Fan blade fatigue crack identification method based on distributed fiber strain response
By organizing the time scale and constructing the oscillation unfolding trajectory of the continuous optical strain response records of wind turbine blades, suspected crack segments are identified and marked, solving the problem of difficulty in identifying the initiation of fatigue cracks in wind turbine blades in the existing technology, and realizing the clear presentation and continuous identification of early crack changes.
Patent Information
- Application Number
- CN202610830730.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-06-10
- Publication Date
- 2026-07-10
AI Technical Summary
In the existing technology, the fatigue crack identification method of wind turbine blades based on distributed optical fiber strain response is difficult to identify crack initiation under long-term low-frequency oscillation conditions, which leads to the missed early warning opportunity and poses a significant operational safety hazard.
By standardizing the time scale of continuous optical strain response records during the operation of wind turbine blades, low-frequency oscillation segments are identified and oscillation development trajectories are constructed. Change segments with amplitude contraction followed by recovery are extracted to form a list of suspected crack segments and mark their time and location. Finally, a crack release trajectory is formed, presenting the crack release process.
It improves the ability to identify the early-stage crack changes in wind turbine blades, clearly presents the local strain change structure under complex background changes, and enhances the continuous identification ability of crack development process.
Smart Images

Figure CN122361441A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of wind power equipment condition monitoring technology, specifically to a method for identifying fatigue cracks in wind turbine blades based on distributed optical fiber strain response. Background Technology
[0002] Fatigue crack identification of wind turbine blades based on distributed fiber optic strain response involves deploying distributed fiber optic sensing elements along the length of the wind turbine blade surface or interior. During wind turbine operation, this continuously acquires strain change information at various locations on the blade and uses fiber optic sensing technology to continuously sense and record minute deformations caused by wind loads, centrifugal forces, and aerodynamic vibrations. Because wind turbine blades are prone to material fatigue under long-term alternating loads, the appearance of microcracks inside or on the surface of the blade alters the local structural stiffness, causing a difference in strain distribution from the normal state. By performing time-series analysis, spatial distribution comparison, or characteristic change identification on the strain response data acquired by the distributed fibers, areas of concentrated abnormal strain can be located. Based on this, the location and development trend of potential fatigue cracks inside the blade can be determined, enabling early identification and continuous monitoring of wind turbine blade structural damage, thus providing a basis for wind turbine operation, maintenance, and safety management.
[0003] The existing technology has the following shortcomings:
[0004] In existing technologies, fatigue crack identification of wind turbine blades based on distributed fiber optic strain response often relies on abrupt changes in the strain waveform to determine crack initiation. However, under prolonged low-frequency oscillation conditions, the overall blade strain exhibits slow, periodic fluctuations, and the strain waveform gradually flattens out. Minor high-frequency changes are easily masked by the overall low-frequency background. At this point, the instantaneous strain release peak amplitude caused by initial crack initiation is small and easily submerged in the gradually changing curve formed by continuous oscillation, making it difficult to detect the true crack release characteristics and creating a monitoring blind zone. As the crack continues to propagate, the anomaly is only identified after a significant decrease in local stiffness, easily missing early warning opportunities and posing a significant operational safety hazard.
[0005] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention
[0006] The purpose of this invention is to provide a method for identifying fatigue cracks in wind turbine blades based on distributed optical fiber strain response, so as to solve the problems mentioned in the background art.
[0007] To achieve the above objectives, the present invention provides the following technical solution: a method for identifying fatigue cracks in wind turbine blades based on distributed optical fiber strain response, comprising the following steps:
[0008] Continuous optical strain response records are collected during the operation of wind turbine blades. The strain change trajectory is organized according to a unified time scale. The difference between adjacent strain amplitudes is calculated on each scale in the strain change trajectory. Slow undulation sections are extracted from the strain change trajectory according to the continuous fluctuation amplitude change law to form a list of low-frequency oscillation sections.
[0009] By using the list of low-frequency oscillation sections, the strain change rhythm is organized step by step in the corresponding time range of the strain change trajectory. The continuous amplitude changes in the corresponding time range are processed by time expansion to form an oscillation expansion trajectory that maintains the order of amplitude changes.
[0010] In the swing vibration development trajectory, the strain changes are compared step by step according to the time progression. The change segments with amplitude first contraction and then rebound are extracted from the swing vibration development trajectory to form a list of suspected crack segments and mark the time position of the suspected crack segments.
[0011] Based on the list of suspected crack segments, the strain change rhythm before and after the suspected crack segments is continuously compared within the corresponding time range in the oscillation unfolding trajectory. The instantaneous amplitude change process of the corresponding time range of the suspected crack segments is unfolded and sorted in the strain change trajectory to form the crack release trajectory.
[0012] Based on the crack release trajectory corresponding to the time range in the strain change trajectory, the crack release trajectory is reorganized and written into the corresponding position of the strain change trajectory. Under the condition of keeping the time progression sequence unchanged, the crack strengthening trajectory is formed, and the crack release change process under the background of low-frequency oscillation is presented through the crack strengthening trajectory.
[0013] The preferred method for forming the list of low-frequency oscillation sections is as follows:
[0014] The continuous optical strain response records are organized with a unified time scale and arranged in the order of time progression to establish a strain change trajectory. For each time scale in the strain change trajectory, a corresponding strain amplitude record is established.
[0015] By comparing the strain amplitudes corresponding to adjacent time scales in the strain change trajectory, a record of strain amplitude changes between adjacent time scales is formed, and a continuously arranged strain amplitude change sequence is formed in the strain change trajectory.
[0016] Based on the strain amplitude change sequence arranged continuously along the time progression, multiple adjacent time scales are divided into continuous time periods, forming continuous fluctuating segments in the strain change trajectory;
[0017] Organize the amplitude changes and rhythms around the continuous undulating segments and extract the slow undulating segments. Record the time scale range corresponding to the slow undulating segments and arrange them in the order of time progression to form a list of low-frequency oscillation segments.
[0018] Preferably, the steps for forming the oscillation unfolding trajectory are as follows:
[0019] Based on the time range of the low-frequency oscillation section list, the corresponding time scale is located in the strain change trajectory. The strain amplitude records of the corresponding time scale are read and continuously arranged along the time progression to form an amplitude change sequence.
[0020] The strain amplitude changes corresponding to the time scale in the amplitude change sequence are sequentially expanded. The amplitude change records are broken down into segments and rearranged in chronological order to form a time expansion sequence.
[0021] The amplitude change records in the time unfolding sequence are arranged continuously, maintaining the sequential connection between the direction and rhythm of amplitude change in the time arrangement to form an amplitude change path;
[0022] By combining the amplitude change paths and arranging them uniformly, and recording the original time scale position, a pendulum expansion trajectory is established in the strain change trajectory while maintaining the corresponding relationship of the amplitude change sequence.
[0023] Preferably, the amplitude change records in the time unfolding sequence are arranged sequentially, and each amplitude change record in the time unfolding sequence is arranged continuously in the order of time progression, while maintaining the original time scale position corresponding to each amplitude change record, forming an amplitude change path in the strain change trajectory and maintaining a consistent relationship of amplitude change order.
[0024] Preferably, the steps for forming the list of suspected crack fragments are as follows:
[0025] In the pendulum expansion trajectory, the strain amplitude records of the corresponding time scale are read one by one in the order of time progression. The strain amplitudes of adjacent time scales are continuously compared and a record of amplitude change is formed. A continuous sequence of amplitude change is established in the pendulum expansion trajectory.
[0026] Based on continuous observation of the amplitude change sequence along the time progression, the continuous time range in which the amplitude change shows a contraction change and a rebound change along the time progression direction is organized into change segments, and the time scale range corresponding to the change segments is recorded in the pendulum unfolding trajectory.
[0027] For the time scale range corresponding to the change segment, the continuous time scale strain amplitude change records are arranged in a centralized manner, maintaining the continuous connection of the amplitude change sequence and forming a change segment;
[0028] The changed segments are summarized and arranged in chronological order according to their position on the pendulum expansion trajectory. The time scale range corresponding to the changed segments is recorded and the position is marked on the pendulum expansion trajectory to form a list of suspected crack segments.
[0029] Preferably, the variable segments in the list of suspected crack segments are located on the corresponding time scale range in the oscillation development trajectory, and the strain amplitude change records corresponding to the continuous time scale within the variable segment are arranged sequentially to maintain the continuous connection between amplitude contraction and amplitude recovery in the time progression direction, and the position of the variable segment is marked in the oscillation development trajectory.
[0030] Preferably, the crack release trajectory formation steps are as follows:
[0031] Based on the list of suspected crack segments, locate the corresponding time scale range in the oscillation development trajectory, read the strain amplitude records of the corresponding time scale in the time progression sequence and arrange them continuously to form the amplitude change sequence of the suspected crack segment corresponding to the time range.
[0032] Based on the amplitude change sequence, the strain amplitude of adjacent time scales is compared scale by scale to form an amplitude change record. The amplitude change record of the middle of the time range corresponding to the suspected crack segment is compared sequentially with the amplitude change record of the time scale before the corresponding time range and the amplitude change record of the time scale after the corresponding time range to form a continuous change path.
[0033] Based on the list of suspected crack fragments, the corresponding time scale strain amplitude records are read from the strain change trajectory and arranged continuously in the order of time progression to form an instantaneous amplitude change sequence.
[0034] The instantaneous amplitude change sequence is continuously connected according to the time progression and the time scale correspondence is maintained to form the crack release trajectory.
[0035] Preferably, when continuously comparing and organizing the amplitude change sequence around the time scale range corresponding to the suspected crack segment, the continuous connection relationship of the amplitude change direction in the time progression sequence is maintained, and the amplitude change records of the corresponding time scale are rearranged step by step in the strain change trajectory, so that the instantaneous amplitude change process within the time range corresponding to the suspected crack segment forms a continuous change structure and runs through the complete time series.
[0036] Preferably, the crack strengthening trajectory formation steps are as follows:
[0037] By combining the crack release trajectory with the corresponding time scale range in the strain change trajectory, the strain amplitude records of the corresponding time scale are read sequentially along the time progression and arranged continuously to form a time series corresponding to the time range of the crack release trajectory.
[0038] Extract strain amplitude records from crack release trajectory, organize them scale by scale, and form amplitude change sequence according to time order. Establish positional correspondence between each time scale in amplitude change sequence and the corresponding time scale in strain change trajectory.
[0039] The amplitude change sequence is written into the time scale position corresponding to the strain change trajectory in the order of time progression, forming an amplitude change record of the time range corresponding to the crack release trajectory in the strain change trajectory;
[0040] The strain amplitude records at the front and back time scales in the amplitude change records and strain change trajectories are continuously arranged to form a crack strengthening trajectory while maintaining the time progression order.
[0041] Preferably, based on the time scale range corresponding to the crack release trajectory, the strain amplitude record of the corresponding time scale is selected and replaced in the strain change trajectory at each scale, and the strain amplitude record of the time scale before the replacement area is kept in a continuous arrangement with the strain amplitude record of the time scale after the replacement area. Under the condition that the time progression order is kept consistent, the amplitude change path of the replacement area is continuously connected, thereby forming a crack strengthening trajectory with a continuous amplitude change structure in the strain change trajectory.
[0042] The technical effects and advantages provided by the present invention in the above technical solution are as follows:
[0043] This invention organizes the continuous optical strain response records generated during the operation of wind turbine blades into a unified time scale and forms a strain change trajectory. Low-frequency oscillation segments are identified within this trajectory, and an oscillation unfolding trajectory is constructed. This allows the amplitude changes, originally superimposed on the low-frequency oscillation background, to be temporally unfolded and rhythmically rearranged, thus clearly revealing the local strain change structure hidden within the oscillation process in a continuous time series. In this way, the minute strain release changes generated by wind turbine blades during long-term operation can be fully presented under complex background changes, enabling the strain change characteristics formed during crack initiation to form a continuous changing structure within the strain change trajectory, thereby improving the ability to identify the early crack change process.
[0044] This invention constructs a list of suspected crack fragments and further traces back to form crack release trajectories. While maintaining the overall temporal progression of the strain change trajectory, it generates crack intensification trajectories, allowing the crack release process to be continuously presented within a complete time series. By continuously organizing the local amplitude changes through crack intensification trajectories, crack release changes generated during wind turbine blade operation can form independent change paths against a low-frequency oscillation background. This ensures that crack release changes are continuously presented within the strain change trajectory, providing a stable reference for wind turbine blade operating status analysis and improving the continuous identification capability of crack development processes. Attached Figure Description
[0045] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.
[0046] Figure 1 This is a flowchart of the method for identifying fatigue cracks in wind turbine blades based on distributed optical fiber strain response, as described in this invention. Detailed Implementation
[0047] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that the description of this disclosure will be more complete and fully convey the concept of the exemplary embodiments to those skilled in the art.
[0048] This invention provides, for example Figure 1 The method for identifying fatigue cracks in wind turbine blades based on distributed optical fiber strain response, as shown, includes the following steps:
[0049] Continuous optical strain response records are collected during the operation of wind turbine blades. The strain change trajectory is organized according to a unified time scale. The difference between adjacent strain amplitudes is calculated on each scale in the strain change trajectory. Slow undulation sections are extracted from the strain change trajectory according to the continuous fluctuation amplitude change law to form a list of low-frequency oscillation sections.
[0050] After the optical strain response records of a continuous time series are standardized with a uniform time scale, the strain amplitude variation relationship can be analyzed layer by layer along the time sequence. This transforms the continuous optical strain response records into strain variation trajectories with a clear variation structure. Within the strain variation trajectory, the amplitude variation process is continuously divided to identify the slow fluctuation segments corresponding to low-frequency oscillation characteristics. The specific implementation process is as follows.
[0051] The continuous optical strain response records generated during the operation of wind turbine blades are organized using a unified time scale. During the organization process, a fixed time interval is selected as the time scale reference. The strain amplitude corresponding to each acquisition moment in the continuous optical strain response records is arranged sequentially according to time, so that all strain amplitude records form a continuous arrangement relationship along the time direction, thereby constructing a complete strain change trajectory. After the strain change trajectory is formed, the strain amplitude of any time scale in the strain change trajectory is compared with the strain amplitude of its subsequent adjacent time scales. The direction of amplitude change and the order of change between the two are recorded in the strain change trajectory, so that the strain change trajectory, while maintaining the original strain amplitude arrangement, also forms a continuous amplitude change relationship sequence.
[0052] In this amplitude change relationship sequence, each time scale corresponds to a unique amplitude change record and maintains a one-to-one correspondence with the time scale in the strain change trajectory, so that the strain change trajectory contains both continuous strain amplitude information and continuous amplitude change direction information.
[0053] After the strain change trajectory and amplitude change sequence are formed, the strain change trajectory is unfolded step by step along the time progression starting from the initial time scale. The amplitude change direction between adjacent time scales is continuously arranged so that the amplitude change direction forms a continuous change chain in chronological order. In this change chain, when the amplitude change direction of multiple consecutive time scales maintains a continuous alternation relationship, that is, when the change process of rising to falling and then rising occurs sequentially between adjacent time scales, the continuous time range is divided into continuous fluctuation segments in the strain change trajectory.
[0054] When the direction of amplitude change is interrupted in the continuous time scale, that is, when two or more consecutive time scales maintain the same direction of change or the direction of amplitude change stops alternating, the position of that time scale is taken as the end position of the continuous fluctuation segment, and the division continues to unfold backward with the end position as the new starting point. Thus, multiple continuous fluctuation segments with clear boundaries are formed in the strain change trajectory. Each continuous fluctuation segment corresponds to a time interval in which the direction of amplitude change is continuous and alternates.
[0055] After forming multiple consecutive fluctuating segments, the amplitude change relationship within each consecutive fluctuating segment is sorted out segment by segment. In the sorting process, the starting time scale of the consecutive fluctuating segment is used as the starting point, and the amplitude change direction within the segment is read step by step along the time progression direction, so that the amplitude change relationship within the segment is fully unfolded in chronological order.
[0056] During the unfolding process, the rhythm of amplitude changes within the continuous fluctuation segment is continuously recorded, so that the direction of amplitude change forms a complete alternating sequence in the time dimension. The frequency of amplitude change direction conversion between adjacent time scales is observed in this sequence. When the direction of amplitude change maintains a uniform alternation relationship in multiple consecutive time scales and there is no interruption between each change, the continuous time range is further marked as a slow fluctuation segment within the continuous fluctuation segment. When the direction of amplitude change repeats continuously or is interrupted at a certain time scale, that time scale is taken as the end boundary of the slow fluctuation segment, and the next slow fluctuation segment is identified again from that position. Thus, one or more slow fluctuation segments with clear time ranges are formed within each continuous fluctuation segment. Each slow fluctuation segment maintains a continuous time scale arrangement and contains a complete sequence of amplitude change directions.
[0057] Based on all the slowly fluctuating segments identified in the strain change trajectory, all slowly fluctuating segments are uniformly organized. The start and end time scales of each slowly fluctuating segment are recorded, and the strain amplitude records and amplitude change relationship sequences within the corresponding time range are kept completely linked, so that each slowly fluctuating segment forms an independent segment record containing time range and amplitude change relationship. Then, all segments are arranged according to the time sequence of the slowly fluctuating segments in the strain change trajectory, so that all slowly fluctuating segments form a continuous segment sequence, and the time connection relationship between each segment is maintained in this sequence.
[0058] After completing the sorting of all sections, all slowly fluctuating sections are compiled into a list of low-frequency oscillation sections. The list of low-frequency oscillation sections contains multiple slowly fluctuating sections arranged in chronological order. Each slowly fluctuating section can be located in a corresponding time range in the strain change trajectory and maintains a consistent correspondence with the time scale inside the strain change trajectory. This allows the list of low-frequency oscillation sections to serve as a unified time reference in subsequent processing, used to locate the time interval related to low-frequency oscillation in the strain change trajectory, and to provide a continuous time basis for the subsequent development of the oscillation change rhythm.
[0059] By using the list of low-frequency oscillation sections, the strain change rhythm is organized step by step in the corresponding time range of the strain change trajectory. The continuous amplitude changes in the corresponding time range are processed by time expansion to form an oscillation expansion trajectory that maintains the order of amplitude changes.
[0060] The various time ranges recorded in the list of low-frequency oscillation segments can be mapped to specific time scale intervals in the strain change trajectory. Using these time scale intervals as the processing objects, the amplitude variation relationships within the strain change trajectory are broken down and sequentially unfolded point by point. This transforms the amplitude variation relationships, originally superimposed during continuous time progression, into independently arranged change paths, thus forming an oscillation unfolding trajectory that maintains the order of amplitude changes. The specific implementation process is as follows.
[0061] Each time range in the low-frequency oscillation section list is located one by one in the strain change trajectory, so that each time range corresponds to a continuously arranged time scale interval. The starting position of the time scale interval is used as the reading starting point, and the strain amplitude record corresponding to each time scale in the interval is read one by one along the time progression direction. At the same time, the strain amplitude of each time scale is arranged in chronological order to form the original amplitude arrangement path.
[0062] During the point-by-point reading process, each time scale and its successor form a group of adjacent amplitude change units. Each group of adjacent amplitude change units contains the strain amplitude value of the previous time scale and the strain amplitude value of the successor time scale. All adjacent amplitude change units are recorded sequentially in chronological order, so that a set of continuously arranged amplitude change units is formed throughout the entire time range. This transforms the continuous amplitude change in the original strain change trajectory into a change structure composed of multiple independent amplitude change units.
[0063] Around the formed set of amplitude change units, each amplitude change unit is broken down one by one, so that each amplitude change unit occupies an independent position in the new arrangement structure. In specific processing, the time progression order is used as the arrangement basis. The first amplitude change unit is taken as the starting point of the expansion and placed at the beginning of the expansion path. The direction of change between the previous time scale and the next time scale in the amplitude change unit is kept unchanged. Then, the second amplitude change unit is arranged in sequence after the first amplitude change unit, so that the two adjacent amplitude change units maintain a continuous arrangement relationship in the expansion path. All subsequent amplitude change units are arranged in the same way.
[0064] By using this method of splitting and arranging units sequentially, the amplitude changes that were originally closely connected in a continuous time scale are transformed into multiple independently arranged change segments. Each change segment corresponds to an amplitude change between the original time scales, thus forming a segmented change structure in the unfolding path.
[0065] The amplitude variation units that have been split and arranged are subjected to unified time expansion processing, so that each amplitude variation unit corresponds to an independent time position in the expansion path and is arranged sequentially according to a unified interval. In this process, the starting position of the expansion path is used as the new time starting point. The first amplitude variation unit is assigned a first time position mark, the second amplitude variation unit is assigned a second time position mark, and subsequent amplitude variation units are sequentially numbered so that all amplitude variation units form a continuously increasing time position sequence. At the same time, the amplitude values before and after the corresponding amplitude variation unit and the direction of change are completely preserved in each time position, so that each time position can reflect an independent amplitude variation process.
[0066] By using a unified time position marking method, all amplitude change units form a new time arrangement structure in the expansion path, and the sequential relationship between each amplitude change unit is kept completely consistent with the original time sequence. This enables the time expansion processing of continuous amplitude changes, allowing the change rhythm that was originally superimposed on the continuous time scale to be expanded segment by segment.
[0067] After the time-spreading process is completed, all amplitude change units are integrated in chronological order, forming a continuous change trajectory in the unfolding path. A correspondence is established between each amplitude change unit and the original time scale in the strain change trajectory. In practice, the original time scale interval from which each amplitude change unit originates is recorded at the time position corresponding to each amplitude change unit, so that each change in the unfolding path can be traced back to the specific time range in the strain change trajectory. At the same time, the arrangement order of all amplitude change units in the unfolding path remains consistent with the original time order.
[0068] This integration method allows the unfolding path to possess both a new time arrangement structure and retain the correspondence of the original time scale, thus forming a complete oscillation unfolding trajectory. This oscillation unfolding trajectory consists of multiple amplitude change units arranged in chronological order. Each amplitude change unit represents an adjacent amplitude change process in the original strain change trajectory. Through time spreading processing, the amplitude change relationship within the low-frequency oscillation segment is unfolded segment by segment in the time dimension, thereby providing a continuous, clear, and traceable change reference path for the subsequent extraction of crack change segments.
[0069] In the swing vibration development trajectory, the strain changes are compared step by step according to the time progression. The change segments with amplitude first contraction and then rebound are extracted from the swing vibration development trajectory to form a list of suspected crack segments and mark the time position of the suspected crack segments.
[0070] By refining and organizing the strain variation relationship on a scale-by-scale around the continuous time intervals of the oscillation trajectory, the amplitude variation process within the trajectory can be broken down and fully presented layer by layer. This allows for the extraction of amplitude contraction followed by recovery segments, forming a list of suspected crack segments, and establishing corresponding time position markers on the oscillation trajectory. The specific implementation process is as follows.
[0071] To enable continuous quantization of the strain change process in the pendulum unfolding trajectory, the discrete scale position index in the pendulum unfolding trajectory is first defined as... ,in and define the first The strain amplitude at each time scale is Based on this, a directed amplitude difference between adjacent time scales is constructed:
[0072]
[0073] in, Indicates the first The change in strain amplitude at each time point relative to the previous time point; Indicates the first The strain amplitude value corresponding to each time scale; Indicates the first The strain amplitude value corresponding to each time scale; This represents the total number of time scale divisions in the oscillation unfolding trajectory. The purpose of this formula is to transform the original strain amplitude sequence into local variations between adjacent scale divisions, providing a unified basic input for subsequent identification of variation direction, division of variation segments, and combination of variation segments.
[0074] Starting from the initial time scale of the pendulum expansion trajectory, the strain amplitude value corresponding to each time scale is read sequentially along the time progression direction. The strain amplitude of the current time scale is compared with the strain amplitude of the previous time scale item by item. When the strain amplitude value of the current time scale is less than that of the previous time scale, the change direction corresponding to that time scale is recorded as amplitude contraction. When the strain amplitude value of the current time scale is greater than that of the previous time scale, the change direction corresponding to that time scale is recorded as amplitude recovery. The above change directions are associated with the corresponding time scale positions and written into the pendulum expansion trajectory, so that each time scale in the pendulum expansion trajectory has a clear change direction indicator.
[0075] After obtaining the directed amplitude difference between adjacent time scales, in order to directly map the amplitude change into two types of directional attributes, namely contraction or recovery, a direction indicator is introduced:
[0076]
[0077] in, Indicates the first The direction of change at each time scale indicates the quantity; when When, the corresponding amplitude contracts; when At that time, the corresponding magnitude rebounded; for The absolute value; The minimum positive adjustment is used to avoid the denominator being zero and to maintain the continuous stability of the mapping expression. The value of is set according to the resolution of the distributed optical fiber strain response data. Its essence is a positive number that is much smaller than the change in strain amplitude. It is used to avoid the denominator being zero and to maintain calculation stability, while not changing the determination result of the amplitude change direction and the inter-segment coupling relationship.
[0078] Specifically, It can be taken as the smallest resolvable change in strain amplitude between adjacent time scales. A ratio value, namely ,in for The proportionality coefficient makes The numerical value is significantly smaller than the actual strain change, thus ensuring numerical stability without substantially affecting the direction judgment and segment coupling criteria.
[0079] The purpose of this formula is to compress the increase or decrease of strain amplitude into a unified directional symbol, so that each time scale has a clear directional attribute and can maintain a direct relationship with the original amplitude change.
[0080] During the process of scaling up the scale, the amplitude change between the current time scale and the previous time scale is recorded simultaneously, and the fixed correspondence between the change process and the time scale sequence is maintained, so that a continuous sequence of change directions is formed inside the swing oscillation trajectory, thereby transforming the original amplitude change into a continuous change path with directional attributes.
[0081] Directional indicators alone are insufficient to fully characterize the amplitude change process; therefore, a directional intensity chain is further constructed based on the directional indicator quantity.
[0082]
[0083] in, Indicates the first The intensity of directional consistency change at each time scale; It is formed by the change quantity and the direction indicator quantity, and is used to uniformly represent the magnitude and direction of change as a single chain quantity. The function of this formula is to ensure that each time scale in the oscillation unfolding trajectory not only has the direction information of contraction or recovery, but also retains the strength of the change in that direction, thereby forming a continuously spliced change path.
[0084] Along the time progression direction of the pendulum unfolding trajectory, the established sequence of changing directions is continuously segmented, forming independent segments where the changing directions are consistent between adjacent time points. During segmentation, the starting time point of the pendulum unfolding trajectory is used as the segmentation starting point. Time intervals where the changing directions are consistent between two or more consecutive time points are divided into the same changing segment, and the starting and ending time points of this segment are recorded. Simultaneously, the strain amplitude change path corresponding to all time points within this segment is recorded. When a time point occurs where the changing direction changes from amplitude contraction to amplitude recovery, or from amplitude recovery to amplitude contraction, the previous changing segment ends, and the time point where the changing direction changes is used as the starting position for a new changing segment, continuing the segmentation process.
[0085] To identify the location where the direction of change changes and establish the boundary of the change segment accordingly, a boundary trigger value is introduced:
[0086]
[0087] in, Indicates the first The boundary response of the change segment at each time scale; and These represent the direction indicators for two adjacent time scales; when the two directions are the same... It remains at a low level; when the two directions switch, the value increases significantly. and These represent the absolute values of directional intensity at the current and previous scale points, respectively, used to enhance the significance of changes in boundary positions. The function of this formula is to utilize both directional change and local intensity to jointly determine segment boundaries, making the division of change segments no longer a simple positive / negative reversal judgment, but rather a boundary recognition capability that combines directional switching and the strength of change.
[0088] By using the above segmentation method, the sequence of changing directions in the pendulum unfolding trajectory is divided into multiple continuously arranged segments. Each segment maintains a uniform direction of change, and the segments are connected sequentially in chronological order, thus forming a clearly structured sequence of segments.
[0089] Once the boundaries of each change segment are determined, in order to represent each change segment as a whole, the first... Segment-level aggregation of each change segment:
[0090]
[0091] in, Indicates the first The directional polymerization intensity of each changing segment; Indicates the segment number; Indicates the first The starting time scale index of each change segment; Indicates the first The formula is used to aggregate the directional intensity within multiple consecutive time scales, thereby compressing the fine-grained, scale-by-scale change path into a segment-level feature quantity that can be used for inter-segment comparison, providing a unified basis for subsequent combination of change segments.
[0092] After the sequence of change segments is formed, the combination relationship between adjacent change segments is compared group by group. Two consecutive change segments are observed as a group. When the change segment in the preceding position is an amplitude contraction change segment and the change segment in the following position is an amplitude recovery change segment, and the two change segments are continuously connected on the time scale, that is, the start time scale of the latter change segment is immediately followed by the end time scale of the former change segment, the two change segments are combined into a complete change segment. During the combination process, the start and end time scales of the amplitude contraction change segment are recorded, and the start and end time scales of the amplitude recovery change segment are also recorded. The strain amplitude change paths corresponding to all time scales within the two change segments are spliced together in chronological order, so that the combined change segment forms a complete change process with a continuous transition from amplitude contraction to amplitude recovery.
[0093] Based on the obtained segment-level aggregation, in order to quantitatively identify the combination relationship of the first segment contraction, the second segment rebound, and the two segments being continuously connected, a segment coupling criterion is constructed:
[0094]
[0095] in, Indicates the first The first change segment and the first The coupling value of a candidate change segment is composed of several change segments; Indicates the directional polymerization intensity of the next change segment; when the first The first change segment is the amplitude contraction change segment, the second... When the change segment is a segment of amplitude recovery, the polymerization intensities of the two have opposite signs. This creates a positive reinforcement; Used to constrain the amplitude of the coupling relationship between the two segments; This is a continuity indicator factor used to constrain two segments to maintain a direct adjacent relationship. The function of this formula is to unify the combination relationship of change direction, segment-level intensity relationship, and temporal continuity relationship into a single expression, giving the extraction of candidate change segments complete structural constraints.
[0096] In this change segment, no other change direction segments are introduced, and the direct connection between the amplitude contraction change segment and the amplitude recovery change segment is maintained, thereby ensuring that the extracted change segment fully reflects the change structure of amplitude first contraction and then recovery.
[0097] To further clarify that there must be no discontinuous continuity between two consecutive change segments, the continuity indicator factor is defined as:
[0098]
[0099] in, Indicates the first The first change segment and the first The continuous connection state between each change segment; Indicates the starting time scale index of the next change segment. When When, it means that the next change segment begins immediately after the end of the previous change segment, and there are no other inserted segments between the two segments; when If this occurs, it indicates that the two segments do not meet the requirement of direct connection. The purpose of this formula is to forcibly limit the composition of the changing segments from the perspective of the time scale structure, ensuring that the finally extracted segments strictly conform to the original technical requirement of directly connecting the contraction segment to the recovery segment.
[0100] All variation segments that satisfy the condition of a continuous combination of amplitude contraction and amplitude recovery variation segments are uniformly organized. Each variation segment is arranged according to its temporal position in the oscillation development trajectory, and the corresponding start and end time scales are recorded for each variation segment, while retaining the complete strain amplitude change path within that time range. During the organization process, all variation segments are compiled into a list of suspected crack segments, which contains multiple variation segment records arranged in chronological order, and each variation segment can be located within a corresponding continuous time range in the oscillation development trajectory.
[0101] After determining the combination of adjacent change segments, the segment pairs that meet the coupling conditions are registered as suspected crack segments, and their time range is defined as follows:
[0102]
[0103] in, Indicates the first The time range corresponding to the suspected crack fragment; Indicates the suspected crack segment number; This indicates the starting time index of the preceding contraction segment that constitutes the suspected crack segment; This indicates the end time index of the subsequent rising change segment that constitutes the suspected crack segment; This indicates that the segment has met the contraction-rebound combination and remains continuously connected. The function of this formula is to unify the results of the aforementioned scale-by-scale identification, segmentation and sorting and inter-segment coupling determination into directly registerable segment boundaries, thereby forming a sortable, traceable and traceable list of suspected crack segments.
[0104] Meanwhile, the start and end time scales of each change segment are marked in the swing vibration development trajectory, so that each change segment in the list of suspected crack segments corresponds to a specific time position in the swing vibration development trajectory and remains consistent with the original time scale in the strain change trajectory. This allows the suspected crack segments to be directly traced back and the amplitude change process of the corresponding time range to be unfolded in subsequent processing.
[0105] After the suspected crack segment boundary has been clearly defined, in order to achieve continuous annotation of all time scales within the segment, a set of location annotations is further defined:
[0106]
[0107] in, Indicates the first A set of location markers for suspected crack fragments in the oscillation unfolding trajectory; This represents any time scale index within the time range of the suspected crack segment. The purpose of this formula is to further expand the start and end boundaries of the suspected crack segment into a complete set of locations, enabling subsequent processing to directly trace back the corresponding original strain change trajectory based on this set of locations and continue to analyze the amplitude change process within this time range.
[0108] Based on the list of suspected crack segments, the strain change rhythm before and after the suspected crack segments is continuously compared within the corresponding time range in the oscillation unfolding trajectory. The instantaneous amplitude change process of the corresponding time range of the suspected crack segments is unfolded and sorted in the strain change trajectory to form the crack release trajectory.
[0109] After a list of suspected crack fragments has been established and the corresponding time ranges can be located in the oscillation development trajectory, the strain change rhythm in the oscillation development trajectory can be continuously compared and organized around the time positions recorded in the list of suspected crack fragments. Simultaneously, the corresponding time ranges can be traced back to the strain change trajectory to unfold the instantaneous amplitude change process layer by layer, thereby forming a complete and continuous crack release trajectory. The specific steps are as follows:
[0110] For each suspected crack segment in the list of suspected crack segments, time positioning is performed. The start time and end time scale recorded in the list of suspected crack segments are searched one by one in the oscillation development trajectory, so that each time range in the list of suspected crack segments forms a corresponding segment in the oscillation development trajectory. After completing the time positioning, the strain amplitude record of the corresponding time scale is read step by step along the time progression of the oscillation development trajectory, using the start time scale recorded in the list of suspected crack segments as the starting point. All strain amplitude records from the start time scale to the end time scale are arranged continuously.
[0111] Based on this, using the starting time scale recorded in the list of suspected crack segments as a reference, multiple time scales are selected consecutively before it in reverse chronological order as the preceding reference time range. At the same time, using the ending time scale recorded in the list of suspected crack segments as a reference, multiple time scales are selected consecutively after it in chronological order as the following reference time range. This makes the preceding reference time range, the corresponding time range of the suspected crack segment, and the following reference time range form a continuous time sequence, and maintains a one-to-one correspondence between the strain amplitude record and the time scale position in the oscillation development trajectory within this continuous time sequence.
[0112] After forming a continuous time series, the strain change rhythm within the continuous time series in the pendulum expansion trajectory is compared and organized step by step. In specific implementation, starting from the starting time scale of the reference time range, the strain amplitude record is read step by step along the time progression direction, and the strain amplitude of the current time scale is compared with the strain amplitude of the previous time scale to form a continuous amplitude change record between adjacent time scales.
[0113] After recording the amplitude changes between all time scales, the amplitude change records within the time range corresponding to the suspected crack segment are sequentially compared with the amplitude change records within the preceding reference time range. Simultaneously, the amplitude change records within the time range corresponding to the suspected crack segment are sequentially compared with the amplitude change records within the subsequent reference time range. This ensures that the amplitude change rhythm within the time range corresponding to the suspected crack segment forms a continuous change chain with the preceding and subsequent reference time ranges in the time dimension. During this comparison process, the direction of amplitude change is kept consistent with the time progression sequence, so that the amplitude change relationship in the continuous time series can form a continuous connection path along the time direction.
[0114] After completing continuous comparison in the swing vibration development trajectory, the corresponding time range recorded in the list of suspected crack segments is traced back to the strain change trajectory for development and organization. In specific implementation, the start time scale and end time scale in the list of suspected crack segments are used as the positioning basis. The strain amplitude record of the corresponding time scale is read step by step in the strain change trajectory, and all strain amplitude records within the time range are arranged continuously in the order of time progression.
[0115] After arranging the time sequence, the amplitude variation relationship between adjacent time scales within the time range is sorted out scale by scale, so that the strain amplitude variation of each time scale can form a continuous connection in the sequence. On this basis, the amplitude variation process within the time range is expanded, so that the amplitude variation relationship originally formed by superposition within the continuous time scale is redistributed according to the time sequence, so that the amplitude variation corresponding to each time scale can be expanded segment by segment in the time sequence and maintain the sequential connection relationship, while keeping the correspondence between each time scale and the corresponding time scale in the oscillation expansion trajectory unchanged.
[0116] After completing the unfolding and sorting of the time range corresponding to the suspected crack segment, the amplitude change sequence formed within the time range is uniformly arranged so that the amplitude change records corresponding to all time scales are continuously connected in the order of time progression, thereby forming a complete and continuous change path. During the sorting process, each time scale within the time range is arranged in chronological order, and the amplitude change relationship between adjacent time scales is kept continuous, while the original time position relationship of each time scale in the strain change trajectory is maintained.
[0117] The crack release trajectory is formed by the above sorting method. The crack release trajectory completely records the instantaneous amplitude change process within the time range corresponding to the suspected crack segment. By rearranging the time sequence, the local amplitude change is expanded in the continuous time dimension, so that the release change process corresponding to the suspected crack segment can be continuously expressed in the strain change trajectory, and provides a continuous time reference basis for the construction of the subsequent crack strengthening trajectory.
[0118] Based on the crack release trajectory corresponding to the time range in the strain change trajectory, the crack release trajectory is reorganized and written into the corresponding position of the strain change trajectory. Under the condition of keeping the time progression sequence unchanged, the crack strengthening trajectory is formed. The crack release change process under the background of low-frequency oscillation is presented through the crack strengthening trajectory.
[0119] After the crack release trajectory has been formed and the corresponding time range can be located within the strain change trajectory, the strain amplitude records in the strain change trajectory can be rearranged scale by scale around the corresponding time range of the crack release trajectory. This allows the instantaneous amplitude changes in the crack release trajectory to be written into the corresponding positions of the strain change trajectory according to a predetermined time scale order, forming a crack strengthening trajectory while maintaining consistent time scale intervals. This results in a continuous and identifiable crack release path under low-frequency oscillation background. The specific implementation process is as follows:
[0120] The time range corresponding to the crack release trajectory is located scale by scale. The start and end time scales of the crack release trajectory record are found in the strain change trajectory, and this time range is defined as a continuous time segment in the strain change trajectory. The original strain amplitude record is read scale by scale within this continuous time segment. At the same time, the strain amplitude records corresponding to the continuous time scales before and after this continuous time segment are recorded, so that the strain change trajectory maintains a continuous arrangement before and after this time segment. The corresponding time scale positions are reserved in this continuous time segment for subsequent writing operations. The time index position of each time scale does not change during the entire time positioning process, so that the time range corresponding to the crack release trajectory forms a fixed time position interval in the strain change trajectory.
[0121] The strain amplitude records within the crack release trajectory are expanded and organized step by step, so that the strain amplitude records of each time scale in the crack release trajectory form a continuous amplitude change sequence in the order of time progression. In this continuous amplitude change sequence, the amplitude change direction between adjacent time scales is recorded step by step, so that the amplitude contraction stage and amplitude recovery stage in the crack release trajectory are arranged continuously in the time series, and the strain amplitude record corresponding to each time scale is consistent with the original time position in the crack release trajectory.
[0122] After completing the continuous amplitude change sequence, each time scale in the sequence is marked to correspond one-to-one with the corresponding time scale in the strain change trajectory. This creates a scale-by-scale mapping relationship between the amplitude change records in the crack release trajectory and the strain change trajectory. It is also ensured that the time scale index remains unchanged during the mapping process, thus establishing a precise positional correspondence for subsequent scale-by-scale writing.
[0123] After establishing a one-to-one correspondence between the amplitude change sequence within the crack release trajectory and the time scale of the strain change trajectory, a scale-by-scale replacement writing operation is performed on the strain amplitude records in the crack release trajectory. In this process, the starting time scale of the crack release trajectory is used as the writing starting point, and the strain amplitude records in the crack release trajectory are written to the corresponding time scale positions along the time advancement direction in the strain change trajectory. The strain amplitude records corresponding to the time scale in the original strain change trajectory are overwritten and replaced, while keeping the strain amplitude records before and after the time scale unchanged.
[0124] Throughout the writing process, the time interval between adjacent time scales is kept consistent, so that the strain change trajectory after writing is still arranged continuously according to the original time scale order. Furthermore, by replacing each scale, the amplitude change structure in the crack release trajectory is continuously embedded within the corresponding time range, thereby forming a local amplitude change reconstruction region in the original strain change trajectory.
[0125] After completing the scale-by-scale replacement and writing operation, the strain change trajectory after writing is sorted out in an overall order so that the strain amplitude records corresponding to all time scales are arranged continuously according to the original time scale index, and the amplitude change records within the time range corresponding to the crack release trajectory are completely preserved. In this sorting process, the amplitude change path within the time range corresponding to the crack release trajectory is continuously connected with the amplitude change path of the preceding and following time scales, so that the entire strain change trajectory forms an uninterrupted continuous change sequence in the time dimension.
[0126] This method forms a crack strengthening trajectory. While maintaining the time progression sequence, the crack strengthening trajectory fully embeds the continuous changes in amplitude contraction and amplitude recovery in the crack release trajectory into the strain change trajectory. This allows the crack release change process to form an independent and continuous change path in the low-frequency oscillation background, thus enabling the crack release change to be continuously presented along the time dimension.
[0127] This invention organizes the continuous optical strain response records generated during the operation of wind turbine blades into a unified time scale and forms a strain change trajectory. Low-frequency oscillation segments are identified within this trajectory, and an oscillation unfolding trajectory is constructed. This allows the amplitude changes, originally superimposed on the low-frequency oscillation background, to be temporally unfolded and rhythmically rearranged, thus clearly revealing the local strain change structure hidden within the oscillation process in a continuous time series. In this way, the minute strain release changes generated by wind turbine blades during long-term operation can be fully presented under complex background changes, enabling the strain change characteristics formed during crack initiation to form a continuous changing structure within the strain change trajectory, thereby improving the ability to identify the early crack change process.
[0128] This invention constructs a list of suspected crack fragments and further traces back to form crack release trajectories. While maintaining the overall temporal progression of the strain change trajectory, it generates crack intensification trajectories, allowing the crack release process to be continuously presented within a complete time series. By continuously organizing the local amplitude changes through crack intensification trajectories, crack release changes generated during wind turbine blade operation can form independent change paths against a low-frequency oscillation background. This ensures that crack release changes are continuously presented within the strain change trajectory, providing a stable reference for wind turbine blade operating status analysis and improving the continuous identification capability of crack development processes.
[0129] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.
Claims
1. A method for identifying fatigue cracks in wind turbine blades based on distributed optical fiber strain response, characterized in that, Includes the following steps: Continuous optical strain response records are collected during the operation of wind turbine blades. The strain change trajectory is organized according to a unified time scale. The difference between adjacent strain amplitudes is calculated on each scale in the strain change trajectory. Slow undulation sections are extracted from the strain change trajectory according to the continuous fluctuation amplitude change law to form a list of low-frequency oscillation sections. By using the list of low-frequency oscillation sections, the strain change rhythm is organized step by step in the corresponding time range of the strain change trajectory. The continuous amplitude changes in the corresponding time range are processed by time expansion to form an oscillation expansion trajectory that maintains the order of amplitude changes. In the swing vibration development trajectory, the strain changes are compared step by step according to the time progression. The change segments with amplitude first contraction and then rebound are extracted from the swing vibration development trajectory to form a list of suspected crack segments and mark the time position of the suspected crack segments. Based on the list of suspected crack segments, the strain change rhythm before and after the suspected crack segments is continuously compared within the corresponding time range in the oscillation unfolding trajectory. The instantaneous amplitude change process of the corresponding time range of the suspected crack segments is unfolded and sorted in the strain change trajectory to form the crack release trajectory. Based on the crack release trajectory corresponding to the time range in the strain change trajectory, the crack release trajectory is reorganized and written into the corresponding position of the strain change trajectory. Under the condition of keeping the time progression sequence unchanged, the crack strengthening trajectory is formed, and the crack release change process under the background of low-frequency oscillation is presented through the crack strengthening trajectory.
2. The method for identifying fatigue cracks in wind turbine blades based on distributed optical fiber strain response according to claim 1, characterized in that, The process for generating the list of low-frequency oscillation sections is as follows: The continuous optical strain response records are organized with a unified time scale and arranged in the order of time progression to establish a strain change trajectory. For each time scale in the strain change trajectory, a corresponding strain amplitude record is established. By comparing the strain amplitudes corresponding to adjacent time scales in the strain change trajectory, a record of strain amplitude changes between adjacent time scales is formed, and a continuously arranged strain amplitude change sequence is formed in the strain change trajectory. Based on the strain amplitude change sequence arranged continuously along the time progression, multiple adjacent time scales are divided into continuous time periods, forming continuous fluctuating segments in the strain change trajectory; Organize the amplitude changes and rhythms around the continuous undulating segments and extract the slow undulating segments. Record the time scale range corresponding to the slow undulating segments and arrange them in the order of time progression to form a list of low-frequency oscillation segments.
3. The method for identifying fatigue cracks in wind turbine blades based on distributed optical fiber strain response according to claim 2, characterized in that, The steps for forming the oscillation unfolding trajectory are as follows: Based on the time range of the low-frequency oscillation section list, the corresponding time scale is located in the strain change trajectory. The strain amplitude records of the corresponding time scale are read and continuously arranged along the time progression to form an amplitude change sequence. The strain amplitude changes corresponding to the time scale in the amplitude change sequence are sequentially expanded. The amplitude change records are broken down into segments and rearranged in chronological order to form a time expansion sequence. Based on the amplitude change records in the time unfolding sequence, they are arranged continuously, maintaining the sequential connection between the direction and rhythm of amplitude change in the time arrangement to form an amplitude change path; By combining the amplitude change paths and arranging them uniformly, and recording the original time scale position, a pendulum expansion trajectory is established in the strain change trajectory while maintaining the corresponding relationship of the amplitude change sequence.
4. The method for identifying fatigue cracks in wind turbine blades based on distributed optical fiber strain response according to claim 3, characterized in that, The amplitude change records in the time unfolding sequence are sorted out sequentially. Each amplitude change record in the time unfolding sequence is arranged continuously in the order of time progression, while maintaining the original time scale position of each amplitude change record. An amplitude change path is formed in the strain change trajectory, and the amplitude change order is kept consistent.
5. The method for identifying fatigue cracks in wind turbine blades based on distributed optical fiber strain response according to claim 3, characterized in that, The steps for creating a list of suspected crack fragments are as follows: In the pendulum expansion trajectory, the strain amplitude records of the corresponding time scale are read one by one in the order of time progression. The strain amplitudes of adjacent time scales are continuously compared and a record of amplitude change is formed. A continuous sequence of amplitude change is established in the pendulum expansion trajectory. Based on continuous observation of the amplitude change sequence along the time progression, the continuous time range in which the amplitude change shows a contraction change and a rebound change along the time progression direction is organized into change segments, and the time scale range corresponding to the change segments is recorded in the pendulum unfolding trajectory. For the time scale range corresponding to the change segment, the continuous time scale strain amplitude change records are arranged in a centralized manner, maintaining the continuous connection of the amplitude change sequence and forming a change segment; The changed segments are summarized and arranged in chronological order according to their position on the pendulum expansion trajectory. The time scale range corresponding to the changed segments is recorded and the position is marked on the pendulum expansion trajectory to form a list of suspected crack segments.
6. The method for identifying fatigue cracks in wind turbine blades based on distributed optical fiber strain response according to claim 5, characterized in that, The variation segments in the suspected crack segment list are located on the corresponding time scale range in the oscillation development trajectory. The strain amplitude change records corresponding to the continuous time scale within the variation segment are arranged sequentially to maintain the continuous connection between amplitude contraction and amplitude recovery in the time progression direction, and the position of the variation segment is marked in the oscillation development trajectory.
7. The method for identifying fatigue cracks in wind turbine blades based on distributed optical fiber strain response according to claim 5, characterized in that, The steps involved in the formation of a crack release trajectory are as follows: Based on the list of suspected crack segments, locate the corresponding time scale range in the oscillation development trajectory, read the strain amplitude records of the corresponding time scale in the time progression sequence and arrange them continuously to form the amplitude change sequence of the suspected crack segment corresponding to the time range. Based on the amplitude change sequence, the strain amplitude of adjacent time scales is compared scale by scale to form an amplitude change record. The amplitude change record of the middle of the time range corresponding to the suspected crack segment is compared sequentially with the amplitude change record of the time scale before the corresponding time range and the amplitude change record of the time scale after the corresponding time range to form a continuous change path. Based on the list of suspected crack fragments, the corresponding time scale strain amplitude records are read from the strain change trajectory and arranged continuously in the order of time progression to form an instantaneous amplitude change sequence. The instantaneous amplitude change sequence is continuously connected according to the time progression and the time scale correspondence is maintained to form the crack release trajectory.
8. The method for identifying fatigue cracks in wind turbine blades based on distributed optical fiber strain response according to claim 7, characterized in that, When continuously comparing and organizing the amplitude change sequence around the time scale range corresponding to the suspected crack segment, the continuous connection relationship of the amplitude change direction in the time progression is maintained, and the amplitude change records of the corresponding time scale are rearranged step by step in the strain change trajectory, so that the instantaneous amplitude change process within the time range corresponding to the suspected crack segment forms a continuous change structure and runs through the complete time series.
9. The method for identifying fatigue cracks in wind turbine blades based on distributed optical fiber strain response according to claim 7, characterized in that, The steps involved in the formation of crack strengthening trajectories are as follows: By combining the crack release trajectory with the corresponding time scale range in the strain change trajectory, the strain amplitude records of the corresponding time scale are read sequentially along the time progression and arranged continuously to form a time series corresponding to the time range of the crack release trajectory. Extract strain amplitude records from crack release trajectory, organize them scale by scale, and form amplitude change sequence according to time order. Establish positional correspondence between each time scale in amplitude change sequence and the corresponding time scale in strain change trajectory. The amplitude change sequence is written into the time scale position corresponding to the strain change trajectory in the order of time progression, forming an amplitude change record of the crack release trajectory within the time range. The strain amplitude records at the front and back time scales in the amplitude change records and strain change trajectories are continuously arranged to form a crack strengthening trajectory while maintaining the time progression order.
10. The method for identifying fatigue cracks in wind turbine blades based on distributed optical fiber strain response according to claim 9, characterized in that, Based on the time scale range corresponding to the crack release trajectory, the strain amplitude record of the corresponding time scale is selected and replaced in the strain change trajectory at each scale. The strain amplitude record of the time scale before the replacement area is kept in a continuous arrangement with the strain amplitude record of the time scale after the replacement area. Under the condition that the time progression order is kept consistent, the amplitude change path of the replacement area is continuously connected, thereby forming a crack strengthening trajectory with a continuous amplitude change structure in the strain change trajectory.
