, the power outage abnormality recognition device further comprises a feedback module, which is configured to:

[0053] The measurement point identifier of the power distribution transformer meeting the second preset abnormal condition is obtained.

[0054] The measurement points with the same measurement point identifier are merged, and the power distribution transformer with the previous time data not being empty, the number of power outage users of the phase loss type being not less than 8, and the power distribution transformer name not containing specific words is fed back to the power grid management platform.

[0055] As a preferred solution, the cluster data acquisition module acquires cluster data from a target cluster, comprising:

[0056] The cluster data acquisition module acquires original data from a metering system, parses the original data, and stores the parsed result in a target cluster based on KAFKA;

[0057] Through Spark Streaming, the data stored in the target cluster is processed in a streaming manner.

[0058] The scheduling center acquires the streaming processing result.

[0059] As a preferred solution, the streaming processing includes abnormal data processing, missing value processing, data type conversion, processing of multiplying three-phase current by its corresponding CT value, and processing of multiplying three-phase voltage by its corresponding PT value.

[0060] Compared with the prior art, the embodiments of the present application have the following beneficial effects:

[0061] The embodiment of the present application provides a power failure abnormality identification method and device, the power failure abnormality identification method comprises the following steps: obtaining cluster data from a target cluster; obtaining power distribution transformer data of each time according to the cluster data; wherein the power distribution transformer data comprises three-phase voltage unbalance degree, three-phase voltage, fluctuation amplitude of corresponding phase current of voltage sag, three-phase current, three-phase instantaneous total active sag amplitude and three-phase current sag amplitude; according to the power distribution transformer data, power distribution transformers meeting a first preset abnormality condition are screened out to realize preliminary abnormality identification; according to the power distribution transformer data, the power distribution transformers meeting the first preset abnormality condition are subjected to deep abnormality identification, and power distribution transformers meeting a second preset abnormality condition are screened out to obtain power failure abnormal power distribution transformers. According to the three-phase voltage unbalance degree, three-phase voltage and three-phase current and other power distribution transformer data, preliminary abnormality identification and deep abnormality identification are performed, compared with the artificial inspection mode, the artificial workload is reduced, the influence factors of the power failure abnormality identification due to the limited energy of the human body are avoided, and the timeliness and accuracy of the power failure abnormality identification are improved. BRIEF DESCRIPTION OF DRAWINGS

[0062] Fig. 1 is a flowchart of one embodiment of the power failure abnormality identification method provided by the present application.

[0063] Fig. 2 is a flowchart of another embodiment of the power failure abnormality identification method provided by the present application.

[0064] Fig. 3 is a flowchart of one embodiment of the preliminary abnormality identification provided by the present application.

[0065] Fig. 4 is a flowchart of one embodiment of the deep abnormality identification provided by the present application.

[0066] Fig. 5 is a flowchart of one embodiment of the suspected power failure single interaction display provided by the present application.

[0067] Fig. 6 is a structural diagram of one embodiment of the power failure abnormality identification device provided by the present application. DETAILED DESCRIPTION

[0068] The technical solutions in the embodiments of the present application will be described clearly and completely in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0069] Embodiment one

[0070] Please refer to Fig. 1, which is a power failure abnormality identification method provided by the embodiment of the present application, comprising steps S101 to S104; wherein,

[0071] Step S101, obtaining cluster data from a target cluster.

[0072] Step S102, obtaining distribution transformer data at each time according to the cluster data; wherein the distribution transformer data includes three-phase voltage unbalance degree, three-phase voltage, fluctuation amplitude of corresponding phase current of voltage sag, three-phase current, sag amplitude of three-phase instantaneous total active power, and sag amplitude of each phase current.

[0073] Step S103, screening out distribution transformers corresponding to the first preset abnormal condition according to the distribution transformer data, to realize preliminary abnormal identification.

[0074] Step S104, performing deep abnormal identification on the distribution transformers meeting the first preset abnormal condition according to the distribution transformer data, screening out distribution transformers corresponding to the second preset abnormal condition, and obtaining power failure abnormal distribution transformers.

[0075] In the embodiment, for the step S101, the cluster data is obtained from the target cluster, including:

[0076] Raw data is obtained from a metering system, the raw data is parsed, and the parsed results are stored in a target cluster based on KAFKA; through Spark Streaming, the data stored in the target cluster is processed in a streaming manner; and the scheduling center obtains the streaming processing results. The streaming processing includes abnormal data processing, missing value processing, data type conversion, three-phase current multiplied by its corresponding CT value processing, and three-phase voltage multiplied by its corresponding PT value processing.

[0077] Specifically, the scheduling center can receive three-phase current and voltage data (collectively referred to as distribution transformer data) of distribution transformers in real time from the power failure KAFKA cluster through a big data component, and parse and clean the data and associated archive information data according to requirements, perform simple calculation and conversion to distribution transformer data meeting the requirements of the power failure abnormality detection and identification model. Then the scheduling center calls the FLASK API interface, transmits the distribution transformer data to the power failure abnormality detection and identification model for detection and identification of the distribution transformer data.

[0078] For the step S103, the three-phase voltage unbalance degree U_ub and the fluctuation amplitude i_tx_idff_pro of the corresponding phase current of the voltage sag of the distribution transformer data can be obtained in real time. When the three-phase voltage unbalance degree is greater than or equal to a preset first unbalance degree threshold u_ub_yz, or the three-phase voltage unbalance degree is equal to zero and the fluctuation amplitude of the corresponding phase current is greater than or equal to a preset first fluctuation amplitude threshold i_tx_pro_yz, the distribution transformer data is marked as preliminary abnormality.

[0079] wherein u ub = (max(u a, u b, u c) - min(u a, u b, u c)) / max(u a, u b, u c);

[0080] i tx idff pro = abs(i tx idff / (i x+0.0001));

[0081] wherein a, b and c refer to three phases, i tx idff pro is a fluctuation amplitude of the corresponding phase current of the voltage sag, i tx idff is a fluctuation value of the corresponding phase current of the voltage sag, and i x is a current value of the corresponding phase of the voltage sag.

[0082] The first unbalance degree threshold u ub_yz is 0.49.

[0083] The first fluctuation amplitude threshold i tx pro_yz is 0.25.

[0084] For three-phase currents, the distribution transformers with three-phase currents less than the preset first current threshold i yz can be selected by comparing the three-phase currents of the distribution data with the preset first current threshold. Preferably, the first current threshold i yz can be 0.4.

[0085] In addition, the distribution transformers with a sag amplitude of the three-phase instantaneous total active power greater than or equal to a first sag amplitude threshold sp_diff_pro or a sag amplitude of a certain phase current in the three-phase current greater than or equal to a second current threshold i_diff_pro are selected to obtain the distribution transformers corresponding to the first preset abnormal condition.

[0086] wherein sp_diff = sp-sp_f;

[0087] sp_diff_pro = sp_diff / (sp_diff+sp+0.0001);

[0088] i_diff = i x-i x_f;

[0089] i_diff_pro = i_diff / (i x+i_diff+0.0001);

[0090] wherein sp is the three-phase instantaneous total active power at the current time point, sp_f is the three-phase instantaneous total active power at the previous time point, sp_diff is the three-phase instantaneous total active power difference between adjacent time points; i x is the current value of a certain phase at the current time point, i x_f is the corresponding phase current value at the previous time point, and i_diff is the current difference between adjacent time points. Exemplarily, the first sag amplitude threshold sp_pro_yz is 0.7. The second current threshold i_pro_yz is 0.9.

[0091] For the above step S104, the distribution transformer data of the first preset abnormal condition corresponding to the distribution transformer in the past three days can be counted and screened out, and the distribution transformer with less than 197 data amount is removed.

[0092] Further, the data amount of three-phase current less than 1A in the past three days is more than 32, or the number of days of three-phase current less than 1A within 90 minutes before and after the same time point in the past three days is greater than 1, and the distribution transformer is removed.

[0093] The data amount of three-phase voltage less than 180V in the past three days is more than 32, or the distribution transformer with three-phase voltage imbalance greater than or equal to 0.4 or equal to 0 in the past three days or within 90 minutes before and after the same time point in the past three days is removed.

[0094] The distribution transformer with the sudden drop amplitude of the current of one phase of the three-phase current greater than or equal to the second current threshold and the data amount greater than or equal to 3 in the past three days is removed.

[0095] The distribution transformer with the current minimum value i_min plus 1 greater than the three-phase current minimum value i_min_f of the past three days is removed.

[0096] The distribution transformer with three-phase current less than 80A in the past three days is removed.

[0097] After the above steps, the preliminary abnormal distribution transformer is uniformly verified for residential electricity distribution characteristics, that is, as a residential electricity distribution area, the distribution transformer has many sub-branch users and various forms of residential electricity, and the three-phase current curve of the distribution transformer shows a large amplitude. Therefore, the distribution transformer with the three-phase current fluctuation amplitude if_diff_pro less than 0.02 in the past three days and the data amount if_pro_counts greater than or equal to 48 is removed.

[0098] The distribution transformer with the number of days of three-phase current less than 1A or three-phase voltage less than 157V in the past month yc_counts greater than 2 is removed.

[0099] The distribution transformer corresponding to the second preset abnormal condition is obtained.

[0100] At the same time, the previous time point without data is marked as sfyk=0, and the subsequent abnormality is not pushed. At the same time, the previous time point with data is marked as sfyk=1.

[0101] According to the preliminary anomaly identification and the deep anomaly identification, a measurement point identifier (ID) of the distribution transformer meeting the second preset anomaly condition can be screened out; measurement points with the same measurement point identifier are merged, and the distribution transformer with the previous moment data not being empty (sfyk=1), the number of users of the power outage area of the phase loss type being not less than 8, and the distribution transformer name not containing specific words (such as "special transformer" and "charging pile") is fed back to the power grid management platform.

[0102] Correspondingly, the embodiment of the application also provides an application example of the power outage anomaly identification method, as shown in FIG. 2. The application example includes three parts. The first part is data access and real-time acquisition, mainly through the big data component to clean the three-phase current, voltage and other data to meet the requirements of the power outage anomaly detection model for the required access data. The second part is the power outage anomaly detection model, including preliminary anomaly identification and deep anomaly identification, to obtain the distribution transformer belonging to the anomaly. The third part is the interaction with the user and the display part of the identification result, mainly to further process the detection and identification result of the power outage anomaly detection model according to the business requirements and push it to the power grid management platform, and accept the related feedback of the offline business personnel to complete the interactive display. The dispatching center is the brain to coordinate the data reception, transmission and task allocation of the three parts.

[0103] Among them, for the first part, the original three-phase current, voltage and other data are transmitted in real time from the metering system in a special file and stored in the KAFKA cluster after related analysis and processing. At the same time, the three-phase current and voltage real-time data in KAFKA are processed by Spark Streaming for stream computing. It includes abnormal data processing, missing value processing, data type conversion, three-phase current multiplied by its corresponding CT value and three-phase voltage multiplied by its corresponding PT value processing. The real-time processed data is input to the power outage anomaly detection model through the dispatching center for detection and identification.

[0104] For the preliminary anomaly detection in the second part, an example is shown in FIG. 3, which specifically includes:

[0105] S1, screening the measurement points corresponding to the distribution transformers with three-phase current less than 600A and three-phase voltage less than 270V; eliminating the measurement points with empty three-phase current or three-phase voltage, and filling other forms of empty with 0.

[0106] S2, calculating the three-phase voltage unbalance degree, screening out the measurement points with three-phase voltage unbalance degree >=0.49 or three-phase voltage all being 0, and the current corresponding to the voltage abnormal phase of these measurement points being accompanied by >=25% sudden drop and the minimum voltage <=50V, and the corresponding phase current being not more than A, and the preliminary abnormal measurement points meeting the above conditions.

[0107] S3, remove S2 measurement points, screen the measurement points with less than 0.4A three-phase current as preliminary abnormal measurement points.

[0108] S4, remove S2 and S3 measurement points, calculate the apparent power and the sudden drop amplitude of three-phase current, screen the measurement points with apparent power sudden drop amplitude >= 70%, or three-phase current sudden drop amplitude >= 90% and the minimum current >= 10A as preliminary abnormal measurement points.

[0109] S5, combine S2, S3 and S4 measurement points and remove duplicates to get all preliminary abnormal measurement points.

[0110] Further, for the second part of the deep abnormal identification, as shown in FIG. 4, the application instance specifically includes:

[0111] S6, through the preliminary abnormal measurement point identification (ID) obtained by S5, obtain 288 three-phase current and voltage data of the preliminary abnormal measurement point in the past three days, and remove the preliminary abnormal measurement points with more than 91 missing data.

[0112] S7, calculate the three-phase current and voltage unbalance degree of each time point of the S6 preliminary abnormal measurement point in the past three days, and obtain the three-phase current and voltage data of the S6 preliminary abnormal measurement point in the 90-minute period before and after the same time point in the past three days.

[0113] S8, screen and count the number of three-phase currents less than 1A in the past three days of each preliminary abnormal measurement point obtained by S2 and S3, and the measurement points with the number greater than or equal to 32 are non-abnormal measurement points.

[0114] S9, count and calculate the number of days with three-phase current less than 1A in the 90-minute period before and after the same time point in the past three days of the preliminary abnormal measurement points obtained by S2 and S3, and the measurement points with two days or more are non-abnormal measurement points.

[0115] S10, screen and count the number of three-phase voltages less than 180V and the number of three-phase voltage unbalance degree >= 0.4 or three-phase voltage all being 0 in the past three days of each preliminary abnormal measurement point obtained by S2, and the measurement points with the number >= 32 are non-abnormal measurement points.

[0116] S11, count and calculate the number of days with three-phase voltage unbalance degree >= 0.4 or three-phase voltage all being 0 in the 90-minute period before and after the same time point in the past three days of the preliminary abnormal measurement points obtained by S2, and the measurement points with two days or more are non-abnormal measurement points.

[0117] S12, screen the measurement points of S3+S4-S8-S9-S10-S11 in the last three days (eliminate the three-phase current that exists 0), calculate the three-phase current sudden drop amplitude, and count the number of each measurement point current sudden drop amplitude greater than or equal to 0.4, wherein the measurement point greater than 2 is a non-exceptional measurement point.

[0118] S13, screen the measurement points of S3+S4-S8-S9-S12 in the last three days, obtain the minimum three-phase current value of each measurement point every day, and count the number of days that each measurement point (the minimum three-phase current of the current time point +1)*1.2 is greater than the minimum three-phase current of each day in the last three days, wherein the measurement point greater than 0 is a non-exceptional measurement point.

[0119] S14, screen out the abnormal measurement points of S4, which exist three-phase current <=80A in the last three days, and are non-exceptional measurement points. Screen out the measurement points in the preliminary abnormal measurement points that have no data at the previous time point of the current time point and mark them as 0, and mark the remaining abnormal measurement points as 1.

[0120] S15, screen out the measurement points in the preliminary abnormal measurement points that are non-exceptional measurement points and have near three days of historical data, obtain the sudden drop amplitude of the three-phase current calculated from the near three days of historical data, and count the number of each measurement point whose current sudden drop amplitude is less than 2% in the last three days, wherein greater than or equal to 48 is a non-exceptional measurement point.

[0121] S16, make a preliminary abnormality judgment on the previous time point of the preliminary abnormal measurement point, and screen out the abnormal measurement points whose three-phase voltage unbalance degree >=0.49 or three-phase voltage is all 0 and three-phase current exists less than 0.4A and mark them as 0, and mark the remaining abnormal measurement points as 1.

[0122] S17, judge whether the previous time point of the abnormal measurement point exists three-phase current less than 10A, the preliminary abnormal measurement point that exists is marked as 0, and the remaining abnormal measurement points are marked as 1.

[0123] S18, combine the above screening to obtain the final abnormal measurement point id_final, select the measurement point result0 whose three-phase current and voltage are less than or equal to 0.4, and delay push combined with the next time point that is also an abnormal measurement point.

[0124] S19, make a judgment on whether there are multiple or regular abnormalities in the last month for the remaining abnormal measurement points of the final abnormal measurement point id_final minus result0, and match with the abnormal measurement point statistics table in the last month; wherein the measurement points whose abnormal times are greater than 2, the power failure type is voltage abnormality and the abnormal type is 'V', and the abnormal times are greater than 2, the power failure type is non-voltage abnormality and the abnormal type is 'A' are eliminated, and the final output result finall is obtained in the form of json string and returned to the dispatch center.

[0125] S20, the dispatching center stores the output result final in the Oracle database in association with the archive information, and only one data with the same mp_id and outage start time will enter the database.

[0126] For the third part, as shown in FIG. 5, the screening and integration of all result data of the day are completed by establishing a stored procedure: there are multiple abnormal data records for the same measurement point identifier id within a day, in order to timely feedback the abnormality of the distribution transformer or the outage, the earliest one of the day should be taken, and the data label sfyk of the one is 1 (the previous time point is not empty or abnormal), the number of users in the outage area of the lack phase type is not less than 8, and the distribution transformer name does not contain words such as "special transformer" and "charging pile".

[0127] Correspondingly, referring to FIG. 6, the embodiment of the application further provides a power outage abnormality identification device, comprising a cluster data acquisition module 201, a distribution transformer data acquisition module 202, a preliminary identification module 203 and a deep abnormality identification module 204; wherein,

[0128] The cluster data acquisition module 201 is configured to acquire cluster data from a target cluster.

[0129] The distribution transformer data acquisition module 202 is configured to acquire distribution transformer data of each time point of each distribution transformer according to the cluster data; wherein the distribution transformer data comprises three-phase voltage unbalance degree, three-phase voltage, fluctuation amplitude of corresponding phase current of voltage sag, three-phase current, sag amplitude of three-phase instantaneous total active power and sag amplitude of each phase current.

[0130] The preliminary identification module 203 is configured to screen out distribution transformers corresponding to the first preset abnormal condition according to the distribution transformer data, so as to realize preliminary abnormality identification.

[0131] The deep abnormality identification module 204 is configured to perform deep abnormality identification on the distribution transformers corresponding to the first preset abnormal condition according to the distribution transformer data, screen out distribution transformers corresponding to the second preset abnormal condition, and obtain the distribution transformer with power outage abnormality.

[0132] As a preferred scheme, the preliminary identification module 203 screens out the distribution transformers corresponding to the first preset abnormal condition, comprising:

[0133] The preliminary identification module 203 selects the distribution transformers with three-phase voltage unbalance degree greater than or equal to a preset first unbalance degree threshold value, or three-phase voltage unbalance degree of zero and fluctuation amplitude of corresponding phase current greater than or equal to a preset first fluctuation amplitude threshold value.

[0134] The preliminary identification module 203 selects the distribution transformers with three-phase current less than a preset first current threshold value.

[0135] select a distribution transformer whose three-phase instantaneous total active sudden drop amplitude is greater than or equal to a first sudden drop amplitude threshold value or whose current of a certain phase in the three-phase current has a sudden drop amplitude greater than or equal to a second current threshold value;

[0136] obtain the distribution transformer that meets the second preset abnormal condition.

[0137] As a preferred solution, the deep abnormality identification module 204 screens out the distribution transformer that meets the second preset abnormal condition, including:

[0138] The deep abnormality identification module 204 screens out the distribution transformer that meets the first preset abnormal condition, including:

[0139] The distribution transformer whose data amount of three-phase current less than 1A exceeds 32 in the data of the last three days or whose number of days of three-phase current less than 1A within 90 minutes before and after the same time point in the last three days is greater than 1 is removed.

[0140] The distribution transformer whose data amount of three-phase voltage less than 180V exceeds 32 in the data of the last three days or whose three-phase voltage unbalance degree is greater than or equal to 0.4 or equal to 0 in the data of the last three days or whose three-phase voltage unbalance degree is greater than or equal to 0.4 or equal to 0 within 90 minutes before and after the same time point in the last three days is removed.

[0141] The distribution transformer whose sudden drop amplitude of current of a certain phase in the three-phase current is greater than or equal to the second current threshold value and whose data amount is greater than or equal to 3 in the data of the last three days is removed.

[0142] The distribution transformer whose current minimum value plus 1 is greater than the current minimum value of the distribution transformer in the data of the last three days is removed.

[0143] The distribution transformer whose three-phase current is less than 80A in the data of the last three days is removed.

[0144] The distribution transformer whose data amount of three-phase current fluctuation amplitude less than 0.02 is greater than or equal to 48 in the data of the last three days is removed.

[0145] The distribution transformer whose number of days of three-phase current less than 1A or three-phase voltage less than 157V in the last month is greater than 2 is removed.

[0146] obtain the distribution transformer that meets the second preset abnormal condition.

[0147] As a preferred solution, the power outage abnormality identification device further includes a feedback module, and the feedback module is configured to:

[0148] obtain the measurement point identifier of the distribution transformer that meets the second preset abnormal condition.

[0149] The measurement points are identified, the same measurement points are merged, and the power distribution transformer whose previous time data is not empty, whose number of users of the power outage area of the missing phase type is not less than 8, and whose transformer substation name does not contain specific words is fed back to the power grid management platform.

[0150] As a preferred solution, the cluster data acquisition module 201 acquires cluster data from a target cluster, including:

[0151] The cluster data acquisition module 201 acquires raw data from a metering system, parses the raw data, and stores the parsed results in a target cluster based on KAFKA;

[0152] Through Spark Streaming, the data stored in the target cluster is processed by stream computing;

[0153] The stream computing result is acquired through a dispatching center.

[0154] As a preferred solution, the stream computing processing includes abnormal data processing, missing value processing, data type conversion, processing of three-phase current multiplied by its corresponding CT value, and processing of three-phase voltage multiplied by its corresponding PT value.

[0155] Compared with the prior art, the embodiment of the application has the following beneficial effects:

[0156] The embodiment of the application provides a power outage anomaly identification method and device, and the power outage anomaly identification method includes: acquiring cluster data from a target cluster; acquiring power distribution transformer data of each time according to the cluster data; wherein the power distribution transformer data includes three-phase voltage unbalance degree, three-phase voltage, fluctuation amplitude of corresponding phase current of voltage sag, three-phase current, sag amplitude of three-phase instantaneous total active power, and sag amplitude of each phase current; according to the power distribution transformer data, power distribution transformers corresponding to a first preset abnormal condition are screened out to realize preliminary anomaly identification; according to the power distribution transformer data, the power distribution transformers meeting the first preset abnormal condition are subjected to deep anomaly identification, and power distribution transformers corresponding to a second preset abnormal condition are screened out to obtain power outage abnormal power distribution transformers. According to the three-phase voltage unbalance degree, three-phase voltage, three-phase current and other power distribution transformer data, preliminary anomaly identification and deep anomaly identification are performed in the embodiment, compared with the manual inspection mode, the artificial workload is reduced, the influence factors of power outage anomaly identification due to limited human energy and the like are avoided, and the timeliness and accuracy of power outage anomaly identification are improved.

[0157] The above-described specific embodiments further illustrate the purpose, technical solutions and beneficial effects of the present application. It should be understood that the above-described specific embodiments are merely examples of the present application and are not intended to limit the protection scope of the present application. It is particularly pointed out that any modification, equivalent replacement, improvement, etc. made by those skilled in the art within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A power failure abnormality identification method characterized by comprising: The method comprises the following steps: obtaining cluster data from a target cluster; obtaining distribution transformer data at each time according to the cluster data, wherein the distribution transformer data comprises three-phase voltage unbalance degree, three-phase voltage, fluctuation amplitude of corresponding phase current of voltage sag, three-phase current, sag amplitude of three-phase instantaneous total active power, and sag amplitude of each phase current; screening out distribution transformers corresponding to a first preset abnormal condition according to the distribution transformer data to realize preliminary abnormal identification; performing deep abnormal identification on the distribution transformers corresponding to the first preset abnormal condition according to the distribution transformer data, screening out distribution transformers corresponding to a second preset abnormal condition, and obtaining distribution transformer with power failure abnormality.

2. The power outage anomaly identification method of claim 1, wherein The screening out of the distribution transformers corresponding to the first preset abnormal condition comprises: selecting a distribution transformer with three-phase voltage unbalance degree greater than or equal to a preset first unbalance degree threshold, or three-phase voltage unbalance degree of zero and fluctuation amplitude of corresponding phase current greater than or equal to a preset first fluctuation amplitude threshold; selecting a distribution transformer with three-phase current less than a preset first current threshold; selecting a distribution transformer with sag amplitude of three-phase instantaneous total active power greater than or equal to a first sag amplitude threshold, or sag amplitude of a certain phase current in three-phase current greater than or equal to a second current threshold; obtaining the distribution transformers corresponding to the first preset abnormal condition.

3. The power outage anomaly identification method of claim 2, wherein The screening out of the distribution transformers corresponding to the second preset abnormal condition comprises: statistically obtaining distribution transformer data of the distribution transformers corresponding to the first preset abnormal condition in the last three days, and eliminating distribution transformers with data amount less than 197; eliminating a distribution transformer with data amount of three-phase current less than 1A exceeding 32 in the last three days, or number of days with three-phase current less than 1A within 90 minutes before and after the same time point in the last three days greater than 1; eliminating a distribution transformer with data amount of three-phase voltage less than 180V exceeding 32 in the last three days, three-phase voltage unbalance degree greater than or equal to 0.4 or equal to 0 in the last three days, or three-phase voltage unbalance degree greater than or equal to 0.4 or equal to 0 within 90 minutes before and after the same time point in the last three days; eliminating a distribution transformer with sag amplitude of a certain phase current in three-phase current greater than or equal to the second current threshold and data amount greater than or equal to 3 in the last three days; eliminating a distribution transformer with current minimum value greater than the minimum value of three-phase current in the last three days by 1; eliminating a distribution transformer with three-phase current less than 80A in the last three days; eliminating a distribution transformer with data amount of three-phase current fluctuation amplitude less than 0.02 greater than or equal to 48 in the last three days; eliminating a distribution transformer with number of days with three-phase current less than 1A or three-phase voltage less than 157V greater than 2 in the last month; obtaining the distribution transformers corresponding to the second preset abnormal condition.

4. The power outage anomaly recognition method of claim 1, wherein The power failure abnormality identification method further comprises: obtaining measurement point identification of the distribution transformers corresponding to the second preset abnormal condition; merging measurement points with the same measurement point identification, and feeding back a distribution transformer with previous time data not empty, number of users in a phase loss type power failure substation area not less than 8, and distribution transformer name not containing specific words to a power grid management platform.

5. The power outage anomaly recognition method of claim 1, wherein The cluster data is obtained from the target cluster, comprising: Obtaining raw data from a metering system, parsing the raw data, and storing the parsed result in a target cluster based on KAFKA; Processing the data stored in the target cluster through Spark Streaming in a streaming manner; Obtaining the streaming processing result through a scheduling center.

6. The power outage anomaly identification method of claim 5, wherein The streaming processing comprises abnormal data processing, missing value processing, data type conversion, processing of three-phase current multiplied by a corresponding CT value, and processing of three-phase voltage multiplied by a corresponding PT value.

7. A power failure abnormality identifying device characterized by comprising: The system comprises a cluster data obtaining module, a distribution transformer data obtaining module, a preliminary identification module, and a deep abnormality identification module, wherein The cluster data obtaining module is configured to obtain cluster data from a target cluster; The distribution transformer data obtaining module is configured to obtain distribution transformer data of each distribution transformer at each time according to the cluster data, wherein the distribution transformer data comprises three-phase voltage unbalance degree, three-phase voltage, fluctuation amplitude of corresponding phase current of voltage sag, three-phase current, three-phase instantaneous total active sag amplitude, and sag amplitude of each phase current; The preliminary identification module is configured to screen out distribution transformers corresponding to a first preset abnormal condition according to the distribution transformer data, and realize preliminary abnormality identification; The deep abnormality identification module is configured to perform deep abnormality identification on the distribution transformers corresponding to the first preset abnormal condition according to the distribution transformer data, screen out distribution transformers corresponding to a second preset abnormal condition, and obtain distribution transformer with power failure abnormality.

8. The power failure abnormality identification device according to claim 7, wherein The preliminary identification module screens out the distribution transformers corresponding to the first preset abnormal condition, comprising: The preliminary identification module selects distribution transformers with three-phase voltage unbalance degree greater than or equal to a preset first unbalance degree threshold value, or three-phase voltage unbalance degree of zero and fluctuation amplitude of corresponding phase current greater than or equal to a preset first fluctuation amplitude threshold value; selects distribution transformers with three-phase current less than a preset first current threshold value; selects distribution transformers with three-phase instantaneous total active sag amplitude greater than or equal to a first sag amplitude threshold value, or sag amplitude of a certain phase current in three-phase current greater than or equal to a second current threshold value; obtains the distribution transformers corresponding to the first preset abnormal condition.

9. The power failure abnormality identification device according to claim 8, wherein The deep abnormality identification module screens out the distribution transformers corresponding to the second preset abnormal condition, comprising: The deep abnormality identification module counts the distribution transformer data of the distribution transformers corresponding to the first preset abnormal condition in the last three days, and eliminates distribution transformers with data amount less than 197; eliminates distribution transformers with data amount of three-phase current less than 1A exceeding 32 in the last three days, or number of days with three-phase current less than 1A within 90 minutes before and after the same time point in the last three days greater than 1; eliminates distribution transformers with data amount of three-phase voltage less than 180V exceeding 32 in the last three days, three-phase voltage unbalance degree greater than or equal to 0.4 or equal to 0 in the last three days, or three-phase voltage unbalance degree greater than or equal to 0.4 or equal to 0 within 90 minutes before and after the same time point in the last three days; The power distribution transformer is removed when the sudden drop of the current of one phase of the three-phase current in the nearly three-day data is greater than or equal to the second current threshold and the number of data is greater than or equal to 3; The power distribution transformer is removed when the current minimum value of the current of the three-phase current is greater than the current minimum value of the three-phase current of the nearly three-day data of the power distribution transformer plus 1; The power distribution transformer is removed when the three-phase current of the nearly three-day data is less than 80 A; The power distribution transformer is removed when the number of data in which the fluctuation amplitude of the three-phase current is less than 0.02 is greater than or equal to 48 in the nearly three-day data; The power distribution transformer is removed when the number of days in which the three-phase current is less than 1 A or the three-phase voltage is less than 157 V is greater than 2 in the nearly one month; The power distribution transformer corresponding to the second preset abnormal condition is obtained.

10. The power failure abnormality identifying device according to Claim 7, wherein The power outage abnormality recognition device further comprises a feedback module, and the feedback module is used for: Obtaining the measurement point identifier of the power distribution transformer corresponding to the second preset abnormal condition; Merging the measurement points with the same measurement point identifier, and feeding back the power distribution transformer, in which the data of the previous moment is not empty, the number of users of the phase loss type power outage area is not less than 8, and the power distribution transformer name does not contain specific words, to the power grid management platform.

Images