Operation fault monitoring method and system for three-phase alternating current load box

Abstract

The invention discloses an operation fault monitoring method and system for a three-phase AC load box, and relates to the technical field of electrical fault detection. The operation fault monitoring method for the three-phase AC load box comprises the following steps: acquiring power monitoring data of the load box; based on the power monitoring data, obtaining the time sequence unbalance degree of the three-phase voltage of the load box; obtaining an initial abnormal moment based on the time sequence unbalance degree; obtaining a fault severity value based on each time sequence unbalance degree after the initial abnormal moment; and adjusting the size of a sample subset of an isolated forest algorithm based on the fault severity value. According to the characteristics of the obtained power monitoring data, the appropriate sample subset size is selected, and the isolated forest algorithm is used to carry out anomaly detection on the load box data, so that the phenomenon that the data characteristics during serious faults cannot be captured due to the fact that the sample subset is too small can be avoided, and the phenomenon that the data characteristics during serious faults cannot be captured due to the fact that the sample subset is too large can also be avoided. And the detection efficiency is relatively low.

Description

Technical Field

[0001] This application relates to the technical field of electrical fault detection, and specifically to an operation fault monitoring method and system for a three-phase AC load box. Background Art

[0002] A three-phase AC load box is a programmable load simulation device used to test and verify various electrical equipment and components in a power system. By simulating actual load conditions, it provides accurate data support for the maintenance, debugging, and optimization of the power system. The application scenarios of three-phase AC load boxes are relatively extensive, including industries such as power, communication, energy, and transportation. In the energy industry, it mainly provides guarantees for the performance testing and safety assessment of key equipment such as inverters and transformers in new energy power generation systems such as wind power generation and solar power generation. Three-phase AC load boxes play an important role in power systems such as wind power generation systems. By monitoring their faults, faults existing in the load box can be detected in a timely manner, such as short circuits, overloads, component damages, etc. The occurrence of these faults will cause changes in the grid voltage and frequency, thereby affecting the stable operation of the entire power system. Therefore, by monitoring the three-phase AC load box, the operation faults of the power system can be detected in a timely manner, effectively avoiding the occurrence of power system safety problems, and thus ensuring the power generation efficiency.

[0003] In the prior art, the isolation forest algorithm is generally used to detect anomalies in three-phase AC load boxes. When using the isolation forest algorithm, it is necessary to select an appropriate sample subset size. If the sample subset size is set too small, it cannot fully represent the distribution of the original data set, resulting in the constructed isolation tree not being able to well reflect the true structure of the data, thereby affecting the accuracy of anomaly detection. If the sample subset is too large, it will increase the calculation cost and time, especially when dealing with large-scale data sets. Summary of the Invention

[0004] The purpose of this application is to provide an operation fault monitoring method and system for a three-phase AC load box to solve the technical problem of being unable to select an appropriate sample subset size when using the isolation forest algorithm to detect anomalies in the three-phase AC load box.

[0005] To achieve the above purpose, this application provides the following technical solutions:

[0006] In a first aspect, this application proposes an operation fault monitoring method for a three-phase AC load box. The operation fault monitoring method for the three-phase AC load box includes:

[0007] Obtain the power monitoring data of the load box; the power monitoring data at least includes the sequential voltage values of each phase in the load box;

[0008] Based on the power monitoring data, obtain the sequential unbalance degree of the three-phase voltages of the load box;

[0009] Based on the timing imbalance degree, obtain the initial abnormal moment;

[0010] Based on each timing imbalance degree after the initial abnormal moment, obtain the fault severity value; the fault severity value is at least used to characterize the magnitude of the fault severity of the load box;

[0011] Based on the fault severity value, adjust the sample subset size of the isolation forest algorithm;

[0012] Based on the sample subset size, use the isolation forest algorithm to monitor the operation faults of the load box.

[0013] As a specific solution in the technical solution of this application, the obtaining of the initial abnormal moment based on the timing imbalance degree includes: if the timing imbalance degree is greater than or equal to the first preset value, and all the timing imbalance degrees within a time period of the second preset value length before the timing of the timing imbalance degree are less than the first preset value, then determine the moment corresponding to the timing imbalance degree as the initial abnormal moment.

[0014] As a specific solution in the technical solution of this application, the obtaining of the fault severity value based on each timing imbalance degree after the initial abnormal moment includes:

[0015] Cluster each timing imbalance degree to obtain multiple clustering clusters; each clustering cluster corresponds to a time period;

[0016] Based on each clustering cluster, obtain the imbalance degree change value corresponding to each time period; the imbalance degree change value is at least used to characterize the fluctuation degree of the timing imbalance degree in the time period corresponding to each clustering cluster;

[0017] Based on the imbalance degree change value, obtain the fault severity value.

[0018] As a specific solution in the technical solution of this application, the calculation formula for obtaining the imbalance degree change value corresponding to each time period based on each clustering cluster is as follows:

[0019]

[0020] where B a represents the imbalance degree change value of the three-phase voltage in the a-th time period; represents the mean value of all timing imbalance degrees in the a-th time period; A1 represents the timing imbalance degree at the initial abnormal moment; j a represents the number of extreme points of the timing imbalance degree in the a-th time period; t a represents the total duration of the a-th time period; ΔAa represents the range of all temporal unbalance degrees within the a-th time period; norm() represents a normalization function used to map the value within the parentheses to the range of [0, 1].

[0021] As a specific solution in the technical solution of this application, obtaining the fault severity value based on the unbalance degree change value includes:

[0022] Based on each temporal unbalance degree, obtain the maximum abnormal phase corresponding to each time sequence; the maximum abnormal phase is the phase with the largest voltage deviation rate in the corresponding time sequence;

[0023] Based on each maximum abnormal phase, obtain the abnormal phase sequence corresponding to each time period;

[0024] Based on the abnormal phase sequence, obtain the fault complexity value corresponding to each time period; the fault complexity value is at least used to characterize the degree of irregularity of the change of each abnormal phase over time in the abnormal phase sequence;

[0025] Based on the unbalance degree change value and the fault complexity value, obtain the fault severity value.

[0026] As a specific solution in the technical solution of this application, the calculation formula for obtaining the fault complexity value corresponding to each time period based on the abnormal phase sequence is as follows:

[0027]

[0028] where C a represents the fault complexity value of the a-th time period; p a represents the overall voltage deviation rate within the a-th time period; z a represents the number of types of phases included in the abnormal phase sequence within the a-th time period; l a represents the number of sequence segments in the abnormal phase sequence of the a-th time period; L a is the duration of the a-th time period.

[0029] As a specific solution in the technical solution of this application, obtaining the fault severity value based on the unbalance degree change value and the fault complexity value includes:

[0030] Based on the abnormal phase sequence, obtain the target phase; the target phase is the phase that appears the most times in the abnormal phase sequence;

[0031] Based on the target phase, obtain the maximum continuous number and the maximum interval number from the abnormal phase sequence; the maximum continuous number is the maximum number of consecutive appearances of the target phase in the abnormal phase sequence; the maximum interval number is the maximum number of other phases between two adjacent target phases;

[0032] Obtain the fault severity value based on the maximum consecutive number, the maximum interval number, the unbalance degree change value, and the fault complexity value.

[0033] As a specific solution in the technical solution of this application, the calculation formula for obtaining the fault severity value based on the maximum consecutive number, the maximum interval number, the unbalance degree change value, and the fault complexity value is as follows:

[0034] D a = B a × C a × q a × exp(-m a )

[0035] In the formula, D a represents the fault severity value in the a-th time period; B a represents the unbalance degree change value of the three-phase voltage in the a-th time period; C a represents the fault complexity value in the a-th time period; q a represents the maximum interval number in the a-th time period; m a represents the maximum consecutive number in the a-th time period.

[0036] As a specific solution in the technical solution of this application, adjusting the sample subset size of the isolation forest algorithm based on the fault severity value includes:

[0037] Obtain a severity sequence based on the fault severity value corresponding to each time period;

[0038] Obtain a severity change value based on the severity sequence;

[0039] Adjust the sample subset size of the isolation forest algorithm based on the severity change value.

[0040] In a second aspect, this application proposes an operating fault monitoring system for a three-phase AC load box. The operating fault monitoring system for a three-phase AC load box includes:

[0041] A collector is used to obtain power monitoring data of the load box; the power monitoring data at least includes the time-series voltage value of each phase in the load box;

[0042] A server is used to obtain the time-series unbalance degree of the three-phase voltage of the load box based on the power monitoring data;

[0043] and obtain an initial abnormal moment based on the time-series unbalance degree;

[0044] And, obtaining a fault severity value based on each of the temporal imbalance degrees after the initial abnormal moment; the fault severity value is at least used to characterize the magnitude of the fault severity of the load box;

[0045] And, adjusting the sample subset size of the isolation forest algorithm based on the fault severity value;

[0046] And, monitoring the operating faults of the load box by using the isolation forest algorithm based on the sample subset size.

[0047] As a specific solution in the technical solution of this application, the server is further configured to, if the temporal imbalance degree is greater than or equal to a first preset value, and all the temporal imbalance degrees within a time period of a second preset value length before the temporal imbalance degree time sequence are less than the first preset value, then determine the moment corresponding to the temporal imbalance degree as the initial abnormal moment.

[0048] As a specific solution in the technical solution of this application, the server is further configured to cluster each of the temporal imbalance degrees to obtain a plurality of clustering clusters; each clustering cluster corresponds to a time period;

[0049] And, obtaining an imbalance degree change value corresponding to each time period based on each of the clustering clusters; the imbalance degree change value is at least used to characterize the magnitude of the fluctuation of the temporal imbalance degree in the time period corresponding to each clustering cluster;

[0050] And, obtaining the fault severity value based on the imbalance degree change value.

[0051] As a specific solution in the technical solution of this application, the formula for the server to obtain the imbalance degree change value corresponding to each time period based on each of the clustering clusters is as follows:

[0052]

[0053] Among them, B a represents the imbalance degree change value of the three-phase voltage in the a-th time period; represents the mean value of all the temporal imbalance degrees in the a-th time period; A1 represents the temporal imbalance degree at the initial abnormal moment; j a represents the number of extreme points of the temporal imbalance degree in the a-th time period; t a represents the total duration of the a-th time period; ΔA a represents the range of all the temporal imbalance degrees in the a-th time period; norm() represents a normalization function, which is used to map the value in the brackets to the interval range of [0, 1].

[0054] As a specific solution in the technical solution of the present application, the server is further configured to obtain the maximum abnormal phase corresponding to each time sequence based on each time sequence unbalance degree; the maximum abnormal phase is the phase with the largest voltage deviation rate in the corresponding time sequence;

[0055] And, based on each maximum abnormal phase, obtain an abnormal phase sequence corresponding to each time period;

[0056] And, based on the abnormal phase sequence, obtain a fault complexity value corresponding to each time period; the fault complexity value is at least used to characterize the degree of irregularity of the change of each abnormal phase with time in the abnormal phase sequence;

[0057] And, based on the unbalance degree change value and the fault complexity value, obtain the fault severity value.

[0058] As a specific solution in the technical solution of the present application, the calculation formula for the server to obtain the fault complexity value corresponding to each time period based on the abnormal phase sequence is as follows:

[0059]

[0060] Where C a represents the fault complexity value of the a-th time period; p a represents the overall voltage deviation rate in the a-th time period; z a represents the number of types of phases included in the abnormal phase sequence in the a-th time period; l a represents the number of sequence segments in the abnormal phase sequence of the a-th time period; L a is the duration of the a-th time period.

[0061] As a specific solution in the technical solution of the present application, the server is further configured to obtain a target phase based on the abnormal phase sequence; the target phase is the phase that appears the most times in the abnormal phase sequence;

[0062] And, based on the target phase, obtain the maximum continuous number and the maximum interval number from the abnormal phase sequence; the maximum continuous number is the maximum number of consecutive appearances of the target phase in the abnormal phase sequence; the maximum interval number is the maximum number of other phases between two adjacent target phases;

[0063] And, based on the maximum continuous number, the maximum interval number, the unbalance degree change value and the fault complexity value, obtain the fault severity value.

[0064] As a specific solution in the technical solution of this application, the server obtains the calculation formula of the fault severity value based on the maximum consecutive number, the maximum interval number, the unbalance degree change value, and the fault complexity value as follows:

[0065] D a =B a ×C a ×q a ×exp(-m a )

[0066] In the formula, D a represents the fault severity value in the a-th time period; B a represents the unbalance degree change value of the three-phase voltage in the a-th time period; C a represents the fault complexity value in the a-th time period; q a represents the maximum interval number in the a-th time period; m a represents the maximum consecutive number in the a-th time period.

[0067] As a specific solution in the technical solution of this application, the server is further configured to obtain a severity sequence based on the fault severity value corresponding to each time period;

[0068] and obtain a severity change value based on the severity sequence;

[0069] and adjust the sample subset size of the isolation forest algorithm based on the severity change value.

[0070] Compared with the prior art, the beneficial effects of this application are:

[0071] This application selects an appropriate sample subset size according to the characteristics of the obtained power monitoring data, and uses the isolation forest algorithm to perform anomaly detection on the load box data, which can not only avoid the phenomenon that the sample subset is too small to capture the data characteristics during a severe fault, but also avoid the phenomenon that the sample subset is too large, resulting in low detection efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] Figure 1 is a schematic flowchart of a method for monitoring operating faults of a three-phase AC load box proposed in this application;

[0073] Figure 2 is a schematic structural diagram of a system for monitoring operating faults of a three-phase AC load box proposed in this application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0074] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative efforts shall fall within the protection scope of the present application.

[0075] In the description of the embodiments of the present application and the above accompanying drawings, the terms "first", "second", etc. are used to distinguish similar objects and do not necessarily need to describe a specific order or sequence. For example, the first preset value and the second preset value mentioned below belong to different preset values. It should be understood that such names can be interchanged under appropriate circumstances so that the embodiments described here can be implemented in an order different from that shown or described here. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or modules does not necessarily have to be limited to those clearly listed steps or modules, but may include other steps or modules not clearly listed or inherent to these processes, methods, products, or devices. The division of modules in the embodiments of the present application is only a logical division, and there may be other division methods in actual implementation. For example, multiple modules can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed coupling, direct coupling, or communication connection between each other may be through some interfaces, and the indirect coupling or communication connection between modules may be electrical or other similar forms, which are not limited in the embodiments of the present application. And the modules or sub-modules described as separate components may or may not be physically separated, may or may not be physical modules, or may be distributed to multiple circuit modules. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of the embodiments of the present application.

[0076] In order to solve the technical problem of not being able to select an appropriate sample subset size when using the isolation forest algorithm to perform anomaly detection on a three-phase AC load box in the background technology, the present application proposes a method for monitoring the operation faults of a three-phase AC load box, as Figure 1 shown, the method for monitoring the operation faults of the three-phase AC load box includes steps S100 to S600.

[0077] Step S100: Obtain the power monitoring data of the load box.

[0078] In this embodiment, the power monitoring data at least includes the sequential voltage values of each phase in the load box. Specifically, the acquisition method of the sequential voltage values of each phase in the load box can be as follows: Connect an intelligent electric meter in parallel to the input end of each phase in the wind power generation system load box, directly measure and record the voltage data of each phase, obtain the voltage of each phase at each moment, and transmit the measured voltage data to the server through the communication interface to participate in subsequent monitoring. By continuously measuring the voltage data of each phase in the three-phase AC load box through the above method, with an acquisition frequency of once per second, the sequential voltage values of each phase are obtained. Of course, in other embodiments of the present application, other acquisition frequencies can also be set according to actual needs, such as once every 2 seconds or once every 3 seconds, etc.

[0079] Step S200: Based on the power monitoring data, obtain the sequential unbalance degree of the three-phase voltages of the load box.

[0080] It should be clear that under normal circumstances, the three-phase voltages in the three-phase AC load box are basically balanced and within the rated voltage range. However, for the three-phase AC load box used in the wind power generation system, due to the relatively harsh outdoor environment of the wind farm, the long-term effects of vibration changes, temperature changes, humidity changes, and rainwater intrusion may damage the insulation layer of the cable, cause the wire joints connecting the load box to loosen and may lead to short circuits. These faults will make the voltage transmission of the load box unstable. Therefore, the balance degree of the three-phase voltages at each moment can be determined first, and whether the three-phase AC load box is abnormal can be judged according to the balance degree.

[0081] In this embodiment, the symmetrical component method is used to calculate the unbalance degree of the three-phase voltages at each moment (i.e., the sequential unbalance degree mentioned above), denoted as ε U , and the symmetrical component method is a well-known technology and will not be elaborated here.

[0082] Step S300: Based on the sequential unbalance degree, obtain the initial abnormal moment.

[0083] In this embodiment, the initial abnormal moment refers to the moment when the sequential unbalance degree of the load box is abnormally high for the first time. Based on this, in step S300, the step of obtaining the initial abnormal moment based on the sequential unbalance degree includes: If the sequential unbalance degree is greater than or equal to the first preset value, and all the sequential unbalance degrees within a time period of the length of the second preset value before the sequential unbalance degree are less than the first preset value, then determine the moment corresponding to the sequential unbalance degree as the initial abnormal moment.

[0084] In the wind power generation system, under normal circumstances, the three-phase voltages are basically balanced, so the sequential unbalance degree should be 0. Therefore, if the unbalance degree ε at a certain moment UWhen it is greater than the first preset value, the current moment can be recorded as an abnormal moment. The abnormal moment may be that all three-phase voltages are abnormal, or it may be that only any one of the three phases is abnormal. In this embodiment, the first preset value can be set according to requirements. For example, the first preset value can be 0.1 or 0.2, etc.

[0085] In this embodiment, if the timing unbalance degree of the load box is abnormal, any timing unbalance degree after this moment may be abnormal, that is, any moment corresponding to the timing unbalance degree after this moment may be considered as the initial abnormal moment. To avoid this phenomenon, the second preset value is set to avoid obtaining too many initial abnormal moments, that is, to avoid excessive subsequent calculation amounts. In this embodiment, the second preset value can be set according to requirements. For example, the second preset value can be 30 minutes or 60 minutes, etc.

[0086] Step S400: Obtain a fault severity value based on each timing unbalance degree after the initial abnormal moment.

[0087] In this embodiment, the fault severity value is at least used to characterize the size of the fault severity of the load box. It should be noted that if the three-phase voltage of the load box is unbalanced, that is, the load box is abnormal. The reason for this abnormality may be that the electrical connection part of the load box is loose or short-circuited, causing voltage fluctuations. It may also be non-fault fluctuations caused by electromagnetic interference or other factors, and it will automatically return to stability when the interference disappears. Since the obtained power monitoring data is obtained in chronological order, when an abnormal moment appears, it is necessary to further observe the changes in the voltage and unbalance degree of each phase after the abnormal moment to determine the severity of the load box fault. It is easy to understand that if it is non-fault fluctuations caused by electromagnetic interference or other factors, the fault severity of the load box is small; if it is fault fluctuations caused by short-circuit or loose electrical connection, the fault severity of the load box is large.

[0088] In this embodiment, the fault severity value can be obtained based on each timing unbalance degree after the initial abnormal moment in any reasonable manner. For example: Step S400, obtaining the fault severity value based on each timing unbalance degree after the initial abnormal moment, may include Step S410 to Step S430.

[0089] Step S410: Cluster each timing unbalance degree to obtain multiple clusters.

[0090] In the embodiments of the present application, any reasonable clustering method can be used to cluster each timing unbalance degree. For example: Euclidean distance clustering algorithm or Manhattan distance clustering algorithm, etc. The clustering algorithm is a mature technology and will not be elaborated here.

[0091] It should be noted that based on step S410, multiple clustering clusters can be obtained. The time corresponding to the data within each clustering cluster is used as a time period, that is, each clustering cluster corresponds to a time period. Thus, the temporal imbalance degrees of multiple time periods can be obtained after the initial abnormal moment.

[0092] Step S420: Based on each clustering cluster, obtain the change value of the imbalance degree corresponding to each time period.

[0093] In this embodiment, the change value of the imbalance degree is at least used to characterize the fluctuation degree of the temporal imbalance degree in the time period corresponding to each clustering cluster. As can be seen from the foregoing, since each time period contains multiple temporal imbalance degrees, the sequence of the three-phase voltage imbalance degrees of each time period after the initial abnormal moment of the load box can be obtained. The imbalance degree may change in different time periods. Therefore, the overall data characteristics (i.e., the change value of the imbalance degree) after the load box is abnormal (i.e., after the initial abnormal moment) can be obtained through the temporal imbalance degrees in each time period after the initial abnormal moment. In the embodiments of the present application, any reasonable method can be used to obtain the change value of the imbalance degree after the load box is abnormal. For example: the average value of the temporal imbalance degrees in each time period can be used as the change value of the imbalance degree of this time period. In another embodiment of the present application, the calculation formula for obtaining the change value of the imbalance degree corresponding to each time period based on each clustering cluster in step S420 can be as follows:

[0094]

[0095] where B a represents the change value of the imbalance degree of the three-phase voltage in the a-th time period; represents the mean value of all temporal imbalance degrees in the a-th time period; A1 represents the temporal imbalance degree at the initial abnormal moment; j a represents the number of extreme points of the temporal imbalance degree in the a-th time period; t a represents the total duration of the a-th time period; ΔA a represents the range of all temporal imbalance degrees in the a-th time period; norm() represents a normalization function, which is used to map the value within the brackets to the interval range of [0, 1].

[0096] It should be clear that represents the difference between the mean value of the temporal imbalance degree of the a-th time period and the imbalance degree at the initial abnormal moment, the larger the value of, the greater the change in the imbalance degree in the a-th time period after the initial abnormal moment compared to the initial abnormal moment. j a / t aIt represents the change frequency of the time - series imbalance degree within the a - th time period. When the duration is fixed, the more the number of changes, the higher the change frequency; ΔA a It represents the range of all time - series imbalance degrees within the a - th time period, which is used to represent the change range of the imbalance degree; that is, (j a / t a ×ΔA a ) represents the specific change characteristics of the imbalance degree within the a - th time period, which is used to adjust the overall change When the change frequency is higher and the change range is larger, it indicates that its change characteristics are greater, corresponding to the imbalance degree change value B a which is also larger.

[0097] Step S430: Obtain the fault severity value based on the imbalance degree change value.

[0098] In the embodiment of the present application, the imbalance degree change value can be directly used as the fault severity value. It is easy to understand that if the imbalance degree change value is larger, it means that the change of the time - series imbalance degree after the initial abnormal moment is greater, that is, the fault generated by the load box is more serious.

[0099] It should be noted that although the imbalance degree change value can intuitively reflect the overall abnormal degree of the three - phase voltage and its development trend, when used to reflect the fault degree of the load box, it is too general and ignores the voltage change of each specific phase. Based on this, when judging the fault severity of the load box, the voltage change of each phase can be further combined for analysis. That is to say, in this embodiment, step S430, obtaining the fault severity value based on the imbalance degree change value, may include steps S440 to S470.

[0100] Step S440: Obtain the maximum abnormal phase corresponding to each time series based on each time - series imbalance degree.

[0101] In this embodiment, the maximum abnormal phase is the phase with the largest voltage deviation rate in the corresponding time series. It should be clear that the voltage of the wind power generation system generally has a fixed standard value and fluctuation range, so the deviation of each phase voltage should also be within a certain limit range. By calculating the voltage deviation rate of each phase at the current moment, it can be found whether the voltage of each phase is abnormal. The calculation formula of the voltage deviation rate can be expressed as:

[0102]

[0103] where R represents the voltage deviation rate of a certain phase at the current moment; represents the measured voltage of a certain phase; U γRepresents the standard voltage value. According to this formula, the voltage deviation rate of each phase at each moment within each time period can be calculated, and the phase with the largest deviation rate is taken as the maximum abnormal phase at the corresponding moment.

[0104] Step S450: Based on each maximum abnormal phase, obtain the abnormal phase sequence corresponding to each time period.

[0105] In this embodiment, according to step S440, the abnormal phase at each moment within each time period can be determined, obtaining an abnormal phase sequence. That is, determine the maximum voltage deviation rate among the three phases at each moment within the time period, obtaining several maximum voltage deviation rates.

[0106] Step S460: Based on the abnormal phase sequence, obtain the fault complexity value corresponding to each time period.

[0107] In this embodiment, the fault complexity value is at least used to characterize the degree of irregularity of the changes of each abnormal phase in the abnormal phase sequence over time. It is easy to understand that if the data in the abnormal phase sequence is more regular, it indicates that the fault generated by the load box is more regular, that is, the severity of the fault generated by the load box is smaller; if the data in the abnormal phase sequence is less regular, it indicates that the fault generated by the load box is more irregular, that is, the severity of the fault generated by the load box is greater.

[0108] In the embodiments of the present application, any reasonable method can be adopted to obtain the fault complexity value corresponding to each time period based on the abnormal phase sequence. For example: in step S460, the calculation formula for obtaining the fault complexity value corresponding to each time period based on the abnormal phase sequence can be as follows:

[0109]

[0110] where C a represents the fault complexity value of the a-th time period; p a represents the overall deviation rate of the voltage within the a-th time period; z a represents the number of types of phases included in the abnormal phase sequence within the a-th time period; l a represents the number of sequence segments in the abnormal phase sequence of the a-th time period; L a is the duration of the a-th time period.

[0111] In this embodiment, the mean value of all the maximum voltage deviation rates within a certain time period is calculated as the overall voltage deviation rate within that time period. Through the clustering algorithm, the temporal imbalance degrees at all moments within each time period are made similar, but the abnormal phases at all moments within the same time period may not be completely similar. That is, in the abnormal phase sequences of each of the above time periods, the changes in the abnormal phases are different. For example, if the three-phase voltages are denoted as 1, 2, and 3 respectively, then the 1, 2, and 3 distributed in the abnormal phase sequence may be different. It is possible that each phase exists, or only one or two of them exist, and the continuity of each phase is also different. Each abnormal phase represents a kind of fault. Therefore, the fault complexity within each time period is also different. That is to say, in this embodiment, the minimum number of types of phases included in the abnormal phase sequence is 1, and the maximum is 3. In this embodiment, one or more consecutive phases of a certain phase in the abnormal phase sequence are recorded as a sequence segment. For example, if an abnormal phase sequence is 1221123, then the number of sequence segments in this abnormal phase sequence is 5, and the respective sequence segments are 1, 22, 11, 2, and 3. The number of sequence segments in other abnormal phase sequences can be obtained by analogy and will not be listed and elaborated later.

[0112] In this embodiment, the greater the overall deviation rate of the voltage within the time period, and the more the abnormal phases and the more the sequence segments in the abnormal phase sequence, the more complex the initial fault in that time period is, and the greater the corresponding fault complexity value is.

[0113] Step S470: Obtain the fault severity value based on the imbalance degree change value and the fault complexity value.

[0114] In the embodiments of the present application, any reasonable method can be adopted to obtain the fault severity value based on the imbalance degree change value and the fault complexity value. For example: the fault severity value can be the sum or product of the imbalance degree change value and the fault complexity value.

[0115] It should be noted that within a certain time period, at the beginning, there may be only one abnormal phase or two abnormal phases alternating. When the third phase is added later, it indicates that the fault gradually worsens, or if all three phases have been abnormal and alternating from the beginning, it also indicates that the abnormality is relatively serious. Therefore, the distribution of the abnormal phases can also be combined to determine the fault severity within each time period. Based on this, in a specific embodiment of the present application, step S470, obtaining the fault severity value based on the imbalance degree change value and the fault complexity value, may include steps S471 to S473.

[0116] Step S471: Obtain the target phase based on the abnormal phase sequence.

[0117] In this embodiment, the target phase is the phase that appears the most times in the abnormal phase sequence.

[0118] Step S472: Based on the target phase, obtain the maximum consecutive number and the maximum interval number from the abnormal phase sequence.

[0119] In this embodiment, the maximum consecutive number is the maximum number of consecutive occurrences of the target phase in the abnormal phase sequence. The maximum interval number is the maximum number of other phases between two adjacent target phases. For example: If an abnormal phase sequence is 112211231, since 1 appears 5 times, 2 appears 3 times, and 3 appears 1 time, the target phase is 1. There is only the 22 sequence segment between the first 11 sequence segment and the second 11 sequence segment, and there are the 2 sequence segment and the 3 sequence segment between the second 11 sequence segment and the last 1 sequence segment. That is to say, both the maximum consecutive number and the maximum interval number are 2. The target phase, the maximum consecutive number, and the maximum interval number of the sequence segments in other abnormal phase sequences are all calculated in this way, and will not be listed and elaborated later.

[0120] Step S473: Based on the maximum consecutive number, the maximum interval number, the unbalance change value, and the fault complexity value, obtain the fault severity value.

[0121] In the embodiments of the present application, any reasonable method can be used to obtain the fault severity value based on the maximum consecutive number, the maximum interval number, the unbalance change value, and the fault complexity value. For example: In step S473, the calculation formula for obtaining the fault severity value based on the maximum consecutive number, the maximum interval number, the unbalance change value, and the fault complexity value can be as follows:

[0122] D a =B a ×C a ×q a ×exp(-m a )

[0123] In the formula, D a represents the fault severity value in the a-th time period; B a represents the unbalance change value of the three-phase voltage in the a-th time period; C a represents the fault complexity value in the a-th time period; q a represents the maximum interval number in the a-th time period; m a represents the maximum consecutive number in the a-th time period.

[0124] Step S500: Based on the fault severity value, adjust the sample subset size of the isolation forest algorithm.

[0125] It should be clear that if the load box fails, causing changes in the three-phase voltage and timing imbalance, the situation after each failure is not necessarily the same. For example, the failure may become more and more serious over time, or it may be a relatively minor failure that does not gradually worsen over time. That is to say, a certain failure remains stable. Therefore, it is necessary to determine the severity of the failure of the load box during the entire failure process based on the severity value of the failure in each time period.

[0126] In this embodiment, the larger the severity value of the failure in the current time period, the more serious the failure generated by the load box. At this time, when monitoring the failure generated by the load box through the Isolation Forest algorithm, the size of the sample subset can be reduced to ensure that the Isolation Forest algorithm can accurately capture data features. The smaller the severity value of the failure in the current time period, the less serious the failure generated by the load box. At this time, when monitoring the failure generated by the load box through the Isolation Forest algorithm, the size of the sample subset can be increased to improve the detection efficiency.

[0127] In the embodiments of the present application, any reasonable method can be used to adjust the size of the sample subset of the Isolation Forest algorithm based on the severity value of the failure. For example: Step S500, adjusting the size of the sample subset of the Isolation Forest algorithm based on the severity value of the failure includes steps S510 to S530.

[0128] Step S510: Obtain a severity sequence based on the severity value of the failure corresponding to each time period.

[0129] As can be seen from the foregoing, a severity value of the failure can be obtained based on the timing imbalance of each time period, and multiple severity values of the failure can form a severity sequence. It is easy to understand that if the failure becomes more and more serious, the severity values of each failure will definitely increase with time.

[0130] Step S520: Obtain a severity change value based on the severity sequence.

[0131] As can be seen from the foregoing, when the severity of the failure gradually increases, the severity values of each failure in the severity sequence also gradually increase. If the severity values of each failure in the severity sequence are always relatively stable, it means that the failure of the load box also remains stable and does not worsen. Based on this, a change line of the failure severity can be drawn, with the horizontal axis being time and the vertical axis being the severity value of the failure. The slope of this change line is obtained and denoted as k, which is used to represent the overall failure change trend during the observation time. The larger k is, the more serious the failure is, and vice versa, it means that the failure is relatively stable or the failure is alleviated.

[0132] In an embodiment of the present application, the slope k can be directly used as the severity change value. In another embodiment of the present application, in step S520, the calculation formula for obtaining the severity change value based on the severity sequence can be as follows:

[0133]

[0134] Where F represents the severity change value of the load box; represents the mean value of the fault severity values of the load box in each time period; k represents the change trend of the fault severity value of the load box, that is, the slope in the previous text; t(D d,max ) represents the time when the maximum fault severity value appears in the severity sequence. The larger the value of t(D d,max ), the later the time when the maximum fault severity appears, indicating that the fault has been intensifying, the larger the change trend, and the later the most severe data appears, indicating that the fault severity is greater and the corresponding severity change value is larger.

[0135] Step S530: Based on the severity change value, adjust the sample subset size of the isolation forest algorithm.

[0136] In an embodiment of the present application, when presetting to use the isolation forest algorithm for load box anomaly monitoring, the initial sample subset size can be 256. When the fault severity is small, the sample subset size can be appropriately increased to improve the detection efficiency. When the fault severity is large, the sample subset size needs to be reduced to ensure that the data characteristics can be accurately captured.

[0137] Therefore, the adjusted sample subset size can be expressed as:

[0138] M = 256 - 256×norm(F)

[0139] M represents the adjusted sample subset size, and norm(F) represents using the normalization function to normalize the fault severity of the load box to the range of [0, 1]. When the fault severity is large, the adjusted sample subset is smaller, and vice versa. Round down the value of 256×norm(F).

[0140] Step S600: Based on the sample subset size, use the isolation forest algorithm to monitor the operation faults of the load box.

[0141] It should be clear that monitoring the operation faults of the load box using the isolation forest algorithm based on the sample subset size is a mature technology and will not be elaborated here. Using the isolation forest algorithm to perform anomaly detection on the acquired data, timely discover the operation faults of the three-phase AC load box, and take timely treatment measures can ensure the safe and stable operation of the power system.

[0142] In the embodiment of the operation fault monitoring method for a three-phase AC load box proposed by this application, according to the characteristics of the obtained power monitoring data, an appropriate sample subset size is selected, and the Isolation Forest algorithm is used to detect anomalies in the load box data. This can avoid the phenomenon that the sample subset is too small to capture the data characteristics during a serious fault, and also avoid the phenomenon that the sample subset is too large, resulting in low detection efficiency. The operation fault monitoring method for a three-phase AC load box proposed by this application can monitor the operation status of the three-phase AC load box in a wind power generation system in real time, accurately judge whether the load box is in a normal working state, timely detect the operation faults of the load box, prevent electrical accidents from occurring, ensure the safe and stable operation of the equipment, reduce the occurrence of shutdowns caused by the operation faults of the load box, and improve work efficiency.

[0143] After introducing the embodiment of the operation fault monitoring method for a three-phase AC load box proposed by this application, the operation fault monitoring system for a three-phase AC load box proposed by this application will be introduced below, as Figure 2 shown. The operation fault monitoring system 10 for a three-phase AC load box includes:

[0144] The collector 11 is used to obtain the power monitoring data of the load box; the power monitoring data at least includes the sequential voltage values of each phase in the load box;

[0145] The server 12 is used to obtain the sequential unbalance degree of the three-phase voltages of the load box based on the power monitoring data;

[0146] and to obtain the initial abnormal moment based on the sequential unbalance degree;

[0147] and to obtain the fault severity value based on the sequential unbalance degrees after the initial abnormal moment; the fault severity value is at least used to characterize the magnitude of the fault severity of the load box;

[0148] and to adjust the sample subset size of the Isolation Forest algorithm based on the fault severity value;

[0149] and to monitor the operation faults of the load box using the Isolation Forest algorithm based on the sample subset size.

[0150] As a specific embodiment in this application, the server 12 is further used to determine that the moment corresponding to the sequential unbalance degree is the initial abnormal moment if the sequential unbalance degree is greater than or equal to a first preset value and all the sequential unbalance degrees within a time period of the length of a second preset value before the sequential unbalance degree time series are less than the first preset value.

[0151] As a specific embodiment in the present application, the server 12 is further configured to cluster each time sequence imbalance degree to obtain a plurality of clustering clusters; each clustering cluster corresponds to a time period;

[0152] and, based on each clustering cluster, obtain an imbalance degree change value corresponding to each time period; the imbalance degree change value is at least used to characterize the fluctuation degree of the time sequence imbalance degree in the time period corresponding to each clustering cluster;

[0153] and, based on the imbalance degree change value, obtain the fault severity value.

[0154] As a specific embodiment in the present application, the formula for the server 12 to obtain the imbalance degree change value corresponding to each time period based on each clustering cluster is as follows:

[0155]

[0156] wherein, B a represents the imbalance degree change value of the three-phase voltage in the a-th time period; represents the mean value of all time sequence imbalance degrees in the a-th time period; A1 represents the time sequence imbalance degree at the initial abnormal moment; j a represents the number of extreme points of the time sequence imbalance degree in the a-th time period; t a represents the total duration of the a-th time period; ΔA a represents the range of all time sequence imbalance degrees in the a-th time period; norm() represents a normalization function, which is used to map the value in the parentheses to the range of [0, 1].

[0157] As a specific embodiment in the present application, the server 12 is further configured to, based on each time sequence imbalance degree, obtain the maximum abnormal phase corresponding to each time sequence; the maximum abnormal phase is the phase with the largest voltage deviation rate in the corresponding time sequence;

[0158] and, based on each maximum abnormal phase, obtain an abnormal phase sequence corresponding to each time period;

[0159] and, based on the abnormal phase sequence, obtain a fault complexity value corresponding to each time period; the fault complexity value is at least used to characterize the irregularity degree of each abnormal phase changing with time in the abnormal phase sequence;

[0160] and, based on the imbalance degree change value and the fault complexity value, obtain the fault severity value.

[0161] As a specific embodiment in the present application, the formula for the server 12 to obtain the fault complexity value corresponding to each time period based on the abnormal phase sequence is as follows:

[0162]

[0163] Among them, C a represents the fault complexity value in the a-th time period; p a represents the overall deviation rate of the voltage in the a-th time period; z a represents the number of types of phases included in the abnormal phase sequence in the a-th time period; l a represents the number of sequence segments in the abnormal phase sequence in the a-th time period; L a is the duration of the a-th time period.

[0164] As a specific embodiment in this application, the server 12 is further configured to obtain a target phase based on the abnormal phase sequence; the target phase is the phase that appears the most times in the abnormal phase sequence;

[0165] and, based on the target phase, obtain the maximum continuous number and the maximum interval number from the abnormal phase sequence; the maximum continuous number is the maximum number of consecutive occurrences of the target phase in the abnormal phase sequence; the maximum interval number is the maximum number of other phases between two adjacent target phases;

[0166] and, based on the maximum continuous number, the maximum interval number, the unbalance change value, and the fault complexity value, obtain the fault severity value.

[0167] As a specific embodiment in this application, the formula for the server 12 to obtain the fault severity value based on the maximum continuous number, the maximum interval number, the unbalance change value, and the fault complexity value is as follows:

[0168] D a = B a × C a × q a × exp(-m a )

[0169] In the formula, D a represents the fault severity value in the a-th time period; B a represents the unbalance change value of the three-phase voltage in the a-th time period; C a represents the fault complexity value in the a-th time period; q a represents the maximum interval number in the a-th time period; m a represents the maximum continuous number in the a-th time period.

[0170] As a specific embodiment in this application, the server 12 is further configured to obtain a severity sequence based on the fault severity value corresponding to each time period;

[0171] And, based on the severity sequence, obtain the severity change value;

[0172] And, based on the severity change value, adjust the sample subset size of the Isolation Forest algorithm.

[0173] In the embodiment of the operation fault monitoring system for the three-phase AC load box proposed in this application, according to the characteristics of the acquired power monitoring data, an appropriate sample subset size is selected, and the Isolation Forest algorithm is used to detect anomalies in the load box data. This can not only avoid the phenomenon that the sample subset is too small to capture the data characteristics during a severe fault, but also avoid the phenomenon that the sample subset is too large, resulting in low detection efficiency. The operation fault monitoring system for the three-phase AC load box proposed in this application can monitor the operation status of the three-phase AC load box in the wind power generation system in real time, accurately judge whether the load box is in a normal working state, timely detect the operation faults of the load box, prevent electrical accidents from occurring, ensure the safe and stable operation of the equipment, reduce the occurrence of shutdowns caused by the operation faults of the load box, and improve work efficiency.

[0174] It should be clear that the computer-readable storage medium in this application includes permanent and non-permanent, removable and non-removable media, and information storage can be achieved by any method or technology. The information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase change memory, static random access memory, dynamic random access memory, other types of random access memory, read-only memory, electrically erasable programmable read-only memory, flash memory, or other memory technologies, compact disc read-only memory, digital versatile disc, or other optical storage, magnetic cassette tapes, disk storage, or other magnetic storage devices, or any other non-transmission medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media do not include transient computer-readable media such as modulated data signals and carrier waves.

[0175] In the above embodiments, the descriptions of each embodiment have their own focuses. For parts not detailed in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0176] Those skilled in the art can clearly understand that for the sake of convenience and brevity of description, the specific working processes of the above-described methods, devices, and equipment can refer to the corresponding processes in the foregoing method embodiments, and will not be elaborated herein.

[0177] In several embodiments provided by the embodiments of the present application, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For example, the division of the modules is only a logical function division. In actual implementation, there may be other division methods. For example, multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection between each other can be through some interfaces. The indirect coupling or communication connection of the devices or modules can be in electrical, mechanical, or other forms.

[0178] The modules described as separate components may or may not be physically separated. The components displayed as modules may or may not be physical modules, that is, they can be located in one place or distributed to multiple network modules. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0179] In addition, in each embodiment of the embodiments of the present application, each functional module can be integrated in a processing module, or each module can exist physically alone, or two or more modules can be integrated in one module. The above integrated modules can be implemented in the form of hardware or in the form of software functional modules. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it can be stored in a computer-readable storage medium.

[0180] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product.

[0181] The computer program product includes one or more computer instructions. When the computer program is loaded and executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wire (such as coaxial cable, optical fiber, digital subscriber line) or wirelessly (such as infrared, wireless, microwave, etc.). The computer-readable storage medium may be any available medium that a computer can store or a data storage device such as a server or data center that includes one or more integrated available media. The available medium may be a magnetic medium (such as a floppy disk, hard disk, magnetic tape), an optical medium (such as a digital video disc), or a semiconductor medium (such as a solid state disk (SSD)), etc.

[0182] Although the embodiments of the present application have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles of the present application.

Claims

1. A method for monitoring operating faults of a three-phase AC load box, characterized in that: include: Obtain power monitoring data from the load box; The power monitoring data at least includes the time-series voltage value of each phase in the load box; Based on the power monitoring data, obtaining the timing imbalance of the three-phase voltage of the load box; Based on the timing imbalance, obtaining an initial abnormal time; Based on each time sequence imbalance after the initial abnormal moment, a fault severity value is obtained; the fault severity value is at least used to characterize the fault severity of the load box; Based on the fault severity value, adjusting the sample subset size of the isolation forest algorithm; Based on the sample subset size, an isolation forest algorithm is used to monitor the operation faults of the load box.

2. The method for monitoring operation failure of a three-phase AC load box according to claim 1, characterized in that: The method of obtaining the initial abnormal moment based on the timing imbalance comprises: if the timing imbalance is greater than or equal to a first preset value, and all timing imbalances within a time period of a second preset value length before the timing imbalance are less than the first preset value, then determining that the moment corresponding to the timing imbalance is the initial abnormal moment.

3. The method for monitoring operation failure of a three-phase AC load box according to claim 1, characterized in that: The obtaining of the fault severity value based on each timing imbalance after the initial abnormal moment includes: Cluster each time series imbalance to obtain multiple clusters; each cluster corresponds to a time period; Based on each cluster, an imbalance change value corresponding to each time period is obtained; the imbalance change value is at least used to characterize the fluctuation degree of the time series imbalance in the time period corresponding to each cluster; The fault severity value is obtained based on the imbalance change value.

4. The method for monitoring operating faults of a three-phase AC load box according to claim 3, characterized in that: The calculation formula for obtaining the imbalance change value corresponding to each time period based on each cluster is as follows: Among them, B a Indicates the change value of the unbalance degree of the three-phase voltage in the ath time period; represents the mean value of all time series imbalances in the ath time period; A1 represents the time series imbalance at the initial abnormal moment; j a Indicates the number of extreme points of time series imbalance in the ath time period; t a Indicates the total duration of the a-th time period; ΔA a It represents the range of all time series imbalances in the ath time period; norm() represents the normalization function, which is used to map the values ​​in the brackets to the interval [0, 1].

5. The method for monitoring operation failure of a three-phase AC load box according to claim 3, characterized in that: The acquiring the fault severity value based on the imbalance change value includes: Based on the imbalance degree of each time sequence, the maximum abnormal phase corresponding to each time sequence is obtained; the maximum abnormal phase is the phase with the largest voltage deviation rate in the corresponding time sequence; Based on each maximum abnormal phase, obtain the abnormal phase sequence corresponding to each time period; Based on the abnormal phase sequence, a fault complexity value corresponding to each time period is obtained; the fault complexity value is at least used to characterize the irregularity of each abnormal phase in the abnormal phase sequence changing with time; The fault severity value is acquired based on the imbalance change value and the fault complexity value.

6. The method for monitoring operating faults of a three-phase AC load box according to claim 5, characterized in that: The calculation formula for obtaining the fault complexity value corresponding to each time period based on the abnormal phase sequence is as follows: Among them, C a represents the fault complexity value of the ath time period; p a represents the overall deviation rate of voltage in the ath time period; z a represents the number of phase types contained in the abnormal phase sequence in the ath time period; l a represents the number of sequence segments in the abnormal phase sequence of the ath time period; L a is the length of the ath time period.

7. The method for monitoring operating faults of a three-phase AC load box according to claim 5, characterized in that: The acquiring the fault severity value based on the imbalance change value and the fault complexity value comprises: Based on the abnormal phase sequence, a target phase is obtained; the target phase is the phase that appears most frequently in the abnormal phase sequence; Based on the target phase, a maximum continuous number and a maximum interval number are obtained from the abnormal phase sequence; the maximum continuous number is the maximum number of consecutive appearances of the target phase in the abnormal phase sequence; the maximum interval number is the maximum number of other phases between two adjacent target phases; The fault severity value is obtained based on the maximum continuous number, the maximum interval number, the imbalance change value and the fault complexity value.

8. The method for monitoring operating faults of a three-phase AC load box according to claim 7, characterized in that: The calculation formula for obtaining the fault severity value based on the maximum continuous number, the maximum interval number, the imbalance change value and the fault complexity value is as follows: D a =B a ×C a ×q a ×exp(-m a ) Where D a Indicates the fault severity value in the ath time period; B a Indicates the change value of the unbalance degree of the three-phase voltage in the ath time period; C a represents the fault complexity value of the ath time period; q a Indicates the maximum number of intervals in the ath time period; m a Indicates the maximum continuous number in the a-th time period.

9. The method for monitoring operating faults of a three-phase AC load box according to claim 7, characterized in that: The adjusting the sample subset size of the isolation forest algorithm based on the fault severity value includes: Based on the fault severity value corresponding to each time period, a severity sequence is obtained; Based on the severity sequence, obtaining a severity change value; Based on the severity change value, the sample subset size of the isolation forest algorithm is adjusted.

10. An operation fault monitoring system for a three-phase AC load box, characterized in that: include: The collector is used to obtain power monitoring data of the load box; the power monitoring data at least includes the time-series voltage value of each phase in the load box; The server is used to obtain the timing imbalance of the three-phase voltage of the load box based on the power monitoring data; and, based on the timing imbalance, obtaining an initial abnormal time; and, obtaining a fault severity value based on each timing imbalance after the initial abnormal moment; The fault severity value is at least used to characterize the fault severity of the load box; and, based on the fault severity value, adjusting a sample subset size of an isolation forest algorithm; And, based on the sample subset size, an isolation forest algorithm is used to monitor the operation fault of the load box.

Images