A wind farm line fault identification method, device, medium and electronic equipment

Abstract

The application discloses a wind farm line fault identification method and device, a storage medium and an electronic device. The method comprises the following steps: acquiring a fault current value set corresponding to each phase line of each wind turbine, wherein the fault current value set is composed of a plurality of fault current values within each preset time length before and after a wind farm line fault moment; performing data processing on each fault current value set by using a preset time-frequency analysis method to obtain a time-frequency matrix and an energy matrix corresponding to each fault current value set; performing statistical analysis by using a predetermined statistical method based on each time-frequency matrix and each energy matrix to obtain a plurality of fault characteristic values corresponding to each fault current value set; and identifying each fault characteristic value based on a preset fault type identification model and a preset fault branch identification model to obtain a wind farm line fault identification result. The method can improve the identification efficiency of the wind farm line fault.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of wind farm power collection line fault identification, and in particular relates to a wind farm line fault identification method and device, a storage medium and an electronic device. BACKGROUND

[0002] Wind energy is abundant and is a rapidly developing renewable energy. However, a wind farm is a multi-branch power transmission system with a complex topology, short-distance transmission lines and a harsh working environment. The harsh operating conditions result in frequent transmission line faults, reducing wind power generation and limiting the development and utilization of wind energy. In recent studies, the main methods for diagnosing transmission line faults can be divided into three categories: impedance methods, traveling wave methods and methods based on intelligent algorithms. Analysis of existing methods shows that the main factors limiting diagnostic performance include fault inception angle, fault resistance, fault location, measurement noise and transmission line length. However, due to the difference in transmission line length between wind farms and distribution networks, wind farms are mostly composed of short-distance transmission lines. When two adjacent short-distance branches each have a line fault, the short-distance line structure will result in similar fault signal characteristics. In addition, the network structure of a wind farm is more complex than that of a distribution network. The complex multi-branch structure and the asymmetry between the power grid and the unit result in the fault signal being suppressed by both the power grid and the wind turbine. The above methods only focus on the transmission lines of the distribution network and rarely focus on wind farms, resulting in the above methods not being directly applicable to wind farms and making the efficiency of wind farm power collection line fault type identification and fault branch identification low. SUMMARY

[0003] Therefore, the present application provides a wind farm line fault identification method, device, storage medium and electronic device, which mainly aims to solve the problem of low efficiency of wind farm power collection line fault type and fault branch identification.

[0004] To solve the above problems, the present application provides a wind farm line fault identification method, comprising:

[0005] Obtain a set of fault current values corresponding to each phase line of each wind turbine, which is composed of a plurality of fault current values within a predetermined time period before and after the wind farm line fault time;

[0006] Perform data processing on each set of fault current values using a predetermined time-frequency analysis method to obtain a time-frequency matrix and an energy matrix corresponding to each set of fault current values;

[0007] Based on each time-frequency matrix and each energy matrix, perform statistical analysis using a predetermined statistical method to obtain a plurality of fault characteristic values corresponding to each set of fault current values;

[0008] The fault feature values are identified based on a preset fault type identification model and a preset fault branch identification model, and a wind farm line fault identification result is obtained.

[0009] Optionally, the preset time-frequency analysis method is used to perform data processing on each of the fault current value sets, to obtain a time-frequency matrix and an energy matrix corresponding to each of the fault current value sets, and specifically includes:

[0010] The Stockwell time-frequency analysis method is used to perform data conversion on each of the fault current value sets, to obtain a time-frequency matrix corresponding to each of the fault current value sets;

[0011] The modulus calculation is performed on each of the time-frequency matrices, to obtain an energy matrix corresponding to each of the fault current value sets.

[0012] Optionally, based on each of the time-frequency matrices and each of the energy matrices, a predetermined statistical method is used for statistical analysis, to obtain a plurality of fault feature values corresponding to each of the fault current value sets, and specifically includes:

[0013] A target time-frequency matrix and a target energy matrix corresponding to a target fault current value set are obtained;

[0014] The target time-frequency matrix and the target energy matrix are compressed, to obtain a plurality of sub-matrices corresponding to the target fault current value set;

[0015] A predetermined statistical method is used to perform statistical analysis on each of the sub-matrices, to obtain a plurality of initial fault feature values corresponding to the target fault current value set, and the predetermined statistical method includes at least one of kurtosis, average value, minimum value, maximum value, standard deviation, deviation, and entropy;

[0016] The Pearson correlation coefficient method is used to screen each of the initial fault feature values, to obtain each target fault current value corresponding to the target fault current value set, to obtain a plurality of fault feature values corresponding to each of the fault current value sets.

[0017] Optionally, the target time-frequency matrix and the target energy matrix are compressed, to obtain a plurality of sub-matrices corresponding to the target fault current value set, and specifically includes:

[0018] Based on the target energy matrix, the absolute values of the fault current modulus values of each column in the target energy matrix are calculated, to obtain a first sub-matrix composed of the maximum values of each fault current modulus absolute value of each column;

[0019] Based on the target time-frequency matrix, the absolute values of the fault current values of each row in the target time-frequency matrix are calculated, to obtain a second sub-matrix composed of the maximum values of each fault current absolute value of each row;

[0020] Based on the target time-frequency matrix, the absolute values of the fault current values of each column in the target time-frequency matrix are calculated to obtain a third sub-matrix composed of the maximum values of the absolute values of each fault current;

[0021] Based on the third sub-matrix, a fourth sub-matrix composed of the phase angles of the fault current values corresponding to the maximum values of the absolute values of each fault current in the third sub-matrix is obtained.

[0022] Optionally, the initial fault characteristic values are screened by using the Pearson correlation coefficient method to obtain target fault current values corresponding to the target fault current value set, so as to obtain a plurality of fault characteristic values corresponding to each fault current value set, specifically including:

[0023] The similarity degrees between the initial fault characteristic values are calculated to obtain a plurality of similarity degrees;

[0024] Each of the similarity degrees is compared with a preset similarity threshold value, and when a target similarity degree is greater than or equal to the preset similarity threshold value, one of the fault characteristic values used to calculate the target similarity degree is randomly deleted to obtain intermediate fault characteristic values;

[0025] The intermediate fault characteristic values are de-duplicated to obtain target fault characteristic values corresponding to the target fault current value set, so as to obtain a plurality of target fault characteristic values corresponding to each fault current value set respectively.

[0026] Optionally, before identifying the fault characteristic values based on a preset fault type identification model and a preset fault branch identification model to obtain a wind farm line fault identification result, the method further includes: constructing the fault type identification model, specifically including:

[0027] Based on the real-time collected historical fault current value sets respectively corresponding to each phase at the outlet side of each wind turbine, data processing is performed to obtain historical fault characteristic values;

[0028] Based on the historical fault characteristic values, a first historical sample set for training the fault type identification model is constructed, and the historical samples in the first historical sample set carry fault type labels;

[0029] Randomly obtaining a preset number of historical samples in the first historical sample set as a training sample set and the remaining historical samples as a test sample set to perform model training on an initial fault type identification model to obtain the fault type identification model.

[0030] Optionally, after the initial fault type identification model is trained by randomly obtaining a preset number of historical samples in the historical sample set as a training sample set and the remaining historical samples as a test sample set, and the fault type identification model is obtained, the method further comprises:

[0031] Based on each historical fault feature value, a second historical sample set for training the fault branch identification model is constructed, and the historical samples in the second historical sample set carry fault branch labels;

[0032] The initial fault branch identification model is trained by randomly obtaining a preset number of historical samples in the second historical sample set as a training sample set and the remaining historical samples as a test sample set, and the fault branch identification model is obtained.

[0033] To solve the above problems, the application provides a wind farm line fault identification device, which comprises:

[0034] The acquisition module is configured to acquire a set of fault current values corresponding to each phase line of each wind turbine, wherein the set of fault current values is composed of a plurality of fault current values within a predetermined time period before and after the wind farm line fault time;

[0035] The data processing module is configured to perform data processing on each set of fault current values by using a predetermined time-frequency analysis method to obtain a time-frequency matrix and an energy matrix corresponding to each set of fault current values;

[0036] The statistical analysis module is configured to perform statistical analysis on each time-frequency matrix and each energy matrix by using a predetermined statistical method to obtain a plurality of fault feature values corresponding to each set of fault current values;

[0037] The fault identification module is configured to identify each fault feature value based on a preset fault type identification model and a preset fault branch identification model to obtain a wind farm line fault identification result.

[0038] To solve the above problems, the application provides a storage medium, which stores a computer program. When the computer program is executed by a processor, the steps of the wind farm line fault identification method described above are implemented.

[0039] To solve the above problems, the application provides an electronic device, which at least comprises a memory and a processor. The memory stores a computer program, and the processor implements the steps of the wind farm line fault identification method described above when executing the computer program stored on the memory.

[0040] The application obtains a fault current value set corresponding to each phase line of each wind turbine, the fault current value set is composed of a plurality of fault current values within each preset time length before and after the wind farm line fault moment; a preset time-frequency analysis method is used to process the data of each fault current value set, to obtain a time-frequency matrix and an energy matrix corresponding to each fault current value set; based on each time-frequency matrix and each energy matrix, a predetermined statistical method is used for statistical analysis, to obtain a plurality of fault characteristic values corresponding to each fault current value set; each fault characteristic value is identified based on a preset fault type identification model and a preset fault branch identification model, to obtain a wind farm line fault identification result. The wind farm line fault identification method of the application can obtain more accurate results of the fault type and fault branch position of the wind farm line, and improve the wind farm line fault identification efficiency.

[0041] The above description is only a summary of the technical scheme of the application, in order to more clearly understand the technical means of the application, the application can be implemented according to the content of the specification, and in order to make the above and other purposes, characteristics and advantages of the application more obvious and easy to understand, the following specific embodiments of the application are described. BRIEF DESCRIPTION OF DRAWINGS

[0042] Various other advantages and benefits will become apparent to those of ordinary skill in the art, upon reading the following detailed description of the preferred embodiments. The accompanying drawings are included to provide a description of preferred embodiments, and are not meant to limit the present application. Furthermore, the same reference numerals are used throughout the several drawings to refer to same or like parts. In the drawings:

[0043] Figure 1 A flowchart of a wind farm line fault identification method provided by an embodiment of the application is shown;

[0044] Figure 2 A flowchart of a wind farm line fault identification method provided by another embodiment of the application is shown;

[0045] Figure 3 A structure block diagram of a wind farm line fault identification device provided by still another embodiment of the application is shown. DETAILED DESCRIPTION

[0046] The various schemes and features of the application are described herein with reference to the accompanying drawings.

[0047] It should be understood that various modifications can be made to the embodiments of the application herein. Therefore, the above description should not be considered as limiting, but only as an example of the embodiments. Those skilled in the art will think of other modifications within the scope and spirit of the application.

[0048] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments of the application and, together with the general description of the application given above, and the detailed description of the embodiments given below, serve to explain the principles of the present application.

[0049] These and other characteristics of the present application will become apparent from the following description of the preferred forms of the application given, by way of non-limiting example, with reference to the accompanying drawings.

[0050] It should also be understood that, although the present application has been described above with reference to particular means, materials and embodiments, the present application is by no means limited to the particulars described and as such extends to all alternative constructions falling within the scope of the application.

[0051] The above and other aspects, features and advantages of the present application will become more apparent from the following detailed description when taken in conjunction with the accompanying drawings, in which:

[0052] Specific embodiments of the present application are described hereinafter, by way of non-limiting example; however, it should be understood that the claimed embodiments are merely examples of the present application, which can be implemented in numerous ways. Well-known and / or redundant functions and structures have not been described in detail to avoid obscuring the present application unnecessarily. Therefore, specific structural and functional details disclosed herein are not intended to limit the claimed embodiments, but merely to set forth representative structures and features of the present application, which can be used in various ways to implement the present application.

[0053] The specification can use phrases such as "in one embodiment", "in another embodiment", "in yet another embodiment", or "in other embodiments", which can refer to one or more of the same or different embodiments of the application.

[0054] The embodiments of the present application provide a wind farm line fault identification method, as shown in the figure, comprising: Figure 1

[0055] Step S101: Obtain a set of fault current values corresponding to each phase line of each wind turbine, which is composed of a plurality of fault current values within a predetermined time period before and after the wind farm line fault time;

[0056] ​In the specific implementation process of this step, the wind farm power collection line has the characteristics of multiple branches and multiple fault types. In order to accurately locate the fault position and fault type of the wind farm power collection line, first, a plurality of fault current values of each branch of the wind farm within a predetermined time interval before and after the fault time are obtained, and the predetermined time interval can be 2S, 3S, etc. The application does not limit the predetermined time interval. Within the predetermined time interval, the fault current value set of each wind turbine phase outlet is collected at a certain frequency. Each wind turbine outlet has three phase lines, namely A phase, B phase and C phase. Each wind turbine corresponds to three sets of fault current value sets, namely the A-phase corresponding fault current value set, the B-phase corresponding fault current value set and the C-phase corresponding fault current value set.

[0057] Step S102: using a preset time-frequency analysis method to process the data of each fault current value set, and obtaining a time-frequency matrix and an energy matrix corresponding to each fault current value set;

[0058] In the specific implementation process of this step, a preset time-frequency analysis method is used to convert the data of each fault current value set to obtain a time-frequency matrix corresponding to each fault current value set. The preset time-frequency analysis method can be a Stockwell time-frequency analysis method, and the application does not limit the time-frequency analysis method. Each time-frequency matrix is calculated by modulus to obtain an energy matrix corresponding to each fault current value set.

[0059] Step S103: based on each time-frequency matrix and each energy matrix, using a predetermined statistical method for statistical analysis to obtain a plurality of fault characteristic values corresponding to each fault current value set;

[0060] In the specific implementation process of this step, a target time-frequency matrix and a target energy matrix corresponding to a target fault current value set are obtained. The target time-frequency matrix and the target energy matrix are compressed to obtain a plurality of sub-matrices corresponding to the target fault current value set. A predetermined statistical method is used to statistically analyze each sub-matrix to obtain a plurality of initial fault characteristic values corresponding to the target fault current value set. The predetermined statistical method includes at least one of kurtosis, average value, minimum value, maximum value, standard deviation, deviation and entropy. The Pearson correlation coefficient method is used to screen each initial fault characteristic value to obtain each target fault current value corresponding to the target fault current value set, so as to obtain a plurality of fault characteristic values corresponding to each fault current value set.

[0061] Step S104: identifying each fault characteristic value based on a preset fault type identification model and a preset fault branch identification model to obtain a wind farm line fault identification result.

[0062] In the specific implementation process of this step, each fault feature value is identified by using the pre-trained preset fault type model and the preset branch identification model to obtain the line fault type and the fault branch position of the wind farm, and an identification result of the wind farm line fault is obtained.

[0063] In the specific implementation process of this step, each fault feature value is identified by using the pre-trained preset fault type model and the preset branch identification model to obtain the line fault type and the fault branch position of the wind farm, and an identification result of the wind farm line fault is obtained.

[0064] In another embodiment of the present application, another wind farm line fault identification method is provided, as shown in Figure 2

[0065] Step S201: Obtain a fault current value set corresponding to each phase line of each wind turbine, the fault current value set being composed of a plurality of fault current values within each preset time length before and after the fault time of the wind farm line;

[0066] In the specific implementation process of this step, the wind farm collection line has the characteristics of multiple branches and multiple fault types. In order to accurately locate the position and fault type of the wind farm collection line fault, first, a plurality of fault current values within each preset time length before and after the fault time of the wind farm branch wind turbine outlet side are obtained, and the preset time length can be 2S, 3S, etc. The present application does not limit the preset time length. Within the preset time length, the fault current value set of each wind turbine phase outlet side is collected at a certain frequency, and the sampling frequency can be 0.001s, 0.002s, etc. The present application does not limit the sampling frequency of the fault current value. Each wind turbine outlet side has three phase lines, and the three phase lines are A phase, B phase and C phase. Each wind turbine corresponds to three sets of fault current value sets, which are A phase corresponding fault current value set, B phase corresponding fault current value set and C phase corresponding fault current value set.

[0067] Step S202: Data processing of each fault current value set is performed by using a preset time-frequency analysis method to obtain a time-frequency matrix and an energy matrix corresponding to each fault current value set.​

[0068] In the implementation process, the preset time-frequency analysis method is used to perform data conversion on each set of fault current values to obtain a time-frequency matrix corresponding to each set of fault current values. The preset time-frequency analysis method can be a Stockwell time-frequency analysis method. The calculation formula of the time-frequency matrix obtained by the Stockwell time-frequency analysis method is shown in the following formula (1):

[0069]

[0070] wherein, i f is any three-phase fault current; n, k, q = 0, 1, 2, …, N-1; N is the number of fault current values, which is determined by the sampling frequency; T is the sampling frequency. When this method is used to analyze fault data, the time domain and frequency domain characteristics of the fault signal can be considered. Then, the modulus calculation is performed on each time-frequency matrix to obtain an energy matrix corresponding to each set of fault current values. The calculation formula of the energy matrix obtained by the modulus calculation on each time-frequency matrix is shown in the following formula (2):

[0071] E m×n = |S m×n | 2 (2)

[0072] wherein: m and n are the frequency number of frequency domain analysis and the time size of time domain analysis, respectively.

[0073] Step S203: Based on each time-frequency matrix and each energy matrix, a predetermined statistical method is used for statistical analysis to obtain a plurality of fault characteristic values corresponding to each set of fault current values.

[0074] In the specific implementation process of the step, the target time-frequency matrix corresponding to the target fault current value set and the target energy matrix are obtained; specifically, the time-frequency matrix and the energy matrix corresponding to each phase line of each wind turbine are statistically analyzed by using a predetermined statistical method to obtain a plurality of fault characteristic values corresponding to each fault current value set. The target time-frequency matrix and the target energy matrix are compressed to obtain a plurality of sub-matrices corresponding to the target fault current value set; specifically, based on the target energy matrix, the absolute values of the fault current modulus values of each column in the target energy matrix are calculated to obtain a first sub-matrix composed of the maximum values of the absolute values of the fault current modulus values of each column; when the target energy matrix is m rows and n columns, the first sub-matrix obtained is 1 row and n columns. Based on the target time-frequency matrix, the absolute values of the fault current values of each row in the target time-frequency matrix are calculated to obtain a second sub-matrix composed of the maximum values of the absolute values of the fault current values of each column; when the target time-frequency matrix is m rows and n columns, the second sub-matrix obtained is m rows and 1 column. Based on the target time-frequency matrix, the absolute values of the fault current values of each column in the target time-frequency matrix are calculated to obtain a third sub-matrix composed of the maximum values of the absolute values of the fault current values of each column; when the target time-frequency matrix is m rows and n columns, the third sub-matrix obtained is 1 row and n columns. Based on the third sub-matrix, a fourth sub-matrix composed of the phase angles of the fault current values corresponding to the maximum values of the absolute values of the fault current in the third sub-matrix is obtained. The fourth sub-matrix is 1 row and n columns. Four sub-matrices corresponding to each fault current value set are obtained through compression processing. A plurality of initial fault characteristic values corresponding to the target fault current value set are obtained by statistically analyzing each sub-matrix using a predetermined statistical method, and the predetermined statistical method includes statistical methods of kurtosis, mean value, minimum value, maximum value, standard deviation, deviation, and entropy for statistical analysis of each sub-matrix; in specific applications, at least one of the statistical methods can be used to statistically analyze each sub-matrix, for example: when seven statistical methods of kurtosis, mean value, minimum value, maximum value, standard deviation, deviation, and entropy are used to statistically analyze each sub-matrix, when a wind farm collection line includes five wind turbines, each wind turbine is divided into three phases, and then the five wind turbines have a total of 15 fault current value sets, one fault current value set corresponds to four sub-matrices, and a total of 15*4=60 sub-matrices are obtained. Each single-phase current of each wind turbine as a monitoring point corresponds to four sub-matrices, seven different statistical calculation methods are applied to four characteristic sub-matrices, seven initial fault characteristic values are generated for each characteristic sub-matrix, four sub-matrices corresponding to each single-phase fault current value set of each wind turbine as a monitoring point generate 4*7=28 initial fault characteristic values, and 60 sub-matrices corresponding to a total of 15 fault current value sets of five wind turbines generate 60*7=420 initial fault characteristic values.The initial fault characteristic values are screened by using a Pearson correlation coefficient method to obtain target fault current values corresponding to a target fault current value set, so as to obtain a plurality of fault characteristic values corresponding to each fault current value set. Specifically, the similarity between each initial fault characteristic value is calculated to obtain a plurality of similarities; and the initial fault characteristic values corresponding to a single phase line in each wind turbine are screened. For example, when one wind turbine has 28 initial fault characteristic values, denoted as T1, T2, …, and T28, the similarity between each initial fault characteristic value is calculated to obtain a plurality of similarities. The similarities are compared with a preset similarity threshold value; when a target similarity is greater than or equal to the preset similarity threshold value, one of the fault characteristic values used to calculate the target similarity is randomly deleted to obtain intermediate fault characteristic values; the preset similarity threshold value can be set to 0.95, 0.98, etc., and the preset similarity threshold value can be set according to actual needs. For example, the similarity between T1 and T2 is 0.96, and when the similarity threshold value is 0.95, one of the fault characteristic values T1 or T2 used to calculate the target similarity is randomly deleted; the intermediate fault characteristic values are obtained after screening; for example, the intermediate fault characteristic values T2, T5, T5, T6, T8, T9, T10, T12, T20, T23, T10, T26, and T27 are obtained after screening.

[0075] The intermediate fault characteristic values are de-duplicated to obtain target fault characteristic values corresponding to a target fault current value set, so as to obtain a plurality of target fault characteristic values corresponding to each fault current value set. For example, T5 and T10 are removed from T2, T5, T5, T6, T8, T9, T10, T12, T20, T23, T10, T26, and T27 to finally obtain 11 target fault characteristic values T2, T5, T6, T8, T9, T10, T12, T20, T23, T26, and T27.

[0076] Step S204: constructing a fault type recognition model and a fault branch recognition model;

[0077] In the implementation process, the classification method of the random forest can be used to construct the fault type identification model and the fault branch identification model. Specifically, based on the historical fault current value set corresponding to each phase line at the outlet side of each wind turbine collected in real time, historical fault characteristic values are obtained. Specifically, a predetermined time-frequency analysis method is used to process the fault current value set to obtain a historical time-frequency matrix and a historical energy matrix corresponding to each fault current value set. Based on the historical time-frequency matrix and the historical energy matrix, a predetermined statistical method is used for statistical analysis to obtain a plurality of historical fault characteristic values corresponding to each historical fault current value set. Based on the historical fault characteristic values, a first historical sample set for training the fault type identification model is constructed, and the historical samples in the first historical sample set carry fault type labels. Specifically, the fault types can be divided into A-phase single-phase grounding fault AG, B-phase single-phase grounding fault BG, C-phase single-phase grounding fault CG, AB-phase short-circuit grounding fault ABG, AC-phase short-circuit grounding fault ACG, BC-phase short-circuit grounding fault BCG, A-phase and B-phase short-circuit fault AB, A-phase and C-phase short-circuit fault AC, and B-phase and C-phase short-circuit fault BC. The label corresponding to the fault type AG is 1, the label corresponding to the fault type BG is 2, the label corresponding to the fault type CG is 3, the label corresponding to the fault type ABG is 4, the label corresponding to the fault type ACG is 5, the label corresponding to the fault type BCG is 6, the label corresponding to the fault type AB is 7, the label corresponding to the fault type AC is 8, and the label corresponding to the fault type BC is 9. When there are 10 branches and 9 fault types, the fault starting angle, the fault resistance, and the fault position are randomly changed, 50 historical fault samples of each fault type are generated for each fault branch, a first historical sample set is obtained, and a total of 4500 historical fault samples are obtained. A predetermined number of historical samples in the first historical sample set are randomly obtained as a training sample set, and the remaining historical samples are obtained as a test sample set to train an initial fault type identification model to obtain the fault type identification model. The predetermined number can be 70% of the historical fault samples in the first historical sample set as training samples, and the remaining 30% of the historical fault samples as test samples are used to train the initial fault type identification model to construct the fault type identification model.

[0078] The method for constructing the fault branch identification model is: first, based on each historical fault characteristic value, a second historical sample set for training the fault branch identification model is constructed, and the historical samples in the second historical sample set carry fault branch labels; a predetermined number of historical samples in the second historical sample set are randomly obtained as a training sample set, and the remaining historical samples are used as a test sample set to train an initial fault branch identification model to obtain the fault branch identification model. Specifically, the fault type identification model can be trained by using a random forest method.

[0079] Step S205: identifying each fault characteristic value based on the fault type identification model and the fault branch identification model to obtain a wind farm line fault identification result;

[0080] In the specific implementation process, each fault characteristic value is identified by using the pre-trained fault type identification model to obtain the line fault type of the wind farm, and each fault characteristic value is identified by using the pre-trained fault branch identification model to obtain the line fault branch of the wind farm, so as to obtain the identification result of the wind farm line fault.

[0081] In the specific implementation process, each fault characteristic value is identified by using the pre-trained fault type identification model to obtain the line fault type of the wind farm, and each fault characteristic value is identified by using the pre-trained fault branch identification model to obtain the line fault branch of the wind farm, so as to obtain the identification result of the wind farm line fault.

[0082] In another embodiment of the present application, a wind farm line fault identification device is provided, as shown in Figure 3 The device comprises:

[0083] The acquisition module 1 is configured to acquire a fault current value set corresponding to each phase line of each wind turbine, and the fault current value set is composed of a plurality of fault current values within each predetermined time length before and after the wind farm line fault time.

[0084] The data processing module 2 is configured to perform data processing on each fault current value set by using a preset time-frequency analysis method to obtain a time-frequency matrix and an energy matrix corresponding to each fault current value set.

[0085] the statistical analysis module 3 is configured to perform statistical analysis on each of the sub-matrices by using a predetermined statistical method, to obtain a plurality of fault feature values corresponding to each of the sets of fault current values;

[0086] the fault identification module 4 is configured to identify each of the fault feature values based on a preset fault type identification model and a preset fault branch identification model, to obtain a wind farm line fault identification result.

[0087] In the implementation process, the data processing module 2 is specifically configured to: perform data conversion on each of the sets of fault current values by using a Stockwell time-frequency analysis method, to obtain a time-frequency matrix corresponding to each of the sets of fault current values; and perform modulus calculation on each of the time-frequency matrices, to obtain an energy matrix corresponding to each of the sets of fault current values.

[0088] In the implementation process, the statistical analysis module 3 is specifically configured to: obtain a target time-frequency matrix and a target energy matrix corresponding to a target set of fault current values; compress the target time-frequency matrix and the target energy matrix, to obtain a plurality of sub-matrices corresponding to the target set of fault current values; perform statistical analysis on each of the sub-matrices by using a predetermined statistical method, to obtain a plurality of initial fault feature values corresponding to the target set of fault current values, the predetermined statistical method including at least one of kurtosis, average value, minimum value, maximum value, standard deviation, deviation, and entropy; and screen each of the initial fault feature values by using a Pearson correlation coefficient method, to obtain each target fault current value corresponding to the target set of fault current values, to obtain a plurality of fault feature values corresponding to each of the sets of fault current values.

[0089] In the implementation process, the statistical analysis module 3 is further configured to: based on the target energy matrix, calculate the absolute values of the fault current modulus values of each column in the target energy matrix, to obtain a first sub-matrix composed of the maximum values of each fault current modulus absolute value of each column; based on the target time-frequency matrix, calculate the absolute values of the fault current values of each row in the target time-frequency matrix, to obtain a second sub-matrix composed of the maximum values of each fault current absolute value of each row; based on the target time-frequency matrix, calculate the absolute values of the fault current values of each column in the target time-frequency matrix, to obtain a third sub-matrix composed of the maximum values of each fault current absolute value of each column; and based on the third sub-matrix, obtain a fourth sub-matrix composed of the phase angles of the fault current values corresponding to the maximum values of each fault current absolute value in the third sub-matrix.

[0090] In the implementation process, the statistical analysis module 3 is further configured to: calculate the similarity degrees between each of the initial fault feature values, to obtain a plurality of similarity degrees.

[0091] The similarity is compared with a preset similarity threshold value, and when the target similarity is greater than or equal to the preset similarity threshold value, one of the fault feature values used to calculate the target similarity is randomly deleted to obtain intermediate fault feature values; the intermediate fault feature values are de-duplicated to obtain target fault feature values corresponding to the target fault current value set, so as to obtain a plurality of target fault feature values corresponding to the fault current value sets respectively.

[0092] In the specific implementation process, the wind farm line fault identification device further includes a construction module, which is specifically configured to: based on the historical fault current value sets respectively corresponding to each phase at the outlet side of each wind turbine that are collected in real time, data processing is performed to obtain historical fault feature values; based on the historical fault feature values, a first historical sample set used to train the fault type identification model is constructed, and the historical samples in the first historical sample set carry fault type labels; a preset number of historical samples in the first historical sample set are randomly obtained as a training sample set, and the remaining historical samples are obtained as a test sample set, and an initial fault type identification model is trained to obtain the fault type identification model.

[0093] In the specific implementation process, the construction module is further configured to: based on the historical fault feature values, a second historical sample set used to train the fault branch identification model is constructed, and the historical samples in the second historical sample set carry fault branch labels; a preset number of historical samples in the second historical sample set are randomly obtained as a training sample set, and the remaining historical samples are obtained as a test sample set, and an initial fault branch identification model is trained to obtain the fault branch identification model.

[0094] The present application obtains a fault current value set corresponding to each phase line of each wind turbine, and the fault current value set is composed of a plurality of fault current values within each preset time length before and after the wind farm line fault time; a preset time-frequency analysis method is used to perform data processing on each fault current value set to obtain a time-frequency matrix and an energy matrix corresponding to each fault current value set; based on each time-frequency matrix and each energy matrix, a predetermined statistical method is used for statistical analysis to obtain a plurality of fault feature values corresponding to each fault current value set; based on a preset fault type identification model and a preset fault branch identification model, each fault feature value is identified to obtain a wind farm line fault identification result. The wind farm line fault identification method of the present application can obtain more accurate results of the fault type and the fault branch position of the wind farm line, and improve the efficiency of wind farm line fault identification.

[0095] Another embodiment of the present application provides a storage medium storing a computer program, and the computer program is executed by a processor to implement the following method steps:

[0096] Step one, obtaining a set of fault current values corresponding to each phase line of each wind turbine, the set of fault current values being composed of a plurality of fault current values within a predetermined time period before and after the wind farm line fault time;

[0097] Step two, using a preset time-frequency analysis method to process the data of each set of fault current values, obtaining a time-frequency matrix and an energy matrix corresponding to each set of fault current values;

[0098] Step three, based on each time-frequency matrix and each energy matrix, using a predetermined statistical method for statistical analysis, obtaining a plurality of fault characteristic values corresponding to each set of fault current values;

[0099] Step four, based on a preset fault type identification model and a preset fault branch identification model, identifying each fault characteristic value to obtain a wind farm line fault identification result.

[0100] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer readable storage medium. When the computer program is executed, it can include the processes of the above-mentioned embodiments. Any reference to memory, storage, database or other medium used in the embodiments provided by the present application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration but not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0101] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the division of the above-mentioned functional units and modules is exemplified. In actual application, the above-mentioned functions can be completed by different functional units and modules according to needs, that is, the internal structure of the device is divided into different functional units or modules to complete all or part of the functions described above.

[0102] The specific implementation process of the above method steps can refer to the embodiments of any of the above wind farm line fault identification methods, which will not be repeated here.

[0103] The present application obtains a set of fault current values corresponding to each phase line of each wind turbine, which is composed of a plurality of fault current values within a predetermined time period before and after the wind farm line fault time; a predetermined time-frequency analysis method is used to process the data of each set of fault current values, to obtain a time-frequency matrix and an energy matrix corresponding to each set of fault current values; based on each time-frequency matrix and each energy matrix, a predetermined statistical method is used for statistical analysis to obtain a plurality of fault characteristic values corresponding to each set of fault current values; and based on a predetermined fault type identification model and a predetermined fault branch identification model, each fault characteristic value is identified to obtain a wind farm line fault identification result. The wind farm line fault identification method of the present application can obtain more accurate results of the fault type and fault branch position of the wind farm line, and improve the efficiency of wind farm line fault identification.

[0104] Another embodiment of the present application provides an electronic device, which can be a server. The electronic device includes a processor, a memory, a network interface and a database connected by a system bus. The processor of the electronic device is used to provide computing and control capabilities. The memory of the electronic device includes a non-volatile and / or volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the electronic device is used to communicate with external clients through network connection. The electronic device program is executed by the processor to implement the functions or steps of a wind farm line fault identification method server side.

[0105] In one embodiment, an electronic device is provided, which can be a client. The electronic device includes a processor, a memory, a network interface, a display screen and an input device connected by a system bus. The processor of the electronic device is used to provide computing and control capabilities. The memory of the electronic device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the electronic device is used to communicate with external servers through network connection. The electronic device program is executed by the processor to implement the functions or steps of a wind farm line fault identification method client side.

[0106] Another embodiment of the present application provides an electronic device comprising at least a memory and a processor, wherein the memory stores a computer program, and the processor implements the following method steps when executing the computer program stored in the memory.

[0107] Step one, obtaining a set of fault current values corresponding to each phase line of each wind turbine, wherein the set of fault current values is composed of a plurality of fault current values within a predetermined time period before and after the wind farm line fault time;

[0108] Step two, performing data processing on each set of fault current values by using a preset time-frequency analysis method to obtain a time-frequency matrix and an energy matrix corresponding to each set of fault current values;

[0109] Step three, performing statistical analysis by using a predetermined statistical method based on each time-frequency matrix and each energy matrix to obtain a plurality of fault characteristic values corresponding to each set of fault current values;

[0110] Step four, identifying each fault characteristic value based on a preset fault type identification model and a preset fault branch identification model to obtain a wind farm line fault identification result.

[0111] The specific implementation process of the above method steps can be referred to the embodiments of any wind farm line fault identification method described above, which will not be repeated here.

[0112] The present application obtains a set of fault current values corresponding to each phase line of each wind turbine, wherein the set of fault current values is composed of a plurality of fault current values within a predetermined time period before and after the wind farm line fault time; performs data processing on each set of fault current values by using a preset time-frequency analysis method to obtain a time-frequency matrix and an energy matrix corresponding to each set of fault current values; performs statistical analysis by using a predetermined statistical method based on each time-frequency matrix and each energy matrix to obtain a plurality of fault characteristic values corresponding to each set of fault current values; and identifies each fault characteristic value based on a preset fault type identification model and a preset fault branch identification model to obtain a wind farm line fault identification result. The wind farm line fault identification method of the present application can obtain more accurate results of the fault type and fault branch position of the wind farm line, and improve the efficiency of wind farm line fault identification.

[0113] The above embodiments are only exemplary embodiments of the present application, and are not used to limit the present application, and the protection scope of the present application is defined by the claims. Those skilled in the art can make various modifications or equivalent replacements to the present application within the spirit and protection scope of the present application, and such modifications or equivalent replacements should also be considered to fall within the protection scope of the present application.

Claims

1. A wind farm line fault identification method, characterized in that, The method comprises the following steps: obtaining a set of fault current values corresponding to each phase line of each wind turbine, the set of fault current values being composed of a plurality of fault current values within a predetermined time period before and after the line fault time of the wind farm; performing data processing on each set of fault current values by using a preset time-frequency analysis method to obtain a time-frequency matrix and an energy matrix corresponding to each set of fault current values; based on each time-frequency matrix and each energy matrix, performing statistical analysis by using a predetermined statistical method to obtain a plurality of fault characteristic values corresponding to each set of fault current values; based on a preset fault type identification model and a preset fault branch identification model, identifying each fault characteristic value to obtain a wind farm line fault identification result; based on each time-frequency matrix and each energy matrix, performing statistical analysis by using a predetermined statistical method to obtain a plurality of fault characteristic values corresponding to each set of fault current values, specifically comprising: obtaining a target time-frequency matrix and a target energy matrix corresponding to a target set of fault current values; compressing the target time-frequency matrix and the target energy matrix to obtain a plurality of sub-matrices corresponding to the target set of fault current values; performing statistical analysis on each sub-matrix by using a predetermined statistical method to obtain a plurality of initial fault characteristic values corresponding to the target set of fault current values, the predetermined statistical method including at least one of kurtosis, mean value, minimum value, maximum value, standard deviation, deviation, and entropy; screening each initial fault characteristic value by using a Pearson correlation coefficient method to obtain each target fault characteristic value corresponding to the target set of fault current values, so as to obtain a plurality of fault characteristic values corresponding to each set of fault current values; compressing the target time-frequency matrix and the target energy matrix to obtain a plurality of sub-matrices corresponding to the target set of fault current values, specifically comprising: based on the target energy matrix, calculating the absolute value of the fault current modulus value of each column in the target energy matrix to obtain a first sub-matrix composed of the maximum value of each fault current modulus absolute value in each column; based on the target time-frequency matrix, calculating the absolute value of the fault current value of each row in the target time-frequency matrix to obtain a second sub-matrix composed of the maximum value of each fault current absolute value in each row; based on the target time-frequency matrix, calculating the absolute value of the fault current value of each column in the target time-frequency matrix to obtain a third sub-matrix composed of the maximum value of each fault current absolute value in each column; based on the third sub-matrix, obtaining a fourth sub-matrix composed of the phase angle of the fault current value corresponding to each fault current absolute value maximum in the third sub-matrix; the screening of each initial fault characteristic value by using the Pearson correlation coefficient method to obtain each target fault current value corresponding to the target set of fault current values, so as to obtain a plurality of fault characteristic values corresponding to each set of fault current values, specifically comprising: calculating the similarity between each initial fault characteristic value to obtain a plurality of similarities; The similarity is compared with a preset similarity threshold value, and when the target similarity is greater than or equal to the preset similarity threshold value, one of the fault feature values used to calculate the target similarity is randomly deleted to obtain intermediate fault feature values; The intermediate fault feature values are de-duplicated to obtain target fault feature values corresponding to the target fault current value set, so as to obtain a plurality of target fault feature values corresponding to the fault current value sets respectively.

2. The method of claim 1, wherein, The data processing module is configured to perform data processing on the fault current value sets by using a preset time-frequency analysis method to obtain time-frequency matrices and energy matrices corresponding to the fault current value sets. The data processing module is configured to perform data conversion on the fault current value sets by using a Stockwell time-frequency analysis method to obtain time-frequency matrices corresponding to the fault current value sets. The data processing module is configured to perform modulus calculation on the time-frequency matrices to obtain energy matrices corresponding to the fault current value sets.

3. The method of claim 1, wherein, Before identifying the fault feature values based on a preset fault type identification model and a preset fault branch identification model to obtain a wind farm line fault identification result, the method further includes: constructing the fault type identification model, specifically including: Performing data processing on the historical fault current value sets corresponding to each phase of each wind turbine outlet based on real-time collection to obtain historical fault feature values; Based on the historical fault feature values, a first historical sample set for training the fault type identification model is constructed, and historical samples in the first historical sample set carry fault type labels; Randomly obtaining a preset number of historical samples in the first historical sample set as a training sample set and the remaining historical samples as a test sample set to perform model training on an initial fault type identification model to obtain the fault type identification model.

4. The method of claim 3, wherein, After randomly obtaining a preset number of historical samples in the first historical sample set as a training sample set and the remaining historical samples as a test sample set to perform model training on an initial fault type identification model to obtain the fault type identification model, the method further includes: Based on the historical fault feature values, a second historical sample set for training the fault branch identification model is constructed, and historical samples in the second historical sample set carry fault branch labels; Randomly obtaining a preset number of historical samples in the second historical sample set as a training sample set and the remaining historical samples as a test sample set to perform model training on an initial fault branch identification model to obtain the fault branch identification model.

5. A wind farm line fault identification device for implementing the wind farm line fault identification method according to any one of claims 1 to 4, characterized in that, The data processing module is configured to perform data processing on the fault current value sets by using a preset time-frequency analysis method to obtain time-frequency matrices and energy matrices corresponding to the fault current value sets. ​ ​ The statistical analysis module is configured to perform statistical analysis on each of the time-frequency matrix and each of the energy matrix by using a predetermined statistical method, and obtain a plurality of fault characteristic values corresponding to each of the set of fault current values. The fault identification module is configured to identify each of the fault characteristic values based on a preset fault type identification model and a preset fault branch identification model, and obtain a wind farm line fault identification result.

6. A storage medium, characterized by The storage medium stores a computer program, and the computer program is executed by the processor to implement the steps of the wind farm line fault identification method in any one of claims 1-4.

7. An electronic device, comprising: The device comprises at least a memory and a processor, and the memory stores a computer program, and the processor implements the steps of the wind farm line fault identification method in any one of claims 1-4 when executing the computer program stored in the memory. The device comprises at least a memory and a processor, and the memory stores a computer program, and the processor implements the steps of the wind farm line fault identification method in any one of claims 1-4 when executing the computer program stored in the memory.

Images