Wind driven generator knowledge graph construction method and system

Abstract

The invention discloses a method and a system for constructing a knowledge graph of a wind driven generator, which are used for analyzing the correlation between the conditions and entity occurrence by investigating the weather and humidity conditions of the wind driven generator when a fault entity occurs. According to the method, which specific conditions are associated with the occurrence of the entity can be determined, and the occurrence probability of the entity under different conditions can be calculated. In this way, the multi-dimensional data can be effectively screened and analyzed.

Description

Technical Field

[0001] The present invention belongs to the field of data processing technology, and in particular relates to a method and system for constructing a knowledge graph of a wind turbine. Background Art

[0002] In power system management, building a knowledge graph is a key process aimed at improving the efficiency and reliability of the system. However, the current traditional method of building a knowledge graph for power system faults mainly relies on the operating parameters of the power equipment itself and lacks comprehensive analysis in multiple fields. For example, the current construction of a knowledge graph for wind turbines only relies on information such as the operating data, fault logs, and maintenance records of wind turbines. This method ignores the impact of external factors such as meteorological conditions, geographical features, and regional load changes on power system faults. Therefore, the current wind turbine knowledge graph construction method only relies on internal operating parameters and fails to effectively integrate data from multiple fields such as meteorology and geography, resulting in the incompleteness and inaccuracy of the knowledge graph in describing and predicting power system faults. The constructed knowledge graph may not fully reflect all potential factors that lead to faults, thereby affecting the effectiveness of fault diagnosis and preventive measures. Summary of the invention

[0003] In order to solve the problems existing in the prior art, the present invention proposes a method and system for constructing a wind turbine knowledge graph.

[0004] The technical solution of the present invention is as follows:

[0005] A method for constructing a wind turbine knowledge graph, comprising:

[0006] Obtaining time series data of multiple wind turbines under multiple conditions and multiple fault types that have occurred;

[0007] Each fault type is regarded as a fault entity, and under each fault entity, the wind turbines are divided into target-type wind turbines representing fault entities and normal-type wind turbines representing no fault entities; according to the correlation index of the target-type wind turbines and the normal-type wind turbines under each fault entity, it is judged whether there is a specific correlation between each fault entity and any condition, and the condition that is specifically correlated with each fault entity is obtained;

[0008] For each entity fault, obtain the time series data of each wind turbine in the target type wind turbine under each condition within a preset time period after each fault, determine the impact association between each fault entity and any condition for each target wind turbine, and obtain the fault entity and condition with impact association corresponding to each target wind turbine, and obtain the ratio of the number of faults occurring in the preset number of faults before the current detection to the preset value of faults occurring for the fault entity and condition with impact association corresponding to each target wind turbine, and at the same time, form the impact combination of each target wind turbine in the target type wind turbine under each fault entity based on the condition with impact association with each fault entity;

[0009] Based on the ratio of the number of faults occurring in the preset number of faults occurring before the current detection to the preset value of faults associated with the fault entities and conditions corresponding to each target wind turbine, a first probability and a second probability of an impact combination of each target wind turbine in the target type wind turbines under each fault entity are obtained; the impact combination is screened according to the first probability and the second probability to obtain an updated impact combination of each target wind turbine in the target type wind turbines of each fault entity;

[0010] The construction of the wind turbine knowledge graph is completed based on the update probability of the update impact combination and the update probability of the condition associated with each fault entity.

[0011] Furthermore, the specific steps of judging whether there is a specific association between each fault entity and any condition according to the inter-group association index of the target type wind turbine generator and the normal type wind turbine generator under each fault entity, and obtaining the condition that is specifically associated with each fault entity include:

[0012] Select any fault entity and a condition, record them as fault entity A and the ath condition, among the target type wind turbines under fault entity A, group any two target wind turbines together, obtain the absolute value of the difference between the mean values ​​of the time series data of the ath condition in each group, record them as the first difference of the ath condition in each group;

[0013] According to the method for obtaining the first difference value of the ath condition in each group of the target type wind turbine generator, the first difference value of the ath condition in each group of the normal type wind turbine generator is obtained;

[0014] Based on the first difference of the ath condition in each group of target-type wind turbines and the first difference of the ath condition in each group of normal-type wind turbines, the correlation index of the ath condition is calculated using the following formula:

[0015]

[0016] In the formula, G ais the correlation index of the ath condition, CY a is the mean of the first difference of the ath condition of all groups in the target type wind turbine, CY' a is the mean of the first difference of the ath condition of all groups in the normal type wind turbine, norm[] is the linear normalization function, and || is the absolute value;

[0017] If the correlation index of the ath condition is greater than the preset threshold, it is determined that there is a specific correlation between the ath condition and the fault entity A;

[0018] Select other conditions and repeat the above steps to determine whether there is a specific association between other conditions and fault entity A;

[0019] Select other fault entities, repeat the above steps, and determine whether there is a specific association between each fault entity and any condition, thereby obtaining the condition under which each fault entity has a specific association.

[0020] Further, for each entity fault, the time series data of each wind turbine in the target type wind turbine under each condition within a preset time period after each fault is obtained, the influence association between each fault entity and any condition of each target wind turbine is determined, and the fault entity and condition with influence association corresponding to each target wind turbine are obtained, and the ratio of the number of faults occurring in the preset number of faults occurring before the current detection to the preset value of faults occurring in the fault entity and condition with influence association corresponding to each target wind turbine is obtained. The specific steps include:

[0021] Choose any fault entity and a condition, record them as fault entity A and the ath condition;

[0022] For fault entity A, the mean of the ath conditional time series data of each target wind turbine within a preset time period after each fault occurs is recorded as the first mean value of each target wind turbine after each fault occurs;

[0023] The mean of the time series data of each target wind turbine generator within a preset time period after each failure occurs is recorded as the second mean;

[0024] In the target type wind turbines, the absolute value of the difference between the first mean and the second mean of each target wind turbine after each fault occurs is calculated; when the absolute value of the difference is greater than one quarter of the second mean, it is determined that there is an influence association between the ath condition and the Ath fault entity of each target wind turbine when each fault occurs;

[0025] For each target wind turbine that has experienced a preset number of failures before the current monitoring, the ratio of the number of failures corresponding to the impact association between the a-th condition and the A-th fault entity to the preset value;

[0026] Select other conditions, repeat the above steps, determine the influence association between each target wind turbine and the fault entity A and any condition, and obtain the condition corresponding to each target wind turbine that has an influence association with the fault entity A, and obtain the ratio of the number of failures of each target wind turbine that has an influence association with the fault entity A to the preset number of failures before the current detection to the preset value;

[0027] Select other fault entities, repeat the above steps, determine the impact association between each fault entity and any condition for each target wind turbine, and obtain the fault entity and condition associated with the existence of the impact for each target wind turbine, and obtain the ratio of the number of faults that occur in the preset number of times before the current detection to the preset value for the fault entity and condition associated with the existence of the impact for each target wind turbine.

[0028] Furthermore, the specific steps of forming the influence combination of each target wind turbine in the target type wind turbine under each fault entity based on the condition that there is an influence association with each fault entity include:

[0029] Choose any faulty entity and record it as faulty entity A;

[0030] All conditions in each target wind turbine generator that are associated with the Ath fault entity form an impact combination;

[0031] Select other fault entities and repeat the above steps to form an impact combination of each target wind turbine in the target type wind turbine under each fault entity.

[0032] Further, based on the ratio of the number of faults occurring in the preset number of faults occurring before the current detection to the preset value of faults associated with the fault entities and conditions corresponding to each target wind turbine, the specific steps of obtaining the first probability and the second probability of the influence combination of each target wind turbine in the target type wind turbine under each fault entity include:

[0033] Choose any faulty entity and record it as faulty entity A;

[0034] The average of the ratios of the number of failures occurring in the preset number of failures before the current detection to the preset value for each condition in the influence combination of each target wind turbine in the target type wind turbine under the fault entity A is recorded as the first probability of the influence combination of each target wind turbine in the target type wind turbine under the fault entity A;

[0035] The average of the first probabilities of all target wind turbines in the target type wind turbines under fault entity A is recorded as the second probability;

[0036] Select other fault entities, repeat the above steps, and obtain the first probability and the second probability of the impact combination of each target wind turbine in the target type wind turbine under each fault entity.

[0037] Furthermore, the specific steps of screening the impact combination according to the first probability and the second probability to obtain the updated impact combination of each target wind turbine in each fault entity target type wind turbine include:

[0038] Choose any faulty entity and record it as faulty entity A;

[0039] When the first probability of the impact combination of each target wind turbine in the target type wind turbine under the fault entity A is greater than the second probability, the impact combination of each target wind turbine is recorded as an updated impact combination.

[0040] Select other fault entities and repeat the above steps to obtain the updated impact combination of each target wind turbine in each fault entity target class wind turbine.

[0041] Furthermore, the specific steps of completing the construction of the wind turbine knowledge graph based on the update probability of the update impact combination and the update probability of the condition associated with each fault entity include:

[0042] Choose any faulty entity and record it as faulty entity A;

[0043] Calculate the average of the ratios of the number of failures of all target wind turbines to the preset number of failures before the current detection for each condition that is associated with the fault entity A, and record it as the updated probability of each condition that is associated with the fault entity A;

[0044] The mean of the correlation indexes of all conditions in each update impact combination is recorded as the correlation index of the update impact combination;

[0045] According to the updated probability of each condition associated with the fault entity A and the associated index of the updated impact combination, the construction of the power knowledge graph of multidimensional data is completed. The specific method is as follows:

[0046] The attribute vector of the node constructed by fault entity A in the knowledge graph is [z, x, c, v, b, n], where z is a combination of all conditions that have an impact on fault entity A, x is a combination of the association indexes of all conditions that have an impact on fault entity A; c is a combination of the update probabilities of all conditions that have an impact on fault entity A; v is a combination of all update impact combinations; b is a combination of the association indexes of all update impact combinations; n is a combination of the first probabilities of all update impact combinations;

[0047] Among all the fault types that have occurred in all wind turbines, each fault type is taken as a node, and an electric power knowledge graph is constructed based on the attribute vectors corresponding to all fault types.

[0048] A wind turbine knowledge graph construction system, comprising a multidimensional data acquisition module, a first data processing module, a second data processing module, a third data processing module and a knowledge graph construction module;

[0049] The multi-dimensional data acquisition module is used to acquire time series data of multiple wind turbines under multiple conditions and multiple fault types that have occurred;

[0050] The first data processing module is used to treat each fault type as a fault entity, and divide the wind turbines into target-type wind turbines representing fault entities and normal-type wind turbines representing no fault entities under each fault entity; according to the correlation index of the target-type wind turbines and the normal-type wind turbines under each fault entity, determine whether there is a specific correlation between each fault entity and any condition, and obtain the condition that is specifically correlated with each fault entity;

[0051] The second data processing module is used to obtain, for each entity fault, the time series data of each wind turbine in the target type wind turbine under each condition within a preset time period after each fault, determine the impact association between each fault entity and any condition for each target wind turbine, and obtain the fault entity and condition with impact association corresponding to each target wind turbine, and obtain the ratio of the number of faults occurring in the preset number of faults before the current detection to the preset value of faults occurring in the fault entity and condition with impact association corresponding to each target wind turbine, and at the same time form an impact combination of each target wind turbine in the target type wind turbine under each fault entity based on the condition with impact association with each fault entity;

[0052] The third data processing module is used to obtain a first probability and a second probability of an influence combination of each target wind turbine in the target type wind turbines under each fault entity based on the ratio of the number of faults occurring in the preset number of faults occurring before the current detection to the preset value of faults associated with the fault entities and conditions corresponding to each target wind turbine; and screen the influence combination according to the first probability and the second probability to obtain an updated influence combination of each target wind turbine in the target type wind turbine for each fault entity;

[0053] The knowledge graph construction module is used to complete the construction of the wind turbine knowledge graph based on the update probability of the update impact combination and the update probability of the condition associated with each fault entity.

[0054] An electronic device comprises a memory and a processor, wherein the memory stores a computer program, and the processor is used to call and run the computer program stored in the memory to execute any of the methods described above.

[0055] A computer-readable storage medium stores a computer program, wherein the computer program implements the steps of any of the above methods when executed by a processor.

[0056] Compared with the prior art, the present invention has the following beneficial effects:

[0057] The present invention proposes a method and system for constructing a knowledge graph of a wind turbine, which can examine in detail the meteorological and humidity conditions when a fault entity occurs, and analyze the correlation between these conditions (including meteorological conditions, temperature conditions, humidity conditions, geographical terrain conditions, and regional load change conditions) and the occurrence of the fault entity. The method can determine which specific conditions are associated with the occurrence of the fault entity, and calculate the probability of the fault entity occurring under different conditions. In this way, the present invention can effectively screen and analyze multidimensional data. In addition, the present invention also integrates the results into the knowledge graph, thereby enriching the content of the constructed wind turbine knowledge graph and optimizing its structure. This integration helps to improve the integrity, accuracy, and real-time performance of the wind turbine knowledge graph, and significantly improves the efficiency of power system management and maintenance. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 This is a flow chart of a method for constructing a wind turbine knowledge graph in an embodiment. DETAILED DESCRIPTION

[0059] The present invention is further explained below in conjunction with the accompanying drawings and specific embodiments. It should be understood that these embodiments are only used to illustrate the present invention and are not used to limit the scope of the present invention. After reading the present invention, various equivalent modifications of the present invention by those skilled in the art all fall within the scope defined by the claims attached to this application.

[0060] Embodiment 1:

[0061] A wind turbine knowledge graph construction method of the present invention, such as Figure 1 As shown, including:

[0062] S1. Obtain the time series data of multiple wind turbines under multiple conditions and multiple fault types that have occurred; the conditions can be meteorological conditions (precipitation, thunderstorm activity), temperature conditions, humidity conditions, geographical conditions (altitude, terrain features, and landforms), and regional load change conditions, etc.; the relevant data on meteorological conditions, temperature conditions, and humidity conditions in the area where the target wind power station is located can be obtained by visiting the meteorological website; the relevant data on geographical conditions in the area where the target wind power station is located can be obtained by visiting the geographic information system (GIS) service provider. Clean the collected time series data combined with multiple conditions, remove erroneous, duplicated, or incomplete records, and obtain time series data under multiple conditions.

[0063] S2. Take each fault type as a fault entity, and divide the wind turbines into target wind turbines with fault entities and normal wind turbines without fault entities under each fault entity; according to the correlation index of the target wind turbines and the normal wind turbines under each fault entity, determine whether there is a specific correlation between each fault entity and any condition, and obtain the conditions that are specifically correlated with each fault entity;

[0064] S3. For each entity fault, obtain the time series data of each wind turbine in the target type wind turbine under each condition within a preset time period after each fault, determine the impact association between each fault entity and any condition for each target wind turbine, and obtain the fault entity and condition with impact association corresponding to each target wind turbine, and obtain the ratio of the number of faults occurring in the preset number of faults before the current detection to the preset value of faults occurring for the fault entity and condition with impact association corresponding to each target wind turbine, and at the same time, form the impact combination of each target wind turbine in the target type wind turbine under each fault entity based on the condition with impact association with each fault entity;

[0065] S4. Based on the ratio of the number of faults occurring in the preset number of faults occurring before the current detection to the preset value for the fault entities and conditions associated with the existence of each target wind turbine, obtain the first probability and the second probability of the influence combination of each target wind turbine in the target type wind turbine under each fault entity; screen the influence combination according to the first probability and the second probability, and obtain the updated influence combination of each target wind turbine in the target type wind turbine for each fault entity;

[0066] S5. Complete the construction of the wind turbine knowledge graph based on the update probability of the update impact combination and the update probability of the conditions associated with each fault entity.

[0067] Embodiment 2:

[0068] The present invention further optionally designs on the basis of the first embodiment that, in this example, according to the inter-group correlation index of the target type wind turbine generator and the normal type wind turbine generator under each fault entity, it is judged whether there is a specific correlation between each fault entity and any condition, and the specific steps of obtaining the condition that is specifically correlated with each fault entity include:

[0069] Select any fault entity and a condition, record them as fault entity A and the ath condition, among the target type wind turbines under fault entity A, group any two target wind turbines together, obtain the absolute value of the difference between the mean values ​​of the time series data of the ath condition in each group, record them as the first difference of the ath condition in each group;

[0070] According to the method for obtaining the first difference value of the ath condition in each group of the target type wind turbine generator, the first difference value of the ath condition in each group of the normal type wind turbine generator is obtained;

[0071] Based on the first difference of the ath condition in each group of target-type wind turbines and the first difference of the ath condition in each group of normal-type wind turbines, the correlation index of the ath condition is calculated using the following formula:

[0072]

[0073] In the formula, G a is the correlation index of the ath condition, CY a is the mean of the first difference of the ath condition of all groups in the target type wind turbine, CY' a is the mean of the first difference of the ath condition of all groups in the normal type wind turbine, norm[] is the linear normalization function, and || is the absolute value;

[0074] If the correlation index of the ath condition is greater than the preset threshold, it is determined that there is a specific correlation between the ath condition and the fault entity A;

[0075] Select other conditions and repeat the above steps to determine whether there is a specific association between other conditions and fault entity A;

[0076] Select other fault entities, repeat the above steps, and determine whether there is a specific association between each fault entity and any condition, thereby obtaining the condition under which each fault entity has a specific association.

[0077] The following is an example to further illustrate the process of obtaining the conditions for the specific association of the above fault entities:

[0078] First, a specific fault of the target wind turbine is determined, and the specific fault is taken as an entity to obtain the analysis target A1;

[0079] Using the data samples of the target wind turbine, we compare the data samples j with and without the analysis target A1, and calculate the differences in their meteorological and geographical conditions i. When there is an analysis target A1, i=1, otherwise i=0;

[0080] It should be noted that when obtaining the differences in meteorological, geographical and other condition values, especially meteorological differences, the meteorological data when the wind turbine fails are selected. For data samples that do not exist as analysis targets, the mean of the regional meteorological data is used as the corresponding meteorological, geographical and other condition values.

[0081] Calculate the mean of the difference between condition i in the two cases separately Right now and

[0082] When there is a mean of the conditional differences in the data sample of the analysis target A1 The larger the value, the more likely it is that the current condition is related to the equipment failure. The smaller the value, the smaller the mean of the conditional difference The smaller the mean value is, the smaller the difference between the conditions in the data sample where the analysis target A1 does not exist. The larger the difference, the more relevant the current condition is to the equipment failure (correlation index). The bigger;

[0083]

[0084] According to different analysis objectives A n And different conditions i, repeatedly obtain the relationship score between condition i and the analysis target That is, condition i has different equipment failure from the target wind turbine (analysis target A n ), and the relationship scores between the analysis target A1 and different conditions can also be obtained repeatedly.

[0085] Set a threshold to filter the relationship scores between the above conditions and specific faults. When condition i is considered to be consistent with the analysis target A n There is a relationship (specific relationship) and the associated conditions with the specific fault are identified. Select any analysis target A n , for the wind turbine data sample in the same area, the evaluation condition i is used to evaluate the analysis target A n the size of the risk.

[0086] Embodiment three:

[0087] The present invention further optionally designs on the basis of the second embodiment, in this example, for each entity fault, obtaining the time series data of each wind turbine in the target type wind turbine under each condition within a preset time period after each fault, determining the influence association between each fault entity and any condition for each target wind turbine, and obtaining the fault entity and condition with influence association corresponding to each target wind turbine, and obtaining the ratio of the number of faults occurring in the preset number of faults occurring before the current detection to the preset value of faults occurring in the fault entity and condition with influence association corresponding to each target wind turbine, the specific steps include:

[0088] Choose any fault entity and a condition, record them as fault entity A and the ath condition;

[0089] For fault entity A, the mean of the ath conditional time series data of each target wind turbine within a preset time period after each fault occurs is recorded as the first mean value of each target wind turbine after each fault occurs;

[0090] The mean of the time series data of each target wind turbine generator within a preset time period after each failure occurs is recorded as the second mean;

[0091] In the target type wind turbines, the absolute value of the difference between the first mean and the second mean of each target wind turbine after each fault occurs is calculated; when the absolute value of the difference is greater than one quarter of the second mean, it is determined that there is an influence association between the ath condition and the Ath fault entity of each target wind turbine when each fault occurs;

[0092] In the preset number of failures of each target wind turbine before the current monitoring, the ratio of the number of failures corresponding to the influence association between the a-th condition and the A-th fault entity to the preset value (which can be recorded as the probability of each condition of each target wind turbine in the target type wind turbine causing the failure to occur, referred to as "the probability of the condition");

[0093] Select other conditions, repeat the above steps, determine the influence association between each target wind turbine and the fault entity A and any condition, and obtain the condition corresponding to each target wind turbine that has an influence association with the fault entity A, and obtain the ratio of the number of failures of each target wind turbine that has an influence association with the fault entity A to the preset number of failures before the current detection to the preset value;

[0094] Select other fault entities, repeat the above steps, determine the impact association between each fault entity and any condition for each target wind turbine, and obtain the fault entity and condition associated with the existence of the impact for each target wind turbine, and obtain the ratio of the number of faults that occur in the preset number of times before the current detection to the preset value for the fault entity and condition associated with the existence of the impact for each target wind turbine.

[0095] The following is an example to further illustrate the above process:

[0096] Analyze the time series data of any wind turbine and compare the occurrence of analysis target A at different times n The time series data differences under each condition at this time (it should be noted that the time series data differences obtained at this time only include those with the analysis target A n Conditions with a specific relationship, that is, conditions with a specific association), evaluate the probability of each condition causing a failure.

[0097] First, search for historical data samples of the current wind turbine to determine the corresponding data (timestamp, meteorological and geographical conditions, etc.) of the analysis target A1 appearing at different times;

[0098] Calculate the data mean H of condition i in the time series i , and obtain the data difference each time a fault occurs (analysis target A1) When the data difference is smaller, it means that the fault corresponding to timestamp t is less related to condition i, and vice versa, the possibility that it is caused by condition i is greater;

[0099] Set the threshold value when the data difference value occurs each time a fault occurs (analysis target A1) When , it is considered that condition i has a greater impact on the occurrence of this fault. Here, the condition with a greater impact can be regarded as a condition with an impact association with the fault entity, and the condition with a smaller impact can be regarded as a condition with no impact association with the fault entity.

[0100] Count the number of times condition i affects analysis target A1, and combine the total number of times analysis target A1 appears in the current wind turbine to obtain the probability of failure caused by condition i.

[0101] Repeat the above steps to obtain the probability of failure (analysis target A1) caused by different conditions

[0102] Embodiment 4:

[0103] The present invention further optionally designs on the basis of the third embodiment, in this example, the specific steps of forming the influence combination of each target wind turbine in the target type wind turbine under each fault entity based on the condition that there is an influence association with each fault entity include:

[0104] Choose any faulty entity and record it as faulty entity A;

[0105] All conditions in each target wind turbine generator that are associated with the Ath fault entity form an impact combination;

[0106] Select other fault entities and repeat the above steps to form an impact combination of each target wind turbine in the target type wind turbine under each fault entity.

[0107] Embodiment five:

[0108] The present invention further optionally designs on the basis of the fourth embodiment, in this example, based on the ratio of the number of faults occurring in the preset number of faults occurring before the current detection to the preset value corresponding to the fault entities and conditions associated with the existence of each target wind turbine, the specific steps of obtaining the first probability and the second probability of the influence combination of each target wind turbine in the target type wind turbine under each fault entity include:

[0109] Choose any faulty entity and record it as faulty entity A;

[0110] The average value of the ratio of the number of failures of each condition in the preset number of failures before the current detection to the preset value (i.e., the probability of each condition) in the influence combination of each target wind turbine in the target type wind turbine under the fault entity A is recorded as the first probability of the influence combination of each target wind turbine in the target type wind turbine under the fault entity A;

[0111] The average of the first probabilities of all target wind turbines in the target type wind turbines under fault entity A is recorded as the second probability;

[0112] Select other fault entities, repeat the above steps, and obtain the first probability and the second probability of the impact combination of each target wind turbine in the target type wind turbine under each fault entity.

[0113] Embodiment six:

[0114] The present invention further optionally designs on the basis of the fifth embodiment, in this example, the influence combination is screened according to the first probability and the second probability, and the specific steps of obtaining the updated influence combination of each target wind turbine in each fault entity target type wind turbine include:

[0115] Choose any faulty entity and record it as faulty entity A;

[0116] When the first probability of the impact combination of each target wind turbine in the target type wind turbine under the fault entity A is greater than the second probability, the impact combination of each target wind turbine is recorded as an updated impact combination.

[0117] Select other fault entities and repeat the above steps to obtain the updated impact combination of each target wind turbine in each fault entity target class wind turbine.

[0118] Embodiment seven:

[0119] The present invention further optionally designs on the basis of the sixth embodiment, in this example, the specific steps of completing the construction of the wind turbine knowledge graph based on the update probability of the update impact combination and the update probability of the condition associated with each fault entity include:

[0120] Choose any faulty entity and record it as faulty entity A;

[0121] Calculate the average value of the ratio of the number of failures of each condition associated with the fault entity A to the preset number of failures before the current detection (i.e., the probability of each condition) of all target wind turbines, and record it as the updated probability of each condition associated with the fault entity A;

[0122] The mean of the correlation indexes of all conditions in each update impact combination is recorded as the correlation index of the update impact combination;

[0123] According to the updated probability of each condition associated with the fault entity A and the associated index of the updated impact combination, the construction of the power knowledge graph of multidimensional data is completed. The specific method is as follows:

[0124] The attribute vector of the node constructed by fault entity A in the knowledge graph is [z, x, c, v, b, n], where z is a combination of all conditions that have an impact on fault entity A, x is a combination of the association indexes of all conditions that have an impact on fault entity A; c is a combination of the update probabilities of all conditions that have an impact on fault entity A; v is a combination of all update impact combinations; b is a combination of the association indexes of all update impact combinations; n is a combination of the first probabilities of all update impact combinations;

[0125] Among all the fault types that have occurred in all wind turbines, each fault type is taken as a node, and an electric power knowledge graph is constructed based on the attribute vectors corresponding to all fault types.

[0126] Embodiment eight:

[0127] A wind turbine knowledge graph construction system of the present invention comprises a multidimensional data acquisition module, a first data processing module, a second data processing module, a third data processing module and a knowledge graph construction module;

[0128] A multi-dimensional data acquisition module, used to acquire time series data of multiple wind turbines under multiple conditions and multiple fault types that have occurred;

[0129] The first data processing module is used to treat each fault type as a fault entity, and divide the wind turbines into target-type wind turbines representing fault entities and normal-type wind turbines representing no fault entities under each fault entity; according to the correlation index of the target-type wind turbines and the normal-type wind turbines under each fault entity, determine whether there is a specific correlation between each fault entity and any condition, and obtain the condition that is specifically correlated with each fault entity;

[0130] A second data processing module is used to obtain, for each entity fault, time series data of each wind turbine in the target type wind turbine under each condition within a preset time period after each fault, determine the impact association between each fault entity and any condition for each target wind turbine, and obtain the fault entity and condition with impact association corresponding to each target wind turbine, and obtain the ratio of the number of faults occurring in a preset number of faults before the current detection to the preset value of faults occurring in the fault entity and condition with impact association corresponding to each target wind turbine, and at the same time form an impact combination of each target wind turbine in the target type wind turbine under each fault entity based on the condition with impact association with each fault entity;

[0131] The third data processing module is used to obtain the first probability and the second probability of the influence combination of each target wind turbine in the target type wind turbine under each fault entity based on the ratio of the number of faults occurring in the preset number of faults occurring before the current detection to the preset value of faults associated with the fault entity and condition corresponding to each target wind turbine; screen the influence combination according to the first probability and the second probability to obtain the updated influence combination of each target wind turbine in the target type wind turbine for each fault entity;

[0132] The knowledge graph construction module is used to complete the construction of the wind turbine knowledge graph based on the update probability of the update impact combination and the update probability of the conditions associated with each fault entity.

[0133] Embodiment nine:

[0134] An electronic device comprises a memory and a processor, wherein the memory stores a computer program, and the processor is used to call and run the computer program stored in the memory to execute any of the methods described above.

[0135] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of any of the above methods are implemented.

[0136] Application examples:

[0137] The wind turbine knowledge graph construction method of the present invention comprises the following specific steps:

[0138] Step S001: within a preset time period, a plurality of conditional time series data and the types of faults that have occurred are obtained for each wind turbine in different areas.

[0139] In a preset time period, multiple conditional time series data of each wind turbine in different areas are obtained, wherein the conditions include meteorological conditions, temperature conditions, humidity conditions, etc.

[0140] Within a preset time period, the type of fault that has occurred for each wind turbine in different areas is obtained.

[0141] In this embodiment, the conditional time series data is obtained from reports from meteorological temperature observation stations, environmental humidity monitoring information systems, and the like.

[0142] It should be noted that this embodiment is described by taking the case where the collection interval of each conditional time series data is 4 hours and the collection duration is 5 days as an example.

[0143] The obtained time series data for each condition are classified and standardized to ensure the consistency of the format, which is convenient for subsequent data comparison and analysis.

[0144] Step S002: Select any fault type as the Ath fault entity, and divide all wind turbines into target type wind turbines and normal type wind turbines according to whether the Ath fault entity exists; and determine whether there is a specific association between each condition and the Ath fault entity according to the time series data of each condition of the target type wind turbines and the normal type wind turbines.

[0145] There are significant differences in the failure conditions and probabilities of wind turbines under different meteorological and geographical conditions. These differences are often due to the combined effects of multiple complex factors, such as climate characteristics, terrain changes, and soil properties. Under different meteorological and geographical conditions, the failure conditions and failure probabilities of wind turbines in different regions are compared and analyzed to obtain the degree of correlation between specific conditions and equipment failures, and to identify and define the relationship between meteorological information and geographical information and entities in the knowledge graph. For example, local equipment failures of wind turbines may be associated with specific weather conditions, or certain geographical features may affect the operating efficiency of equipment in wind turbines.

[0146] Among all the fault types that have occurred in all wind turbines, any one fault type is recorded as the Ath fault entity. The fault type may be blade damage, gearbox failure, etc.

[0147] According to whether there is an Ath fault entity, all wind turbines in different areas are divided into two categories. The wind turbines in the category with the Ath fault entity are recorded as target wind turbines, and the wind turbines in the category without the Ath fault entity are recorded as normal wind turbines.

[0148] Among the target type wind turbines, any two target wind turbines are grouped together, and the absolute value of the difference between the mean values ​​of the a-th condition time series data of each group is obtained, which is recorded as the first difference of the a-th condition of each group.

[0149] According to the method for obtaining the first difference value of the ath condition in each group of the target type wind turbine generators, the first difference value of the ath condition in each group of the normal type wind turbine generators is obtained.

[0150] In order to quantify the strength of the relationship between a condition and a fault, this embodiment provides a calculation method of a condition correlation index as follows:

[0151]

[0152] Where: G a is the correlation index of the ath condition, CY a is the mean of the first difference of the ath condition of all groups in the target type wind turbine, CY' a is the mean of the first difference of the ath condition of all groups in the normal type wind turbine, norm[] is the linear normalization function, and || is the absolute value.

[0153] In the above formula, G a The closer the value is to 1, the stronger the relationship between the ath condition and the Ath fault entity is; conversely, the closer the value is to 0, the weaker the relationship is.

[0154] If the correlation index of the ath condition is greater than a preset threshold, it is determined that there is a specific correlation between the ath condition and the Ath fault entity.

[0155] In this embodiment, the preset threshold is 0.6, which is used as an example for description. In other implementations, it can be set to other values, which are not limited in this embodiment.

[0156] Similarly, determine whether there is a specific association between other conditions and the Ath fault entity.

[0157] It should be noted that the correlation analysis in this embodiment is only used to reveal the relationship between meteorological and geographical conditions and equipment failures, but cannot directly explain the impact of a specific condition on equipment failures.

[0158] Step S003: Obtain the time series data of each condition within a preset time period after each failure of all target wind turbines in the target type wind turbines; obtain the probability of each condition of each target wind turbine in the target type wind turbines according to the time series data of each condition within a preset time period after each failure of each target wind turbine in the target type wind turbines.

[0159] In the target type wind turbine, the ath conditional time series data of the Lth target wind turbine within the preset time period after each fault occurs is obtained. In this embodiment, the preset value is 6, which is taken as an example. It should be noted that the ath conditional time series data when a fault occurs only obtains the time series data within one week when the fault occurs.

[0160] The mean value of the ath conditional time series data after a preset time period after each failure of the Lth target wind turbine generator is recorded as the first mean value of the Lth target wind turbine generator when each failure occurs.

[0161] Furthermore, the mean value of the time series data of the Lth target wind turbine within a preset time period after each failure occurs is recorded as a second mean value.

[0162] In the target type wind turbines, the absolute value of the difference between the first mean and the second mean of the Lth target wind turbine when a fault occurs each time is calculated. When the absolute value of the difference is greater than one quarter of the second mean, it is determined that there is an influence association between the ath condition and the Ath fault entity of the Lth target wind turbine when a fault occurs each time.

[0163] It should be noted that the smaller the absolute value of the above difference is, the smaller the correlation between the fault and the condition at a certain moment is, and vice versa.

[0164] Within a preset time period after each failure of the Lth target wind turbine, the ratio of the number of failures corresponding to the influence association between the ath condition and the Ath fault entity to a preset value is recorded as the probability of each condition of each target wind turbine in the target type wind turbine.

[0165] Similarly, the probability of each condition of the Lth target wind turbine in the target type wind turbine is obtained.

[0166] Step S004: Based on whether there is a specific association between each condition and the Ath fault entity, obtain the first probability of the impact combination of each target wind turbine in the target type wind turbine; obtain the second probability based on the first probability of the impact combination of all target wind turbines in the target type wind turbine; based on the first probability and the second probability of the impact combination of each target wind turbine in the target type wind turbine, filter out the updated impact combination from the impact combination.

[0167] When analyzing the probability of a condition causing a failure, the failure may occur due to the combined influence of multiple conditions, or it may be caused by only a single condition. Therefore, when two conditions are neither completely mutually exclusive nor completely independent, the two conditions should be merged into one condition in the process of constructing the knowledge graph and used as a node in the graph.

[0168] Similarly, determine whether there is an impact association between each condition and the Ath fault entity. All conditions with impact associations are combined into an impact combination, which is recorded as the impact combination of the Lth target wind turbine in the target type wind turbine.

[0169] The probability mean of each condition in the impact combination is obtained, and recorded as the first probability of the impact combination of the Lth target wind turbine in the target type wind turbine.

[0170] Similarly, the first probability of the impact combination of each target wind turbine in the target type wind turbine is obtained.

[0171] The average of the first probabilities of the impact combinations of all target wind turbines in the target type wind turbines is recorded as the second probability.

[0172] When the first probability of the influence combination of the Lth target wind turbine in the target type wind turbine is greater than the second probability, it is judged that each condition in the influence combination of the Lth target wind turbine belongs to a common influence condition, and the influence combination of the Lth target wind turbine is recorded as an updated influence combination, and it is divided into the same node in the knowledge graph.

[0173] Thus, some of the target wind turbines in the target type have the target wind turbines with the update influence combination.

[0174] Step S005: According to the probability of each condition of each wind turbine and the first probability of updating the impact combination, the construction of the power knowledge graph of the multidimensional data is completed.

[0175] When performing fault analysis on wind turbines, it is extremely important to have a deep understanding of the probability of failure under specific conditions or combinations of conditions. This approach can provide a more comprehensive data analysis, leading to more accurate prediction and prevention of potential failures.

[0176] Obtain the probability of each condition and the first probability of the impact combination for all wind turbines.

[0177] It should be noted that the larger the probability mean, the greater the risk caused by the corresponding conditions and the higher the risk assessment score.

[0178] According to the conditions and combinations of conditions in the above analysis process, and according to the relationship between the conditions and the faults and the probability of their occurrence, corresponding attribute vectors are added to the nodes in the known knowledge graph.

[0179] The probability mean of each condition associated with the Ath fault entity for all target wind turbines is calculated and recorded as the updated probability of each condition associated with the Ath fault entity.

[0180] The mean of the correlation indexes of all conditions in each update impact combination is recorded as the correlation index of the update impact combination;

[0181] For the Ath fault entity, if it has a relationship with any condition and combination of conditions, the attribute vector of the node constructed in the knowledge graph is [z, x, c, v, b, n], where z is a combination of all conditions that have an impact association with the Ath fault entity, x is a combination of the association indexes of all conditions that have an impact association with the Ath fault entity; c is a combination of the update probabilities of all conditions that have an impact association with the Ath fault entity; v is a combination of all updated impact combinations; b is a combination of the association indexes of all updated impact combinations; n is a combination of the first probabilities of all updated impact combinations.

[0182] Similarly, the attribute vector corresponding to each fault type is obtained. Among all the fault types that have occurred in all wind turbines, each fault type is taken as a node, and the power knowledge graph is constructed according to the attribute vectors corresponding to all fault types. The construction of the knowledge graph is a well-known technology and will not be described in detail here.

[0183] At this point, the construction of the power knowledge graph of multidimensional data is completed.

[0184] It should be noted that the update of the knowledge graph first relies on continuous monitoring and collection of data sources and environmental changes to determine the update requirements. Continuously monitor the operating status and environmental changes of wind turbines, regularly check and update new data collected from external data sources such as meteorology and geography, clean the obtained data, and identify entity targets in the new data. According to the above method, extract the relationship between entities from the new data, and update or add the attribute information of the entity. Finally, perform a quality check on the data and knowledge to ensure that the updated knowledge graph is accurate.

[0185] So far, the present invention is completed.

[0186] The above description is only a specific implementation of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by a technician familiar with the technical field within the technical scope disclosed by the present invention should be covered by the protection scope of the present invention.

Claims

1. A method for constructing a wind turbine knowledge graph, characterized in that: include: Obtaining time series data of multiple wind turbines under multiple conditions and multiple fault types that have occurred; Each fault type is regarded as a fault entity, and under each fault entity, the wind turbines are divided into target-type wind turbines representing fault entities and normal-type wind turbines representing no fault entities; according to the correlation index of the target-type wind turbines and the normal-type wind turbines under each fault entity, it is judged whether there is a specific correlation between each fault entity and any condition, and the condition that is specifically correlated with each fault entity is obtained; For each entity fault, obtain the time series data of each wind turbine in the target type wind turbine under each condition within a preset time period after each fault, determine the impact association between each fault entity and any condition for each target wind turbine, and obtain the fault entity and condition with impact association corresponding to each target wind turbine, and obtain the ratio of the number of faults occurring in the preset number of faults before the current detection to the preset value of faults occurring for the fault entity and condition with impact association corresponding to each target wind turbine, and at the same time, form the impact combination of each target wind turbine in the target type wind turbine under each fault entity based on the condition with impact association with each fault entity; Based on the ratio of the number of faults occurring in the preset number of faults occurring before the current detection to the preset value of faults associated with the fault entities and conditions corresponding to each target wind turbine, a first probability and a second probability of an impact combination of each target wind turbine in the target type wind turbines under each fault entity are obtained; the impact combination is screened according to the first probability and the second probability to obtain an updated impact combination of each target wind turbine in the target type wind turbines of each fault entity; The construction of the wind turbine knowledge graph is completed based on the update probability of the update impact combination and the update probability of the condition associated with each fault entity.

2. The method for constructing a wind turbine knowledge graph according to claim 1, characterized in that: The specific steps of judging whether there is a specific association between each fault entity and any condition according to the inter-group association index of the target type wind turbine generator and the normal type wind turbine generator under each fault entity, and obtaining the condition that is specifically associated with each fault entity include: Select any fault entity and a condition, record them as fault entity A and the ath condition, among the target type wind turbines under fault entity A, group any two target wind turbines together, obtain the absolute value of the difference between the mean values ​​of the time series data of the ath condition in each group, record them as the first difference of the ath condition in each group; According to the method for obtaining the first difference value of the ath condition in each group of the target type wind turbine generators, the first difference value of the ath condition in each group of the normal type wind turbine generators is obtained; Based on the first difference of the ath condition in each group of target-type wind turbines and the first difference of the ath condition in each group of normal-type wind turbines, the correlation index of the ath condition is calculated using the following formula: In the formula, G a is the correlation index of the ath condition, CY a is the mean of the first difference of the ath condition of all groups in the target type wind turbine, CY' a is the mean of the first difference of the ath condition of all groups in the normal type wind turbine, norm[] is the linear normalization function, and | | is the absolute value; If the correlation index of the ath condition is greater than the preset threshold, it is determined that there is a specific correlation between the ath condition and the fault entity A; Select other conditions and repeat the above steps to determine whether there is a specific association between other conditions and fault entity A; Select other fault entities, repeat the above steps, and determine whether there is a specific association between each fault entity and any condition, thereby obtaining the condition under which each fault entity has a specific association.

3. The method for constructing a wind turbine knowledge graph according to claim 2, characterized in that: The specific steps of obtaining, for each entity fault, time series data of each wind turbine in the target type wind turbine under each condition within a preset time period after each fault, determining the influence association between each fault entity and any condition for each target wind turbine, obtaining the fault entity and condition with influence association corresponding to each target wind turbine, and obtaining the ratio of the number of faults occurring in the preset number of faults occurring before the current detection to the preset value of faults occurring in the fault entity and condition with influence association corresponding to each target wind turbine include: Choose any fault entity and a condition, record them as fault entity A and the ath condition; For fault entity A, the mean of the ath conditional time series data of each target wind turbine within a preset time period after each fault occurs is recorded as the first mean value of each target wind turbine after each fault occurs; The mean of the time series data of each target wind turbine generator within a preset time period after each failure occurs is recorded as the second mean; In the target type wind turbines, the absolute value of the difference between the first mean and the second mean of each target wind turbine after each fault occurs is calculated; when the absolute value of the difference is greater than one quarter of the second mean, it is determined that there is an influence association between the ath condition and the Ath fault entity of each target wind turbine when each fault occurs; For each target wind turbine that has experienced a preset number of failures before the current monitoring, the ratio of the number of failures corresponding to the impact association between the a-th condition and the A-th fault entity to the preset value; Select other conditions, repeat the above steps, determine the influence association between each target wind turbine and the fault entity A and any condition, and obtain the condition corresponding to each target wind turbine that has an influence association with the fault entity A, and obtain the ratio of the number of failures of each target wind turbine that has an influence association with the fault entity A to the preset number of failures before the current detection to the preset value; Select other fault entities, repeat the above steps, determine the impact association between each fault entity and any condition for each target wind turbine, and obtain the fault entity and condition associated with the existence of the impact for each target wind turbine, and obtain the ratio of the number of faults that occur in the preset number of times before the current detection to the preset value for the fault entity and condition associated with the existence of the impact for each target wind turbine.

4. The method for constructing a wind turbine knowledge graph according to claim 3, characterized in that: The specific steps of forming the influence combination of each target wind turbine in the target type wind turbine under each fault entity based on the condition that there is an influence association with each fault entity include: Choose any faulty entity and record it as faulty entity A; All conditions in each target wind turbine generator that are associated with the Ath fault entity form an impact combination; Select other fault entities and repeat the above steps to form an impact combination of each target wind turbine in the target type wind turbine under each fault entity.

5. The method for constructing a wind turbine knowledge graph according to claim 4, characterized in that: The specific steps of obtaining the first probability and the second probability of the influence combination of each target wind turbine in the target type wind turbine under each fault entity based on the ratio of the number of faults occurring in the preset number of faults occurring before the current detection to the preset value of faults associated with the fault entity and condition corresponding to each target wind turbine include: Choose any faulty entity and record it as faulty entity A; The average of the ratios of the number of failures occurring in the preset number of failures before the current detection to the preset value for each condition in the influence combination of each target wind turbine in the target type wind turbine under the fault entity A is recorded as the first probability of the influence combination of each target wind turbine in the target type wind turbine under the fault entity A; The average of the first probabilities of all target wind turbines in the target type wind turbines under fault entity A is recorded as the second probability; Select other fault entities, repeat the above steps, and obtain the first probability and the second probability of the impact combination of each target wind turbine in the target type wind turbine under each fault entity.

6. The method for constructing a wind turbine knowledge graph according to claim 5, characterized in that: The specific steps of screening the impact combination according to the first probability and the second probability to obtain the updated impact combination of each target wind turbine in each fault entity target type wind turbine include: Choose any faulty entity and record it as faulty entity A; When the first probability of the impact combination of each target wind turbine in the target type wind turbine under the fault entity A is greater than the second probability, the impact combination of each target wind turbine is recorded as an updated impact combination; Select other fault entities, repeat the above steps, and obtain the updated impact combination of each target wind turbine in each fault entity target class wind turbine.

7. The method for constructing a wind turbine knowledge graph according to claim 6, characterized in that: The specific steps of completing the construction of the wind turbine knowledge graph based on the update probability of the update impact combination and the update probability of the condition associated with each fault entity include: Choose any faulty entity and record it as faulty entity A; Calculate the average of the ratios of the number of failures of all target wind turbines to the preset number of failures before the current detection for each condition that is associated with the fault entity A, and record it as the updated probability of each condition that is associated with the fault entity A; The mean of the correlation indexes of all conditions in each update impact combination is recorded as the correlation index of the update impact combination; According to the updated probability of each condition associated with the fault entity A and the associated index of the updated impact combination, the construction of the power knowledge graph of multidimensional data is completed. The specific method is as follows: The attribute vector of the node constructed by fault entity A in the knowledge graph is [z, x, c, v, b, n], where z is a combination of all conditions that have an impact on fault entity A, x is a combination of the association indexes of all conditions that have an impact on fault entity A; c is a combination of the update probabilities of all conditions that have an impact on fault entity A; v is a combination of all update impact combinations; b is a combination of the association indexes of all update impact combinations; n is a combination of the first probabilities of all update impact combinations; Among all the fault types that have occurred in all wind turbines, each fault type is taken as a node, and an electric power knowledge graph is constructed based on the attribute vectors corresponding to all fault types.

8. A wind turbine knowledge graph construction system, characterized in that: It includes a multidimensional data acquisition module, a first data processing module, a second data processing module, a third data processing module and a knowledge graph construction module; The multi-dimensional data acquisition module is used to acquire time series data of multiple wind turbines under multiple conditions and multiple fault types that have occurred; The first data processing module is used to treat each fault type as a fault entity, and divide the wind turbines into target-type wind turbines representing fault entities and normal-type wind turbines representing no fault entities under each fault entity; according to the correlation index of the target-type wind turbines and the normal-type wind turbines under each fault entity, determine whether there is a specific correlation between each fault entity and any condition, and obtain the condition that is specifically correlated with each fault entity; The second data processing module is used to obtain, for each entity fault, the time series data of each wind turbine in the target type wind turbine under each condition within a preset time period after each fault, determine the impact association between each fault entity and any condition for each target wind turbine, and obtain the fault entity and condition with impact association corresponding to each target wind turbine, and obtain the ratio of the number of faults occurring in the preset number of faults before the current detection to the preset value of faults occurring in the fault entity and condition with impact association corresponding to each target wind turbine, and at the same time form an impact combination of each target wind turbine in the target type wind turbine under each fault entity based on the condition with impact association with each fault entity; The third data processing module is used to obtain a first probability and a second probability of an influence combination of each target wind turbine in the target type wind turbines under each fault entity based on the ratio of the number of faults occurring in the preset number of faults occurring before the current detection to the preset value of faults associated with the fault entities and conditions corresponding to each target wind turbine; and screen the influence combination according to the first probability and the second probability to obtain an updated influence combination of each target wind turbine in the target type wind turbine for each fault entity; The knowledge graph construction module is used to complete the construction of the wind turbine knowledge graph based on the update probability of the update impact combination and the update probability of the condition associated with each fault entity.

9. An electronic device, characterized in that: The electronic device comprises a memory and a processor, wherein the memory stores a computer program, and the processor is configured to call and run the computer program stored in the memory to execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method as claimed in any one of claims 1 to 7 are implemented.

Images