Abnormality detection method, device and equipment of power equipment and storage medium

Abstract

The invention discloses an anomaly detection method and device for power equipment, equipment and a storage medium, relates to the technical field of power equipment detection, is applied to an anomaly detection system, and comprises the following steps: intercepting multi-parameter original data of target power equipment in a target time window to obtain target parameter sequences corresponding to target parameters; determining a target slope based on each target parameter sequence and the time attenuation weight by using a target weighted slope calculation formula, determining a target parameter change trend corresponding to each target parameter based on each target slope, and determining a target association relationship among the target parameter change trends; and determining a corresponding preset association relationship between the target parameters from a preset target comparison strategy library, judging whether the preset association relationship is matched with the target association relationship or not, and generating an anomaly detection result representing that the change trend of each target parameter is abnormal when the preset association relationship is not matched with the target association relationship. According to the invention, the fault omission ratio of the power equipment is reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power equipment detection, and in particular relates to an abnormality detection method, device and equipment of power equipment and a storage medium. BACKGROUND

[0002] In the field of intelligent operation and maintenance of power equipment, trend correlation anomaly detection based on multi-parameter association is a core technology to ensure the safe operation of key equipment such as transformers and water turbines. With the advancement of new power system construction, the operating conditions of equipment are becoming more complex and variable, and traditional monitoring solutions are facing severe challenges. For example, when the cooling system of a transformer fails, the load rises but the temperature abnormally decreases; when the water turbine cavitation occurs, the gate opening increases but the power generation power decreases. Such implicit faults that deviate from the trend of associated parameters cannot be effectively identified by traditional systems that rely on single-parameter threshold values or fixed numerical association models (such as pre-set linear equations), resulting in missed detection of implicit faults; the association between equipment parameters (such as water and electricity gate opening and power generation power) changes dynamically with external environment (such as water level) and internal state. Fixed mathematical models (such as linear equations with fixed slope k and intercept b) lack adaptability and require frequent manual parameter adjustment, which violates the real-time requirements of power systems and causes dynamic response lag.

[0003] To sum up, how to reduce the fault missed detection rate of power equipment and improve the dynamic response efficiency is a problem to be solved at present. SUMMARY

[0004] Therefore, the purpose of the present application is to provide an abnormality detection method, device and equipment of power equipment and a storage medium, which can reduce the fault missed detection rate of power equipment and improve the dynamic response efficiency. The specific scheme is as follows:

[0005] In a first aspect, the present application discloses an abnormality detection method of power equipment, applied to an abnormality detection system, comprising:

[0006] determining a target time window, and intercepting the multi-parameter original data of the target power equipment in the target time window to obtain each target parameter sequence corresponding to each target parameter; the window length of the target time window is determined based on the device type and configuration information of the target power equipment, and the target time window is a time window that automatically slides with system time;

[0007] determining a target slope based on each target parameter sequence and a time decay weight by using a target weighted slope calculation formula, and determining a target parameter change trend corresponding to each target parameter based on each target slope, and determining a target association relationship between each target parameter change trend;

[0008] The preset correlation relationship between each target parameter is determined from the preset target comparison strategy library. It is then determined whether the preset correlation relationship matches the target correlation relationship. If the preset correlation relationship does not match the target correlation relationship, an anomaly detection result is generated to indicate that the changing trend of each target parameter is abnormal.

[0009] Optionally, the formula for calculating the target weighted slope is:

[0010] ;

[0011] in, The slope of the change of the target parameter within the target time window. The total number of data sampling points within the target time window. The time decay weight is... For the first The timestamp of each sampling point The average time of all timestamps within the target time window. For the target parameter in the first... The actual value of the parameter at each sampling point It is the average of the actual values ​​of all parameters of the target parameter within the target time window.

[0012] Optionally, the trend of the target parameter change includes an upward trend, a downward trend, and remaining unchanged;

[0013] Accordingly, determining whether the preset association relationship matches the target association relationship includes:

[0014] If the preset correlation indicates that there is a positive correlation between the target parameters and the target parameters corresponding to each target parameter have the same trend of change, then it is determined that the preset correlation matches the target correlation.

[0015] If the preset correlation relationship indicates that there is a positive correlation between the target parameters, and the trend of change of the target parameters corresponding to each target parameter is different, then it is determined that the preset correlation relationship does not match the target correlation relationship.

[0016] If the preset correlation indicates that there is a negative correlation between the target parameters and the change trends of the target parameters corresponding to each target parameter are opposite to each other, then it is determined that the preset correlation matches the target correlation.

[0017] If the preset correlation indicates that there is a negative correlation between the target parameters, and the change trends of the target parameters are the same, or if any target parameter does not change while the other target parameters change, then it is determined that the preset correlation does not match the target correlation.

[0018] Optionally, the anomaly detection method for the power equipment further includes:

[0019] If none of the target parameters change, then the preset association relationship is determined to match the target association relationship.

[0020] Optionally, the anomaly detection method for the power equipment further includes:

[0021] If the anomaly detection result indicates that the changing trend of each of the target parameters is abnormal, and there is currently no corresponding anomaly alarm, then the anomaly alarm is generated.

[0022] If the anomaly detection result indicates that the changing trend of each of the target parameters is abnormal, and an anomaly alarm has already been issued, then the number of consecutive cycles of the anomaly alarm is recorded.

[0023] If the anomaly detection results indicate that the changing trends of each of the target parameters are not abnormal, and an anomaly alarm has already been issued, then a corresponding regression prompt will be generated.

[0024] Optionally, the anomaly detection method for the power equipment further includes:

[0025] Based on the target parameter sequences and the anomaly detection results, a multidimensional diagnostic matrix corresponding to the target parameters is generated, and the multidimensional diagnostic matrix is ​​displayed on the corresponding visualization diagnostic platform.

[0026] The multidimensional diagnostic matrix includes a real-time curve comparison chart, a historical abnormal data record table, an alarm source tracing view, and a periodic statistical floating prompt.

[0027] Optionally, generating a multidimensional diagnostic matrix corresponding to the target parameters based on each of the target parameter sequences and the anomaly detection results includes:

[0028] Based on each of the target parameter sequences, the real-time data of the main parameter and the comparison parameter in the target parameters are superimposed and displayed in the target coordinate system in the form of curves; the horizontal axis time interval of the target coordinate system is automatically matched with the smallest sampling interval between the sampling interval of the main parameter and the sampling interval of the comparison parameter;

[0029] The time intervals corresponding to the abnormal detection results that characterize the abnormal trend of each of the target parameters are determined as abnormal time intervals, and the abnormal time intervals are marked in the target coordinate system to obtain the real-time curve comparison chart corresponding to the target parameters.

[0030] The historical anomaly data record table includes the first trigger time of the anomaly, the main parameter and the comparison parameter, the segment length of the anomaly time segment, and the change trend of the target parameter.

[0031] Secondly, this application discloses an anomaly detection device for power equipment, applied to an anomaly detection system, comprising:

[0032] The parameter sequence acquisition module is used to determine the target time window and extract the original data of multiple parameters of the target power equipment within the target time window to obtain the target parameter sequence corresponding to each target parameter; the window length of the target time window is determined based on the equipment type and configuration information of the target power equipment, and the target time window is a time window that automatically slides with the system time;

[0033] The trend determination module is used to determine the target slope based on the target parameter sequence and time decay weight using the target weighted slope calculation formula, and to determine the target parameter change trend corresponding to each target parameter based on the target slope.

[0034] The detection result generation module is used to determine the preset correlation between each target parameter from the preset target comparison strategy library, determine whether the preset correlation matches the target correlation, and generate an anomaly detection result indicating that the changing trend of each target parameter is abnormal when the preset correlation does not match the target correlation.

[0035] Thirdly, this application discloses an electronic device, including:

[0036] Memory, used to store computer programs;

[0037] A processor is used to execute the computer program to implement the aforementioned method for detecting abnormalities in power equipment.

[0038] Fourthly, this application discloses a computer-readable storage medium for storing a computer program, wherein the computer program, when executed by a processor, implements the aforementioned method for detecting abnormalities in power equipment.

[0039] In this application, when performing anomaly detection on power equipment, the anomaly detection system determines a target time window and extracts the original multi-parameter data of the target power equipment within the target time window to obtain the target parameter sequence corresponding to each target parameter. The window length of the target time window is determined based on the equipment type and configuration information of the target power equipment, and the target time window is a time window that automatically slides with system time. Using the target weighted slope calculation formula, the target slope is determined based on each target parameter sequence and time decay weight, and the target parameter change trend corresponding to each target parameter is determined based on each target slope, as well as the target correlation relationship between the change trends of each target parameter. The preset correlation relationship between each target parameter is determined from the preset target comparison strategy library, and it is determined whether the preset correlation relationship matches the target correlation relationship. When the preset correlation relationship does not match the target correlation relationship, an anomaly detection result is generated indicating that the change trend of each target parameter is abnormal. As can be seen, this application captures the target parameter sequence corresponding to each target parameter by utilizing a dynamic target time window, and then uses a target slope weighted formula to determine the target slope of each target parameter to determine its changing trend, thereby capturing the direction of change and determining the target correlation between the changing trends of each target parameter. Subsequently, it combines a preset target comparison strategy library to determine whether the target correlation matches the preset correlation, thereby accurately identifying trend deviation anomalies, significantly reducing the failure rate of power equipment and improving dynamic response efficiency. Attached Figure Description

[0040] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0041] Figure 1 This is a flowchart of an anomaly detection method for power equipment disclosed in this application;

[0042] Figure 2 This is a signaling diagram of a specific method for detecting anomalies in power equipment disclosed in this application;

[0043] Figure 3 This is a schematic diagram of a specific trend state extraction process disclosed in this application;

[0044] Figure 4 This is a schematic diagram of a specific alarm decision-making process disclosed in this application;

[0045] Figure 5This is a schematic diagram of the structure of an anomaly detection device for power equipment disclosed in this application;

[0046] Figure 6 This is a structural diagram of an electronic device disclosed in this application. Detailed Implementation

[0047] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0048] With the advancement of new power system construction, equipment operating conditions are becoming increasingly complex and variable, posing a severe challenge to traditional monitoring solutions. For example, when a transformer cooling system malfunctions, the load increases but the temperature drops abnormally; when a hydroelectric turbine experiences cavitation, the gate opening increases but the power generation decreases. Traditional systems relying on single-parameter thresholds or fixed numerical correlation models (such as preset linear equations) cannot effectively identify these latent faults, leading to missed detections. Furthermore, the correlation between equipment parameters (such as the relationship between hydroelectric gate opening and power generation) dynamically changes with external environmental factors (such as water level) and internal conditions. Fixed mathematical models (such as linear equations with fixed slope k and intercept b) lack adaptability, requiring frequent manual parameter adjustments, violating the real-time requirements of power systems, and exhibiting dynamic response lag. To address these technical problems, this application discloses an anomaly detection method for power equipment, which can reduce the missed fault rate and improve dynamic response efficiency.

[0049] See Figure 1 As shown, this embodiment of the invention discloses an anomaly detection method for power equipment, applied to an anomaly detection system, comprising:

[0050] Step S11: Determine the target time window and extract the original data of multiple parameters of the target power equipment within the target time window to obtain the target parameter sequence corresponding to each target parameter; the window length of the target time window is determined based on the equipment type and configuration information of the target power equipment, and the target time window is a time window that automatically slides with the system time.

[0051] In this embodiment, the anomaly detection system can be deployed in the power monitoring backend or edge computing node to realize functions such as data acquisition, dynamic policy library management, trend status calculation, policy matching, intelligent alarm management, visual diagnosis, and configuration management. Figure 2As shown, users need to define the main parameters to be monitored for key equipment (such as the active power on the high-voltage side of #T1_Load), the list of comparison parameters (such as the winding temperature of #T1_WindingTemp and the oil temperature of #T1_OilTemp), the correlation strategy for each parameter pair (such as the winding temperature vs. load: positive correlation; oil temperature vs. load: positive correlation), and the suggested value for the dynamic time window (such as 30 minutes for transformers) through the configuration management interface. The system also associates equipment topology information according to user needs. The anomaly detection system can adapt to changes in correlation relationships under different equipment (such as a 60-minute window for slow transformer response and a 5-minute window for fast turbine response) and different operating conditions, eliminating the need for manual parameter adjustment.

[0052] In this embodiment, as Figure 3 As shown, to detect anomalies in target power equipment, the anomaly detection system needs to acquire the raw data of multiple parameters of the target power equipment for status analysis. Specifically, the window length of the target time window can be determined based on the equipment type and configuration information of the target power equipment. The target time window, which automatically slides with system time, is used to extract the raw data of multiple parameters for each target parameter within the corresponding time period, thereby obtaining the sequence of each target parameter. It is understood that the window length of the target time window can be dynamically adjusted within a certain range, which can be set between 1 minute and 60 minutes. For example, the anomaly detection system can use this target time window to automatically acquire the load and temperature data of the target power equipment 60 minutes before the current time point to determine whether the target power equipment has any anomalies.

[0053] Step S12: Using the target weighted slope calculation formula, determine the target slope based on each target parameter sequence and time decay weight, determine the target parameter change trend corresponding to each target parameter based on each target slope, and determine the target correlation relationship between the change trends of each target parameter.

[0054] In this embodiment, the anomaly detection system calculates the target slope for each target parameter sequence within the target time window. The formula for calculating the target weighted slope is as follows:

[0055] ;

[0056] in, The slope of the target parameter's change within the target time window, also known as the target slope, is used to determine the trend of the target parameter, i.e., the trend of the target parameter's change. The total number of data sampling points within the target time window is determined by the dynamic target time window and the sampling frequency. For time decay weight, For the first The timestamps of each sampling point need to be normalized when used (e.g., converted to seconds from the start of the window). This is the average time of all timestamps within the target time window. . For the target parameter in the first The actual values ​​of parameters (such as temperature and load) at each sampling point are derived from the real-time monitoring data stream. This represents the average of all actual values ​​of the target parameter within the target time window. In one specific implementation, ,in, For the current time, The decay factor is configurable. By adjusting the decay factor, recent data can be given higher weight, thereby enhancing real-time performance.

[0057] In this embodiment, as Figure 3 As shown, after determining each target slope, the target parameter change trend corresponding to each target parameter can be determined based on each target slope, as well as the target correlation relationship between the change trends of each target parameter. The target parameter change trend includes an upward trend, a downward trend, and remaining unchanged. When determining the trend state, the anomaly detection system will base its judgment on the target slope. The system automatically classifies the changing trends of target parameters based on their numerical range. For example, setting a threshold range of 0.001, when the target slope value is greater than 0.001, the target parameter corresponding to that slope is considered to have an upward trend; when the target slope value is less than -0.001, the target parameter corresponding to that slope is determined to have a downward trend; and when the absolute value of the target slope is not greater than 0.001, the target parameter corresponding to that slope is considered to have a constant trend. The changing trends of each target parameter can then be used to determine the target correlation between them.

[0058] Step S13: Determine the preset correlation relationship between each target parameter from the preset target comparison strategy library, determine whether the preset correlation relationship matches the target correlation relationship, and generate an anomaly detection result indicating that the change trend of each target parameter is abnormal when the preset correlation relationship does not match the target correlation relationship.

[0059] In this embodiment, the anomaly detection system determines the preset correlation relationships between each target parameter from a preset target comparison strategy library, and then determines whether the preset correlation relationship matches the target correlation relationship. Specifically, this can include: if the preset correlation relationship indicates a positive correlation between each target parameter and the target parameter changes in the same direction, then the preset correlation relationship matches the target correlation relationship; if the preset correlation relationship indicates a positive correlation between each target parameter and the target parameter changes in different directions, then the preset correlation relationship does not match the target correlation relationship; if the preset correlation relationship indicates a negative correlation between each target parameter and the target parameter changes in opposite directions, then the preset correlation relationship matches the target correlation relationship; if the preset correlation relationship indicates a negative correlation between each target parameter and the target parameter changes in the same direction, or if any target parameter remains unchanged while the others change, then the preset correlation relationship does not match the target correlation relationship. It can be understood that if none of the target parameters change, then the preset correlation relationship matches the target correlation relationship.

[0060] In this embodiment, if the preset correlation does not match the target correlation, an anomaly detection result is generated indicating that the changing trends of each target parameter are abnormal. Conversely, if the preset correlation matches the target correlation, an anomaly detection result is generated indicating that the changing trends of each target parameter are not abnormal. This improves the numerical comparison in the traditional anomaly detection system to state matching, reducing the false alarm rate for power equipment faults and improving dynamic response efficiency. In a specific implementation, through a trend state machine and a strategy matching engine, this embodiment can accurately identify trend deviation anomalies such as "load increase and temperature decrease" that cannot be captured by traditional numerical thresholds or fixed models. Engineering measurements (hydropower station) show that the false alarm rate has decreased from 23% in the traditional scheme to 6.7%, and the false alarm rate has decreased significantly.

[0061] like Figure 4As shown, to address the alarm storm problem, this embodiment discloses an alarm decision tree, which can include three alarm decision mechanisms: initial trigger latch, continuous anomaly accumulation, and automatic recovery mechanism. Specifically, initial trigger latch means that for the same detected abnormal event (referring to the same set of parameter pairs and the same abnormal cause), the system only generates and sends an alarm message when it is first detected as not conforming to the policy. In subsequent scan cycles, as long as the anomaly persists, no alarm is generated again. Continuous anomaly accumulation means that the system maintains a counter to record the number of scan cycles in which the abnormal event continues to occur. This value can be used to quantify the severity and duration of the fault and is displayed in a visual interface (such as a floating prompt). The automatic recovery mechanism means that when the system detects in a subsequent scan that the trend state of the set of parameters once again conforms to the configured policy, it determines that the anomaly has disappeared. The system immediately generates and pushes a recovery notification (such as "XXX anomaly has been recovered"), while simultaneously clearing the alarm status and accumulator count of the abnormal event. In other words, if the anomaly detection results indicate an abnormal trend in the changing trends of each target parameter, and there is currently no corresponding anomaly alarm, then an anomaly alarm is generated; if the anomaly detection results indicate an abnormal trend in the changing trends of each target parameter, and there is already an anomaly alarm, then the number of consecutive cycles of the anomaly alarm is recorded; if the anomaly detection results indicate no abnormal trend in the changing trends of each target parameter, and there is already an anomaly alarm, then a corresponding recovery prompt is generated. Initial trigger latching ensures that each anomaly event generates only one alarm, thus resolving alarm storms; continuous anomaly accumulation allows the severity of the fault to be quantified by recording the number of consecutive anomaly cycles; the automatic recovery mechanism proactively pushes a recovery notification when the anomaly disappears. For example, if the trend analysis program of the anomaly detection system starts at 9:00 AM, and each detection cycle is one hour, then 9:00 AM to 10:00 AM constitutes one detection cycle. The trend analysis program analyzes the anomalies of the target electronic device in real time and generates corresponding anomaly detection results. When the anomaly detection results indicate a change in the state of the target electronic device (from normal to anomaly, or from anomaly to normal), the corresponding anomaly detection results are proactively pushed to the alarm program. Meanwhile, the alarm program proactively acquires trend analysis results (i.e., anomaly detection results) once per detection cycle. If the anomaly persists, the anomaly time is accumulated and the severity level of the anomaly is increased; if the anomaly recovers, a corresponding reset prompt is issued, and the accumulated level is reset to zero. Under this premise, if the anomaly lasts from 9:30 to 10:30, spanning two detection cycles, in one specific implementation, the intelligent alarm suppression mechanism (first trigger + automatic reset) ensures the uniqueness and clarity of alarm signals, reducing the average daily number of duplicate alarms from 17 to 2, preventing critical signals from being overwhelmed.

[0062] In this embodiment, as Figure 2As shown, a multidimensional diagnostic matrix corresponding to the target parameters can also be generated based on the sequence of each target parameter and the anomaly detection results, and displayed on the corresponding visualization diagnostic platform. The multidimensional diagnostic matrix includes a real-time curve comparison chart, a historical anomaly data record table, an alarm source tracing view, and a periodic statistical floating prompt. Generating the multidimensional diagnostic matrix based on the sequence of each target parameter and the anomaly detection results includes: displaying the real-time data of the main parameter and the comparison parameter in the target parameters as curves overlaid on the target coordinate system based on the sequence of each target parameter; automatically matching the minimum sampling interval between the sampling interval of the main parameter and the sampling interval of the comparison parameter on the horizontal axis of the target coordinate system; identifying the time segment corresponding to the anomaly detection results that characterize the changing trend of each target parameter as the abnormal time segment, and marking the abnormal time segment in the target coordinate system to obtain the real-time curve comparison chart corresponding to the target parameter; wherein, the historical anomaly data record table includes the first trigger time of the anomaly, the main parameter and the comparison parameter, the segment length of the abnormal time segment, and the changing trend of the target parameter.

[0063] In one specific embodiment, a real-time curve comparison chart can be obtained through the following steps:

[0064] a) Display the real-time data of the main parameter (such as load) and the comparative parameter (such as temperature) as a curve on the same coordinate system.

[0065] b) Dynamic coordinate adaptation: The horizontal axis time interval is automatically matched to the smallest sampling interval between the two parameters.

[0066] c) Highlighting abnormal time periods: Use a prominent color (such as red) to mark the time periods where the strategy failed to match.

[0067] In one specific embodiment, the historical anomaly data record table displays records of historical anomalies occurring in the target electronic device in list or table format. Key fields include the first trigger time of the anomaly (timestamp of the first trigger), the main parameter and the comparison parameter (parameter pair), the length of the anomaly time segment (number of consecutive cycles, used to record the total number of scan cycles for the anomaly event), and the target parameter change trend (i.e., a snapshot of the effective strategy, used to record the positive / negative correlation strategies in effect during analysis). Figure 4 As shown, the historical anomaly data record table also has a double-click reproduction function. If the user double-clicks a record row, the system will automatically call up historical data to reconstruct the analysis scenario and graph at that time.

[0068] In this embodiment, the alarm tracing view, combined with power grid topology information, visually displays the location of alarm devices and their upstream and downstream related devices on a single-line diagram or topology diagram. This alarm tracing view allows users to quickly locate the real-time monitoring or historical analysis interface of an alarm device by clicking on its node. Furthermore, a periodic statistics hover tooltip displays the duration of the abnormal event (based on an accumulator) and other statistical information by hovering the mouse over the abnormal period / record in the curve graph or anomaly list. The multi-dimensional diagnostic matrix, composed of the real-time curve comparison chart, historical anomaly data record table, alarm tracing view, and periodic statistics hover tooltip, provides maintenance personnel with a comprehensive and efficient analysis tool. In practical applications, the dimensions of the multi-dimensional diagnostic matrix can be added or removed as needed. In one specific implementation, the multi-dimensional visual diagnostic matrix, especially the double-click historical penetration reproduction function, reduces fault location time from an average of 45 minutes to 8 minutes (a decrease of 82%), and historical data analysis time is reduced from several hours manually to 30 seconds automatically, resulting in a significant leap in analysis efficiency.

[0069] In a specific implementation, such as a transformer monitoring scenario, anomaly detection is typically performed through the following steps:

[0070] 1) Configuration parameters:

[0071] a) Main parameter telemetry ID: #T1_Load;

[0072] b) Comparison parameter telemetry ID: #T1_WindingTemp;

[0073] c) Comparison strategy: positive correlation;

[0074] d) Time window: 30 minutes;

[0075] 2) Abnormal triggering:

[0076] a) Scenario: In a certain scan window, the load (#T1_Load) data shows an upward trend (load increases by about 10%), while the winding temperature (#T1_WindingTemp) data shows a downward trend (temperature decreases by about 2℃).

[0077] b) Strategy matching: The configured strategy is "positive correlation" (requiring changes in the same direction), but the actual state is "mainly rising vs. relative falling" → does not conform to the strategy (abnormal).

[0078] c) Alarm generation: When the system detects this anomaly for the first time, an alarm is generated: #T1_Load / #T1_WindingTemp trend comparison anomaly.

[0079] d) Visualization: The 4D diagnostic interface automatically pops up (or highlights a prompt), displaying a hyperbola comparison graph of #T1_Load and #T1_WindingTemp, with abnormal periods (load increase, temperature decrease) highlighted in red. The continuous anomaly accumulator starts counting.

[0080] 3) Historical Analysis:

[0081] a) Maintenance personnel checked the records of transformer #T1 for the past 7 days and found 3 similar "load increase and temperature decrease" abnormal events.

[0082] b) Double-click to reproduce: Double-click one of the records → the system will automatically reconstruct the analysis scenario and graph at that time.

[0083] c) Locating the root cause: Based on the occurrence time, it was found that all three anomalies occurred during the period when the cooling system fan was malfunctioning, allowing for a quick identification of the root cause of the fault.

[0084] 4) Expanding application scenarios:

[0085] a) Transformer: Load current and oil temperature (positive correlation); tap changer operation frequency and winding temperature (specific strategy).

[0086] b) Hydropower station: Gate opening and power generation (negative correlation - power should increase with increasing opening; if power decreases with increasing opening, cavitation may occur).

[0087] c) Photovoltaic station: Irradiance and inverter output power (positively correlated); Ambient temperature vs. inverter efficiency (specific model).

[0088] As can be seen, this application captures the target parameter sequence corresponding to each target parameter by utilizing a dynamic target time window, and then uses a target slope weighted formula to determine the target slope of each target parameter to determine its changing trend, thereby capturing the direction of change and determining the target correlation between the changing trends of each target parameter. Subsequently, it combines a preset target comparison strategy library to determine whether the target correlation matches the preset correlation, thereby accurately identifying trend deviation anomalies, significantly reducing the failure rate of power equipment and improving dynamic response efficiency.

[0089] See Figure 5 As shown, this application discloses an anomaly detection device for power equipment, applied in an anomaly detection system, comprising:

[0090] The parameter sequence acquisition module 11 is used to determine the target time window and extract the original data of multiple parameters of the target power equipment within the target time window to obtain the target parameter sequence corresponding to each target parameter; the window length of the target time window is determined based on the equipment type and configuration information of the target power equipment, and the target time window is a time window that automatically slides with the system time.

[0091] The trend determination module 12 is used to determine the target slope based on the target parameter sequence and time decay weight using the target weighted slope calculation formula, and to determine the target parameter change trend corresponding to each target parameter based on the target slope.

[0092] The detection result generation module 13 is used to determine the preset correlation relationship between each target parameter from the preset target comparison strategy library, determine whether the preset correlation relationship matches the target correlation relationship, and generate an anomaly detection result indicating that the change trend of each target parameter is abnormal when the preset correlation relationship does not match the target correlation relationship.

[0093] As can be seen, this application captures the target parameter sequence corresponding to each target parameter by utilizing a dynamic target time window, and then uses a target slope weighted formula to determine the target slope of each target parameter to determine its changing trend, thereby capturing the direction of change and determining the target correlation between the changing trends of each target parameter. Subsequently, it combines a preset target comparison strategy library to determine whether the target correlation matches the preset correlation, thereby accurately identifying trend deviation anomalies, significantly reducing the failure rate of power equipment and improving dynamic response efficiency.

[0094] In one specific implementation, the trend of the target parameter change may specifically include an upward trend, a downward trend, or remaining unchanged;

[0095] Accordingly, the detection result generation module 13 may specifically include:

[0096] The first judgment unit is used to determine that the preset correlation relationship matches the target correlation relationship if the preset correlation relationship represents a positive correlation between the target parameters and the target parameters corresponding to the target parameters have the same trend of change.

[0097] The second judgment unit is used to determine that the preset correlation relationship does not match the target correlation relationship if the preset correlation relationship represents a positive correlation between each of the target parameters and the target parameter change trends corresponding to each of the target parameters are not the same.

[0098] The third judgment unit is used to determine that the preset correlation relationship matches the target correlation relationship if the preset correlation relationship represents a negative correlation between each of the target parameters and the change trends of the target parameters corresponding to each of the target parameters are opposite to each other.

[0099] The fourth judgment unit is used to determine that the preset correlation relationship does not match the target correlation relationship if the preset correlation relationship represents a negative correlation between the target parameters and the change trends of the target parameters are the same or any target parameter does not change while the other target parameters change.

[0100] In one specific embodiment, the device may further include:

[0101] The matching relationship determination module is used to determine that the preset association relationship matches the target association relationship if none of the target parameters change.

[0102] In one specific embodiment, the device may further include:

[0103] The first alarm operation module is used to generate the abnormal alarm prompt if the abnormal detection result indicates that the changing trend of each of the target parameters is abnormal and there is no corresponding abnormal alarm prompt at present.

[0104] The second alarm operation module is used to record the number of consecutive cycles of the abnormal alarm if the abnormal detection result indicates that the changing trend of each of the target parameters is abnormal and the abnormal alarm prompt already exists.

[0105] The regression module is used to generate a corresponding regression prompt if the anomaly detection result indicates that the changing trend of each of the target parameters is not abnormal, and the anomaly alarm prompt already exists.

[0106] In one specific embodiment, the device may further include:

[0107] The diagnostic matrix display module is used to generate a multidimensional diagnostic matrix corresponding to each target parameter based on each target parameter sequence and the anomaly detection result, and to display the multidimensional diagnostic matrix on the corresponding visualization diagnostic platform.

[0108] The multidimensional diagnostic matrix includes a real-time curve comparison chart, a historical abnormal data record table, an alarm source tracing view, and a periodic statistical floating prompt.

[0109] In one specific implementation, the diagnostic matrix display module may include:

[0110] The curve display unit is used to overlay and display the real-time data of the main parameter and the comparison parameter in the target parameters in the form of curves in the target coordinate system based on each target parameter sequence; the horizontal axis time interval of the target coordinate system is automatically matched with the smallest sampling interval between the sampling interval of the main parameter and the sampling interval of the comparison parameter;

[0111] The curve annotation unit is used to determine the time segment corresponding to the abnormal detection result that represents the abnormal change trend of each of the target parameters as the abnormal time segment, and to annotate the abnormal time segment in the target coordinate system to obtain the real-time curve comparison chart corresponding to the target parameter.

[0112] The historical anomaly data record table includes the first trigger time of the anomaly, the main parameter and the comparison parameter, the segment length of the anomaly time segment, and the change trend of the target parameter.

[0113] Furthermore, embodiments of this application also disclose an electronic device, Figure 6 This is a structural diagram of an electronic device 20 according to an exemplary embodiment. The content of the diagram should not be construed as limiting the scope of this application.

[0114] Figure 6 This is a schematic diagram of the structure of an electronic device 20 provided in an embodiment of this application. Specifically, the electronic device 20 may include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input / output interface 25, and a communication bus 26. The memory 22 stores a computer program, which is loaded and executed by the processor 21 to implement the relevant steps in the abnormal detection method for power equipment disclosed in any of the foregoing embodiments. Alternatively, the electronic device 20 in this embodiment may specifically be an electronic computer.

[0115] In this embodiment, the power supply 23 is used to provide operating voltage for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and external devices, and the communication protocol it follows can be any communication protocol applicable to the technical solution of this application, and is not specifically limited here; the input / output interface 25 is used to acquire external input data or output data to the outside world, and its specific interface type can be selected according to specific application needs, and is not specifically limited here.

[0116] In addition, the memory 22, as a carrier for resource storage, can be a read-only memory, random access memory, disk, or optical disk, etc. The resources stored thereon can include an operating system 221, computer programs 222, etc., and the storage method can be temporary storage or permanent storage.

[0117] The operating system 221 is used to manage and control the various hardware devices on the electronic device 20 and the computer program 222, which may be Windows Server, Netware, Unix, Linux, etc. In addition to including a computer program capable of performing the power equipment anomaly detection method executed by the electronic device 20 as disclosed in any of the foregoing embodiments, the computer program 222 may further include a computer program capable of performing other specific tasks.

[0118] Furthermore, this application also discloses a computer-readable storage medium for storing a computer program; wherein, when the computer program is executed by a processor, it implements the aforementioned method for detecting anomalies in power equipment. Specific steps of this method can be found in the corresponding content disclosed in the foregoing embodiments, and will not be repeated here.

[0119] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to in the method section.

[0120] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0121] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented directly by hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0122] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0123] The technical solutions provided in this application have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for detecting anomalies in power equipment, characterized in that, Applications include anomaly detection systems, including: A target time window is determined, and the original data of multiple parameters of the target power equipment within the target time window are truncated to obtain the target parameter sequence corresponding to each target parameter; the window length of the target time window is determined based on the equipment type and configuration information of the target power equipment, and the target time window is a time window that automatically slides with the system time; Using the target weighted slope calculation formula, the target slope is determined based on the sequence of each target parameter and the time decay weight, and the target parameter change trend corresponding to each target parameter is determined based on each target slope, as well as the target correlation relationship between the change trends of each target parameter; The preset correlation relationship between each target parameter is determined from the preset target comparison strategy library. It is then determined whether the preset correlation relationship matches the target correlation relationship. If the preset correlation relationship does not match the target correlation relationship, an anomaly detection result is generated to indicate that the changing trend of each target parameter is abnormal.

2. The method for detecting abnormalities in power equipment according to claim 1, characterized in that, The formula for calculating the target weighted slope is: ； in, The slope of the change of the target parameter within the target time window. The total number of data sampling points within the target time window. The time decay weight is... For the first The timestamp of each sampling point The average time of all timestamps within the target time window. For the target parameter in the first... The actual value of the parameter at each sampling point It is the average of the actual values ​​of all parameters of the target parameter within the target time window.

3. The method for detecting abnormalities in power equipment according to claim 1, characterized in that, The trend of the target parameter includes an upward trend, a downward trend, and remaining unchanged; Accordingly, determining whether the preset association relationship matches the target association relationship includes: If the preset correlation indicates that there is a positive correlation between the target parameters and the target parameters corresponding to each target parameter have the same trend of change, then it is determined that the preset correlation matches the target correlation. If the preset correlation relationship indicates that there is a positive correlation between the target parameters, and the trend of change of the target parameters corresponding to each target parameter is different, then it is determined that the preset correlation relationship does not match the target correlation relationship. If the preset correlation indicates that there is a negative correlation between the target parameters and the change trends of the target parameters corresponding to each target parameter are opposite to each other, then it is determined that the preset correlation matches the target correlation. If the preset correlation indicates that there is a negative correlation between the target parameters, and the change trends of the target parameters are the same, or if any target parameter does not change while the other target parameters change, then it is determined that the preset correlation does not match the target correlation.

4. The method for detecting abnormalities in power equipment according to claim 1, characterized in that, Also includes: If none of the target parameters change, then the preset association relationship is determined to match the target association relationship.

5. The method for detecting abnormalities in power equipment according to claim 1, characterized in that, Also includes: If the anomaly detection result indicates that the changing trend of each of the target parameters is abnormal, and there is currently no corresponding anomaly alarm, then the anomaly alarm is generated. If the anomaly detection result indicates that the changing trend of each of the target parameters is abnormal, and an anomaly alarm has already been issued, then the number of consecutive cycles of the anomaly alarm is recorded. If the anomaly detection results indicate that the changing trends of each of the target parameters are not abnormal, and an anomaly alarm has already been issued, then a corresponding regression prompt will be generated.

6. The method for detecting anomalies in power equipment according to any one of claims 1 to 5, characterized in that, Also includes: Based on the target parameter sequences and the anomaly detection results, a multidimensional diagnostic matrix corresponding to the target parameters is generated, and the multidimensional diagnostic matrix is ​​displayed on the corresponding visualization diagnostic platform. The multidimensional diagnostic matrix includes a real-time curve comparison chart, a historical abnormal data record table, an alarm source tracing view, and a periodic statistical floating prompt.

7. The method for detecting abnormalities in power equipment according to claim 6, characterized in that, The step of generating a multidimensional diagnostic matrix corresponding to each target parameter based on the target parameter sequence and the anomaly detection result includes: Based on each of the target parameter sequences, the real-time data of the main parameter and the comparison parameter in the target parameters are superimposed and displayed in the target coordinate system in the form of curves; the horizontal axis time interval of the target coordinate system is automatically matched with the smallest sampling interval between the sampling interval of the main parameter and the sampling interval of the comparison parameter; The time intervals corresponding to the abnormal detection results that characterize the abnormal trend of each of the target parameters are determined as abnormal time intervals, and the abnormal time intervals are marked in the target coordinate system to obtain the real-time curve comparison chart corresponding to the target parameters. The historical anomaly data record table includes the first trigger time of the anomaly, the main parameter and the comparison parameter, the segment length of the anomaly time segment, and the change trend of the target parameter.

8. An anomaly detection device for power equipment, characterized in that, Applications include anomaly detection systems, including: The parameter sequence acquisition module is used to determine the target time window and extract the original data of multiple parameters of the target power equipment within the target time window to obtain the target parameter sequence corresponding to each target parameter; the window length of the target time window is determined based on the equipment type and configuration information of the target power equipment, and the target time window is a time window that automatically slides with the system time; The trend determination module is used to determine the target slope based on the target parameter sequence and time decay weight using the target weighted slope calculation formula, and to determine the target parameter change trend corresponding to each target parameter based on the target slope. The detection result generation module is used to determine the preset correlation between each target parameter from the preset target comparison strategy library, determine whether the preset correlation matches the target correlation, and generate an anomaly detection result indicating that the changing trend of each target parameter is abnormal when the preset correlation does not match the target correlation.

9. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor for executing the computer program to implement the abnormal detection method for power equipment as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, Used to store a computer program, wherein the computer program, when executed by a processor, implements the abnormal detection method for power equipment as described in any one of claims 1 to 7.

Images