Underwater broadband acoustic tomography detection method for wind power pile foundation

Abstract

The application discloses a wind power pile foundation underwater broadband acoustic tomography imaging detection method and relates to the technical field of structure detection, and comprises the following steps: in the rising and falling tide alternation stage, a continuous echo time sequence is acquired through broadband acoustic tomography imaging detection, a time of arrival variation curve corresponding to the period is constructed, and a sound velocity variation mark representing the sound velocity variation trend with the tide is extracted from the time of arrival variation curve; the continuous echo time sequence is sequentially rearranged in the time of arrival around the sound velocity variation mark, a propagation path offset track corresponding to the sound velocity variation mark is formed, and a path offset scale is extracted according to the propagation path offset track. The application extracts the sound velocity variation mark and constructs the propagation path offset track, identifies the path mismatch entrance and the imaging adjustment starting point in sections, realizes the front control of the propagation deflection influence, disperses the time offset caused by the tide variation through the imaging time mapping rhythm rearrangement and the period expansion and contraction adjustment, and improves the imaging stability and the abnormal interpretation reliability.

Description

Technical Field

[0001] This invention relates to the field of structural inspection technology, specifically to a method for underwater broadband acoustic tomography detection of wind turbine pile foundations. Background Technology

[0002] Underwater broadband acoustic tomography detection of wind turbine pile foundations refers to the process of transmitting acoustic signals covering multiple frequency bands to the surrounding area and interior of the pile foundation using a broadband acoustic excitation device deployed in the water during the construction or operation phase of offshore wind turbine foundations. The acoustic waves propagate, reflect, and transmit between the water, the pile concrete, and interfaces with potential defects. A receiving device records the arrival time, energy attenuation, phase changes, and spectral characteristics. Subsequently, based on multiple sets of acoustic response data formed along different propagation paths, inversion calculations are performed to reconstruct the continuous distribution of materials within the pile foundation and the spatial location of abnormal areas, thus forming a two-dimensional or three-dimensional structural image similar to a tomographic scan. The core of this method lies in utilizing the differences in the response of broadband acoustic waves to defects of different scales. Through comprehensive analysis of multi-path propagation information, it achieves the visual identification of voids, cracks, cavities, or loose sections within the underwater wind turbine pile foundation. This represents a comprehensive application of underwater acoustic non-destructive testing and acoustic tomography technology.

[0003] The existing technology has the following shortcomings: In existing technologies, underwater broadband acoustic tomography of wind turbine pile foundations typically relies on a pre-established sound velocity distribution model for propagation path inversion calculations. However, at the ebb and flow of tides, water temperature and salinity distributions undergo short-term adjustments, and water density stratification is reconstructed accordingly, leading to changes in the sound velocity gradient and causing the sound wave propagation path to bend and shift. If the inversion process still uses the original propagation path model for temporal positioning and spatial mapping, the actual propagation distance will not match the calculated path. This can easily lead to misinterpretation of changes in echo arrival time as abnormalities in the pile's internal materials, resulting in images of clearly defined but actually non-existent internal cavities. This can mislead subsequent reinforcement decisions.

[0004] 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

[0005] The purpose of this invention is to provide an underwater broadband acoustic tomography detection method for wind turbine pile foundations to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: an underwater broadband acoustic tomography detection method for wind turbine pile foundations, comprising the following steps: During the alternation of high and low tides, continuous echo time series are obtained by broadband acoustic tomography, and arrival time variation curves for the corresponding time periods are constructed. Sound speed variation markers that characterize the trend of sound speed variation with tides are extracted from the arrival time variation curves. The arrival time sequence of the continuous echo is rearranged around the sound speed change marker to form the propagation path offset trajectory corresponding to the sound speed change marker, and the path offset scale is extracted based on the propagation path offset trajectory. By comparing and analyzing the preset propagation path model with the path offset scale, the time segment in which the propagation direction continues to deflect is determined, and the time segment in which the propagation direction continues to deflect is determined as the path mismatch entry point. Tracing back the corresponding continuous echo time series along the path mismatch entry, comparing the arrival time change curve with the sound speed change mark in the synchronous interval, extracting the synchronous change interval and determining the synchronous change interval as the imaging adjustment starting point; The imaging time mapping relationship is rearranged rhythmically around the imaging adjustment starting point, progressive time reversal processing is implemented within the synchronous change interval, and the time mapping parameters of the preset propagation path model are periodically stretched and adjusted to complete the dynamic imaging adjustment during the alternation of high and low tides.

[0007] Preferably, the steps for extracting sound speed change markers are as follows: Broadband acoustic tomography was used to detect and collect the arrival times of echoes from each propagation direction. The echoes were assigned a unified time identifier and arranged in chronological order to form a continuous echo time series. Based on the continuous echo time series, the arrival time difference between adjacent time markers is calculated according to the propagation direction, a time change gradient sequence is constructed and aligned to form an arrival time change curve; By comparing the time change gradients of different propagation directions under the same time marker around the arrival time change curve, splicing time intervals with the same direction to form a continuous trend segment and recording the corresponding time range and propagation direction set; The continuous trend segments are numbered and bound to the time range and the time change gradient amplitude to form sound speed change markers, which are arranged in chronological order and established to correspond with the continuous echo time series.

[0008] Preferably, the path offset scale extraction steps are as follows: Based on the time intervals corresponding to the changes in sound speed, the continuous echo time series is segmented and located to form a segmented continuous echo time series structure. Based on the segmented continuous echo time series structure, the arrival times of echoes in all propagation directions are sorted and assigned sequential numbers under the same time identifier, and a chain of sequential position changes after the propagation direction is rearranged is constructed. By combining the chain of sequential position changes after the propagation direction is rearranged, the sequence number offset segment is extracted, and multiple propagation directions are compared laterally within the same time interval to form the propagation path offset trajectory. The path offset scale is formed by statistically numbering the number of changes in the propagation path offset trajectory and then arranging the path offset scale in chronological order.

[0009] Preferably, the sequential number offset segments are merged based on the unidirectional changes of the sequential numbers in the continuous time identifier to form the propagation path offset trajectory segments. The propagation path offset trajectory segments include a time range and a set of propagation directions, and the total number of sequential number changes in each propagation direction within the propagation path offset trajectory segments is used as the path offset scale.

[0010] Preferably, the steps for determining the path mismatch entry point are as follows: Extract the time segment corresponding to the path offset scale and locate the same time range in the preset propagation path model. Retrieve the standard path sequence expression information of the corresponding propagation direction and compare it with the number of sequence number changes in the path offset scale. Based on the above comparison results, record the sequential numbering change direction of the same propagation direction under continuous time markers, and merge time segments with the same direction to form a segment with continuously changing sequential numbers; By combining the sequential numbering of all propagation directions and continuously changing segments, a horizontal comparison is made to select time segments in which multiple propagation directions are simultaneously in a state of continuous change within the same time range to form time segments in which the propagation direction continues to deflect, and the cumulative value of the corresponding path offset scale is calculated. The time nodes in the time intervals during which the propagation direction continuously deflects are extracted, and the first consecutive increases in the cumulative value of the path offset scale are identified as the path mismatch entry points and arranged in chronological order.

[0011] Preferably, when determining the path mismatch entry point, the starting time marker of the cumulative value of the path offset scale that continues to increase in one direction within the time interval of continuous deflection of the propagation direction is used as the path mismatch entry point, and the path mismatch entry point is bound and labeled with the corresponding propagation direction set and the corresponding time interval.

[0012] Preferably, the steps for determining the imaging adjustment starting point are as follows: The time marker corresponding to the path mismatch entry point is located and the continuous echo time series is traced back along the preceding time axis to form the traceback interval before the path mismatch entry point. Extract the arrival time change curves corresponding to the time markers around the backtracking interval, and retrieve the sound speed change markers covering the backtracking interval, and align them on a unified time axis; Based on the alignment results, the arrival time change curve and the sound speed change mark are compared point by point. Continuous time marks with the same splicing direction are formed to form a synchronous change interval and the start time mark and end time mark are recorded. Extract the start time identifier of the synchronous change interval and establish a time correspondence with the path mismatch entry point to determine the imaging adjustment starting point.

[0013] Preferably, the steps for rhythmically rearranging the imaging time mapping relationship around the imaging adjustment starting point, implementing progressive time reversal processing within the synchronous change interval, and periodically scaling and adjusting the time mapping parameters of the preset propagation path model are as follows: Locate the time marker corresponding to the starting point of the imaging adjustment and divide the reference segment and adjustment segment in the original imaging time mapping relationship, and record the correspondence between the time marker and the mapping sequence within the adjustment segment; The time markers are grouped around the adjustment section according to the time span of the synchronous change interval, and the mapping sequence within each group is alternately arranged to form a rhythmic rearrangement structure. By combining the rhythm rearrangement structure, the synchronous change interval is divided into continuous sub-segments, and the mapping order of each sub-segment is reversed to form a progressive time foldback structure. Based on the path offset scale, the time mapping parameters of the preset propagation path model within the adjustment section are periodically divided and the time interval is adjusted to complete the dynamic imaging adjustment of the ebb and flow phase.

[0014] The technical effects and advantages provided by the present invention in the above technical solution are as follows: This invention extracts sound velocity change markers and constructs propagation path offset trajectories during the alternating ebb and flow of tides, enabling a continuous expression of the sound velocity variation trend with the tide in both the time and propagation direction dimensions. Before imaging processing, it segments and quantifies the potential deflection process of the propagation path, thus avoiding directly mapping propagation time changes to anomalies in the pile's internal materials. By determining the path mismatch inlet and the imaging adjustment starting point in a layered manner, the impact of propagation path deflection on the imaging results is pre-identified and systematically mitigated, improving the imaging stability and consistency of underwater broadband acoustic tomography during tidal disturbance phases.

[0015] This invention implements rhythm rearrangement and progressive time reversal processing on the imaging time mapping relationship around the imaging adjustment starting point, and combines it with the periodic scaling adjustment of the time mapping parameters to form a dynamic adjustment structure for the time mapping expression during the ebb and flow of tides. This disperses the time offset caused by changes in the sound velocity field on the propagation path to multiple time segments for transition processing, thereby reducing the impact of concentrated accumulation of local time errors on the imaging spatial mapping, improving the accuracy of anomaly identification and the reliability of engineering interpretation, and enhancing the adaptability of the underwater detection process of wind power pile foundations in complex sea conditions. Attached Figure Description

[0016] 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.

[0017] Figure 1 This is a flowchart of the underwater broadband acoustic tomography detection method for wind turbine pile foundations according to the present invention. Detailed Implementation

[0018] 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.

[0019] This invention provides, for example Figure 1 The underwater broadband acoustic tomography method for detecting wind turbine pile foundations, as shown, includes the following steps: During the alternation of high and low tides, continuous echo time series are obtained by broadband acoustic tomography, and arrival time variation curves for the corresponding time periods are constructed. Sound speed variation markers that characterize the trend of sound speed variation with tides are extracted from the arrival time variation curves. In the marine environment, the aquatic environment during the ebb and flow of tides is constantly changing. The internal temperature and salinity distributions of the water body recombine rapidly, leading to continuous changes in the spatial distribution of sound velocity. By organizing acoustic response data in stages, focusing on the formation of continuous echo time series and the extraction of sound velocity change markers, the temporal trend of sound velocity variation with tides can be fully characterized. The specific implementation steps are as follows: Broadband acoustic tomography was continuously implemented during the ebb and flow of the tides. Acoustic signals covering different frequency ranges were emitted into the waters surrounding the pile foundation using a preset frequency band combination, creating multiple propagation paths in multiple directions. After the acoustic waves propagated and returned, the arrival time of the echoes corresponding to each propagation path was collected sequentially, and each echo arrival time was assigned a time marker under a unified time reference. The acquisition rhythm was kept consistent across multiple consecutive sampling periods, ensuring that each propagation direction had a corresponding echo arrival time record under the same time marker. Subsequently, the echo arrival times of all propagation directions were arranged vertically according to the order of the time markers, forming a continuous echo time series covering the entire ebb and flow of the tides. During the formation of the continuous echo time series, an independent sequence was established for each propagation direction, ensuring that the echo arrival times of the same propagation direction under different time markers form a continuously changing chain. Simultaneously, it was ensured that different propagation directions could form a horizontal alignment relationship at the same time marker, thus maintaining both temporal continuity and information correlation along the propagation direction dimension in the continuous echo time series.

[0020] After the continuous echo time series is formed, the arrival times of echoes in the same propagation direction are compared point by point under adjacent time markers, based on the propagation direction. The difference in arrival time between adjacent time markers is calculated, and these differences are arranged in chronological order to form a time variation gradient sequence. After the time variation gradient sequence for a single propagation direction is constructed, the same process is repeated for all propagation directions to form an independent time variation gradient sequence for each propagation direction. Subsequently, the time variation gradient sequences of each propagation direction are uniformly mapped to the same time axis and aligned one by one according to the time markers, so that the time variation gradients of different propagation directions under the same time markers can be compared laterally. On this basis, the time variation gradients corresponding to each time marker on the time axis are continuously connected to form the arrival time variation curve for the corresponding time period. During the formation of the arrival time variation curve, the positions of the high tide stage and the low tide stage on the time axis are marked so that the arrival time variation curve fully reflects the continuous change of echo arrival time during the alternation of high and low tides in the time dimension.

[0021] After the arrival time variation curve is formed, the time variation gradients of different propagation directions under the same time marker are compared centrally to filter out time intervals in which consistent changes occur simultaneously in multiple propagation directions. During the filtering process, the directions of the time variation gradients within each time interval are classified, and time intervals with consistent directions are continuously spliced ​​together to form trend continuous segments. Subsequently, the start and end positions of the trend continuous segments on the time axis are marked, and the trend continuous segments are correlated with the time nodes of the ebb and flow of tides, thus establishing a time correlation between the trend continuous segments and the time periods of tidal changes. After completing the time positioning of the trend continuous segments, the amplitude of the time variation gradient in each propagation direction within the trend continuous segments is recorded one by one, so that each trend continuous segment contains information on time range, propagation direction set, and time variation gradient amplitude, thereby forming a basic data set representing the trend of sound speed changing with the tide.

[0022] After obtaining the basic data set of the aforementioned continuous trend segments, each continuous trend segment is uniformly numbered, and the number is bound to the corresponding time range, propagation direction set, and time change gradient amplitude information to form a sound speed change marker. Subsequently, all sound speed change markers are arranged in time axis order to form a continuous sound speed change marker sequence, so that the sound speed change trend of the entire ebb and flow phase forms a complete chain in the time dimension. After the sound speed change marker sequence is formed, it is matched with the continuous echo time series, so that the echo arrival time under any time marker can find the corresponding change segment in the sound speed change marker sequence. Through this time correspondence, each arrival time change in the continuous echo time series has a clear sound speed change attribute, thereby completing the systematic extraction of the sound speed change trend with the tide during the ebb and flow phase, and providing a continuous, traceable, and structurally complete time basis for subsequent time rearrangement and propagation path offset analysis based on the sound speed change markers.

[0023] The arrival time sequence of the continuous echo is rearranged around the sound speed change marker to form the propagation path offset trajectory corresponding to the sound speed change marker, and the path offset scale is extracted based on the propagation path offset trajectory. After the sound speed variation marker sequence has been formed and established a temporal correspondence with the continuous echo time series, in order to transform the sound speed variation trend with the tide into a temporal variation expression at the propagation path level, the continuous echo time series is partitioned and sequentially reconstructed. This allows the arrival time variation within the time domain to form a continuous offset trajectory in the propagation direction dimension, which is further quantified as a path offset scale. The specific implementation steps are as follows: Based on the time interval corresponding to each sound speed change marker in the sound speed change marker sequence, the continuous echo time series is segmented and located. All echo arrival time records belonging to the same sound speed change marker time interval are extracted, while keeping the time identifier and propagation direction index information corresponding to each echo arrival time record unchanged. After extracting a single sound speed change marker time interval, the echo arrival time records within that time interval are arranged according to the time identifier order, so that echo arrival times under the same time identifier but different propagation directions are in the same horizontal position. Subsequently, the time intervals between adjacent sound speed change markers are segmented and arranged in the same way, so that the continuous echo time series of the entire ebb and flow phase is divided into multiple time-continuous and clearly defined segments. Each segment is bounded by a sound speed change marker, thus constructing a segmented continuous echo time series structure with sound speed change markers as boundaries.

[0024] Within each segmented time interval, the arrival times of echoes from all propagation directions under the same time marker are centrally sorted. Specifically, the arrival time values ​​of echoes from all propagation directions are extracted under a fixed time marker, rearranged in chronological order from earliest to latest, and assigned sequential numbers to the rearranged positions, with the sequential numbers increasing continuously from the beginning. After assigning the sequential numbers, the sequential numbers are bound to the original propagation direction index, so that any propagation direction has both the original sequential position and the rearranged sequential position under that time marker. Then, along the time axis, the rearranged sequential positions of the same propagation direction under consecutive time markers are connected point by point, so that the rearranged sequential positions form a continuous chain of change in the time dimension. By repeating the above operation for all propagation directions, each propagation direction forms an independent chain of changes in the rearranged sequential positions, thereby completing the sequential rearrangement of the arrival time of the continuous echo time series guided by the sound speed change marker.

[0025] After the rearranged sequence of positions is formed, the changes in sequence numbers between adjacent time markers in the same propagation direction are compared one by one. The time nodes where the sequence numbers change are extracted, and the difference in sequence numbers before and after the change is recorded. Segments with the same direction of sequence number change in multiple consecutive time markers are merged to form sequence number offset segments. Within each sequence number offset segment, the number of sequence number changes is accumulated and recorded, along with the start and end time markers of the segment. Subsequently, the sequence number offset segments in all propagation directions are compared horizontally to select the time range in which multiple propagation directions simultaneously have sequence number offset segments within the same time interval. This time range is then bound to the corresponding set of propagation directions to form the propagation path offset trajectory. The propagation path offset trajectory thus contains time range information, propagation direction set information, and sequence number change information, allowing the offset process of the propagation path in the time dimension to be continuously expressed.

[0026] After determining the propagation path offset trajectory, the number of sequence number changes within each propagation path offset trajectory segment is quantified, and the total number of sequence number changes for the same propagation direction within that segment is recorded as a unidirectional change value. Subsequently, the unidirectional change values ​​for all propagation directions belonging to the same propagation path offset trajectory segment are summarized to form a comprehensive change value for the corresponding time segment. This comprehensive change value is used as the numerical expression of the path offset scale, and the path offset scale is associated and labeled with the corresponding time segment and the corresponding propagation direction set. Then, all path offset scales are arranged in chronological order, so that the path offset scale forms a continuous change sequence throughout the ebb and flow of the tide. Through the temporal correspondence between the path offset scale sequence and the sound speed change marker sequence, the trend of sound speed change with the tide can be quantitatively expressed at the propagation path level. This completes the entire transformation process from continuous echo time series to arrival time sequence rearrangement, and then to propagation path offset trajectory and path offset scale extraction, providing a temporally continuous, hierarchically clear, and fully expressive basic data for subsequent propagation path model comparative analysis based on the path offset scale.

[0027] By comparing and analyzing the preset propagation path model with the path offset scale, the time segment in which the propagation direction continues to deflect is determined, and the time segment in which the propagation direction continues to deflect is determined as the path mismatch entry point. After the path offset scale sequence has been arranged chronologically, to ensure that the path offset scale forms a comparable deflection expression within the framework of the preset propagation path model, a layered approach is taken regarding the temporal and propagation direction relationships between the path offset scale and the preset propagation path model. Through continuous time mapping and sequential comparison, the time segment of continuous propagation direction deflection is fully defined, and this time segment is further identified as the path mismatch entry point. The specific implementation steps are as follows: Each time segment corresponding to a path offset scale in the path offset scale sequence is extracted one by one, and a time range completely consistent with that time segment is located in the preset propagation path model. After the time range is located, all propagation direction indices involved in the time segment are extracted one by one and arranged in order of propagation direction number. Then, the standard path order expression information of the corresponding propagation direction in the initial modeling state is retrieved from the preset propagation path model, so that the same propagation direction has both path offset scale data and standard path order expression data in the same time segment. On this basis, the number of changes in the order number corresponding to the path offset scale is arranged side by side with the original order position in the standard path order expression, so that each propagation direction forms a comparison data table between the standard order position and the rearranged order position in the same time segment. Through the above dual mapping processing of time and propagation direction, the path offset scale can obtain a clear reference position in the preset propagation path model framework.

[0028] After completing the comparison data arrangement between the path offset scale and the preset propagation path model, the direction of change of the sequential number under continuous time markers for the same propagation direction is recorded one by one. When the sequential number changes in the same direction within three or more consecutive time markers, the time period is marked as a unidirectional deflection segment, and the start and end time markers of the unidirectional deflection segment are recorded. Subsequently, the unidirectional deflection segments of the same propagation direction within the entire path offset scale sequence time range are summarized and organized, and the unidirectional deflection segments that are adjacent in time and have the same direction of change are merged, so that the sequential number change of the propagation direction on the time axis forms a complete continuous segment. Through this process, each propagation direction forms a corresponding continuous sequential number change segment, providing a continuous time basis for the identification of continuous deflection of the propagation direction.

[0029] After a single propagation direction continuously changing segment is formed, all propagation direction continuously changing segments are compared horizontally to filter out time segments where two or more propagation directions show continuously changing sequential numbers within the same time range. During the filtering process, point-by-point comparison is performed using time markers as units. When multiple propagation directions are all within a continuously changing sequential number segment at the same time marker, that time marker is included in the candidate time interval. Subsequently, multiple consecutive candidate time markers are spliced ​​together to form a time segment where the propagation direction continuously deviates. After forming a time segment where the propagation direction continuously deviates, the path offset scale values ​​corresponding to each propagation direction within the time segment are accumulated and recorded, and the accumulated values ​​are bound to the time segment. This gives the time segment where the propagation direction continuously deviates both time range attributes and path offset scale accumulation attributes, thereby quantitatively expressing the overall deviation state of the propagation path in the time dimension.

[0030] After determining the time intervals of continuous propagation direction deflection, the time nodes at which the cumulative path offset scale value first increases continuously within each time interval are extracted and used as the starting nodes for the path to transition from the standard state to the deflection state. These starting nodes are defined as path mismatch entry points, and the path mismatch entry points are bound and recorded with the corresponding propagation direction set and the corresponding time interval. Subsequently, all path mismatch entry points are arranged in chronological order, forming a continuous sequence of identifiers on the time axis. Through the temporal correspondence between the path mismatch entry point sequence and the preset propagation path model, the time intervals of continuous propagation direction deflection can be accurately located within the framework of the preset propagation path model. This completes the entire transformation process from path offset scale to continuous propagation direction deflection time intervals and then to the determination of path mismatch entry points, providing clear temporal boundaries and propagation direction basis for subsequent echo time series backtracking and imaging adjustment starting point extraction around the path mismatch entry points.

[0031] Tracing back the corresponding continuous echo time series along the path mismatch entry, comparing the arrival time change curve with the sound speed change mark in the synchronous interval, extracting the synchronous change interval and determining the synchronous change interval as the imaging adjustment starting point; After the path mismatch ingress has been identified by time, to ensure that the propagation path deflection has an adjustable starting reference in the time dimension, the continuous echo time series is traced back point by point around the time node corresponding to the path mismatch ingress. The arrival time variation curve and the sound velocity variation marker are compared item by item on a unified time axis to extract the synchronous variation interval and further determine the imaging adjustment starting point. The specific implementation steps are as follows: Using the time marker corresponding to the path mismatch entry point as the starting position for backtracking, the echo arrival time record corresponding to this time marker is located in the continuous echo time series, and the data is expanded point by point in the preceding direction of the time axis with this time marker as the center. During the expansion process, the echo arrival time data corresponding to the preceding time marker is retrieved one by one in descending order of the time marker, while maintaining the consistent association between each echo arrival time record and the corresponding propagation direction index. While backtracking point by point, the echo arrival times of all propagation directions under each time marker are arranged in a centralized manner, so that each propagation direction is in the same horizontal position under the same time marker. Then, the backtracking range is limited to a continuous number of time markers before the path mismatch entry point, so that the continuous echo time series before the path mismatch entry point forms a complete backtracking interval, establishing a time basis for the subsequent synchronous arrangement with the arrival time change curve.

[0032] After determining the backtracking interval of the continuous echo time series, the arrival time variation curve data corresponding one-to-one with the time markers of the backtracking interval are extracted and arranged vertically according to the same time marker order, so that the continuous echo time series and arrival time variation curves within the backtracking interval are aligned point by point on the time axis. After alignment, the sound speed variation markers covering the time range of the backtracking interval are extracted one by one from the sound speed variation marker sequence and arranged next to the arrival time variation curve data according to the time marker order, so that there is a correspondence between the continuous echo time series data, the arrival time variation curve data, and the sound speed variation marker data under each time marker. Through the unified arrangement of the above three, the changes in echo arrival time and the expression of sound speed variation trend can be observed simultaneously at any time marker on the time axis, thus providing a complete data foundation for synchronous interval comparison.

[0033] After the three times are aligned, the arrival time change curve at each time marker is compared point by point with the change direction expressed by the corresponding sound speed change mark, using the time marker as the smallest comparison unit. When the arrival time change curve maintains the same direction of change within three or more consecutive time markers, and this direction of change is consistent with the change direction expressed by the sound speed change mark, the continuous time marker segment is recorded as a candidate synchronization interval. For the candidate synchronization interval, the change process of the arrival time change amplitude within it is further observed point by point. The time period in which the change amplitude is continuously increasing or decreasing within the continuous time marker is extracted, and this time period is compared again with the change segment of the corresponding sound speed change mark. When the change direction and change process are consistent throughout the entire time period, the time period is officially determined as the synchronization change interval, and its start time marker and end time marker are recorded, so that the synchronization change interval has a clear boundary on the time axis.

[0034] After determining the synchronous change intervals, the start time identifier of each synchronous change interval is extracted, and the time interval between the start time identifier and the time identifier of the path mismatch entry is marked. When the synchronous change interval is located before the path mismatch entry and there is no change segment in the opposite direction between it and the path mismatch entry, the start time identifier of the synchronous change interval is defined as the imaging adjustment starting point. After the imaging adjustment starting point is determined, the continuous echo time series data, arrival time change curve data, and sound velocity change marker data under the time identifier corresponding to the imaging adjustment starting point are centrally labeled, so that the imaging adjustment starting point has a complete data correspondence in the time dimension and the propagation direction dimension. By determining the imaging adjustment starting point, the propagation path deflection forms a clear adjustment starting position at the imaging time mapping level, thereby completing the continuous organization process from the path mismatch entry back to the extraction of the synchronous change interval and then to the determination of the imaging adjustment starting point, providing a clear time boundary and change basis for the subsequent rearrangement of the imaging time mapping rhythm around the imaging adjustment starting point.

[0035] The imaging time mapping relationship is rearranged rhythmically around the imaging adjustment starting point, progressive time reversal processing is implemented within the synchronous change interval, and the time mapping parameters of the preset propagation path model are periodically stretched and adjusted to complete the dynamic imaging adjustment during the alternating tide stage. After the imaging adjustment starting point has been time-positioned and established with the synchronous change interval, in order to ensure a controllable rhythmic change in the imaging time mapping relationship during the ebb and flow phases, a segmented time organization is implemented around the imaging adjustment starting point. This is achieved through time mapping rhythm rearrangement, progressive time foldback processing within the synchronous change interval, and periodic scaling adjustments to the time mapping parameters of the preset propagation path model, resulting in a continuous transition in the imaging representation over time. The specific implementation steps are as follows: Using the time marker corresponding to the imaging adjustment start point as a benchmark, the mapping sequence corresponding to this time marker is located in the original imaging time mapping relationship. This mapping sequence is then used as a boundary to divide the entire time axis into a benchmark segment and an adjustment segment. The benchmark segment encompasses the entire range of time markers before the imaging adjustment start point, while the adjustment segment encompasses the entire range of time markers from the imaging adjustment start point to the path mismatch entry point. After segmentation, the mapping sequence of each time marker within the adjustment segment in the original imaging time mapping relationship is recorded line by line, forming a lookup table of time markers and mapping sequences. Based on this, the time markers within the adjustment segment are continuously grouped according to the time span of the synchronous change interval. Each group contains a fixed number of continuous time markers, allowing the adjustment segment to form multiple continuous sub-segments on the time axis, thus establishing a clear temporal structure boundary for subsequent rhythm rearrangement.

[0036] After grouping the adjustment segments, the mapping sequence corresponding to each group's time markers is rearranged within the group. Specifically, while maintaining the time order between groups, the mapping sequence within each group is reordered according to a preset alternating arrangement rule, so that the mapping sequence, which was originally increasing in time order, forms an alternating arrangement structure within the group. During the rearrangement process, each time marker is re-bound to a new mapping sequence, forming an updated table of correspondence between time markers and mapping sequences. Subsequently, the same intra-group rearrangement is performed on all groups in sequence, so that the imaging time mapping relationship within the entire adjustment segment forms a rhythmic variation structure. Through the above rhythmic rearrangement, the time mapping sequence alternates within local segments, thereby reducing the concentrated accumulation of path deflection in the imaging time expression.

[0037] After the rhythm rearrangement is completed, a progressive time reversal process is implemented within the synchronous change interval. Specifically, the time markers within the synchronous change interval are divided into multiple consecutive sub-segments according to a fixed number, and the mapping sequence of each sub-segment is reversed so that the mapping sequence within the sub-segment corresponds in reverse time. After completing the reverse arrangement of the first sub-segment, it is spliced ​​with the forward arrangement of the next sub-segment, so that the order direction of the two sub-segments is switched. Then, the third sub-segment is reversed again, so that multiple sub-segments form a forward and reverse alternating arrangement structure. After completing the above progressive time reversal process within the entire synchronous change interval, the time mapping relationship within the synchronous change interval presents a continuous back-and-forth change pattern, thereby releasing the time mapping offset caused by the change in sound speed within this interval in segments, so that the imaging time expression forms a smooth transition on the time axis.

[0038] After completing the rhythm rearrangement and progressive time rewinding processing, the time mapping parameters of the preset propagation path model are periodically scaled. Specifically, using the time length of the synchronous change interval as the period unit, the adjustment segment is divided into several period segments, each covering a fixed number of continuous time markers. Within each period segment, the time interval corresponding to the time mapping parameters is proportionally adjusted based on the cumulative change value of the path offset scale within the corresponding time segment, so that some time intervals are extended and some time intervals are compressed within the period segment. After completing the time interval adjustment of a single period segment, the time intervals of multiple period segments are... The mapping parameters are sequentially spliced ​​in chronological order, forming a periodic scaling structure for the time mapping parameters within the adjustment segment. Through periodic scaling adjustment, the time mapping expression of the preset propagation path model during the ebb and flow phase can be periodically corrected according to the trend of sound speed change. This completes the continuous process of rearranging the rhythm of the imaging time mapping relationship around the imaging adjustment starting point, progressive time reversal processing within the synchronous change interval, and periodic scaling adjustment of the time mapping parameters. This achieves dynamic imaging adjustment during the ebb and flow phase, ensuring that the imaging expression remains continuous in the time dimension and dispersing the time mapping offset effect caused by propagation path deflection.

[0039] This invention extracts sound velocity change markers and constructs propagation path offset trajectories during the alternating ebb and flow of tides, enabling a continuous expression of the sound velocity variation trend with the tide in both the time and propagation direction dimensions. Before imaging processing, it segments and quantifies the potential deflection process of the propagation path, thus avoiding directly mapping propagation time changes to anomalies in the pile's internal materials. By determining the path mismatch inlet and the imaging adjustment starting point in a layered manner, the impact of propagation path deflection on the imaging results is pre-identified and systematically mitigated, improving the imaging stability and consistency of underwater broadband acoustic tomography during tidal disturbance phases.

[0040] This invention implements rhythm rearrangement and progressive time reversal processing on the imaging time mapping relationship around the imaging adjustment starting point, and combines it with the periodic scaling adjustment of the time mapping parameters to form a dynamic adjustment structure for the time mapping expression during the ebb and flow of tides. This disperses the time offset caused by changes in the sound velocity field on the propagation path to multiple time segments for transition processing, thereby reducing the impact of concentrated accumulation of local time errors on the imaging spatial mapping, improving the accuracy of anomaly identification and the reliability of engineering interpretation, and enhancing the adaptability of the underwater detection process of wind power pile foundations in complex sea conditions.

[0041] 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 underwater broadband acoustic tomography detection of wind turbine pile foundations, characterized in that, Includes the following steps: During the ebb and flow of the tide, continuous echo time series are obtained by broadband acoustic tomography, and arrival time variation curves for the corresponding time periods are constructed. Sound speed variation markers that characterize the trend of sound speed variation with the tide are extracted from the arrival time variation curves. The arrival time sequence of the continuous echo is rearranged around the sound speed change marker to form the propagation path offset trajectory corresponding to the sound speed change marker, and the path offset scale is extracted based on the propagation path offset trajectory. By comparing and analyzing the preset propagation path model with the path offset scale, the time segment in which the propagation direction continues to deflect is determined, and the time segment in which the propagation direction continues to deflect is determined as the path mismatch entry point. Tracing back the corresponding continuous echo time series along the path mismatch entry, comparing the arrival time change curve with the sound speed change mark in the synchronous interval, extracting the synchronous change interval and determining the synchronous change interval as the imaging adjustment starting point; The imaging time mapping relationship is rearranged rhythmically around the imaging adjustment starting point, progressive time reversal processing is implemented within the synchronous change interval, and the time mapping parameters of the preset propagation path model are periodically stretched and adjusted to complete the dynamic imaging adjustment during the alternation of high and low tides.

2. The underwater broadband acoustic tomography detection method for wind turbine pile foundations according to claim 1, characterized in that, The steps for extracting sound speed change markers are as follows: Broadband acoustic tomography was used to detect and collect the arrival times of echoes from each propagation direction. The echoes were assigned a unified time identifier and arranged in chronological order to form a continuous echo time series. Based on the continuous echo time series, the arrival time difference between adjacent time markers is calculated according to the propagation direction, a time change gradient sequence is constructed and aligned to form an arrival time change curve; By comparing the time change gradients of different propagation directions under the same time marker around the arrival time change curve, splicing time intervals with the same direction to form a continuous trend segment and recording the corresponding time range and propagation direction set; The continuous trend segments are numbered and bound to the time range and the time change gradient amplitude to form sound speed change markers, which are arranged in chronological order and established to correspond with the continuous echo time series.

3. The underwater broadband acoustic tomography detection method for wind turbine pile foundations according to claim 2, characterized in that, The steps for extracting path offset ticks are as follows: Based on the time intervals corresponding to the changes in sound speed, the continuous echo time series is segmented and located to form a segmented continuous echo time series structure. Based on the segmented continuous echo time series structure, the arrival times of echoes in all propagation directions are sorted and assigned sequential numbers under the same time identifier, and a chain of sequential position changes after the propagation direction is rearranged is constructed. By combining the chain of sequential position changes after the propagation direction is rearranged, the sequence number offset segment is extracted, and multiple propagation directions are compared laterally within the same time interval to form the propagation path offset trajectory. The path offset scale is formed by statistically numbering the number of changes in the propagation path offset trajectory and then arranging the path offset scale in chronological order.

4. The underwater broadband acoustic tomography detection method for wind turbine pile foundations according to claim 3, characterized in that, The sequential number offset segment is formed by merging the sequential number unidirectional changes in the continuous time identifier to form the propagation path offset trajectory segment. The propagation path offset trajectory segment includes a time range and a set of propagation directions, and the total number of sequential number changes in each propagation direction within the propagation path offset trajectory segment is used as the path offset scale.

5. The underwater broadband acoustic tomography detection method for wind turbine pile foundations according to claim 3, characterized in that, The steps for determining the path mismatch entry point are as follows: Extract the time segment corresponding to the path offset scale and locate the same time range in the preset propagation path model. Retrieve the standard path sequence expression information of the corresponding propagation direction and compare it with the number of sequence number changes in the path offset scale. Based on the above comparison results, record the sequential numbering change direction of the same propagation direction under continuous time markers, and merge time segments with the same direction to form a segment with continuously changing sequential numbers; By combining the sequential numbering of all propagation directions and continuously changing segments, a horizontal comparison is made to select time segments in which multiple propagation directions are simultaneously in a state of continuous change within the same time range to form time segments in which the propagation direction continues to deflect, and the cumulative value of the corresponding path offset scale is calculated. Extract the time nodes in the time interval where the propagation direction continuously deflects, and determine them as the entry points of path mismatch, arranging them in chronological order.

6. The underwater broadband acoustic tomography detection method for wind turbine pile foundations according to claim 5, characterized in that, When determining the path mismatch entry point, the starting time of the cumulative path offset scale value that continues to increase in one direction within the time interval of continuous propagation direction deflection is taken as the path mismatch entry point, and the path mismatch entry point is bound and labeled with the corresponding propagation direction set and the corresponding time interval.

7. The underwater broadband acoustic tomography detection method for wind turbine pile foundations according to claim 5, characterized in that, The steps for determining the imaging adjustment starting point are as follows: The time marker corresponding to the path mismatch entry point is located and the continuous echo time series is traced back along the preceding time axis to form the traceback interval before the path mismatch entry point. Extract the arrival time change curves corresponding to the time markers around the backtracking interval, and retrieve the sound speed change markers covering the backtracking interval, and align them on a unified time axis; Based on the alignment results, the arrival time change curve and the sound speed change mark are compared point by point. Continuous time marks with the same splicing direction are formed to form a synchronous change interval and the start time mark and end time mark are recorded. Extract the start time identifier of the synchronous change interval and establish a time correspondence with the path mismatch entry point to determine the imaging adjustment starting point.

8. The underwater broadband acoustic tomography detection method for wind turbine pile foundations according to claim 7, characterized in that, The imaging time mapping relationship is rhythmically rearranged around the imaging adjustment starting point, progressive time rewinding is implemented within the synchronous change interval, and the time mapping parameters of the preset propagation path model are periodically scaled and adjusted as follows: Locate the time marker corresponding to the starting point of the imaging adjustment and divide the reference segment and adjustment segment in the original imaging time mapping relationship, and record the correspondence between the time marker and the mapping sequence within the adjustment segment; The time markers are grouped around the adjustment section according to the time span of the synchronous change interval, and the mapping sequence within each group is alternately arranged to form a rhythmic rearrangement structure. By combining the rhythm rearrangement structure, the synchronous change interval is divided into continuous sub-segments, and the mapping order of each sub-segment is reversed to form a progressive time foldback structure. Based on the path offset scale, the time mapping parameters of the preset propagation path model within the adjustment section are periodically divided and the time interval is adjusted to complete the dynamic imaging adjustment of the ebb and flow phase.

Images