Underwater pile foundation state detection method and system

Abstract

The invention discloses an underwater pile foundation state detection method and system, and the method comprises the following steps: determining an underwater pile foundation detection region, and arranging an annular sonar array around an underwater pile foundation in the detection region; the annular sonar array is controlled to scan around the underwater pile foundation, and echo signals are returned after scanning; collecting an echo signal, and processing the echo signal to obtain a modeling signal; based on the modeling signal, a three-dimensional sound field model around the underwater pile foundation is established, and various underwater pile foundation state related parameters are calculated according to the three-dimensional sound field model around the underwater pile foundation; and the various underwater pile foundation state related parameters obtained through multiple times of scanning are fused, various final underwater pile foundation state related parameters are obtained, and a pile foundation state detection report is generated based on the various final underwater pile foundation state related parameters. The method has the advantages of being wide in detection range, high in precision, high in adaptability and capable of achieving underwater pile foundation detection of multi-parameter synchronous monitoring, the structural safety of the underwater pile foundation can be effectively guaranteed, and the operation and maintenance cost is reduced.

Description

Technical Field

[0001] This invention relates to the field of marine engineering testing technology, and in particular to a method and system for underwater pile foundation condition testing. Background Technology

[0002] With the rapid development of the wind power industry, the number of offshore and inland river wind power projects is constantly increasing. As the core support structure of wind power equipment, underwater pile foundations are in a complex hydrological environment for a long time and are easily affected by factors such as water flow erosion, siltation, tidal action and wave impact, resulting in problems such as underwater pile foundation erosion, uneven siltation around the pile foundation and underwater pile foundation tilting, which seriously threaten the structural safety and stable operation of wind power equipment.

[0003] Relevant underwater pile foundation testing technologies mainly include underwater robot testing, sonar single-point testing, and ultrasonic testing. Among these, underwater robot testing requires manual operation, resulting in low efficiency, high cost, and a risk of collision in complex water flow environments. Sonar single-point testing can only detect localized areas and cannot comprehensively cover the area surrounding the underwater pile foundation, limiting its detection range. Ultrasonic testing is significantly affected by water quality, with accuracy dropping drastically in turbid water, and it cannot simultaneously monitor erosion, sedimentation, and tilting. Furthermore, these technologies mostly employ single-scan methods, making them susceptible to environmental interference and leading to large detection errors, which cannot meet the high-precision requirements for long-term underwater pile foundation monitoring. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to address at least one defect of the related technologies mentioned in the background: the underwater pile foundation detection range of related wind power underwater pile foundation detection technologies is limited, the detection accuracy is low, and multi-parameter synchronous monitoring cannot be achieved. The present invention provides an underwater pile foundation condition detection method and system based on a ring sonar array.

[0005] The technical solution adopted by this invention to solve its technical problem is: to provide a method for detecting the condition of underwater pile foundations, comprising the following steps: S1: Determine the underwater pile foundation testing area, and deploy a ring-shaped sonar array around the underwater pile foundation within the testing area; S2: Control the ring sonar array to scan around the underwater pile foundation, and return the echo signal after scanning; S3: Acquire echo signals, process the echo signals, and obtain modeling signals; S4: Based on the modeling signal, establish a three-dimensional acoustic field model around the underwater pile foundation, and calculate various underwater pile foundation state-related parameters based on the three-dimensional acoustic field model around the underwater pile foundation. S5: The various underwater pile foundation state-related parameters obtained from multiple scans are fused to obtain multiple final underwater pile foundation state-related parameters, and a pile foundation state inspection report is generated based on these multiple final underwater pile foundation state-related parameters.

[0006] In some embodiments, controlling a ring-shaped sonar array to scan around an underwater pile foundation includes: Each scan uses at least two different frequency combinations. The scanning method involves scanning around the underwater pile foundation in both clockwise and counterclockwise directions. The scanning interval does not exceed a preset time, and the number of scans is not less than a preset number.

[0007] In some embodiments, step S3 involves processing the echo signal, including: The echo signal is preprocessed, enhanced, and its features are extracted to remove interference signals from water flow, bubbles, and marine organisms.

[0008] In some embodiments, step S2 further includes: controlling the ring sonar array to monitor environmental parameters in real time; Preprocessing of the echo signal includes: An adaptive filtering algorithm is used to eliminate dynamic noise; this includes: using environmental parameters as input, establishing a noise prediction model, dynamically adjusting the filter gain, and performing targeted filtering on echo signals of different frequencies.

[0009] In some embodiments, signal enhancement of the echo signal includes: The preprocessed echo signal is decomposed several times using wavelet transform algorithm. Each decomposition yields a set of low-frequency coefficients and a set of high-frequency coefficients, where the low-frequency coefficients correspond to the overall contour and the high-frequency coefficients correspond to local details. The low-frequency coefficients of the array are enhanced by threshold shrinkage to suppress residual noise; the high-frequency coefficients of the array are amplified by coefficient reconstruction to highlight the interface abrupt change characteristics.

[0010] In some embodiments, the step of performing feature extraction processing on the echo signal includes: A deep learning model is introduced to convert the enhanced echo signal into a two-dimensional time-frequency map, which is then input into a trained CNN network to extract spatial features from the two-dimensional time-frequency map. The feature vectors output by the CNN network are then input into the LSTM network to capture the time-series features of the signal and obtain the modeled signal.

[0011] In some embodiments, step S4 includes: Using the center of the flange at the top of the underwater pile foundation as the origin, a spatial rectangular coordinate system is established. The echo signal is converted into three-dimensional cloud points in the spatial rectangular coordinate system. Then, by comparing the weighted fusion of the features of the modeling signals at different frequencies, the identification of the siltation interface, the basement rock layer interface, and the surface contour of the underwater pile foundation is achieved. Based on the three-dimensional cloud points, a three-dimensional acoustic field model around the underwater pile foundation is established by combining the siltation interface, the basement rock layer interface, and the surface contour of the underwater pile foundation. Various underwater pile foundation state-related parameters are calculated based on the three-dimensional acoustic field model around the underwater pile foundation.

[0012] In some embodiments, step S4, which calculates various underwater pile foundation state-related parameters based on the three-dimensional acoustic field model surrounding the underwater pile foundation, includes: Calculate the underwater pile foundation siltation thickness; this includes: at each preset angle of the ring sonar array 1, calculate the vertical distance from the water surface to the siltation interface in the vertical direction to obtain the underwater pile foundation siltation thickness; Calculate the erosion depth of underwater pile foundations; this includes: obtaining the erosion depth of underwater pile foundations by calculating the difference between the interface of the base rock layer and the designed base elevation; Calculate the inclination angle of the underwater pile foundation; it includes: extracting multiple horizontal sections at preset intervals along the height direction of the underwater pile foundation, uniformly selecting multiple sampling points in each horizontal section, fitting the center coordinates of each horizontal section using the least squares method, connecting the multiple center coordinates to obtain the actual axis of the underwater pile foundation, and calculating the angle between the actual axis and the vertical line in the spatial rectangular coordinate system to obtain the inclination angle of the underwater pile foundation. The parameters related to the final underwater pile foundation status in step S5 are the final underwater pile foundation siltation thickness, the final underwater pile foundation erosion depth, and the final underwater pile foundation tilt angle.

[0013] In some embodiments, a pile foundation condition inspection report is generated based on various parameters related to the final underwater pile foundation condition, including: Multiple parameters related to the final underwater pile foundation status are compared with preset alarm thresholds. If any parameter related to the final underwater pile foundation status exceeds its preset alarm threshold, an early warning message is triggered. The pile foundation status detection report includes early warning information, a visualized three-dimensional sound field model around the underwater pile foundation, and change curves of multiple underwater pile foundation status parameters.

[0014] The present invention also provides an underwater pile foundation condition detection system, comprising: The detection determination module is used to determine the detection area of ​​underwater pile foundations; A ring-shaped sonar array is deployed around the underwater pile foundation within the detection area. The scanning control module is used to control the ring sonar array to scan around the underwater pile foundation and return echo signals after scanning; The signal acquisition and processing module is used to acquire echo signals, process the echo signals, and obtain modeling signals. The visualization and calculation module is used to establish a three-dimensional acoustic field model around the underwater pile foundation based on the modeling signal, and to calculate various underwater pile foundation state-related parameters based on the three-dimensional acoustic field model around the underwater pile foundation. The data fusion and analysis module is used to fuse various underwater pile foundation state-related parameters obtained from multiple scans to obtain various final underwater pile foundation state-related parameters, and generate a pile foundation state inspection report based on these parameters.

[0015] By implementing this invention, the following beneficial effects are achieved: This invention has the advantages of wide detection range, high accuracy, strong adaptability, and the ability to simultaneously monitor various underwater pile foundation status parameters, which can effectively ensure the structural safety of underwater pile foundations and reduce operation and maintenance costs. Attached Figure Description

[0016] The present invention will be further described below with reference to the accompanying drawings and embodiments. In the accompanying drawings: Figure 1 A flowchart of an embodiment of the underwater pile foundation condition detection method of the present invention is shown; Figure 2 A top view of a ring sonar array is shown in an embodiment of the underwater pile foundation condition detection system of the present invention; Figure 3 A front view of a height-adjustable mounting structure used in a ring sonar array according to an embodiment of the present invention is shown; Figure 4 A system architecture diagram of an embodiment of the underwater pile foundation condition detection system of the present invention is shown; Figure 5 A schematic diagram of three-dimensional data modeling for underwater pile foundation monitoring according to an embodiment of the present invention is shown. Detailed Implementation

[0017] To provide a clearer understanding of the technical features, objectives, and effects of the present invention, specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0018] It should be noted that the flowcharts shown in the accompanying drawings are merely illustrative and do not necessarily include all content and operations / steps, nor do they necessarily have to be performed in the described order. For example, some operations / steps can be broken down, while others can be combined or partially combined; therefore, the actual execution order may change depending on the specific circumstances.

[0019] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor devices and / or microcontroller devices.

[0020] like Figure 1 As shown, some embodiments of the present invention disclose a method for detecting the condition of underwater pile foundations, including the following steps: S1: Determine the detection area of ​​underwater pile foundation 2, and deploy a ring sonar array 1 around the underwater pile foundation within the detection area; S2: Control the ring sonar array 1 to scan around the underwater pile foundation 2, and return the echo signal after scanning; S3: Acquire echo signals, process the echo signals, and obtain modeling signals; S4: Based on the modeling signal, establish a three-dimensional acoustic field model around the underwater pile foundation, and calculate various underwater pile foundation state-related parameters based on the three-dimensional acoustic field model around the underwater pile foundation. S5: The various underwater pile foundation state-related parameters obtained from multiple scans are fused to obtain multiple final underwater pile foundation state-related parameters, and a pile foundation state inspection report is generated based on these multiple final underwater pile foundation state-related parameters.

[0021] This embodiment has the advantages of wide detection range, high accuracy, strong adaptability and the ability to simultaneously monitor various underwater pile foundation status parameters, which can effectively ensure the structural safety of underwater pile foundations at sea and inland waterways and reduce operation and maintenance costs.

[0022] In some embodiments, determining the detection area of ​​the underwater pile foundation 2 includes: scanning the terrain within a preset range around the underwater pile foundation 2 using an underwater topographic mapping device to ensure that there are no interfering factors affecting the deployment and signal transmission of the ring sonar array 1 in the detection area, so as to determine the detection area of ​​the underwater pile foundation 2.

[0023] Specifically, underwater topographic mapping equipment (such as side-scan sonar and multibeam echo sounder) is used to scan the topography within a 50m radius around the underwater pile foundation 2 to determine the flatness of the riverbed or seabed, the distribution of obstacles (such as reefs and shipwrecks), and the initial state of sediment accumulation. Areas with dense obstacles are avoided to ensure that there are no interfering factors in the detection area that could affect sensor installation and signal transmission.

[0024] In some embodiments, deploying a ring-shaped sonar array 1 around the underwater pile foundation 2 within the detection area includes: determining the diameter of the underwater pile foundation 2; determining the radius of the ring-shaped sonar array 1 based on the diameter of the underwater pile foundation 2; formulating a deployment time window based on the environmental constraints of the detection area; and deploying the ring-shaped sonar array 1 around the underwater pile foundation 2 within the detection area based on the radius of the ring-shaped sonar array 1 and the set time window.

[0025] Specifically, the diameter of underwater pile 2 is determined based on the design drawings of underwater pile 2 (the diameter of underwater piles for conventional offshore wind power is mostly 3-6m, and the diameter of underwater piles for inland river wind power is mostly 2-4m). The radius of the ring sonar array 1 is determined according to the calculation formula "ring radius = underwater pile radius + 2-5m". For example, for an underwater pile with a diameter of 6m, the ring radius is set to 5-8m. This ensures that there is sufficient detection space between the sensor and the underwater pile 2, while avoiding signal attenuation due to excessive distance. Based on the water depth (0-50m), water flow velocity (≤3m / s), and tidal cycle (if it is a marine environment) of the detection area, a deployment time window is determined. Deployment operations are carried out during periods when the water flow velocity is ≤1.5m / s and the tide is at its slack to reduce the impact of water flow on the installation operation. Based on the radius of the ring sonar array 1 and the set time window, the ring sonar array 1 is deployed around the underwater pile foundation 2 within the detection area.

[0026] In some embodiments, the sonar array includes at least two sets of digital sonar sensors with different frequencies, which are installed at intervals on the liftable mounting structure 3 of the annular guide rail 4, and the included angle between the centers of adjacent digital sonar sensors is no greater than a preset angle. Furthermore, the transmission and reception directions of each set of sensors are all oriented towards the center of the annular circle (i.e., the center of the underwater pile foundation 2), forming a 360° annular detection area without blind spots.

[0027] Specifically, such as Figure 2 As shown, the ring sonar array 1 includes a total of 8 sets of digital sonar sensors, which are installed at 45° intervals on the liftable mounting structure 3 of the ring guide rail 4. They are divided into 4 frequency types, with 2 sets of each frequency. The ring multi-frequency digital sonar array is deployed 3m around the underwater pile foundation 2, with 8 sets of sensors installed, namely 2 sets of each of 50kHz digital sonar sensor 1.4, 100kHz digital sonar sensor 1.3, 200kHz digital sonar sensor 1.2 and 500kHz digital sonar sensor 1.1. The sensor installation depth is 5m underwater.

[0028] The sensors are configured with frequencies of 50kHz (detection depth 10-50m, used for deep foundation erosion detection), 100kHz (detection depth 5-30m, used for medium-deep siltation detection), 200kHz (detection depth 2-15m, used for shallow-medium siltation and underwater pile foundation surface contour identification), and 500kHz (detection depth 0.5-8m, used for shallow siltation and underwater pile foundation surface minute defect identification). Eight sets of sensors are evenly distributed along the preset circular sonar array 1 trajectory. The angle between the centers of adjacent sensors is strictly controlled at 45°, and the transmission and reception directions of each set of sensors are all directed towards the center of the ring (i.e., the center of the underwater pile foundation 2), forming a 360° ring detection area with no blind spots.

[0029] It should be noted that the specific numbers in the number of groups, installation intervals, and frequency types included in the above-mentioned ring sonar array 1 are for illustrative purposes only and are not intended to limit this application; other types are also possible.

[0030] The deployment of the ring sonar array 1 specifically includes mounting each group of sensors on an independent liftable installation structure 3.

[0031] Among them, such as Figure 3 As shown, the independent liftable mounting structure 3 includes: a bottom fixed base 3.1, a middle lifting rod 3.2, and a top sensor fixed base 3.3.

[0032] The bottom fixed base 3.1 and the top sensor fixed base 3.3 are connected by the central lifting rod 3.2. It should be noted that the connection method includes fixed connection and detachable connection.

[0033] The bottom fixed base 3.1 is placed flat on the riverbed or seabed surface, and the anchor claw is inserted into the mud and sand layer to a depth of 0.5-1m through the hydraulic device to ensure that the base does not move under the impact of water flow.

[0034] The bottom mounting base 3.1 is made of cast iron weighing 50-100kg, with anti-slip texture and detachable anchor claws (4 in total, arranged in a cross shape) on the bottom; the middle lifting rod 3.2 is made of stainless steel, with a rod diameter of 80-120mm, and its length can be adjusted from 0.5-5m via a built-in electric push rod, with an adjustment accuracy of ±1cm. The lifting rod surface is engraved with scale markings for easy real-time reading of the installation depth; the top sensor mounting base 3.3 is made of waterproof aluminum alloy, and is sealed to the sensor with a waterproof sealing ring (made of fluororubber, with a temperature resistance range of -20℃ to 200℃). The mounting base has horizontal adjustment knobs on both sides, which can be used to adjust the horizontal angle of the sensor to ensure that the detection direction of all sensors is accurately oriented towards the center of the ring, with a deviation of no more than ±1°.

[0035] A ring-shaped sonar array 1 is deployed 3m around the underwater pile foundation 2, with 8 sets of sensors installed, namely 2 sets each of 50kHz digital sonar sensor 1.4, 100kHz digital sonar sensor 1.3, 200kHz digital sonar sensor 1.2 and 500kHz digital sonar sensor 1.1. The sensor installation depth is 5m underwater.

[0036] In some embodiments, the deployment of the ring sonar array 1 further includes: setting sensor installation depths in layers, with no less than one set of sensors in each layer, and each layer being installed at a different underwater depth; after installation, using satellite positioning signals and underwater measurement data to calibrate the position and attitude of each set of sensors.

[0037] Specifically, the sensor installation depth is set in two layers according to the detection requirements. Four sets of sensors (50kHz and 100kHz, deep-sea detection group) are installed at a depth of 3-8m underwater (based on the water surface), while four sets of sensors (200kHz and 500kHz, shallow-sea detection group) are installed at a depth of 1-3m underwater. These two layers of sensors form a layered detection structure in the vertical direction, covering the entire depth range from the water surface to the substrate. After installation, the position and attitude of each sensor set need to be calibrated using satellite positioning signals obtained through a GPS positioning module (positioning accuracy ≤1m) and underwater measurement data obtained through underwater attitude sensors (measurement accuracy ±0.1°). Record the latitude and longitude coordinates of each set of sensors and compare them with the coordinates of the preset circular trajectory. If the deviation exceeds 0.3m, the base position needs to be readjusted. The horizontal tilt angle and vertical pitch angle of the sensor are detected by the attitude sensor. The horizontal tilt angle should be controlled within ±0.5° and the vertical pitch angle should be controlled within ±2°. If they are outside the range, they can be corrected by adjusting the horizontal knob of the fixed base and the height of the lifting rod.

[0038] Acquiring underwater measurement data also includes: starting a digital sonar sensor to perform a trial scan, collecting the initial echo signal on the surface of the underwater pile foundation, judging whether the sensor's detection direction is accurate by the signal intensity distribution, and if the signal intensity of one set of sensors is significantly lower than that of other sensors of the same frequency, the orientation of that sensor needs to be recalibrated.

[0039] In some embodiments, the deployment of the ring sonar array 1 further includes: each group of sensors and the control terminal are connected through dual channels, the first channel being a data transmission channel and the second channel being a power supply channel, and both the first and second channels are equipped with protective measures.

[0040] Specifically, each sensor group is connected to the control terminal on the shore or offshore platform via two cables. The first channel can be a shielded data transmission cable (armored twisted pair, transmission rate 100Mbps, transmission distance ≤1000m) used to transmit sonar echo signals and control commands. The second channel can be a waterproof power supply cable (multi-core copper cable, rated voltage 24VDC, rated current 5A) used to provide power to the sensors. After the cables are led out from the sensor mounting base, they are laid down along the lifting rod and connected to the main cable at the base via a waterproof connector. The main cable extends along the riverbed or seabed surface towards the shore, with a cable fixing stake (made of plastic, inserted into the mud and sand to a depth of 0.3m) set every 5m to prevent the cable from drifting under the action of water currents. At the same time, the main cable is wrapped with a wear-resistant waterproof sleeve (made of polyethylene, 5mm thick, water pressure resistant to 5MPa). When crossing reefs or obstacle areas, an additional stainless steel protective pipe (50mm in diameter) is installed to prevent the cable from being scratched by sharp objects, which could lead to water leakage or signal interruption.

[0041] In some embodiments, the deployment of the ring sonar array 1 further includes: adding at least one set of backup sensors in each of the four directions of the ring sonar array 1 (east, south, west, and north); the backup sensors are at a preset distance from the main sensors and have the same installation structure as the main sensors; the working status of the main sensors is monitored in real time, and when one or more sets of main sensors fail, the system automatically switches to the corresponding backup sensors.

[0042] Specifically, to improve system reliability, a set of backup sensors is added in each of the four directions (east, south, west, and north) of the ring sonar array 1. The frequency of the backup sensors is consistent with that of the adjacent main sensors (e.g., if the main sensor on the east side is 50kHz, then the backup sensor on the east side is also 50kHz). The distance between the backup sensors and the main sensors is 1-2m, and their installation structure is exactly the same as that of the main sensors. The backup sensors monitor the working status of the main sensors in real time. When one or more sets of main sensors experience signal interruption, data abnormality, or failure, the system automatically switches to the corresponding backup sensor within 10 seconds to ensure the continuity of the detection operation and avoid detection interruption due to the failure of a single sensor.

[0043] In some embodiments, controlling the ring sonar array 1 to scan around the underwater pile foundation 2 includes: Each scan employs at least two different frequency combinations, and the scanning method involves bidirectional scanning around the underwater pile foundation 2 in both clockwise and counterclockwise directions. The scanning interval does not exceed a preset time, and the number of scans is not less than a preset number. Understandably, "at least two" can refer to two, three, or any number of frequencies.

[0044] Specifically, scanning parameters are set according to detection requirements. A dual-frequency mode is used, including a first frequency and a second frequency, with the second frequency being higher than the first. The first frequency is used to detect deep siltation and base erosion, while the second frequency is used to identify details on the underwater pile foundation surface. For example, the first frequency is 50kHz and the second frequency is 200kHz; or the first frequency is 100kHz and the second frequency is 500kHz. The scanning angle is a full 360° range, and the scanning depth covers from the water surface to 1m below the base. The number of scans is set to 6: 3 clockwise scans and 3 counterclockwise scans, with a 20-minute interval between each scan to avoid errors from a single scan. It should be noted that the specific numbers of the scanning parameters set for the above-mentioned ring sonar array 1 are for illustrative purposes only and are not intended to limit this application; other parameters may also be used.

[0045] In some embodiments, controlling the ring sonar array 1 to scan around the underwater pile foundation 2 further includes: During the scanning process, adjacent sensors employ a synchronous triggering mechanism to ensure that the transmitted sonar signal is synchronized with the opening of the receiving window, thereby reducing phase difference errors.

[0046] In some embodiments, step S2 further includes: The ring sonar array 1 is controlled to monitor environmental parameters in real time; specifically, it monitors environmental parameters such as water flow velocity, water temperature, and water turbidity in real time, providing environmental compensation basis for subsequent signal processing.

[0047] Step S3 involves processing the echo signal, including: The echo signal is preprocessed, enhanced, and its features are extracted to remove interference signals from water flow, bubbles, and marine organisms.

[0048] During underwater transmission, sonar echo signals are easily affected by factors such as water flow noise, bubble interference, scattering of suspended sediment, and marine biological activity, resulting in a reduced signal-to-noise ratio and blurred feature information. Therefore, a multi-stage processing procedure is required to achieve signal optimization and feature extraction.

[0049] In some embodiments, the echo signal is preprocessed, including: An adaptive filtering algorithm is used to eliminate dynamic noise; this includes: using environmental parameters as input, establishing a noise prediction model, dynamically adjusting the filter gain, and performing targeted filtering on echo signals of different frequencies.

[0050] Specifically, in the preprocessing stage, an adaptive Kalman filter algorithm is used to eliminate dynamic noise. Using environmental parameters such as water flow velocity and water temperature as input, a noise prediction model is established, and the filter gain is dynamically adjusted to perform targeted filtering on echo signals of different frequencies from 50kHz to 500kHz. For example, for the first frequency signal (50kHz, 100kHz), the focus is on suppressing low-frequency water flow turbulence noise, and the filter window is set to 20ms; for the second frequency signal (200kHz, 500kHz), the focus is on eliminating high-frequency bubble bursting noise, and the filter window is shortened to 5ms, ensuring that signal details are preserved while reducing noise. After this processing, the signal-to-noise ratio can be improved by 30%-40%.

[0051] In some embodiments, signal enhancement of the echo signal includes: The preprocessed echo signal is decomposed several times using wavelet transform algorithm. Each decomposition yields a set of low-frequency coefficients and a set of high-frequency coefficients, where the low-frequency coefficients correspond to the overall contour and the high-frequency coefficients correspond to local details. The low-frequency coefficients of the array are enhanced by threshold shrinkage to suppress residual noise; the high-frequency coefficients of the array are amplified by coefficient reconstruction to highlight the interface abrupt change characteristics.

[0052] Specifically, after preprocessing, the signal enhancement stage begins. A wavelet transform algorithm is used to decompose the signal several times: selecting the db4 wavelet basis function, the signal is decomposed eight times, resulting in eight sets of low-frequency coefficients and eight sets of high-frequency coefficients. The low-frequency coefficients correspond to the overall contour of the signal (such as the macroscopic features of the underwater pile foundation surface and siltation interface), while the high-frequency coefficients correspond to local details (such as subtle erosion marks and interface textures). Threshold contraction is used to enhance the eight sets of low-frequency coefficients to suppress residual noise; coefficient reconstruction is used to amplify the eight sets of high-frequency coefficients to highlight interface abrupt changes. For example, at the siltation interface, the echo signal amplitude may show a 10-20 dB abrupt change; enhancing the high-frequency coefficients makes this abrupt change more significant, facilitating subsequent identification. It should be noted that the specific numbers in the number of groups, installation intervals, and frequency types included in the above-mentioned ring sonar array 1 are merely illustrative and not intended to limit this application; other types are also possible.

[0053] In some embodiments, the step of performing feature extraction processing on the echo signal includes: A deep learning model is introduced to convert the enhanced echo signal into a two-dimensional time-frequency map, which is then input into a trained CNN network to extract spatial features from the two-dimensional time-frequency map. The feature vectors output by the CNN network are then input into the LSTM network to capture the time-series features of the signal and obtain the modeled signal.

[0054] Specifically, the final step is feature extraction, which introduces a CNN-LSTM-based deep learning model. CNN-LSTM is a hybrid deep learning model that combines convolutional neural networks (CNN) and long short-term memory networks (LSTM). It uses CNN to extract spatial features (such as image texture and local patterns) and then uses LSTM to capture long-term dependencies in time series, making it particularly suitable for processing data with spatiotemporal characteristics, such as video analysis and time series signal prediction. The model converts the processed signal into a two-dimensional time-frequency graph (horizontal axis for time, vertical axis for frequency, and color for signal intensity), which is then input into a trained CNN network (containing 3 convolutional layers and 2 pooling layers) to extract spatial features from the time-frequency graph (such as the continuity of the interface and abrupt change locations). The feature vector output by the CNN is then input into the LSTM network (containing 2 hidden layers) to capture the time series features of the signal (such as the duration and intensity change trend of the interface signal), resulting in a modeled signal that enables intelligent recognition of siltation interfaces, basement rock layer interfaces, and the surface contours of underwater pile foundations. During model training, 1,000 sets of measured sonar data under different environments (covering different water depths, water quality, and siltation conditions) were used for training. The cross-validation accuracy reached over 95%, which can effectively avoid the recognition bias of a single algorithm in complex environments.

[0055] In some embodiments, step S4 includes: Using the center of the flange at the top of the underwater pile foundation 2 as the origin, a spatial rectangular coordinate system is established. The echo signal is converted into three-dimensional cloud points in the spatial rectangular coordinate system. Then, by comparing the weighted fusion of the features of the modeling signals at different frequencies, the identification of the siltation interface, the basement rock layer interface and the surface contour of the underwater pile foundation is realized. Based on the three-dimensional cloud points, a three-dimensional acoustic field model around the underwater pile foundation is established by combining the siltation interface, the basement rock layer interface, and the surface contour of the underwater pile foundation. Various underwater pile foundation state-related parameters are calculated based on the three-dimensional acoustic field model around the underwater pile foundation.

[0056] Specifically, a three-dimensional sound field model was constructed, and OpenGL technology was used to achieve the visualization fusion of multi-source data. OpenGL (Open Graphics Library) is a cross-language, cross-platform graphics rendering API (Application Programming Interface) used to render 2D and 3D vector graphics. It communicates directly with the GPU (Graphics Processing Unit) to achieve hardware-accelerated graphics rendering. A spatial rectangular coordinate system was established with the center of the flange at the top of the underwater pile foundation as the origin (X-axis along the east-west direction, Y-axis along the north-south direction, and Z-axis along the vertical direction). The data from six scans by eight sensors (a total of 48 data sets) were converted into a three-dimensional point cloud. The time delay of the echo signal from each sensor corresponds to the spatial distance (distance = speed of sound × time delay / 2, with the speed of sound corrected in real time according to water temperature and salinity, with a correction accuracy of ±0.5 m / s), and the signal intensity corresponds to the grayscale value of the point cloud (the higher the intensity, the larger the grayscale value). A weighted fusion strategy was adopted for different frequency data. The weight of the first frequency data (50kHz, 100kHz) was set to 0.6 to determine the deep basement contour; the weight of the second frequency data (200kHz, 500kHz) was set to 0.4 to optimize the surface details of shallow siltation and underwater pile foundation. After fusion, the model resolution can reach 0.1m×0.1m×0.05m (X×Y×Z), which can clearly present the siltation thickness variation above 0.05m, realize the intelligent identification of siltation interface, basement rock layer interface and underwater pile foundation surface contour, establish a three-dimensional sound field model around the underwater pile foundation, and calculate various underwater pile foundation state-related parameters based on the three-dimensional sound field model around the underwater pile foundation.

[0057] In some embodiments, step S4 calculates various underwater pile foundation state-related parameters based on the three-dimensional acoustic field model surrounding the underwater pile foundation, including: Calculate the underwater pile foundation siltation thickness; this includes: at each preset angle of the ring sonar array 1, calculate the vertical distance from the water surface to the siltation interface in the vertical direction to obtain the underwater pile foundation siltation thickness; Specifically, based on the model, the sediment thickness is calculated. At each angle of the ring sonar array 1 (1° interval, 360 sampling points in total), the vertical distance from the water surface (Z=0) to the sediment interface (where the gray value changes abruptly) is identified along the vertical direction (Z-axis direction), which is the sediment thickness at that point.

[0058] To eliminate local interference, the thickness values ​​of 360 sampling points were smoothed using the moving average method (window size of 5°) to obtain an annular siltation thickness distribution curve. For example, on the north side of underwater pile foundation 2, the siltation thickness may show a gradient change of 0.8-1.2m due to water erosion. The curve can intuitively reflect the non-uniformity of siltation.

[0059] Calculate the erosion depth of underwater pile foundations; this includes: obtaining the erosion depth of underwater pile foundations by calculating the difference between the interface of the base rock layer and the designed base elevation; Specifically, the erosion depth calculation is based on the design base elevation of underwater pile foundation 2. The actual base rock layer interface (where the gray value is strongly reflected, and the signal strength is ≥80dB) is identified by the first frequency signal (50kHz). The Z coordinate value of the interface is read, and the difference between it and the design base elevation (obtained from the design drawings of underwater pile foundation 2) is the erosion depth. If the actual Z coordinate is lower than the design value, it indicates that erosion exists, and the larger the difference, the more severe the erosion. If the actual Z coordinate is higher than the design value, it indicates that there is base siltation.

[0060] Calculate the inclination angle of the underwater pile foundation; it includes: extracting multiple horizontal sections at preset intervals along the height direction of the underwater pile foundation 2, uniformly selecting multiple sampling points in each horizontal section, fitting the center coordinates of each horizontal section by the least squares method, connecting the multiple center coordinates to obtain the actual axis of the underwater pile foundation 2, and calculating the angle between the actual axis and the vertical line in the spatial rectangular coordinate system to obtain the inclination angle of the underwater pile foundation. Specifically, the underwater pile foundation inclination angle is calculated using the axis fitting method: In the three-dimensional model, along the height direction (Z-axis direction) of the underwater pile foundation 2, a point cloud of the underwater pile foundation surface of a horizontal section is extracted every 0.5m. For each section, 36 evenly distributed points (interval of 10°) are selected, and the coordinates of the center of each section are fitted by the least squares method. The coordinates of the center of each section are connected to obtain the actual axis of the underwater pile foundation 2. The angle between the actual axis and the design axis (the vertical line along the Z-axis) is calculated to obtain the inclination angle of the underwater pile foundation.

[0061] To improve accuracy, the spatial vector method is adopted: the actual axis and the design axis are represented as vectors respectively, the angle between the two vectors is calculated, which is the tilt angle, and the tilt direction is recorded at the same time.

[0062] The parameters related to the final underwater pile foundation status in step S5 are the final underwater pile foundation siltation thickness, the final underwater pile foundation erosion depth, and the final underwater pile foundation tilt angle.

[0063] Specifically, a Bayesian data fusion algorithm was used to fuse the parameter results of the six scans, eliminating the random errors of a single scan, and obtaining the final underwater pile foundation siltation thickness, final underwater pile foundation erosion depth, and final underwater pile foundation tilt angle.

[0064] In some embodiments, a pile foundation condition inspection report is generated based on various parameters related to the final underwater pile foundation condition, including: Multiple parameters related to the final underwater pile foundation status are compared with preset alarm thresholds. If any parameter related to the final underwater pile foundation status exceeds its preset alarm threshold, an early warning message is triggered. The pile foundation status detection report includes early warning information, a visualized three-dimensional sound field model around the underwater pile foundation, and change curves of multiple underwater pile foundation status parameters.

[0065] Specifically, if the final underwater pile foundation siltation thickness exceeds 1.5m, the final underwater pile foundation erosion depth exceeds 0.5m, or the final underwater pile foundation tilt angle exceeds 0.5°, an early warning message will be triggered. The pile foundation status detection report includes the early warning message (if any), a visualized three-dimensional sound field model around the underwater pile foundation, and various underwater pile foundation status-related parameter change curves.

[0066] In some embodiments, after triggering the early warning information, the pile foundation status monitoring report is sent to the operation and maintenance personnel. For example, the pile foundation status monitoring report can be transmitted to the operation and maintenance center, and SMS and email notifications can be sent to the operation and maintenance personnel. The above notification methods are only examples and are not intended to limit this application; other methods may also be used.

[0067] In some embodiments, the warning information can be compared and graded according to various parameters related to the final underwater pile foundation status, with a preset alarm threshold as the benchmark, and the grading is not less than 2 levels. Corresponding warning signals are set and triggered when the parameters related to the final underwater pile foundation status are greater than the preset alarm threshold.

[0068] Specifically, the warning information can be divided into three levels based on comparisons of various parameters related to the final underwater pile foundation status with preset alarm thresholds. The first level is a blue warning, triggered when the parameters related to the final underwater pile foundation status reach 90% of the preset alarm threshold; the second level is a yellow warning, triggered when the parameters related to the final underwater pile foundation status exceed 10% of the preset alarm threshold; and the third level is a red warning, triggered when the parameters related to the final underwater pile foundation status exceed 30% of the preset alarm threshold. It should be noted that the specific numbers for the above warning information levels and the colors corresponding to the warning signals at each level are for illustrative purposes only and are not intended to limit this application; other levels are also possible.

[0069] like Figure 2 and Figure 4As shown, some embodiments of the present invention disclose an underwater pile foundation condition detection system, comprising: The detection and determination module is used to determine the detection area of ​​underwater pile foundation 2; Ring sonar array 1 is deployed around the underwater pile foundation 2 within the detection area; The scanning control module is used to control the ring sonar array 1 to scan around the underwater pile foundation 2 and return the echo signal after scanning; The signal acquisition and processing module is used to acquire echo signals, process the echo signals, and obtain modeling signals. The visualization and calculation module is used to establish a three-dimensional acoustic field model around the underwater pile foundation based on the modeling signal, and to calculate various underwater pile foundation state-related parameters based on the three-dimensional acoustic field model around the underwater pile foundation. The data fusion and analysis module is used to fuse various underwater pile foundation state-related parameters obtained from multiple scans to obtain various final underwater pile foundation state-related parameters, and generate a pile foundation state inspection report based on these parameters.

[0070] This embodiment has the advantages of wide detection range, high accuracy, strong adaptability and the ability to simultaneously monitor various underwater pile foundation status parameters, which can effectively ensure the structural safety of underwater pile foundations at sea and inland waterways and reduce operation and maintenance costs.

[0071] In some embodiments, determining the detection area of ​​the underwater pile foundation 2 includes: scanning the terrain within a preset range around the underwater pile foundation 2 using an underwater topographic mapping device to ensure that there are no interfering factors affecting the deployment and signal transmission of the ring sonar array 1 in the detection area, so as to determine the detection area of ​​the underwater pile foundation 2.

[0072] Specifically, underwater topographic mapping equipment (such as side-scan sonar and multibeam echo sounder) is used to scan the topography within a 50m radius around the underwater pile foundation 2 to determine the flatness of the riverbed or seabed, the distribution of obstacles (such as reefs and shipwrecks), and the initial state of sediment accumulation. Areas with dense obstacles are avoided to ensure that there are no interfering factors in the detection area that could affect sensor installation and signal transmission.

[0073] In some embodiments, deploying a ring-shaped sonar array 1 around the underwater pile foundation within the detection area includes: determining the diameter of the underwater pile foundation 2; determining the radius of the ring-shaped sonar array 1 based on the diameter of the underwater pile foundation 2; formulating a deployment time window based on the environmental constraints of the detection area; and deploying the ring-shaped sonar array 1 around the underwater pile foundation 2 within the detection area based on the radius of the ring-shaped sonar array 1 and the set time window.

[0074] Specifically, the diameter of underwater pile 2 is determined based on the design drawings of underwater pile 2 (the diameter of underwater piles for conventional offshore wind power is mostly 3-6m, and the diameter of underwater piles for inland river wind power is mostly 2-4m). The radius of the ring sonar array 1 is determined according to the calculation formula "ring radius = underwater pile radius + 2-5m". For example, for an underwater pile with a diameter of 6m, the ring radius is set to 5-8m. This ensures that there is sufficient detection space between the sensor and the underwater pile 2, while avoiding signal attenuation due to excessive distance. Based on the water depth (0-50m), water flow velocity (≤3m / s), and tidal cycle (if it is a marine environment) of the detection area, a deployment time window is determined. Deployment operations are carried out during periods when the water flow velocity is ≤1.5m / s and the tide is at its slack to reduce the impact of water flow on the installation operation. Based on the radius of the ring sonar array 1 and the set time window, a ring sonar array 1 is set up around the underwater pile foundation 2 within the detection area.

[0075] In some embodiments, the system further includes: multiple liftable mounting structures 3 and annular guide rails 4.

[0076] The liftable mounting structure 3 is used to mount the sensor; the ring rail 4 is used to connect multiple liftable mounting structures 3.

[0077] The ring sonar array 1 includes at least two sets of digital sonar sensors with different frequencies, which are installed at intervals on the liftable mounting structure 3 of the ring guide rail 4, and the included angle between the centers of adjacent digital sonar sensors is no greater than a preset angle. Furthermore, the transmission and reception directions of each set of sensors are all oriented towards the center of the ring (i.e., the center of the underwater pile foundation 2), forming a 360° ring detection area without blind spots.

[0078] Specifically, such as Figure 2 As shown, the annular sonar array 1 includes a total of 8 digital sonar sensors, which are installed at 45° intervals on the liftable mounting structure 3 of the annular guide rail 4. They are divided into 4 frequency types, with 2 sets of each frequency. The annular multi-frequency digital sonar array 1 is deployed 3m around the underwater pile foundation 2, with 8 sets of sensors installed, namely 2 sets of each of 50kHz digital sonar sensor 1.4, 100kHz digital sonar sensor 1.3, 200kHz digital sonar sensor 1.2 and 500kHz digital sonar sensor 1.1. The sensor installation depth is 5m underwater.

[0079] The sensors are configured with frequencies of 50kHz (detection depth 10-50m, used for deep foundation erosion detection), 100kHz (detection depth 5-30m, used for medium-deep siltation detection), 200kHz (detection depth 2-15m, used for shallow-medium siltation and underwater pile foundation surface contour identification), and 500kHz (detection depth 0.5-8m, used for shallow siltation and underwater pile foundation surface minute defect identification). Eight sets of sensors are evenly distributed along the preset circular sonar array 1 trajectory. The angle between the centers of adjacent sensors is strictly controlled at 45°, and the transmission and reception directions of each set of sensors are all directed towards the center of the ring (i.e., the center of the underwater pile foundation), forming a 360° circular detection area with no blind spots.

[0080] It should be noted that the specific figures in the number of groups, installation intervals, and frequency types included in the above-mentioned ring sonar array are for illustrative purposes only and are not intended to limit this application; other types are also possible.

[0081] The deployment of the ring sonar array 1 specifically includes mounting each group of sensors on an independent liftable installation structure 3.

[0082] Among them, such as Figure 3 As shown, the independent liftable mounting structure 3 includes: a bottom fixed base 3.1, a middle lifting rod 3.2, and a top sensor fixed base 3.3.

[0083] The bottom fixed base 3.1 and the top sensor fixed base 3.3 are connected by the central lifting rod 3.2. It should be noted that the connection method includes fixed connection and detachable connection.

[0084] The bottom fixed base 3.1 is placed flat on the riverbed or seabed surface, and the anchor claw is inserted into the mud and sand layer to a depth of 0.5-1m through the hydraulic device to ensure that the base does not move under the impact of water flow.

[0085] The bottom mounting base 3.1 is made of cast iron weighing 50-100kg, with anti-slip texture and detachable anchor claws (4 in total, arranged in a cross shape) on the bottom; the middle lifting rod 3.2 is made of stainless steel, with a rod diameter of 80-120mm, and its length can be adjusted from 0.5-5m via a built-in electric push rod, with an adjustment accuracy of ±1cm. The lifting rod surface is engraved with scale markings for easy real-time reading of the installation depth; the top sensor mounting base 3.3 is made of waterproof aluminum alloy, and is sealed to the sensor with a waterproof sealing ring (made of fluororubber, with a temperature resistance range of -20℃ to 200℃). The mounting base has horizontal adjustment knobs on both sides, which can be used to adjust the horizontal angle of the sensor to ensure that the detection direction of all sensors is accurately oriented towards the center of the ring, with a deviation of no more than ±1°.

[0086] A ring-shaped sonar array 1 is deployed 3m around the underwater pile foundation 2, with 8 sets of sensors installed, namely 2 sets each of 50kHz digital sonar sensor 1.4, 100kHz digital sonar sensor 1.3, 200kHz digital sonar sensor 1.2 and 500kHz digital sonar sensor 1.1. The sensor installation depth is 5m underwater.

[0087] In some embodiments, deploying the ring sonar array 1 further includes: setting sensor installation depths in layers, with at least one set of sensors in each layer, and each layer being installed at a different underwater depth. After installation, the position and attitude of each set of sensors are calibrated using satellite positioning signals and underwater measurement data.

[0088] Specifically, the sensor installation depth is set in two layers according to the detection requirements. Four sets of sensors (50kHz and 100kHz, deep-sea detection group) are installed at a depth of 3-8m underwater (based on the water surface), while four sets of sensors (200kHz and 500kHz, shallow-sea detection group) are installed at a depth of 1-3m underwater. These two layers of sensors form a layered detection structure in the vertical direction, covering the entire depth range from the water surface to the substrate. After installation, the position and attitude of each sensor set need to be calibrated using satellite positioning signals obtained through a GPS positioning module (positioning accuracy ≤1m) and underwater measurement data obtained through underwater attitude sensors (measurement accuracy ±0.1°). Record the latitude and longitude coordinates of each set of sensors and compare them with the coordinates of the preset circular trajectory. If the deviation exceeds 0.3m, the base position needs to be readjusted. The horizontal tilt angle and vertical pitch angle of the sensor are detected by the attitude sensor. The horizontal tilt angle should be controlled within ±0.5° and the vertical pitch angle should be controlled within ±2°. If they are outside the range, they can be corrected by adjusting the horizontal knob of the fixed base and the height of the lifting rod.

[0089] Acquiring underwater measurement data also includes: starting a digital sonar sensor to perform a trial scan, collecting the initial echo signal on the surface of the underwater pile foundation, judging whether the sensor's detection direction is accurate by the signal intensity distribution, and if the signal intensity of one set of sensors is significantly lower than that of other sensors of the same frequency, the orientation of that sensor needs to be recalibrated.

[0090] In some embodiments, the deployment of the ring sonar array 1 further includes: each group of sensors and the control terminal are connected through dual channels, the first channel being a data transmission channel and the second channel being a power supply channel, and both the first and second channels are equipped with protective measures.

[0091] Specifically, each sensor group is connected to the control terminal on the shore or offshore platform via two cables. The first channel can be a shielded data transmission cable (armored twisted pair, transmission rate 100Mbps, transmission distance ≤1000m) used to transmit sonar echo signals and control commands. The second channel can be a waterproof power supply cable (multi-core copper cable, rated voltage 24VDC, rated current 5A) used to provide power to the sensors. After the cables are led out from the sensor mounting base, they are laid down along the lifting rod and connected to the main cable at the base via a waterproof connector. The main cable extends along the riverbed or seabed surface towards the shore, with a cable fixing stake (made of plastic, inserted into the mud and sand to a depth of 0.3m) set every 5m to prevent the cable from drifting under the action of water currents. At the same time, the main cable is wrapped with a wear-resistant waterproof sleeve (made of polyethylene, 5mm thick, water pressure resistant to 5MPa). When crossing reefs or obstacle areas, an additional stainless steel protective pipe (50mm in diameter) is installed to prevent the cable from being scratched by sharp objects, which could lead to water leakage or signal interruption.

[0092] In some embodiments, the deployment of the ring sonar array 1 further includes: adding at least one set of backup sensors in each of the four directions of the ring sonar array 1 (east, south, west, and north); the backup sensors are at a preset distance from the main sensors and have the same installation structure as the main sensors; the working status of the main sensors is monitored in real time, and when one or more sets of main sensors fail, the system automatically switches to the corresponding backup sensors.

[0093] Specifically, to improve system reliability, a set of backup sensors is added in each of the four directions (east, south, west, and north) of the ring sonar array 1. The frequency of the backup sensors is consistent with that of the adjacent main sensors (e.g., if the main sensor on the east side is 50kHz, then the backup sensor on the east side is also 50kHz). The distance between the backup sensors and the main sensors is 1-2m, and their installation structure is exactly the same as that of the main sensors. The backup sensors monitor the working status of the main sensors in real time. When one or more sets of main sensors experience signal interruption, data abnormality, or failure, the system automatically switches to the corresponding backup sensor within 10 seconds to ensure the continuity of the detection operation and avoid detection interruption due to the failure of a single sensor.

[0094] In some embodiments, controlling the ring sonar array 1 to scan around the underwater pile foundation 2 includes: Each scan employs at least two different frequency combinations, and the scanning method involves bidirectional scanning around the underwater pile foundation 2 in both clockwise and counterclockwise directions. The scanning interval does not exceed a preset time, and the number of scans is not less than a preset number. Understandably, "at least two" can refer to two, three, or any number of frequencies.

[0095] Specifically, scanning parameters are set according to detection requirements. A dual-frequency mode is used, including a first frequency and a second frequency, with the second frequency being higher than the first. The first frequency is used to detect deep siltation and base erosion, while the second frequency is used to identify details on the underwater pile foundation surface. For example, the first frequency is 50kHz and the second frequency is 200kHz; or the first frequency is 100kHz and the second frequency is 500kHz. The scanning angle is a full 360° range, and the scanning depth covers from the water surface to 1m below the base. The number of scans is set to 6: 3 clockwise scans and 3 counterclockwise scans, with a 20-minute interval between each scan to avoid errors from a single scan. It should be noted that the specific numbers for the scanning parameters set for the above-mentioned ring sonar array are for illustrative purposes only and are not intended to limit this application; other parameters may also be used.

[0096] In some embodiments, controlling the ring sonar array 1 to scan around the underwater foundation pile 2 further includes: During the scanning process, adjacent sensors employ a synchronous triggering mechanism to ensure that the transmitted sonar signal is synchronized with the opening of the receiving window, thereby reducing phase difference errors.

[0097] In some embodiments, the scanning control module is also used to control the ring sonar array 1 to monitor environmental parameters in real time; specifically, to monitor environmental parameters such as water flow velocity, water temperature, and water turbidity in real time, so as to provide environmental compensation basis for subsequent signal processing.

[0098] In some embodiments, the echo signal is processed, including: The echo signal is preprocessed, enhanced, and its features are extracted to remove interference signals from water flow, bubbles, and marine organisms.

[0099] During underwater transmission, sonar echo signals are easily affected by factors such as water flow noise, bubble interference, scattering of suspended sediment, and marine biological activity, resulting in a reduced signal-to-noise ratio and blurred feature information. Therefore, a multi-stage processing procedure is required to achieve signal optimization and feature extraction.

[0100] In some embodiments, the echo signal is preprocessed, including: An adaptive filtering algorithm is used to eliminate dynamic noise; this includes: using environmental parameters as input, establishing a noise prediction model, dynamically adjusting the filter gain, and performing targeted filtering on echo signals of different frequencies.

[0101] Specifically, in the preprocessing stage, an adaptive Kalman filter algorithm is used to eliminate dynamic noise. Using environmental parameters such as water flow velocity and water temperature as input, a noise prediction model is established, and the filter gain is dynamically adjusted to perform targeted filtering on echo signals of different frequencies from 50kHz to 500kHz. For example, for the first frequency signal (50kHz, 100kHz), the focus is on suppressing low-frequency water flow turbulence noise, and the filter window is set to 20ms; for the second frequency signal (200kHz, 500kHz), the focus is on eliminating high-frequency bubble bursting noise, and the filter window is shortened to 5ms, ensuring that signal details are preserved while reducing noise. After this processing, the signal-to-noise ratio can be improved by 30%-40%.

[0102] In some embodiments, signal enhancement of the echo signal includes: The preprocessed echo signal is decomposed several times using wavelet transform algorithm. Each decomposition yields a set of low-frequency coefficients and a set of high-frequency coefficients, where the low-frequency coefficients correspond to the overall contour and the high-frequency coefficients correspond to local details. The low-frequency coefficients of the array are enhanced by threshold shrinkage to suppress residual noise; the high-frequency coefficients of the array are amplified by coefficient reconstruction to highlight the interface abrupt change characteristics.

[0103] Specifically, after preprocessing, the signal enhancement stage begins. Wavelet transform is used to decompose the signal several times: the db4 wavelet basis function is selected, and the signal is decomposed eight times, resulting in eight sets of low-frequency coefficients and eight sets of high-frequency coefficients. The low-frequency coefficients correspond to the overall contour of the signal (such as the macroscopic features of underwater pile foundation surfaces and siltation interfaces), while the high-frequency coefficients correspond to local details (such as subtle erosion marks and interface textures). Threshold contraction is used to enhance the eight sets of low-frequency coefficients to suppress residual noise; coefficient reconstruction is used to amplify the eight sets of high-frequency coefficients to highlight interface abrupt changes. For example, at siltation interfaces, the echo signal amplitude may show a 10-20 dB abrupt change; enhancing the high-frequency coefficients makes this abrupt change more significant, facilitating subsequent identification. It should be noted that the specific numbers in the above-mentioned ring sonar array, including the number of groups, installation intervals, and frequency types, are merely illustrative and not intended to limit this application; other types are also possible.

[0104] In some embodiments, the step of performing feature extraction processing on the echo signal includes: A deep learning model is introduced to convert the enhanced echo signal into a two-dimensional time-frequency map, which is then input into a trained CNN network to extract spatial features from the two-dimensional time-frequency map. The feature vectors output by the CNN network are then input into the LSTM network to capture the time-series features of the signal and obtain the modeled signal.

[0105] Specifically, the final step is feature extraction, which introduces a CNN-LSTM-based deep learning model. CNN-LSTM is a hybrid deep learning model that combines convolutional neural networks (CNN) and long short-term memory networks (LSTM). It uses CNN to extract spatial features (such as image texture and local patterns) and then uses LSTM to capture long-term dependencies in time series, making it particularly suitable for processing data with spatiotemporal characteristics, such as video analysis and time series signal prediction. The model converts the processed signal into a two-dimensional time-frequency graph (horizontal axis for time, vertical axis for frequency, and color for signal intensity), which is then input into a trained CNN network (containing 3 convolutional layers and 2 pooling layers) to extract spatial features from the time-frequency graph (such as the continuity of the interface and abrupt change locations). The feature vector output by the CNN is then input into the LSTM network (containing 2 hidden layers) to capture the time series features of the signal (such as the duration and intensity change trend of the interface signal), resulting in a modeled signal that enables intelligent recognition of siltation interfaces, basement rock layer interfaces, and the surface contours of underwater pile foundations. During model training, 1,000 sets of measured sonar data under different environments (covering different water depths, water quality, and siltation conditions) were used for training. The cross-validation accuracy reached over 95%, which can effectively avoid the recognition bias of a single algorithm in complex environments.

[0106] In some embodiments, a three-dimensional acoustic field model around the underwater pile foundation is established based on the modeling signal. Various underwater pile foundation state-related parameters are calculated based on this model, including: Using the center of the flange at the top of the underwater pile foundation 2 as the origin, a spatial rectangular coordinate system is established. The echo signal is converted into three-dimensional cloud points in the spatial rectangular coordinate system. Then, by comparing the weighted fusion of the features of the modeling signals at different frequencies, the identification of the siltation interface, the basement rock layer interface and the surface contour of the underwater pile foundation is realized. Based on the three-dimensional cloud points, a three-dimensional acoustic field model around the underwater pile foundation is established by combining the siltation interface, the basement rock layer interface, and the surface contour of the underwater pile foundation. Various underwater pile foundation state-related parameters are calculated based on the three-dimensional acoustic field model around the underwater pile foundation.

[0107] Specifically, a three-dimensional sound field model was constructed, and OpenGL technology was used to achieve the visualization fusion of multi-source data. OpenGL (Open Graphics Library) is a cross-language, cross-platform graphics rendering API (Application Programming Interface) used to render 2D and 3D vector graphics. It communicates directly with the GPU (Graphics Processing Unit) to achieve hardware-accelerated graphics rendering. Taking the center of the flange at the top of the underwater pile foundation 2 as the origin, a spatial rectangular coordinate system was established (X-axis along the east-west direction, Y-axis along the north-south direction, and Z-axis along the vertical direction). The 6 scan data from 8 sets of sensors (a total of 48 sets of data) were converted into a three-dimensional point cloud. The time delay of the echo signal of each set of sensors corresponds to the spatial distance (distance = speed of sound × time delay / 2, the speed of sound is corrected in real time according to water temperature and salinity, with a correction accuracy of ±0.5m / s), and the signal intensity corresponds to the gray value of the point cloud (the higher the intensity, the larger the gray value). A weighted fusion strategy was adopted for different frequency data. The weight of the first frequency data (50kHz, 100kHz) was set to 0.6 to determine the deep basement contour; the weight of the second frequency data (200kHz, 500kHz) was set to 0.4 to optimize the surface details of shallow siltation and underwater pile foundation. After fusion, the model resolution can reach 0.1m×0.1m×0.05m (X×Y×Z), which can clearly present the siltation thickness variation above 0.05m, realize the intelligent identification of siltation interface, basement rock layer interface and underwater pile foundation surface contour, establish a three-dimensional sound field model around the underwater pile foundation, and calculate various underwater pile foundation state-related parameters based on the three-dimensional sound field model around the underwater pile foundation.

[0108] In some embodiments, various underwater pile foundation state-related parameters are calculated based on a three-dimensional acoustic field model surrounding the underwater pile foundation, including: Calculate the underwater pile foundation siltation thickness; this includes: at each preset angle of the ring sonar array 1, calculate the vertical distance from the water surface to the siltation interface in the vertical direction to obtain the underwater pile foundation siltation thickness; Specifically, based on the model, the sediment thickness is calculated. At each angle of the ring sonar array 1 (1° interval, 360 sampling points in total), the vertical distance from the water surface (Z=0) to the sediment interface (where the gray value changes abruptly) is identified along the vertical direction (Z-axis direction), which is the sediment thickness at that point.

[0109] To eliminate local interference, the thickness values ​​of 360 sampling points were smoothed using the moving average method (window size of 5°) to obtain an annular siltation thickness distribution curve. For example, on the north side of underwater pile foundation 2, the siltation thickness may show a gradient change of 0.8-1.2m due to water erosion. The curve can intuitively reflect the non-uniformity of siltation.

[0110] Calculate the erosion depth of underwater pile foundations; this includes: obtaining the erosion depth of underwater pile foundations by calculating the difference between the interface of the base rock layer and the designed base elevation; Specifically, the erosion depth calculation is based on the design base elevation of underwater pile foundation 2. The actual base rock layer interface (where the gray value is strongly reflected, and the signal strength is ≥80dB) is identified by the first frequency signal (50kHz). The Z coordinate value of the interface is read, and the difference between it and the design base elevation (obtained from the design drawings of underwater pile foundation 2) is the erosion depth. If the actual Z coordinate is lower than the design value, it indicates that erosion exists, and the larger the difference, the more severe the erosion. If the actual Z coordinate is higher than the design value, it indicates that there is base siltation.

[0111] Calculate the inclination angle of the underwater pile foundation; it includes: extracting multiple horizontal sections at preset intervals along the height direction of the underwater pile foundation 2, uniformly selecting multiple sampling points in each horizontal section, fitting the center coordinates of each horizontal section by the least squares method, connecting the multiple center coordinates to obtain the actual axis of the underwater pile foundation 2, and calculating the angle between the actual axis and the vertical line in the spatial rectangular coordinate system to obtain the inclination angle of the underwater pile foundation. Specifically, the underwater pile foundation inclination angle is calculated using the axis fitting method: In the three-dimensional model, along the height direction (Z-axis direction) of the underwater pile foundation 2, the surface point cloud of the underwater pile foundation 2 is extracted at every 0.5m horizontal section. 36 evenly distributed points (10° interval) are selected for each section, and the center coordinates of each section are fitted by the least squares method. The center coordinates of different heights are connected to obtain the actual axis of the underwater pile foundation 2. The angle between the actual axis and the design axis (the vertical line along the Z-axis) is calculated to obtain the underwater pile foundation inclination angle.

[0112] To improve accuracy, the spatial vector method is adopted: the actual axis and the design axis are represented as vectors respectively, the angle between the two vectors is calculated, which is the tilt angle, and the tilt direction is recorded at the same time.

[0113] In some embodiments, the parameters related to the final underwater pile foundation state are the final underwater pile foundation siltation thickness, the final underwater pile foundation erosion depth, and the final underwater pile foundation tilt angle.

[0114] Specifically, a Bayesian data fusion algorithm was used to fuse the parameter results of the six scans, eliminating the random errors of a single scan, and obtaining the final underwater pile foundation siltation thickness, final underwater pile foundation erosion depth, and final underwater pile foundation tilt angle.

[0115] In some embodiments, a pile foundation condition inspection report is generated based on various parameters related to the final underwater pile foundation condition, including: Multiple parameters related to the final underwater pile foundation status are compared with preset alarm thresholds. If any parameter related to the final underwater pile foundation status exceeds its preset alarm threshold, an early warning message is triggered. The pile foundation status detection report includes early warning information, a visualized three-dimensional sound field model around the underwater pile foundation, and change curves of multiple underwater pile foundation status parameters.

[0116] Specifically, if the final underwater pile foundation siltation thickness exceeds 1.5m, the final underwater pile foundation erosion depth exceeds 0.5m, or the final underwater pile foundation tilt angle exceeds 0.5°, an early warning message will be triggered. The pile foundation status detection report includes the early warning message (if any), a visualized three-dimensional sound field model around the underwater pile foundation, and various underwater pile foundation status-related parameter change curves.

[0117] In some embodiments, after triggering the early warning information, the pile foundation status monitoring report is sent to the operation and maintenance personnel. For example, the pile foundation status monitoring report can be transmitted to the operation and maintenance center, and SMS and email notifications can be sent to the operation and maintenance personnel. The above notification methods are only examples and are not intended to limit this application; other methods may also be used.

[0118] In some embodiments, the warning information can be compared and graded according to various parameters related to the final underwater pile foundation status, with a preset alarm threshold as the benchmark, and the grading is not less than 2 levels. Corresponding warning signals are set and triggered when the parameters related to the final underwater pile foundation status are greater than the preset alarm threshold.

[0119] Specifically, the warning information can be divided into three levels based on comparisons of various parameters related to the final underwater pile foundation status with preset alarm thresholds. The first level is a blue warning, triggered when the parameters related to the final underwater pile foundation status reach 90% of the preset alarm threshold; the second level is a yellow warning, triggered when the parameters related to the final underwater pile foundation status exceed 10% of the preset alarm threshold; and the third level is a red warning, triggered when the parameters related to the final underwater pile foundation status exceed 30% of the preset alarm threshold. It should be noted that the specific numbers for the above warning information levels and the colors corresponding to the warning signals at each level are for illustrative purposes only and are not intended to limit this application; other levels are also possible.

[0120] In some embodiments, the ring sonar array 1 adopts an IP68 waterproof and pressure-resistant design, and the sensor housing is made of titanium alloy, which can withstand 5MPa water pressure and is suitable for water depth environments of 0-50m. Each sensor has a built-in temperature compensation module to dynamically adjust the sound velocity value in real time according to water temperature and salinity, ensuring signal stability within a temperature range of -20℃ to 60℃, with a measurement error of less than ±2%.

[0121] The scanning control module, based on the STM32 microcontroller, supports RS485 and CAN bus communication. It can remotely control the scanning parameters of the ring sonar array 1 and has a fault self-diagnosis function. When a sensor fails, it automatically switches to a backup sensor to ensure continuous detection.

[0122] The signal acquisition and processing module employs a 16-bit high-precision AD converter with a sampling rate of up to 1MHz, ensuring the integrity of signal acquisition. An integrated FPGA chip enables real-time filtering, noise reduction, and other processing, with a processing latency of less than 100ms.

[0123] The data fusion and analysis module uses an industrial-grade computer with a built-in SQL Server database, capable of storing over 10 years of testing data. It utilizes MATLAB toolboxes to implement data fusion algorithms and trend analysis, supporting the generation of underwater pile foundation status change reports on weekly, monthly, and yearly cycles.

[0124] The visualization and calculation module uses dedicated visualization software that supports rotation, scaling, and slicing of 3D models, and can intuitively display the state of siltation, erosion, and tilt.

[0125] To verify the effectiveness of the present invention, a 2.5MW underwater pile foundation of a wind turbine was selected for field testing at an offshore wind farm. The test environment was as follows: water depth 15m, water flow velocity 1.2m / s, water turbidity (NTU) 80, underwater pile foundation diameter 3m, and design tilt angle allowable value 0.5°.

[0126] like Figure 2 As shown, a ring-shaped multi-frequency digital sonar array 1 is deployed 3m around the underwater pile foundation 2, with 8 sets of sensors installed, namely 2 sets each of 50kHz digital sonar sensor 1.4, 100kHz digital sonar sensor 1.3, 200kHz digital sonar sensor 1.2 and 500kHz digital sonar sensor 1.1. The sensor installation depth is 5m underwater.

[0127] Set the scanning parameters: the scanning frequency combination is 50kHz+200kHz, 100kHz+500kHz, the number of scans is 6 (3 clockwise and 3 counterclockwise), the interval is 20 minutes, and the scanning depth range is 0-20m.

[0128] The system was activated for automated scanning, acquiring sonar signals from six scans. Signal processing and data fusion were then performed. The results showed that the siltation thickness around the underwater pile foundation ranged from 0.8 to 1.2 meters, with an average of 1.0 meter; the erosion depth at the base was 0.3 meters; and the tilt angle of the underwater pile foundation was 0.3 degrees. All these values ​​were within the safe threshold range (siltation ≤ 1.5 meters, erosion ≤ 0.5 meters, tilt ≤ 0.5 degrees). The system did not trigger any warnings, and a 3D visualization model was output as follows: Figure 5 As shown.

[0129] By implementing this invention, the following beneficial effects are achieved: This invention achieves 360° full-range scanning through a ring sonar array, combined with multi-frequency sensors, covering the entire depth range from the water surface to the base, avoiding local detection blind spots, and comprehensively understanding the surrounding conditions of underwater pile foundations; By employing multiple bidirectional scans and data fusion algorithms, single-scan errors are eliminated; multi-frequency signal combination and deep learning interface recognition improve the measurement accuracy of siltation, erosion, and tilt angles, meeting the requirements of high-precision monitoring. Simultaneous monitoring of multiple parameters enables simultaneous detection of sediment thickness, erosion depth, and tilt angle, eliminating the need for multiple tests, thus improving detection efficiency and reducing operation and maintenance costs. The ring sonar array is waterproof and pressure resistant, and has a compensation mechanism to adapt to different water depths, temperatures and water quality environments. It can work stably under complex hydrological conditions and is not affected by weather, light and other factors. By enabling automated scanning, real-time signal processing, intelligent early warning, and trend prediction, manual intervention is reduced, operational difficulty is lowered, and the long-term online monitoring needs of underwater pile foundations are met.

[0130] It is understood that the above embodiments only illustrate some implementation methods of the present invention, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can freely combine the above embodiments or technical features without departing from the concept of the present invention, and can also make several modifications and improvements, all of which fall within the protection scope of the present invention. That is, the embodiments described "in some embodiments" can be freely combined with any of the preceding and following embodiments. Therefore, all equivalent transformations and modifications made within the scope of the claims of the present invention should be covered by the claims of the present invention.

Claims

1. A method for detecting the condition of underwater pile foundations, characterized in that, Includes the following steps: S1: Determine the detection area of ​​the underwater pile foundation (2), and deploy a ring sonar array (1) around the underwater pile foundation (2) within the detection area. S2: Control the ring sonar array (1) to scan around the underwater pile foundation (2), and return the echo signal after scanning; S3: Acquire the echo signal, process the echo signal, and obtain the modeling signal; S4: Based on the modeling signal, establish a three-dimensional acoustic field model around the underwater pile foundation, and calculate various underwater pile foundation state-related parameters according to the three-dimensional acoustic field model around the underwater pile foundation. S5: The various underwater pile foundation state-related parameters obtained from several scans are fused to obtain a variety of final underwater pile foundation state-related parameters, and a pile foundation state detection report is generated based on the various final underwater pile foundation state-related parameters.

2. The underwater pile foundation condition detection method according to claim 1, characterized in that, The control of the ring sonar array (1) to scan around the underwater pile foundation (2) includes: Each scan uses at least two different frequency combinations. The scanning method is to perform a clockwise and counterclockwise bidirectional scan around the underwater pile foundation (2). The scanning interval does not exceed a preset time, and the number of scans is not less than a preset number.

3. The underwater pile foundation condition detection method according to claim 1, characterized in that, The processing of the echo signal in step S3 includes: The echo signal is preprocessed, enhanced, and its features are extracted to remove interference signals from water flow, bubbles, and marine organisms.

4. The underwater pile foundation condition detection method according to claim 3, characterized in that, Step S2 further includes: controlling the ring sonar array (1) to monitor environmental parameters in real time; The preprocessing of the echo signal includes: An adaptive filtering algorithm is used to eliminate dynamic noise; it includes: using the environmental parameters as input, establishing a noise prediction model, dynamically adjusting the filtering gain, and performing targeted filtering on echo signals of different frequencies.

5. The underwater pile foundation condition detection method according to claim 3, characterized in that, The signal enhancement of the echo signal includes: The preprocessed echo signal is decomposed several times using a wavelet transform algorithm. Each decomposition yields a set of low-frequency coefficients and a set of high-frequency coefficients, where the low-frequency coefficients correspond to the overall contour and the high-frequency coefficients correspond to local details. The low-frequency coefficients in the array are enhanced using a threshold shrinkage method to suppress residual noise; the high-frequency coefficients in the array are amplified using a coefficient reconstruction method to highlight the interface abrupt change characteristics.

6. The underwater pile foundation condition detection method according to claim 3, characterized in that, The step of performing feature extraction processing on the echo signal includes: A deep learning model is introduced to convert the enhanced echo signal into a two-dimensional time-frequency map, which is then input into a trained CNN network to extract spatial features from the two-dimensional time-frequency map. The feature vector output by the CNN network is then input into the LSTM network to capture the time-series features of the signal, thus obtaining the modeled signal.

7. The underwater pile foundation condition detection method according to claim 6, characterized in that, Step S4 includes: Using the center of the flange at the top of the underwater pile foundation (2) as the origin of the coordinate system, a spatial rectangular coordinate system is established. The echo signal is converted into three-dimensional cloud points in the spatial rectangular coordinate system. Then, by comparing the weighted fusion of the features of the modeling signal at different frequencies, the identification of the siltation interface, the basement rock layer interface and the surface contour of the underwater pile foundation is realized. Based on the three-dimensional cloud points, a three-dimensional acoustic field model around the underwater pile foundation is established by combining the siltation interface, the basement rock layer interface, and the surface contour of the underwater pile foundation. Various parameters related to the state of the underwater pile foundation are calculated based on the three-dimensional acoustic field model around the underwater pile foundation.

8. The underwater pile foundation condition detection method according to claim 7, characterized in that, The calculation of various underwater pile foundation state-related parameters based on the three-dimensional acoustic field model surrounding the underwater pile foundation in step S4 includes: Calculate the underwater pile foundation siltation thickness; which includes: at each preset angle of the annular sonar array (1), calculate the vertical distance from the water surface to the siltation interface in the vertical direction to obtain the underwater pile foundation siltation thickness; Calculating the erosion depth of underwater pile foundations includes: obtaining the erosion depth of underwater pile foundations by calculating the difference between the interface of the base rock layer and the designed base elevation; Calculate the inclination angle of the underwater pile foundation; it includes: extracting multiple horizontal sections at preset intervals along the height direction of the underwater pile foundation (2), uniformly selecting multiple sampling points in each horizontal section, fitting the center coordinates of each horizontal section by the least squares method, connecting the multiple center coordinates to obtain the actual axis of the underwater pile foundation (2), and calculating the angle between the actual axis and the vertical line in the spatial rectangular coordinate system to obtain the inclination angle of the underwater pile foundation; The various parameters related to the final underwater pile foundation state mentioned in step S5 are the final underwater pile foundation siltation thickness, the final underwater pile foundation erosion depth, and the final underwater pile foundation tilt angle.

9. The underwater pile foundation condition detection method according to claim 1, characterized in that, The generation of a pile foundation condition inspection report based on various parameters related to the final underwater pile foundation condition includes: The various parameters related to the final underwater pile foundation status are compared with preset alarm thresholds. If any parameter related to the final underwater pile foundation status exceeds its preset alarm threshold, an early warning message is triggered. The pile foundation status detection report includes the early warning message, a visualized three-dimensional sound field model around the underwater pile foundation, and change curves of various parameters related to the underwater pile foundation status.

10. An underwater pile foundation condition monitoring system, characterized in that, include: The detection determination module is used to determine the detection area of ​​the underwater pile foundation (2); A ring sonar array (1) is deployed around the underwater pile foundation (2) within the detection area. The scanning control module is used to control the ring sonar array (1) to scan around the underwater pile foundation (2) and return the echo signal after scanning; The signal acquisition and processing module is used to acquire the echo signal, process the echo signal, and obtain the modeling signal; The visualization and calculation module is used to establish a three-dimensional acoustic field model around the underwater pile foundation based on the modeling signal, and to calculate various underwater pile foundation state-related parameters based on the three-dimensional acoustic field model around the underwater pile foundation. The data fusion and analysis module is used to fuse multiple underwater pile foundation state-related parameters obtained from several scans to obtain multiple final underwater pile foundation state-related parameters, and generate a pile foundation state detection report based on the multiple final underwater pile foundation state-related parameters.

Images