A Method and System for Monitoring the Condition of Wind Turbine Blades Based on Acoustic Mode

Abstract

This invention provides a method and system for monitoring the condition of wind turbine blades based on acoustic signature modalities, relating to the field of data processing technology. During wind turbine operation, when damage such as cracks or delamination occurs inside or outside the blades, the acoustic signals will change accordingly. This invention collects acoustic signature signals through an acoustic signature sensor, and the edge computing terminal performs CEEMDAN decomposition and reconstruction on the collected acoustic signature signals. It combines ICA and beamforming to suppress directional noise, and uses an unsupervised pre-training combined with supervised fine-tuning strategy to construct a feature extraction model. Based on this model, deep feature extraction is performed on the noise-reduced acoustic signature signals, and then transmitted to the background software system for comparison and analysis, realizing intelligent identification of early damage inside and outside the blades. Specifically, this invention performs secondary matching on the wind turbines to be detected in the target wind turbine area to obtain a target comparison group, and then compares and analyzes the operating sounds of each similar wind turbine in the target comparison group to identify abnormal equipment.

Description

Technical Field

[0001] This invention relates to data processing technology, and more particularly to a method and system for monitoring the condition of wind turbine blades based on acoustic signature modalities. Background Technology

[0002] Wind turbines operate in outdoor environments for extended periods. During operation, the turbine blades are subjected to aerodynamic loads, gravity, inertial forces, and other excitations, generating broadband acoustic signals. When damage such as cracks or delamination occurs inside or outside the blades, their structural dynamic characteristics change, leading to corresponding changes in the frequency components, energy distribution, and modulation characteristics of the acoustic signals.

[0003] Acoustic fingerprint monitoring technology collects and analyzes acoustic fingerprint signals generated during the operation of wind turbines to determine whether there is damage to the turbine blades. It has become a primary technology for monitoring the condition of wind turbine blades. Currently, most acoustic fingerprint monitoring technologies for wind turbine blades employ threshold monitoring or multi-turbine comparison methods. For threshold monitoring, when the acoustic fingerprint signal of a single wind turbine exceeds a threshold, an abnormality is determined in the turbine blades. This method can only identify wind turbines whose acoustic fingerprint signals have already changed significantly; it is difficult to detect wind turbines whose acoustic fingerprint signals are still within the threshold range but are already damaged or abnormal. For multi-turbine comparison, abnormal equipment is identified by the differences in acoustic fingerprint signals between multiple wind turbines. However, the terrain and usage time of each wind turbine vary, leading to significant biases in direct acoustic fingerprint signal comparison, making it difficult to reflect the actual differences in blade condition.

[0004] Therefore, how to eliminate the interference of environmental conditions and usage time differences on acoustic fingerprint monitoring and realize the early identification of wind turbine blade anomalies has become a key problem that urgently needs to be solved. Summary of the Invention

[0005] This invention provides a method and system for monitoring the condition of wind turbine blades based on acoustic signature modalities. It can eliminate the interference of environmental condition differences and usage duration differences on acoustic signature monitoring, and realize the early identification of wind turbine blade anomalies.

[0006] A first aspect of the present invention provides a method for monitoring the condition of wind turbine blades based on acoustic signature modes, comprising: A secondary matching process is performed on the wind turbines to be tested in the target wind turbine area to obtain the target comparison group; The operating sounds of various similar fans within the target comparison group are compared and analyzed to identify abnormal equipment.

[0007] Optionally, in one possible implementation of the first aspect, the secondary matching of the wind turbines to be tested in the target wind turbine area to obtain a target comparison group includes: Environmental condition matching is performed on the wind turbines to be tested in the target wind turbine area to obtain regional comparison groups; The wind turbines in the region comparison group are matched for wind turbine status to obtain the target comparison group.

[0008] Optionally, in one possible implementation of the first aspect, the environmental state matching of the wind turbines to be tested in the target wind turbine area to obtain a region comparison group includes: Obtain the detection points of the fans to be tested in the target fan area; Using the detection points as centers, an environmental state zone corresponding to each fan to be tested is constructed based on a preset influence distance; The fixed and variable elements within the environmental state region are compared to obtain the environmental similarity, and the region comparison group is determined based on the environmental similarity.

[0009] Optionally, in one possible implementation of the first aspect, comparing fixed elements and variable elements within the environmental state region to obtain environmental similarity, and determining a region comparison group based on the environmental similarity, includes: Identify the contour lines of fixed elements and the variation regions of variable elements within the environmental state region; The contour lines within the environmental state area are compared in orientation to obtain static similarity. By comparing the changing regions within the environmental state region, dynamic similarity is obtained. The environmental similarity is obtained based on the static similarity and the dynamic similarity. The wind turbines whose environmental similarity is greater than the preset similarity are identified as regional wind turbines. The corresponding regional wind turbines are counted to obtain regional comparison groups.

[0010] Optionally, in one possible implementation of the first aspect, the step of performing orientation comparison on contour lines within the environmental state area to obtain static similarity includes: Retrieve multiple preset comparison positions and construct azimuth rays corresponding to the preset comparison positions, starting from the detection point; Based on the detection point, each of the directional rays is rotated to both sides according to a preset rotation angle to obtain the region division line; The environmental state area is segmented based on the region segmentation line to obtain the comparison area corresponding to each preset comparison position. The coverage range of contour lines within the same comparison area is compared to obtain the static similarity of the two wind turbines to be tested.

[0011] Optionally, in one possible implementation of the first aspect, comparing the coverage areas of contour lines within the same comparison region to obtain the static similarity between the two wind turbines to be tested includes: Retrieve the preset extraction dimension corresponding to the preset comparison position, perform coordinate processing on the environmental state area, and obtain the coordinate extreme points corresponding to the contour lines in the comparison area as occlusion points based on the preset extraction dimension. Connect the occlusion point with the detection point to obtain the occlusion area, and obtain the angle range of the detection point at the occlusion area as the corresponding preset occlusion range of the position. Calculate the directional similarity of the coverage area of ​​the two fans to be tested at the corresponding preset positions; The mean of the azimuth similarity of each preset comparison position is calculated to obtain the static similarity of the two fans to be tested.

[0012] Optionally, in one possible implementation of the first aspect, the step of performing region comparison on the changing regions within the environmental state region to obtain dynamic similarity includes: Obtain the changing regions within each environmental state zone, and obtain the intersection of the corresponding changing regions of two fans to be tested to obtain the intersection region; Obtain the union of the corresponding change regions of the two fans to be tested to obtain the union region; The dynamic similarity is obtained by the ratio of the area of ​​the intersection region to the area of ​​the union region.

[0013] Optionally, in one possible implementation of the first aspect, it also includes: Retrieve the historical area of ​​the changed region and obtain the current area of ​​the changed region in real time; The changed area is obtained by calculating the absolute value of the difference between the current area and the historical area. The degree of change is obtained based on the ratio of the changed area to the historical area. When the degree of change is determined to be greater than the preset degree value, the dynamic similarity is recalculated.

[0014] Optionally, in one possible implementation of the first aspect, the step of matching the wind turbine status of the regional wind turbines in the regional comparison group to obtain the target comparison group includes: Obtain the operating time of the fans in each region of the regional comparison group; The operating time of the regional fans is calculated to obtain the difference in operating time between the two regional fans. Fans in areas where the difference duration is less than a preset difference duration are classified as similar fans, and the similar fans are statistically analyzed to obtain a target comparison group.

[0015] A second aspect of the present invention provides a wind turbine blade condition monitoring system based on acoustic signature modalities, comprising: The matching module is used to perform secondary matching on the wind turbines to be tested in the target wind turbine area to obtain the target comparison group; The analysis module is used to compare and analyze the operating sounds of various similar fans within the target comparison group to identify abnormal equipment.

[0016] The beneficial effects of this invention are as follows: 1. This invention performs secondary matching on the wind turbines to be tested in the target wind turbine area to obtain a target comparison group. The environmental conditions and usage time of each wind turbine of the same type in the target comparison group are similar. The operating sound of each wind turbine of the same type in the target comparison group is compared and analyzed to identify abnormal equipment. This avoids the problem of large deviation in soundprint comparison caused by the difference in environmental conditions and usage time of each wind turbine when comparing multiple wind turbines, and can identify abnormal blades at an early time.

[0017] 2. This invention performs environmental state matching on the fans to be tested in the target fan area to obtain a regional comparison group, and then performs fan state matching on the regional fans in the regional comparison group to obtain a target comparison group. During environmental state matching, the detection points of the fans to be tested are obtained, and an environmental state area with the detection points as the center is constructed. Fixed elements and variable elements within the environmental state area are compared to obtain environmental similarity, and regional fans are screened based on environmental similarity. During fan state matching, the fan usage time of regional fans is calculated to obtain the difference duration, and regional fans with a difference duration less than the preset difference duration are screened as similar fans. This invention eliminates the interference of environmental state differences and equipment usage time differences before performing voiceprint signal comparison.

[0018] 3. This invention processes fixed and variable elements within the environmental state area separately. For fixed elements, contour lines are first extracted, and then azimuth rays corresponding to multiple preset comparison positions are constructed. These azimuth rays are rotated by a preset angle to obtain region segmentation lines. Based on these region segmentation lines, the environmental state area is segmented to obtain comparison regions corresponding to each preset comparison position. Coordinate extreme points are extracted from the contour lines within the comparison regions as masking points. These masking points are connected to the detection points to obtain the masking range. Static similarity is calculated based on the masking range. For variable elements, the intersection and union of the variable regions corresponding to two wind turbines to be detected are obtained. Dynamic similarity is obtained based on the ratio of the intersection area to the union area. This invention obtains environmental similarity based on static and dynamic similarity, and filters wind turbines within the region based on environmental similarity, thus achieving the filtering of wind turbines to be detected in similar environmental states. Attached Figure Description

[0019] Figure 1A flowchart of a wind turbine blade condition monitoring method based on acoustic signature modalities provided by the present invention; Figure 2 This is a schematic diagram of the region division lines in this invention; Figure 3 This is a schematic diagram of the comparison area in this invention; Figure 4 This is a schematic diagram of the covering point in the present invention; Figure 5 A schematic diagram of the structure of a wind turbine blade condition monitoring system based on acoustic signature modalities provided by the present invention; Figure 6 This is a schematic diagram of the hardware structure of an electronic device provided by the present invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein.

[0022] It should be understood that in the various embodiments of the present invention, the sequence number of each process does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0023] It should be understood that in this invention, "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion, for example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or units that are not explicitly listed or that are inherent to such process, method, product, or device.

[0024] It should be understood that in this invention, "multiple" refers to two or more. "And / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "Contains A, B, and C", "Contains A, B, and C" means that all three A, B, and C are contained; "Contains A, B, or C" means that one of A, B, and C is contained; "Contains A, B, and / or C" means that any one, two, or three of A, B, and C are contained.

[0025] It should be understood that in this invention, "B corresponding to A", "B corresponding to A", "A and B correspond", or "B and A correspond" means that B is associated with A, and B can be determined based on A. Determining B based on A does not mean determining B solely based on A; B can also be determined based on A and / or other information. Matching A and B is defined as a similarity between A and B that is greater than or equal to a preset threshold.

[0026] Depending on the context, "if" as used here can be interpreted as "when," "when," "in response to determination," or "in response to detection."

[0027] The technical solution of the present invention will be described in detail below with reference to specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0028] This invention provides a method for monitoring the condition of wind turbine blades based on acoustic signature modes, such as... Figure 1 As shown, it includes: S1, perform secondary matching on the wind turbines to be tested in the target wind turbine area to obtain the target comparison group.

[0029] It should be noted that in existing technologies, the method of collecting acoustic signature signals from individual wind turbines and comparing them with thresholds can only identify anomalies after the acoustic signature signal has exceeded a preset threshold range. This makes it difficult to detect wind turbines with damaged or abnormal acoustic signatures that are still within the threshold range. While differences in acoustic signature signals between wind turbines can help identify abnormal equipment early, the terrain of wind turbines in different locations varies, and the usage time of each turbine is also different. Therefore, to eliminate the interference of environmental and usage time differences on acoustic signature monitoring and achieve early identification of wind turbine blade anomalies, this invention performs a secondary matching of the wind turbines to be tested. First, regional wind turbines in similar environmental conditions are selected. Then, similar wind turbines with similar usage times are selected from the regional wind turbines. The acoustic signature signals of these similar wind turbines are compared to ensure that the compared wind turbines have similar environmental conditions and similar usage times.

[0030] Among them, the target wind turbine area refers to the wind farm area where voiceprint comparison is performed to identify abnormal wind turbines; the wind turbine to be tested refers to the wind turbine generator in the target wind turbine area that requires blade status monitoring; secondary matching refers to the process of performing environmental status matching and wind turbine status matching on the wind turbine to be tested in the target wind turbine area in sequence; regional wind turbines refer to the wind turbines to be tested that are in similar environmental conditions and are selected through the environmental status matching process; similar wind turbines refer to regional wind turbines that have been used for a similar period of time and are selected through the wind turbine status matching process; and the target comparison group refers to the collection of similar wind turbines.

[0031] In some embodiments, step S1 (performing secondary matching of the wind turbines to be detected in the target wind turbine area to obtain the target comparison group) includes S11-S12: S11, perform environmental condition matching on the wind turbines to be tested in the target wind turbine area to obtain the area comparison group.

[0032] Understandably, this step will screen out the wind turbines to be tested that are in similar environmental conditions and form regional comparison groups.

[0033] Among them, environmental condition matching refers to the process of screening the wind turbines to be tested based on the similarity of environmental conditions; regional comparison group refers to the collection of wind turbines in a region.

[0034] In some embodiments, step S11 (performing environmental state matching of the wind turbines to be tested in the target wind turbine area to obtain a region comparison group) includes S111-S113: S111, Obtain the detection point of the fan to be tested in the target fan area.

[0035] It is understandable that the installation locations of each fan to be tested in the target fan area are used as test points.

[0036] S112, with the detection point as the center, construct an environmental state zone corresponding to each fan to be tested based on a preset influence distance.

[0037] Among them, the preset influence distance refers to the maximum distance value that is pre-set to affect the acoustic fingerprint collection of the fan to be tested.

[0038] It is understandable that a circular area is constructed with the detection point as the center and the preset influence distance as the radius, and this circular area is used as the environmental state zone of the corresponding fan to be tested.

[0039] S113, compare the fixed elements and variable elements in the environmental state area to obtain the environmental similarity, and determine the region comparison group based on the environmental similarity.

[0040] It should be noted that this step categorizes geographical targets within the environmental state area into fixed elements and variable elements. Fixed elements refer to geographical targets like mountains, whose states are relatively stable and do not change in the short term. Variable elements refer to geographical targets like trees and forests, whose states are easily altered by factors such as fires and logging. Therefore, environmental state matching involves comparing and matching the fixed and variable elements within the environmental state area separately, and identifying the wind turbines to be tested whose similarity to both fixed and variable elements meets the corresponding threshold conditions as regional wind turbines.

[0041] In some embodiments, step S113 (comparing fixed elements and variable elements within the environmental state region to obtain environmental similarity, and determining a region comparison group based on the environmental similarity) includes S1131-S1135: S1131, Identify the contour lines of fixed elements and the variation regions of variable elements within the environmental state area.

[0042] Understandably, in this step, the geographic information map corresponding to the target wind turbine area is retrieved, and the environmental state area of ​​each wind turbine to be tested is obtained, identifying fixed and variable elements within the environmental state area. For fixed elements, a reference height is pre-set based on the influence range of the acoustic signature signal acquisition, and the contour lines corresponding to the fixed elements at the reference height are extracted as the contour lines of the corresponding fixed elements. For variable elements, the area covered by the variable elements is taken as the variable area.

[0043] Among them, a geographic information map refers to an electronic map that contains topographic data of the target wind turbine area.

[0044] S1132, compare the orientation of contour lines within the environmental state area to obtain static similarity.

[0045] Static similarity refers to the degree of similarity of fixed elements in the environmental state regions of any two wind turbines to be tested.

[0046] It should be noted that the distribution of fixed elements in different environmental state areas may differ. This step divides the environmental state areas by orientation, then compares the contour lines of fixed elements in the same orientation to obtain the orientation similarity, and finally determines the static similarity of any two wind turbines to be tested based on the orientation similarity corresponding to each orientation.

[0047] In some embodiments, step S1132 (performing orientational comparison of contour lines within the environmental state area to obtain static similarity) includes SA1-SA4: SA1 retrieves multiple preset comparison positions and constructs an azimuth ray corresponding to the preset comparison positions, starting from the detection point.

[0048] Among them, the preset comparison orientation refers to the geographical orientation that is set in advance to divide the environmental state zone.

[0049] Understandably, multiple preset reference directions are set in advance. These preset reference directions can be the four directions of east, south, west, and north. Starting from the detection point, rays are constructed along each preset reference direction to obtain the directional rays corresponding to each preset reference direction.

[0050] SA2, using the detection point as a reference, rotates each of the directional rays to both sides based on a preset rotation angle to obtain the region division line.

[0051] The preset rotation angle refers to the angle that is set in advance to rotate the azimuth ray to both sides.

[0052] Understandably, a preset rotation angle, such as 45°, is set in advance. Using the detection point as a reference, each azimuth ray is rotated by the preset rotation angle in both clockwise and counterclockwise directions, and the resulting ray serves as the region dividing line.

[0053] For example, such as Figure 2 As shown, the preset orientation is east, south, west, and north. When the preset rotation angle is 45°, each orientation ray rotates 45° to both sides. The angle between the two area dividing lines obtained after the rotation is 90°. At this time, the area dividing lines generated by the relative rotation of two adjacent orientation rays coincide, and finally four area dividing lines are formed in the environmental state area.

[0054] SA3, based on the region segmentation line, divides the environmental state area to obtain the comparison area corresponding to each preset comparison position.

[0055] It is understandable that, such as Figure 3 As shown, the environmental state area is segmented based on the region dividing lines to obtain fan-shaped areas corresponding to each preset comparison direction. These fan-shaped areas are the comparison areas. For example, referring to the example in step SA2, the environmental state area is segmented based on four region dividing lines to obtain four fan-shaped areas, corresponding to the four directions of east, south, west, and north, respectively. These fan-shaped areas are the comparison areas for the corresponding preset comparison directions.

[0056] SA4 compares the coverage of contour lines within the same comparison area to obtain the static similarity of the two wind turbines to be tested.

[0057] It should be noted that the coverage area of ​​contour lines within the comparison area determines the influence range of fixed elements in the corresponding preset comparison location on the voiceprint signal. The closer the coverage areas of two wind turbines in the same preset comparison location, the closer the influence of the terrain in the corresponding preset comparison location on the voiceprint. This step compares the comparison areas in the same preset comparison location to determine the static similarity of the two wind turbines to be tested.

[0058] In some embodiments, step SA4 (comparing the coverage areas of contour lines within the same comparison area to obtain the static similarity of the two wind turbines to be tested) includes SA41-SA44: SA41 retrieves the preset extraction dimension corresponding to the preset comparison position, performs coordinate processing on the environmental state area, and obtains the coordinate extreme points corresponding to the contour lines in the comparison area as the masking points based on the preset extraction dimension.

[0059] Among them, the preset extraction dimension refers to the direction that is pre-set and used to extract the extreme values ​​of coordinates on the contour lines in the comparison area.

[0060] It should be noted that, in order to determine the coverage of contour lines in each comparison area, this step pre-sets the preset extraction dimension corresponding to each preset comparison location, and extracts the boundary points of contour lines from the comparison area according to the preset extraction dimension.

[0061] Understandably, a Cartesian coordinate system is established for the environmental state area, using the detection point as the origin. For each comparison area, the extreme coordinate points on both sides of the contour line are extracted based on the corresponding preset extraction dimension, and these extracted extreme coordinate points are used as the occlusion points of the corresponding comparison area. For example... Figure 4 As shown, referring to the example in step SA2, the vertical axis of the coordinate system is constructed based on the two preset directions of north and south, and the horizontal axis of the coordinate system is constructed based on the two preset directions of east and west. The preset extraction dimensions corresponding to the two preset directions of north and south are set as the direction of the horizontal axis, and the preset extraction dimensions corresponding to the two preset directions of east and west are set as the direction of the vertical axis. The occlusion points corresponding to the two preset directions of north and south are the coordinate extreme points on the horizontal axis, and the occlusion points corresponding to the two preset directions of east and west are the coordinate extreme points on the vertical axis.

[0062] SA42 connects the occlusion point and the detection point to obtain the occlusion area, and obtains the angle range of the detection point at the occlusion area as the corresponding preset occlusion range.

[0063] Understandably, by connecting the two occlusion points and the detection point within each comparison area, two connecting line segments are formed. The area enclosed by these two connecting line segments and the contour line is the occlusion area within the corresponding comparison area. Furthermore, using the detection point as the vertex and selecting true north as the zero-degree reference, the angles formed by the two connecting line segments and true north within the comparison area are measured respectively. The angle interval formed by the two angles is used as the occlusion range of the corresponding preset comparison direction.

[0064] For example, taking due north as the zero-degree reference and clockwise as the direction of angle increase, for the comparison area under the preset comparison position of east, if the angle between one of the connecting line segments and due north is 75° and the angle between the other connecting line segment and due north is 120°, then the coverage range of the preset comparison position of east is the angle interval between 75° and 120°.

[0065] SA43 calculates the directional similarity of the coverage area of ​​the two fans to be tested at the corresponding preset positions.

[0066] Understandably, for any two wind turbines to be tested, the coverage area of ​​the two wind turbines in the same preset comparison position is extracted, the intersection angle and the union angle of the two coverage areas are obtained, the ratio of the intersection angle to the union angle is calculated, and the ratio is used as the orientation similarity of the two wind turbines to be tested in the corresponding preset comparison position.

[0067] For example, the coverage range of the first fan to be tested is 25° to 70°, and the coverage range of the second fan to be tested is 30° to 70°. The intersection range of these two coverage ranges is 30° to 70°, and the intersection angle is 40°. The union range is 25° to 70°, and the union angle is 45°. The directional similarity is the ratio of 40 to 45.

[0068] SA44 calculates the mean of the azimuth similarity of each preset comparison position to obtain the static similarity of the two fans to be tested.

[0069] Understandably, for any two fans to be tested, the directional similarity between the two fans is calculated under all preset comparison positions. The directional similarity of the two fans to be tested is summed, and the sum is divided by the number of preset comparison positions. The mean value is the static similarity between the two fans to be tested.

[0070] S1133, compare the changing regions within the environmental state region to obtain dynamic similarity.

[0071] It should be noted that this step obtains the intersection and union regions between the corresponding changing regions of any two wind turbines to be tested, and calculates the ratio of the area of ​​the intersection region to the area of ​​the union region to obtain the dynamic similarity.

[0072] Dynamic similarity refers to the degree of similarity of changing elements in the environmental state zones of any two wind turbines to be tested.

[0073] In some embodiments, step S1133 (comparing the changing regions within the environmental state region to obtain dynamic similarity) includes SB1-SB3: SB1: Obtain the changing regions within each environmental state zone, and obtain the intersection of the changing regions of the two fans to be tested, thus obtaining the intersection region.

[0074] It is understandable that, for any two fans to be tested, the intersection of the corresponding change regions of the two fans to be tested is obtained, and this intersection is taken as the intersection region.

[0075] SB2, obtain the union of the corresponding change regions of the two wind turbines to be tested, and obtain the union region.

[0076] It is understandable that, based on step SB1, the union of the corresponding change regions of the two wind turbines to be tested is obtained, and this union is used as the union region.

[0077] SB3: The dynamic similarity is obtained based on the ratio of the area of ​​the intersection region to the area of ​​the union region.

[0078] It is understandable that the area of ​​the intersection region and the area of ​​the union region are obtained, and the ratio of the area of ​​the intersection region to the area of ​​the union region is calculated. This ratio is the dynamic similarity between the two wind turbines to be tested.

[0079] In some embodiments, SC1-SC4 are also included: SC1 retrieves the historical area of ​​the changed region and obtains the current area of ​​the changed region in real time.

[0080] It is understandable that the area of ​​the changed region at a historical point in time is retrieved as the historical area, and the area of ​​the changed region at the current point in time is obtained as the current area.

[0081] SC2, the changed area is obtained based on the absolute value of the difference between the current area and the historical area.

[0082] Understandably, the change in area is obtained by calculating the absolute value of the difference between the current area and the historical area.

[0083] SC3, based on the ratio of the changed area to the historical area, obtains the degree of change value.

[0084] Understandably, the ratio of the changed area to the historical area is used as the degree of change value. The larger the degree of change value, the greater the magnitude of change in the changed area.

[0085] SC4, when the degree of change is determined to be greater than the preset degree value, the dynamic similarity is recalculated.

[0086] It should be noted that when the change range of the variable region is small, the dynamic similarity calculated based on the historical variable region can still accurately represent the similarity between the two wind turbine variable elements to be detected, and there is no need to repeat the area calculation of the intersection and union regions. When the change range of the variable region is large, the original dynamic similarity cannot represent the current state of the variable region, and it is necessary to recalculate the intersection and union areas based on the current variable region.

[0087] The preset degree value refers to the threshold value of change that is set in advance to determine whether the dynamic similarity needs to be recalculated.

[0088] Understandably, when the degree of change is greater than the preset degree value, it indicates that the area of ​​change has undergone significant changes, and steps SB1-SB3 are re-executed to calculate the dynamic similarity between the two wind turbines to be tested based on the current area of ​​change; when the degree of change is less than or equal to the preset degree value, it indicates that the change in the area of ​​change has not exceeded the preset degree value, and the dynamic similarity calculated in the past is called.

[0089] S1134, Based on the static similarity and the dynamic similarity, the environmental similarity is obtained.

[0090] It should be noted that this step determines the environmental state similarity of the two wind turbines to be tested only when both static similarity and dynamic similarity meet the corresponding conditions.

[0091] It is understandable that environmental similarity includes static similarity and dynamic similarity.

[0092] S1135, the wind turbines to be detected with an environmental similarity greater than the preset similarity are taken as regional wind turbines, and the corresponding regional wind turbines are counted to obtain regional comparison groups.

[0093] Among them, the preset similarity refers to the environmental similarity threshold set in advance to screen the wind turbines under test that are in similar environmental states. It can be that there is a preset similarity for static similarity and dynamic similarity respectively.

[0094] It should be noted that environmental similarity includes static similarity and dynamic similarity. This step performs threshold judgment on static similarity and dynamic similarity respectively. When both static similarity and dynamic similarity are greater than the preset similarity, the corresponding wind turbine to be detected is regarded as the regional wind turbine.

[0095] Understandably, a regional comparison group is constructed for each fan to be tested. For example, if the fan to be tested is A, the environmental similarity of fan A is compared with that of other fans to be tested. All other fans to be tested whose static and dynamic similarities are greater than a preset similarity are included together with fan A as regional fans. All regional fans are then grouped together to form the regional comparison group corresponding to fan A. The above process is performed on all other fans to be tested in the target fan area to obtain a one-to-one regional comparison group for each fan to be tested. For example, for fan A, the static and dynamic similarities of fan A with fans B, C, D, and E are compared respectively. If the environmental similarity between A and B, and between A and C is greater than a preset similarity, and the environmental similarity between A and D, and between A and E is less than a preset similarity, then the regional comparison group corresponding to A is obtained, which includes A, B, and C.

[0096] S12, perform fan status matching on the regional fans in the regional comparison group to obtain the target comparison group.

[0097] It should be noted that if two wind turbines have similar environments but significantly different usage durations—for example, one has been used for 5 years while the other has only been used for 1 year—the acoustic signatures resulting from long-term wear and tear on their mechanical structures will also differ considerably. This step, after matching environmental conditions, further filters regional wind turbines by usage duration, selecting similar turbines with similar usage durations to form a target comparison group.

[0098] Among them, wind turbine status matching refers to the process of screening regional wind turbines based on the similarity of their own usage time.

[0099] In some embodiments, step S12 (matching the wind turbine status of the regional wind turbines in the regional comparison group to obtain the target comparison group) includes S121-S123: S121, obtain the operating time of the wind turbines in each region of the region comparison group.

[0100] It is understandable that the usage time of the wind turbines in each region of the regional comparison group is retrieved as the wind turbine usage time.

[0101] S122, calculate the operating time of the regional fans to obtain the difference in operating time between the two regional fans.

[0102] Understandably, for each area comparison group corresponding to the fan to be tested, the difference in fan usage time between the corresponding fan to be tested and the fans in other areas of the corresponding area comparison group is calculated, and the absolute value of the difference is taken to obtain the difference in usage time between the fan to be tested and the fans in other areas.

[0103] S123, the regional wind turbines whose difference duration is less than the preset difference duration are regarded as the same type of wind turbines, and the same type of wind turbines are counted to obtain the target comparison group.

[0104] Among them, the preset difference duration refers to the pre-set threshold for the difference duration when filtering similar wind turbines.

[0105] Understandably, for each area comparison group corresponding to the fan to be tested, the difference duration between the corresponding fan to be tested and the fans in other areas is obtained, and each difference duration is compared with the preset difference duration. Fans in other areas with a difference duration less than the preset difference duration and the corresponding fan to be tested are regarded as the same type of fan. All fans of the same type are gathered to obtain the target comparison group corresponding to the fan to be tested.

[0106] S2, compare and analyze the operating sounds of each similar fan in the target comparison group to identify abnormal equipment.

[0107] It is understandable that the environmental conditions and usage durations of the various types of fans in the target comparison group are similar, so the acoustic signatures of the fans should be quite similar. By collecting the operating sounds generated by each type of fan during operation, when the operating sound of a particular fan in the target comparison group deviates significantly from the operating sounds of other fans, it can be determined that the blade condition of that fan is different from that of the others, and that fan can be identified as an abnormal device.

[0108] Among them, operating sound refers to the acoustic signal generated by the fan during operation.

[0109] In one optional implementation, wireless acoustic signature sensors can be installed inside the blades of each wind turbine under test, while wired acoustic signature sensors are installed at the tower nacelle doors. The acoustic signature sensors deployed on a single wind turbine under test collectively constitute the acoustic signature acquisition device for that turbine. During the operation of the wind turbine under test, the acoustic signature acquisition device collects acoustic signature signals from different locations inside and outside the blades. The acoustic signature signals collected from multiple locations collectively constitute the operating sound of the wind turbine under test. An edge computing terminal is deployed inside the nacelle of each wind turbine under test. The edge computing terminal connects to the acoustic signature acquisition device of the corresponding wind turbine. The edge computing terminal performs CEEMDAN decomposition and reconstruction on the collected operating sound, combines ICA and beamforming to suppress directional noise, and then performs feature extraction processing on the noise-reduced operating sound. The feature-extracted operating sound data is then transmitted to the backend software system. It is worth noting that a feature extraction model based on deep neural networks, such as Transformer networks or Graph Neural Networks (GNNs), can be built into the edge computing terminal. The feature extraction model is constructed using a strategy of unsupervised pre-training combined with supervised fine-tuning. This model performs deep feature extraction processing on the operating sounds collected by the acoustic signature acquisition device. The backend software system compares and analyzes the operating sound data uploaded by the edge computing terminal to identify abnormal equipment. The backend software system can be deployed using a B / S architecture, providing a human-machine interface to centrally display the monitoring results of internal and external blade defects, and supports integration with existing wind farm monitoring systems via API interfaces.

[0110] See Figure 5 This is a schematic diagram of a wind turbine blade condition monitoring system based on acoustic signature modalities provided in an embodiment of the present invention. The system includes: The matching module is used to perform secondary matching on the wind turbines to be tested in the target wind turbine area to obtain the target comparison group; The analysis module is used to compare and analyze the operating sounds of various similar fans within the target comparison group to identify abnormal equipment.

[0111] See Figure 6 This is a schematic diagram of the hardware structure of an electronic device provided in an embodiment of the present invention. The electronic device 60 includes: a processor 61, a memory 62, and a computer program; wherein... The memory 62 is used to store the computer program, and the memory may also be flash memory. The computer program is, for example, an application program or functional module that implements the above method.

[0112] The processor 61 is configured to execute the computer program stored in the memory to implement the various steps performed by the device in the above method. For details, please refer to the relevant descriptions in the preceding method embodiments.

[0113] Alternatively, the memory 62 can be either standalone or integrated with the processor 61.

[0114] When the memory 62 is a device independent of the processor 61, the device may further include: Bus 63 is used to connect the memory 62 and the processor 61.

[0115] The present invention also provides a readable storage medium storing a computer program, which, when executed by a processor, is used to implement the methods provided in the various embodiments described above.

[0116] The readable storage medium can be a computer storage medium or a communication medium. A communication medium includes any medium that facilitates the transfer of computer programs from one location to another. A computer storage medium can be any available medium accessible to a general-purpose or special-purpose computer. For example, a readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an Application-Specific Integrated Circuit (ASIC). Alternatively, the ASIC can be located in a user equipment. Of course, the processor and the readable storage medium can also exist as discrete components in a communication device. The readable storage medium can be a read-only memory (ROM), random access memory (RAM), CD-ROM, magnetic tape, floppy disk, and optical data storage device, etc.

[0117] The present invention also provides a program product including executable instructions stored in a readable storage medium. At least one processor of the device can read the executable instructions from the readable storage medium, and the at least one processor executes the executable instructions to cause the device to implement the methods provided in the various embodiments described above.

[0118] In the embodiments of the above-described device, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly manifested as execution by a hardware processor, or execution by a combination of hardware and software modules within the processor.

[0119] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for monitoring the condition of wind turbine blades based on acoustic signature modes, characterized in that, include: A secondary matching process is performed on the wind turbines to be tested in the target wind turbine area to obtain the target comparison group; The operating sounds of various similar fans within the target comparison group are compared and analyzed to identify abnormal equipment.

2. The method according to claim 1, characterized in that, The secondary matching of the wind turbines to be tested in the target wind turbine area to obtain the target comparison group includes: Environmental condition matching is performed on the wind turbines to be tested in the target wind turbine area to obtain regional comparison groups; The wind turbines in the region comparison group are matched for wind turbine status to obtain the target comparison group.

3. The method according to claim 2, characterized in that, The environmental condition matching of the wind turbines to be tested in the target wind turbine area to obtain the area comparison group includes: Obtain the detection points of the fans to be tested in the target fan area; Using the detection points as centers, an environmental state zone corresponding to each fan to be tested is constructed based on a preset influence distance; The fixed and variable elements within the environmental state region are compared to obtain the environmental similarity, and the region comparison group is determined based on the environmental similarity.

4. The method according to claim 3, characterized in that, The step of comparing fixed and variable elements within the environmental state region to obtain environmental similarity, and determining a region comparison group based on the environmental similarity, includes: Identify the contour lines of fixed elements and the variation regions of variable elements within the environmental state region; The contour lines within the environmental state area are compared in orientation to obtain static similarity. By comparing the changing regions within the environmental state region, dynamic similarity is obtained. The environmental similarity is obtained based on the static similarity and the dynamic similarity. The wind turbines whose environmental similarity is greater than the preset similarity are identified as regional wind turbines. The corresponding regional wind turbines are counted to obtain regional comparison groups.

5. The method according to claim 4, characterized in that, The step of performing orientation comparison of contour lines within the environmental state area to obtain static similarity includes: Retrieve multiple preset comparison positions and construct azimuth rays corresponding to the preset comparison positions, starting from the detection point; Based on the detection point, each of the directional rays is rotated to both sides according to a preset rotation angle to obtain the region division line; The environmental state area is segmented based on the region segmentation line to obtain the comparison area corresponding to each preset comparison position. The coverage range of contour lines within the same comparison area is compared to obtain the static similarity of the two wind turbines to be tested.

6. The method according to claim 5, characterized in that, The comparison of the coverage area of ​​contour lines within the same comparison area to obtain the static similarity of the two wind turbines to be tested includes: Retrieve the preset extraction dimension corresponding to the preset comparison position, perform coordinate processing on the environmental state area, and obtain the coordinate extreme points corresponding to the contour lines in the comparison area as occlusion points based on the preset extraction dimension. Connect the occlusion point with the detection point to obtain the occlusion area, and obtain the angle range of the detection point at the occlusion area as the corresponding preset occlusion range of the position. Calculate the directional similarity of the coverage area of ​​the two fans to be tested at the corresponding preset positions; The mean of the azimuth similarity of each preset comparison position is calculated to obtain the static similarity of the two fans to be tested.

7. The method according to claim 4, characterized in that, The step of performing region comparison within the environmental state zone to obtain dynamic similarity includes: Obtain the changing regions within each environmental state zone, and obtain the intersection of the corresponding changing regions of two fans to be tested to obtain the intersection region; Obtain the union of the corresponding change regions of the two fans to be tested to obtain the union region; The dynamic similarity is obtained by the ratio of the area of ​​the intersection region to the area of ​​the union region.

8. The method according to claim 7, characterized in that, Also includes: Retrieve the historical area of ​​the changed region and obtain the current area of ​​the changed region in real time; The changed area is obtained by calculating the absolute value of the difference between the current area and the historical area. The degree of change is obtained based on the ratio of the changed area to the historical area. When the degree of change is determined to be greater than the preset degree value, the dynamic similarity is recalculated.

9. The method according to claim 8, characterized in that, The step of matching the wind turbine status of the wind turbines in the regional comparison group to obtain the target comparison group includes: Obtain the operating time of the fans in each region of the regional comparison group; The operating time of the regional fans is calculated to obtain the difference in operating time between the two regional fans. Fans in areas where the difference duration is less than a preset difference duration are classified as similar fans, and the similar fans are statistically analyzed to obtain a target comparison group.

10. A wind turbine blade condition monitoring system based on acoustic signature modalities, characterized in that, include: The matching module is used to perform secondary matching on the wind turbines to be tested in the target wind turbine area to obtain the target comparison group; The analysis module is used to compare and analyze the operating sounds of various similar fans within the target comparison group to identify abnormal equipment.

Images