A power system dynamic recording wave analysis method based on fault feature self-adaption

Abstract

The application discloses a kind of power system dynamic recording wave analysis methods based on fault feature self-adaption, it is related to dynamic recording wave analysis technical field. Including: through synchronous sampling device, dynamic data in power system is recorded, obtains dynamic data information item, based on dynamic information data item, obtains the fluctuation information of dynamic data, obtains fluctuation information generation data, obtains fluctuation information set;Based on fluctuation information set carries out time stamp alignment, obtains mark time item, based on mark time item obtains corresponding state of power system, obtains power feature item.The application obtains power feature information by recording dynamic data in power system, and carries out fault determination by this, obtains fault position and fault name, by taking power feature information and fault information as training data to carry out model training, creates fault fluctuation matching model, and then realizes power system analysis based on fault feature self-regulation type.

Description

Technical Field

[0001] This invention relates to the field of dynamic waveform analysis technology, specifically to a dynamic waveform analysis method for power systems based on fault characteristic adaptation. Background Technology

[0002] Dynamic waveform analysis of power systems refers to a set of methods for recording, extracting, and analyzing the dynamic behavior of power systems during operation. Dynamic waveform recording uses sampling devices such as data acquisition units, relay protection devices, protection and control systems, and phasor measurement units (PMUs) to record signals such as voltage, current, and frequency in the system at high-speed time resolution. The waveform data is generally a continuous time series, covering the entire process from fault occurrence and transient response to stable operation. The waveform data undergoes preprocessing such as denoising, baseline correction, and phase alignment, followed by time-domain and frequency-domain analysis to reveal the system's oscillation modes, modal characteristics, impedance changes, etc.

[0003] A method for inter-station time difference analysis based on fault recording data, disclosed in patent publication number CN106950445A, describes a process where, after a fault occurs on a line, the fault signal propagates along the lines TL_a and TL_b to the two end stations A and B. Once the fault signal reaches the respective station recorders Tr_a and Tr_b, the recorders record waveforms according to predefined logic conditions. The recorders continuously collect and buffer electrical quantities, including voltage and current, but do not create waveform files. When a fault occurs, the recorders also store data from a period prior to the fault time. Different recorders have different response times to fault recording, resulting in different timestamps from the fault point to the start of the fault. By locating the relative time of the fault point within the waveform recording, precise time synchronization between the two ends of the line can be achieved.

[0004] In the process of waveform analysis, the above-mentioned and similar technical solutions often rely on pre-set parameters and thresholds. These parameters and thresholds are usually based on experience summaries of specific fault types and system operating conditions. However, the complexity of power systems and the diversity of operating environments make it difficult for single parameters and thresholds to adapt to all situations. If fixed parameters and thresholds are still used for analysis, it may lead to misjudgment, omission of faults, or even wrong decisions, thereby affecting the safe and stable operation of the power system. Summary of the Invention

[0005] The purpose of this invention is to provide a dynamic waveform analysis method for power systems based on fault characteristics to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a power system dynamic waveform analysis method based on fault characteristic adaptation, comprising:

[0007] Dynamic data in the power system is recorded by a synchronous sampling device to obtain dynamic data information items. Based on the dynamic information data items, the fluctuation information of the dynamic data is obtained, and the fluctuation information generates data to obtain a fluctuation information set.

[0008] Timestamp alignment is performed based on the fluctuation information set to obtain marked time items. The corresponding state of the power system is obtained based on the marked time items to obtain power feature items.

[0009] Fault determination is based on power feature items. When the power system is in fault determination, fault data is acquired to obtain power fault items. Power fault items correspond to the fluctuation information set corresponding to the marked time items. The fluctuation items in the fluctuation information set are weighted and then the fault information weight items are obtained.

[0010] When the fluctuation information of dynamic data is obtained again, the fluctuation information update set, the marker time update set, the power feature update item and the power fault update item are obtained. The fluctuation items in the fluctuation information update set are weighted and then the fault information weight update item is obtained.

[0011] The matching attributes of power feature items and power feature update items are determined to obtain feature matching items. Fault information weight items and fault information weight update items are merged and filtered to select and save weights, resulting in weight saving items. The fluctuation items corresponding to the saved weights are obtained as the corresponding fluctuations of power feature items and power feature update items. These are used as training data for model training to create a fault fluctuation matching model. Dynamic analysis is performed based on the fluctuation information of the fault fluctuation matching model and dynamic data, thereby realizing power system analysis based on fault feature self-adjustment.

[0012] Furthermore, the dynamic data includes voltage, current, and frequency signals, and the method for obtaining dynamic data information items includes:

[0013] The synchronous sampling device is configured to consist of a high-precision synchronous phasor measurement unit, a data acquisition and storage unit, and a data interface terminal for the protection and control system.

[0014] The high-precision synchronous phasor measurement unit provides three-phase voltage, three-phase current and frequency output. The data acquisition and storage unit performs high-speed sampling, buffering and local storage of field voltage, current and frequency signals to provide raw waveform data. The protection and control system data interface provides alarm, event recording, trip protection action time point and local measurement data, thereby recording dynamic data in the power system and obtaining dynamic data information items.

[0015] Furthermore, the method for obtaining the fluctuation information set includes:

[0016] Set a time window value, which is a fixed time value. Based on the time window value, obtain the average value of the dynamic data in the dynamic data information item to obtain the dynamic data average item.

[0017] A fluctuation threshold is set, which is a fixed range of fluctuations in the dynamic data. The fluctuation information of the average item of the dynamic data is judged based on the fluctuation threshold. The average item value of the dynamic data that exceeds the fluctuation threshold is obtained, and the corresponding dynamic data is obtained, thus obtaining the fluctuation information set.

[0018] Furthermore, the corresponding state of the power system includes environmental state and operating state, and the method for obtaining power characteristic items includes:

[0019] The status acquisition module is configured, which includes an environment status acquisition unit and a runtime status acquisition unit.

[0020] Using the location of the power system as the determination point, the environmental state information of the determination point is obtained based on the environmental state acquisition unit, including environmental temperature, environmental humidity and environmental vibration, to obtain environmental feature items;

[0021] Based on the operation status acquisition unit, the power system state stability information is obtained, including current stability, voltage stability and frequency stability, and operation status characteristic items are obtained. The environmental characteristic items and operation status characteristic items are combined to obtain power characteristic items.

[0022] Furthermore, the method for obtaining the environmental state information of the determination point includes:

[0023] Obtain the attribute information of the target power system, including model information and production batch information, to obtain power attribute items;

[0024] Based on power attribute items, information on the impact of power attributes is obtained through historical data and big data acquisition methods, thereby obtaining the system's affected items;

[0025] Based on the affected items of the system, an affected threshold is set. The affected threshold is a fixed percentage threshold. The affected items of the system are screened based on the affected threshold to obtain the retained affected information. The retained affected information is used as the environmental state information of the decision point.

[0026] Furthermore, the fault data includes the fault location and fault name, and the method for obtaining power fault items includes:

[0027] Based on power characteristic items, fault characteristics corresponding to power characteristic items are obtained through big data acquisition and historical data acquisition, and fault name information is obtained;

[0028] Based on power characteristic items, the location of the characteristic is obtained through fault analysis equipment to obtain fault location information. The fault name information and fault location information are combined to obtain power fault items.

[0029] Furthermore, the method for obtaining the fault information weight term includes:

[0030] The fluctuation range of the fluctuation items in the fluctuation information set is obtained, and at least one fluctuation ratio item is obtained. A division weight value is set based on the fluctuation ratio item. The division weight value is the same as the proportion of the fluctuation ratio item. The fluctuation items are weighted based on the division weight value, and then the fault information weight item is obtained.

[0031] Furthermore, the matching attributes between the power feature items and the power feature update items include the degree of matching between fault location and fault name, and the method for obtaining the feature matching items includes:

[0032] Based on power feature items and power feature update items, fault location information and fault name information are compared separately. A range expansion value is set, which is a fixed ratio value. Based on the range expansion value, the fault location information of power feature items and power feature update items are combined to obtain a first fault range and a second fault range. It is then determined whether the fault location information of power feature items and power feature update items matches the first fault range and the second fault range, and the matching attributes of power feature items and power feature update items are judged.

[0033] Furthermore, the method for obtaining the weight storage item includes:

[0034] Set a minimum weight retention value, which is a fixed percentage. Based on the minimum weight retention value, merge and filter the fault information weight items and fault information weight update items, and remove the merged items that are lower than the minimum weight retention value to obtain the weight retention items.

[0035] Compared with the prior art, the beneficial effects of the present invention are:

[0036] This fault-feature-adaptive dynamic waveform analysis method for power systems records dynamic data in the power system to obtain power feature information, which is then used to determine faults, obtain fault locations and names, and train a model using power feature information and fault information as training data to create a fault fluctuation matching model. Based on the fault fluctuation matching model and the fluctuation information of dynamic data, dynamic analysis is performed. When a fault of a certain type or location occurs, the fault fluctuation matching model can quickly locate the power feature of interest. Conversely, when an abnormal fluctuation of a certain power feature occurs, the fault fluctuation matching model can quickly locate the fault type and fault location, thereby realizing a fault feature-based self-adjusting power system analysis. Attached Figure Description

[0037] Figure 1 This is a schematic diagram of the overall process of the present invention;

[0038] Figure 2 This is a schematic diagram illustrating the relationship between dynamic data and the synchronous sampling device of the present invention;

[0039] Figure 3 This is a schematic diagram of the dynamic data averaging process of the present invention;

[0040] Figure 4 This is a schematic diagram of the process for obtaining environmental state information at the determination point in this invention.

[0041] Figure 5 This is a schematic diagram illustrating the matching attribute judgment between the power feature items and the power feature update items of the present invention. Detailed Implementation

[0042] 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.

[0043] The complexity of power systems and the diversity of operating environments make it difficult for single parameters and thresholds to adapt to all situations. For example, different fault types, such as short circuits, grounding faults, and open circuits, have different electrical characteristics, and different equipment, such as transformers, generators, and transmission lines, have different response characteristics to faults. Furthermore, different system conditions, such as load levels and operating modes, also affect the presentation of fault characteristics. If fixed parameters and thresholds are still used for analysis, it may lead to misjudgment, missed judgment, or even incorrect decisions, affecting the safe and stable operation of the power system. The technical solution provided in this application obtains power characteristic information by recording dynamic data in the power system, and uses this information to determine faults, obtain fault locations and names, and trains a fault fluctuation matching model using power characteristic information and fault information as training data. Based on the fault fluctuation matching model and the fluctuation information of dynamic data, dynamic analysis is performed. When a fault of a certain type or location occurs, the fault fluctuation matching model quickly locates the power characteristics of interest. Conversely, when an abnormal fluctuation of a certain power characteristic occurs, the fault fluctuation matching model quickly locates the fault type and location, thereby realizing a fault characteristic-based self-regulating power system analysis. Figure 1 As shown, it includes steps S100-S900.

[0044] Step S100: Record the dynamic data in the power system through a synchronous sampling device to obtain dynamic data information items.

[0045] It is important to note that, such as Figure 2 As shown, the dynamic data includes voltage, current, and frequency signals. The method for acquiring dynamic data information items includes: setting up a synchronous sampling device, including a high-precision synchronous phasor measurement unit, a data acquisition and storage unit, and a protection and control system data interface terminal; providing three-phase voltage, three-phase current, and frequency outputs based on the high-precision synchronous phasor measurement unit; performing high-speed sampling, buffering, and local storage of field voltage, current, and frequency signals based on the data acquisition and storage unit to provide raw waveform data; and providing alarm, event recording, trip protection action time points, and local measurement data based on the protection and control system data interface terminal, thereby recording the dynamic data in the power system to obtain dynamic data information items.

[0046] Specifically, the high-precision synchronous phasor measurement unit is used to provide three-phase voltage, three-phase current, phase, frequency, and phasor data. The time synchronization accuracy is usually in the microsecond range. At the same time, models with self-diagnosis, data quality scoring, and time alignment verification should be selected as much as possible to facilitate subsequent adaptive analysis. The sampling rate should be at least 60 Hz baseline, and 256Hz, 512Hz, or even 1kHz can be selected for transients. The data acquisition and storage unit is used to sample, buffer, and store field voltage, current, and other sensor signals at high speed, providing raw waveform data. The sampling rate should be compatible with or higher than that of the high-precision synchronous phasor measurement unit, at least 1–2kHz, for short-term transients. The protection and control system data interface is used to provide alarm, event recording, trip, protection action time point, and local measurement data. The interface should be compatible with IEC 61850, MODBUS, and DNP3, with priority given to IEC 61850 MMS / GOOSE.

[0047] Step S200: Based on the dynamic information data items, obtain the fluctuation information of the dynamic data, obtain the data generated by the fluctuation information, and obtain the fluctuation information set.

[0048] It should be noted that the method for obtaining the fluctuation information set includes: setting a time window value, which is a fixed time value; based on the time window value, obtaining the average value of the dynamic data in the dynamic data information item to obtain the dynamic data average item; setting a fluctuation threshold, which is a fixed fluctuation range value of the dynamic data; based on the fluctuation threshold, judging the fluctuation information of the dynamic data average item, obtaining the value of the dynamic data average item that exceeds the fluctuation threshold, obtaining the corresponding dynamic data, and thus obtaining the fluctuation information set.

[0049] Specifically, such as Figure 3As shown, the set time window value is 5s. The average value of different dynamic data in the dynamic data information items within 5s is obtained, and at least two dynamic data average items are obtained. For example, when the dynamic data is voltage, current and frequency signals, there are three dynamic data average items, namely voltage data average item, current data average item and frequency signal data average item. The set fluctuation threshold is ±4%. When the value of the dynamic data average item exceeds the set fluctuation threshold, the corresponding dynamic data is obtained, and thus the fluctuation information set is obtained.

[0050] In the specific implementation process, it is now necessary to obtain the fluctuation information of dynamic data in a certain power system. The dynamic data of the power system to be obtained includes voltage, current, and frequency signals. At the same time, the specified voltage value is 345kV, the specified current value is 1.2kA, and the frequency fluctuation threshold is 50Hz. According to the set time window, the average values ​​of voltage, current, and frequency signals are obtained as shown in Table 1.

[0051] Table 1:

[0052] Time window Voltage / kV Current / kA Frequency / Hz determination 1 348.2 1.18 50.02 / 2 332.5 1.25 49.97 Current trigger 3 341.7 1.32 49.72 Current trigger 4 362.8 1.21 50.03 Voltage trigger

[0053] Since the set fluctuation threshold is ±5%, there is no trigger in time window 1. In time window 2, the current fluctuation value is 4.2%, exceeding the fluctuation threshold. In time window 3, the current fluctuation is 10%, exceeding the fluctuation threshold. In time window 4, the voltage fluctuation is 5.1%, exceeding the fluctuation threshold. At this time, there are three fluctuation information sets, corresponding to time window 2, time window 3 and time window 4 respectively. The dynamic data in fluctuation information set 1 is current, the dynamic data in fluctuation information set 2 is current, and the dynamic data in fluctuation information set 3 is voltage.

[0054] Step S300: Perform timestamp alignment based on the fluctuation information set to obtain the marked time item, and obtain the corresponding state of the power system based on the marked time item to obtain the power feature item.

[0055] It is important to note that the corresponding state of the power system includes both environmental state and operational state. The method for obtaining power characteristic items includes: setting up a state acquisition module, which includes an environmental state acquisition unit and an operational state acquisition unit; taking the location of the power system as the determination point, acquiring environmental state information of the determination point based on the environmental state acquisition unit, including environmental temperature, environmental humidity, and environmental vibration, to obtain environmental characteristic items; acquiring power system state stability information based on the operational state acquisition unit, including current stability, voltage stability, and frequency stability, to obtain operational state characteristic items; and combining environmental characteristic items and operational state characteristic items to obtain power characteristic items.

[0056] Specifically, the environmental status acquisition unit includes a temperature sensor, a humidity sensor, and a vibration sensor. The temperature sensor, humidity sensor, and vibration sensor acquire the ambient temperature, ambient humidity, and ambient vibration, respectively, to obtain environmental characteristic items. The operating status acquisition unit includes a voltage sensor, a current sensor, and a frequency sensor to acquire power system state stability information, including current stability, voltage stability, and frequency stability, to obtain operating status characteristic items.

[0057] It is important to note that, such as Figure 4 As shown, the method for obtaining environmental status information at the determination point includes: obtaining attribute information of the target power system, including model information and production batch information, to obtain power attribute items; based on the power attribute items, obtaining information on the impact of power attributes through historical data and big data acquisition, thereby obtaining system-affected items; based on the system-affected items, setting an impact threshold, which is a fixed percentage threshold, filtering the system-affected items based on the impact threshold, and obtaining retained impact information, which is used as the environmental status information at the determination point.

[0058] Specifically, since different power systems are affected by different attributes, for example, one power system may be affected by temperature and humidity, while another power system may not be affected by temperature but only by humidity. Therefore, by obtaining the attribute information of the target power system, including model information and production batch information, and by using past data and big data to obtain the affected items of the target model information and production batch information, an affected threshold is set. The affected threshold is 10%. For example, in the environmental impact of a system, the temperature impact accounts for 5%, the humidity impact accounts for 45%, and the vibration impact accounts for 50%. At this time, the temperature impact does not reach the affected threshold, so the affected information is retained as the humidity impact and vibration impact, and this is used as the environmental state information.

[0059] Step S400: Based on power characteristic items, perform fault determination. When the power system is in fault determination, obtain fault data to obtain power fault items.

[0060] It is important to note that the power fault items correspond to the fluctuation information sets corresponding to the marked time items. The fault data includes the fault location and fault name. The methods for obtaining power fault items include: based on power feature items, obtaining fault features corresponding to power feature items through big data acquisition and historical data acquisition, and obtaining fault name information; based on power feature items, obtaining the location of the feature through fault analysis equipment, obtaining fault location information, and combining the fault name information and fault location information to obtain the power fault item.

[0061] Specifically, when different power characteristics appear in the power system, such as voltage fluctuations, current fluctuations, or frequency fluctuations, the fault characteristics corresponding to these power characteristics are obtained through big data acquisition and historical data acquisition, and the fault name information is obtained. For example, when the voltage to ground drops significantly, the voltages of the other two phases are relatively normal or fluctuate slightly. At this time, the fault characteristic corresponding to this power characteristic is determined to be a single-phase ground fault through big data acquisition and historical data acquisition. At the same time, the fault location is obtained through analysis equipment, and then the power fault item is obtained.

[0062] Step S500: Assign weights to the fluctuation items in the fluctuation information set to obtain the fault information weight items.

[0063] It should be noted that the method for obtaining the fault information weight item includes: obtaining the fluctuation range of the fluctuation items in the fluctuation information set, obtaining at least one fluctuation ratio item, setting a division weight value based on the fluctuation ratio item, the division weight value being the same as the proportion of the fluctuation ratio item, and then dividing the fluctuation items into weights based on the division weight value, thereby obtaining the fault information weight item.

[0064] Specifically, the division weight values ​​are set according to the fluctuation range of the fluctuation items in the fluctuation information set. For example, when the fluctuation items in the fluctuation information set are voltage, current and frequency signals, and the fluctuation ranges of voltage, current and frequency signals are 10%, 5% and 5% respectively, the division weight values ​​set at this time are 50%, 25% and 25% respectively, thereby obtaining the fault information weight items.

[0065] Step S600: Obtain the fluctuation information of dynamic data again.

[0066] It is important to note that after obtaining the fluctuation information update set, the marker time update set, the power characteristic update item, and the power fault update item, the fluctuation items in the fluctuation information update set are weighted to obtain the fault information weight update item. Since the average value of the dynamic data in the dynamic data information item is obtained according to the set time window, the dynamic data average item is obtained. The time windows are ordered sequentially. When the fluctuation information of the dynamic data is obtained again, the fluctuation information update set can be obtained. The above steps are repeated to obtain the marker time update set, the power characteristic update item, and the power fault update item in sequence. The fluctuation items in the fluctuation information update set are weighted to obtain the fault information weight update item.

[0067] Step S700: Determine the matching attributes between the power feature item and the power feature update item to obtain the feature matching item.

[0068] It is important to note that, such as Figure 5As shown, the matching attributes of power feature items and power feature update items include the degree of matching of fault location and the degree of matching of fault name. The method for obtaining feature matching items includes: based on power feature items and power feature update items, comparing fault location information and fault name information respectively, setting a range expansion value, the range expansion value being a fixed ratio value, combining the range expansion value with the fault location information of power feature items and power feature update items respectively to obtain a first fault range and a second fault range, determining whether the fault location information of power feature items and power feature update items matches the first fault range and the second fault range, and then judging the matching attributes of power feature items and power feature update items.

[0069] Specifically, based on power feature items and power feature update items, fault location information and fault name information are compared separately. The range expansion value is set to 10%. Based on the range expansion value, the fault location information of power feature items and power feature update items are combined to obtain the first fault range and the second fault range. For example, the fault location information of power feature items and power feature update items are points A and B, respectively. Point A is a circular area with a radius of 10, and point B is a circular area with a radius of 5. In this case, the first fault range is the circular area with a radius of 11 located at point A, and the second fault range is the circular area with a radius of 5.5 located at point B. By delineating the range in this way, it is possible to better determine whether the first range and the second range intersect while dividing the range. At this time, it is determined whether the fault location information of power feature items and power feature update items matches the first fault range and the second fault range, and then the matching attributes of power feature items and power feature update items are judged.

[0070] Step S800: Merge and filter the fault information weight items and fault information weight update items, filter out the weights to be saved, and obtain the weight saving items.

[0071] It should be noted that the method for obtaining the weighted items includes: setting a minimum weighted value, which is a fixed percentage; merging and filtering the fault information weighted items and fault information weighted update items based on the minimum weighted value; removing merged items that are lower than the minimum weighted value; and thus obtaining the weighted items.

[0072] Specifically, the minimum weight retention value is set to 20%. Based on this minimum weight retention value, fault information weight items and fault information weight update items are merged and filtered. Items with weights lower than the minimum weight retention value are removed, thus obtaining the weight retention items. For example, when the fluctuation items in the fluctuation information set are fluctuation item 1, fluctuation item 2, fluctuation item 3, and fluctuation item 4, and the fluctuation range of fluctuation item 1, fluctuation item 2, fluctuation item 3, and fluctuation item 4 is all 5%, the set division weight values ​​are 25%, 25%, 25%, and 25% respectively, thus obtaining the fault information weight items. When the fluctuation items in the fluctuation information update set are fluctuation item 1, fluctuation item 2, fluctuation item 3, and fluctuation item 5, and the fluctuation range of fluctuation item 1, fluctuation item 2, fluctuation item 3, and fluctuation item 5 is all 5%, the set division weight values ​​are 25%, 25%, 25%, and 25% respectively. The weight values ​​are 25%, 25%, 25%, and 25% respectively, thus obtaining the fault information weight update item. The fault information weight item and the fault information weight update item are merged and filtered. At this time, the weight ratios of fluctuation item 1, fluctuation item 2, fluctuation item 3, fluctuation item 4 and fluctuation item 5 are 50%, 50%, 50%, 25% and 25% respectively. The weights are then recalculated, and the weight ratios of fluctuation item 1, fluctuation item 2, fluctuation item 3, fluctuation item 4 and fluctuation item 5 are 25%, 25%, 25%, 12.5% ​​and 12.5% ​​respectively. Since the weight ratios of fluctuation item 4 and fluctuation item 5 do not exceed 20%, the merged items that are lower than the minimum weight retention value are removed, thus obtaining the weight retention items, namely fluctuation item 1, fluctuation item 2 and fluctuation item 3.

[0073] Step S900: Obtain the fluctuation items corresponding to the saved weights as the corresponding fluctuations of the power feature items and the power feature update items, and create a fault fluctuation matching model.

[0074] It is important to note that after the fluctuation matching model is created, dynamic analysis is performed based on the fluctuation information of the fault fluctuation matching model and dynamic data. When a fault of a certain type or location occurs, the fault fluctuation matching model can quickly locate the power characteristics that need attention. Conversely, when an abnormal fluctuation of a certain power characteristic occurs, the fault fluctuation matching model can quickly locate the fault type and fault location, thereby realizing a self-regulating power system analysis based on fault characteristics.

[0075] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended embodiments and their equivalents.

Claims

1. A dynamic waveform analysis method for power systems based on fault characteristic adaptation, comprising: Dynamic data in the power system is recorded by a synchronous sampling device to obtain dynamic data information items. Based on the dynamic information data items, the fluctuation information of the dynamic data is obtained, and the fluctuation information generates data to obtain a fluctuation information set. Its characteristic is that it further includes: Timestamp alignment is performed based on the fluctuation information set to obtain marked time items. The corresponding state of the power system is obtained based on the marked time items to obtain power feature items. Fault determination is based on power feature items. When the power system is in fault determination, fault data is acquired to obtain power fault items. Power fault items correspond to the fluctuation information set corresponding to the marked time items. The fluctuation items in the fluctuation information set are weighted and then the fault information weight items are obtained. When the fluctuation information of dynamic data is obtained again, the fluctuation information update set, the marker time update set, the power feature update item and the power fault update item are obtained. The fluctuation items in the fluctuation information update set are weighted and then the fault information weight update item is obtained. The matching attributes of power feature items and power feature update items are determined to obtain feature matching items. Fault information weight items and fault information weight update items are merged and filtered to select and save weights, resulting in weight saving items. The fluctuation items corresponding to the saved weights are obtained as the corresponding fluctuations of power feature items and power feature update items. These are used as training data for model training to create a fault fluctuation matching model. Dynamic analysis is performed based on the fluctuation information of the fault fluctuation matching model and dynamic data, thereby realizing power system analysis based on fault feature self-adjustment.

2. The power system dynamic waveform analysis method based on fault characteristic adaptation according to claim 1, characterized in that: The dynamic data includes voltage, current, and frequency signals, and the methods for obtaining dynamic data information items include: The synchronous sampling device is configured to consist of a high-precision synchronous phasor measurement unit, a data acquisition and storage unit, and a data interface terminal for the protection and control system. The high-precision synchronous phasor measurement unit provides three-phase voltage, three-phase current and frequency output. The data acquisition and storage unit performs high-speed sampling, buffering and local storage of field voltage, current and frequency signals to provide raw waveform data. The protection and control system data interface provides alarm, event recording, trip protection action time point and local measurement data, thereby recording dynamic data in the power system and obtaining dynamic data information items.

3. The power system dynamic waveform analysis method based on fault characteristic adaptation according to claim 1, characterized in that: The method for obtaining the fluctuation information set includes: Set a time window value, which is a fixed time value. Based on the time window value, obtain the average value of the dynamic data in the dynamic data information item to obtain the dynamic data average item. A fluctuation threshold is set, which is a fixed range of fluctuations in the dynamic data. The fluctuation information of the average item of the dynamic data is judged based on the fluctuation threshold. The average item value of the dynamic data that exceeds the fluctuation threshold is obtained, and the corresponding dynamic data is obtained, thus obtaining the fluctuation information set.

4. The power system dynamic waveform analysis method based on fault characteristic adaptation according to claim 1, characterized in that: The power system's corresponding state includes environmental state and operational state, and the methods for obtaining power characteristic items include: The status acquisition module is configured, which includes an environment status acquisition unit and a runtime status acquisition unit. Using the location of the power system as the determination point, the environmental state information of the determination point is obtained based on the environmental state acquisition unit, including environmental temperature, environmental humidity and environmental vibration, to obtain environmental feature items; Based on the operation status acquisition unit, the power system state stability information is obtained, including current stability, voltage stability and frequency stability, and operation status characteristic items are obtained. The environmental characteristic items and operation status characteristic items are combined to obtain power characteristic items.

5. The power system dynamic waveform analysis method based on fault characteristic adaptation according to claim 4, characterized in that: The method for obtaining the environmental state information of the determination point includes: Obtain the attribute information of the target power system, including model information and production batch information, to obtain power attribute items; Based on power attribute items, information on the impact of power attributes is obtained through historical data and big data acquisition methods, thereby obtaining the system's affected items; Based on the affected items of the system, an affected threshold is set. The affected threshold is a fixed percentage threshold. The affected items of the system are screened based on the affected threshold to obtain the retained affected information. The retained affected information is used as the environmental state information of the decision point.

6. The power system dynamic waveform analysis method based on fault characteristic adaptation according to claim 1, characterized in that: The fault data includes the fault location and fault name. Methods for obtaining power fault items include: Based on power characteristic items, fault characteristics corresponding to power characteristic items are obtained through big data acquisition and historical data acquisition, and fault name information is obtained; Based on power characteristic items, the location of the characteristic is obtained through fault analysis equipment to obtain fault location information. The fault name information and fault location information are combined to obtain power fault items.

7. The power system dynamic waveform analysis method based on fault characteristic adaptation according to claim 1, characterized in that: The methods for obtaining the fault information weight items include: The fluctuation range of the fluctuation items in the fluctuation information set is obtained, and at least one fluctuation ratio item is obtained. A division weight value is set based on the fluctuation ratio item. The division weight value is the same as the proportion of the fluctuation ratio item. The fluctuation items are weighted based on the division weight value, and then the fault information weight item is obtained.

8. The power system dynamic waveform analysis method based on fault characteristic adaptation according to claim 1, characterized in that: The matching attributes between the power feature items and the power feature update items include the degree of matching between fault location and fault name. The methods for obtaining the feature matching items include: Based on power feature items and power feature update items, fault location information and fault name information are compared separately. A range expansion value is set, which is a fixed ratio value. Based on the range expansion value, the fault location information of power feature items and power feature update items are combined to obtain a first fault range and a second fault range. It is then determined whether the fault location information of power feature items and power feature update items matches the first fault range and the second fault range, and the matching attributes of power feature items and power feature update items are judged.

9. The power system dynamic waveform analysis method based on fault characteristic adaptation according to claim 1, characterized in that: The method for obtaining the weight storage item includes: Set a minimum weight retention value, which is a fixed percentage. Based on the minimum weight retention value, merge and filter the fault information weight items and fault information weight update items, and remove the merged items that are lower than the minimum weight retention value to obtain the weight retention items.

Images