Ocean engineering pile foundation scouring experiment maximum scouring depth estimation method based on integrated underwater imager

Abstract

The invention provides an ocean engineering pile foundation scouring experiment maximum scouring depth estimation method based on an integrated underwater imager. According to the integrated underwater imager, a self-adaptive algorithm is matched with a multi-wavelength laser technology, the laser intensity, the wavelength and the scanning speed are adjusted in real time according to the transparency and the turbidity of a water body so as to adapt to different water body conditions, and scour depth data are efficiently and accurately obtained in a complex underwater environment through data fusion processing in combination with a sonic sensor. In the experiment implementation process, the integrated imager carries out data scanning at regular time, the obtained depth-of-impact elevation data is rapidly processed, and the maximum depth of impact and the change rate of the maximum depth of impact are judged in real time. And when the change rate is less than 5% continuously twice, the experiment is judged to reach a balance state, so that time and manpower are saved. Afterwards, the ARIMA model is utilized to analyze and predict the maximum depth data obtained through previous scanning, a more accurate maximum depth value is obtained, and the defects of the experiment are further made up. The scientificity and practicability of the ocean engineering pile foundation scouring experiment are remarkably enhanced, and the ocean engineering pile foundation scouring experiment device has wide application prospects.

Description

Technical Field

[0001] The present invention relates to a method for evaluating and predicting local scour of foundations in indoor water tank experiments, and in particular to a method for predicting the maximum scour depth of a marine engineering pile foundation scour experiment based on an integrated underwater imager, which is suitable for foundation model experiments in marine engineering and port engineering. Background Art

[0002] In marine engineering, port engineering and offshore infrastructure design, local scouring of foundations is crucial to the safety and long-term stability of structures. When foundation structures are affected by water flow, the scouring conditions around them are often complex and changeable. Traditional scouring monitoring methods (such as mechanical measurement or underwater photography) usually have problems such as poor real-time performance, low accuracy, and cumbersome operation. It is difficult to efficiently and accurately capture the scouring and silting changes around the foundation. Especially in foundation tests, the monitoring work is more complicated and it is difficult to obtain dynamic change information of multiple foundations at the same time.

[0003] In addition, existing maximum punching depth assessment methods usually rely on subjective judgment or manual analysis to determine when the experiment reaches equilibrium, and lack objective and real-time monitoring standards. This makes the experimental process prone to lengthy time and inaccurate results, increasing the cost and complexity of the experiment. Summary of the invention

[0004] The present invention provides a method for estimating the maximum scour depth of a marine engineering pile foundation scour experiment based on an integrated underwater imager. This method, combined with an integrated underwater imager, is significantly superior to traditional mechanical measurement and underwater photography techniques. The integrated imager is equipped with a transparency sensor and a turbidity sensor, which can monitor the transparency and turbidity of the water body in real time. It uses an adaptive algorithm to automatically optimize scanning parameters according to the transparency and turbidity of the water body, adjust the laser intensity and scanning speed, so as to achieve intelligent switching of different scanning parameters and improve the accuracy of data acquisition. At the same time, combined with multi-wavelength laser technology, the optimal wavelength is dynamically selected for scanning under different water conditions, so as to effectively respond to changes in water turbidity and improve the stability and reliability of the measurement. In addition, the integrated imager also integrates a sonar sensor, performs sonar scanning at the same time as laser scanning, and fuses the data to provide comprehensive data analysis and enhance performance in complex underwater environments.

[0005] The characteristic of this technology is that it can save calculation amount by pre-screening the points on the foundation, and then quickly screen out the maximum scour points around the foundation according to the elevation changes of the remaining points, which greatly improves the data processing efficiency and has great flexibility in experimental design. It can not only analyze the scour change trend of the foundation through multiple measurements, but also flexibly judge whether the experiment has reached the scour equilibrium state according to the preset time period, ΔZ change amplitude or specific experimental objectives, thereby saving experimental time while ensuring the accuracy of the data and the reliability of the experimental results. After determining the experimental balance, the ARIMA model is used to analyze and estimate the maximum scour values ​​obtained from previous scans to obtain a more accurate maximum scour value, further making up for the shortcomings of the experiment.

[0006] The present invention is achieved through the following technical scheme: a method for estimating the maximum scour depth of a marine engineering pile foundation scour experiment based on an integrated underwater imager, comprising the following steps:

[0007] S1. Initial terrain data measurement and preparation: determine the test section, obtain water conditions, transparency and turbidity data through the equipped transparency sensor and turbidity sensor, and adjust the imager parameters to set the benchmark. After adjusting the parameters, formally scan to obtain the 3D point cloud data of the initial sand pool surface after data fusion processing, and save it as the benchmark for subsequent experiments;

[0008] S2. Subsequent data acquisition and preprocessing: Rapidly monitor water quality changes before each scan. If significant changes are detected, automatically adjust the laser intensity, wavelength and scanning speed, and save the fused data;

[0009] S3. Calculation of the maximum scouring depth around the foundation: remove the points located on the foundation, calculate the elevation changes of the remaining points, and select the maximum scouring points around the foundation;

[0010] S4. Experimental equilibrium state judgment: By analyzing the dynamic changes of the maximum punching depth, it is judged whether the experiment has reached the equilibrium state, thereby deciding whether to end the experiment;

[0011] S5. Maximum punching depth estimation: After the experimental balance is completed, the ARIMA model is used to analyze the maximum punching depth value obtained from each scan and predict the maximum value to improve the scientificity and practicality of the experiment.

[0012] Wherein, the step S1 includes the following sub-steps:

[0013] S101. After leveling the sand pool and allowing it to settle naturally to ensure the terrain is stable, determine the test section, including the scanning starting point, scanning range, scanning accuracy, and the location, centroid, and plane size of the foundation;

[0014] S102. Move the integrated underwater imager to the designated location, obtain water condition, transparency and turbidity data through the transparency sensor and turbidity sensor, perform the initial scan after the imager adjusts the scanning parameters and fuses the data to obtain the initial three-dimensional point cloud data of the sand pool surface after fusion in the entire scanning area;

[0015] S103. The initial scanning data is saved in the form of a time-named dat file, the file content includes the plane coordinates (X, Y) and the corresponding elevation value (Z) for subsequent experimental comparison.

[0016] Wherein, the step S2 includes the following sub-steps:

[0017] S201. Scan regularly according to preset time intervals or experimental requirements, and automatically adjust the imager parameters (laser intensity, wavelength, scanning speed) according to changes in water quality, and record the point cloud data of the sand pool surface at each moment. During each scan, the position, scanning range and accuracy of the imager remain consistent with the initial scan to ensure the uniformity of the measurement area and conditions;

[0018] S202. After the scanning is completed, the fused data file is named by time and transferred to the computer, and the format remains consistent (the first two lines are the coordinate name and the scanning range, and the third line onwards are the measurement point coordinates and the corresponding elevation values).

[0019] Wherein, the step S3 includes the following sub-steps:

[0020] S301. Using the center coordinates and known radius of each foundation, first filter out the data points that fall on the foundation, and only keep the measurement points around each foundation. This can greatly reduce the amount of subsequent calculations;

[0021] Calculate the distance between the measuring point and the center of the foundation. The principle is the following geometric distance formula:

[0022]

[0023] Where L is the distance from the point to the centroid of the foundation. If the distance is less than or equal to the radius of the foundation, that is, L≤R, the point is located on the foundation and needs to be excluded; otherwise, the points around the foundation are retained.

[0024] S302. Calculate the Z value difference between the screened points and the corresponding points of the initial terrain at each time point to obtain the elevation change ΔZ of each point (positive value is siltation, negative value is scour), the principle is the following formula:

[0025] ΔZ=Z i -Z0

[0026] Among them, Z iis the elevation value corresponding to the point at the i-th scan, and Z0 is the elevation value of the corresponding point at the initial scan;

[0027] In order to enhance the stability of the experimental results, the average elevation change of multiple scans can be further calculated using the time integration method:

[0028]

[0029] in, represents the average elevation change within time t, ΔZ(τ) is the elevation change value at time point τ, and t represents the total time from the start of the experiment to the current moment;

[0030] S303. Arrange the ΔZ values ​​of all remaining points in ascending order (maximum scour to minimum scour) and retain the corresponding coordinates (X, Y);

[0031] S304. In the data arranged in ascending order, the first point that appears is the maximum scouring point around the foundation, and the corresponding value is the maximum scouring depth around the foundation.

[0032] Wherein, the step S4 includes the following sub-steps:

[0033] S401. After multiple measurements, the ΔZ obtained from each measurement is dynamically analyzed. If the maximum scour depth around the foundation changes within 5% for two consecutive times, it can be determined that the experiment has reached the scour equilibrium state. The principle is the following formula:

[0034]

[0035] Among them, Z (t) is the current maximum punching depth value, Z (t-1) The maximum punching depth value of the last time;

[0036] S402. According to the experimental design requirements, after reaching the equilibrium state, the measurement can be stopped and the final maximum punching depth data can be recorded;

[0037] In order to enhance the description of real-time analysis, a sliding window method is used to determine the changes in drawing depth in real time:

[0038]

[0039] Among them, ω is the window size (such as the last three scan data), ΔZ i is the elevation change of the i-th scan.

[0040] The step S5 comprises the following sub-steps:

[0041] S501. Collect the maximum depth data obtained from each scan to form a time series. Process missing values ​​and outliers and perform data smoothing (such as moving average);

[0042] S502. Use the autocorrelation function ACF and the partial autocorrelation function PACF to determine the order p, d, q of the model and perform a stationarity test.

[0043]

[0044] p is the number of autoregressive terms, d is the number of differences, q is the number of moving average terms, t is the time point, n is the total number of samples, k is the number of lags, X is the time series data, and Φ is the AR model parameter;

[0045] Then, the ARIMA model is fitted to generate a forecast model.

[0046] X t =μ+φ1X t-1 +...+φ p X t-p +θ 1t-1 +...+θ q ∈ t-q +∈ t

[0047] Among them, μ is the constant term, φ is the autoregressive coefficient, θ is the moving average coefficient, and ε t is the white noise term;

[0048] S503. Calculate AIC and BIC to select the optimal model. Use the fitted model to predict the future maximum impact depth value and generate a confidence interval.

[0049]

[0050] Where L is the log-likelihood value of the model. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 It is a structural method flow chart of the present invention;

[0052] Figure 2 It is a three-dimensional view of the terrain imager and data processor of the present invention;

[0053] Figure 3 It is the maximum drawing depth estimation result of the embodiment of the present invention. DETAILED DESCRIPTION

[0054] The technical solution of the present invention is further described in detail below in conjunction with the accompanying drawings:

[0055] like Figure 1 As shown, a method for estimating the maximum scour depth of a marine engineering pile foundation scour experiment based on an integrated underwater imager includes the following steps:

[0056] S1. Initial terrain data measurement and preparation: Obtain the 3D point cloud data of the initial sand pool surface and save it as a benchmark for subsequent experiments;

[0057] The step S1 comprises the following sub-steps:

[0058] S101. After leveling the sand pool and allowing it to settle naturally to ensure that the terrain is stable, determine the test section, including the scanning starting point, scanning range, scanning accuracy, and the location, centroid, and plane size of the foundation; in the embodiment of the present application, the scanning range is 2400mm*1600mm, and the scanning accuracy is d s =4mm, the scanning results include 240,000 coordinate positions and elevations, the centroid positions of the three foundations are (1578, 937), (976, 1320) and (944, 608), respectively, and the foundation diameter D = 240mm;

[0059] S102. Move the integrated underwater imager to the specified location, and obtain water condition, transparency and turbidity data through the transparency sensor and turbidity sensor. In this embodiment, the measured transparency T = 95%, turbidity C = 1NTU, according to the data in Table 1, get Similarly, Available get according to Select wavelength λ=625, adjust parameters, perform initial scanning and perform data fusion processing, and obtain the initial three-dimensional point cloud data of the sand pool surface after fusion processing in the entire scanning area;

[0060] Table 1 Selection of empirical values ​​of imager parameters

[0061]

[0062] S103. The initial scanning data is saved in the form of a time-named dat file, the file content includes the plane coordinates (X, Y) and the corresponding elevation value (Z) for subsequent experimental comparison.

[0063] S2. Subsequent data acquisition and preprocessing: Scan the sand pool regularly, calculate the difference between the data obtained each time and the initial data, and perform adaptive noise processing according to the scanning noise level to ensure data consistency and quality;

[0064] The step S2 comprises the following sub-steps:

[0065] S201. Scan regularly according to preset time intervals or experimental requirements, and automatically adjust the imager parameters (laser intensity, wavelength, scanning speed) according to changes in water quality, and record the point cloud data of the sand pool surface at each moment. During each scan, the position, scanning range and accuracy of the imager remain consistent with the initial scan to ensure the uniformity of the measurement area and conditions;

[0066] S202. After the scanning is completed, the fused data file is named by time and transferred to the computer, and the format remains consistent (the first two lines are the coordinate name and the scanning range, and the third line onwards are the measurement point coordinates and the corresponding elevation values).

[0067] S3. Calculation of the maximum scouring depth around the foundation: remove the points located on the foundation, calculate the elevation changes of the remaining points, and select the maximum scouring points around the foundation;

[0068] The step S3 comprises the following sub-steps:

[0069] S301. Using the center coordinates and known radius of each foundation, first filter out the data points that fall on the foundation, and only keep the measurement points around each foundation. This can greatly reduce the amount of subsequent calculations;

[0070] Calculate the distance between the measuring point and the center of the foundation. The principle is the following geometric distance formula;

[0071]

[0072] Where L is the distance from the point to the centroid of the foundation. If the distance is less than or equal to the radius of the foundation, that is, L≤R, the point is located on the foundation and needs to be excluded; otherwise, the points around the foundation are retained.

[0073] In a specific embodiment, the geometric distance calculation is performed on each measurement point by using the is_outside_all_circles function in the PyCharm software, and the all() method is used to ensure that the point is outside all bases, and the points that meet this condition will be retained. This can greatly reduce the amount of subsequent calculations;

[0074] S302. Calculate the Z value difference between the eliminated points and the corresponding points of the initial terrain at each time point to obtain the elevation change ΔZ of each point (positive value is siltation, negative value is scour), the principle is the following formula;

[0075] ΔZ=Z i -Z0

[0076] Among them, Z i is the elevation value corresponding to the point at the i-th scan, and Z0 is the elevation value of the corresponding point at the initial scan.

[0077] In order to enhance the stability of the experimental results, the average elevation change of multiple scans can be further calculated using the time integration method:

[0078]

[0079] in, represents the average elevation change within time t, ΔZ(τ) is the elevation change value at time point τ, and t represents the total time from the start of the experiment to the current moment;

[0080] S303. Arrange the ΔZ values ​​of all remaining points in ascending order (maximum scour to minimum scour) and retain the corresponding coordinates (X, Y);

[0081] S304. In the data arranged in ascending order, the first point that appears is the maximum scouring point around the foundation, and the corresponding value is the maximum scouring depth around the foundation;

[0082] In a specific embodiment, the distance from each point to the center of the foundation is calculated by the distance_to_circle function, and then the z value is sorted in ascending order using the find_nearest_points function. The ascending order is used to find the point with the smallest z value (i.e., the point with the largest flushing). Then, the point closest to the foundation is selected as the result;

[0083] S4. Experimental equilibrium state judgment: through the dynamic change analysis of scouring and sedimentation, judge whether the experiment has reached the equilibrium state, so as to determine the end time of the experiment;

[0084] The step S4 comprises the following sub-steps:

[0085] S401. After multiple measurements, the ΔZ obtained from each measurement is dynamically analyzed. If the maximum scour depth around the foundation changes within 5% for two consecutive times, it can be determined that the experiment has reached the scour equilibrium state. The principle is the following formula:

[0086]

[0087] Among them, Z (t) is the current maximum punching depth value, Z (t-1) The maximum punching depth value of the last time;

[0088] In a specific embodiment, the calculate_z_change_rate function is used to traverse each basic point, calculate the difference between the current z value and the last z value, and calculate the change rate according to the above formula. If the last z value is zero, the change rate is set to zero to avoid division by zero errors;

[0089] S402. According to the experimental design requirements, after reaching the equilibrium state, the measurement can be stopped and the final maximum punching depth data can be recorded;

[0090] In order to enhance the description of real-time analysis, a sliding window method is used to determine the change in drawing depth in real time:

[0091]

[0092] Among them, ω is the window size (such as the last three scan data), ΔZ i is the elevation change of the ith scan;

[0093] In a specific embodiment, the check_equilibrium function is used to compare the calculated z-value change rates of the maximum scouring and silting to see whether these change rates are all less than a threshold of 5%. This method not only analyzes the changing trends of scouring and silting in real time, but also predicts the arrival of the equilibrium state in advance.

[0094] The step S5 comprises the following sub-steps:

[0095] S501. Collect the maximum depth data obtained from each scan to form a time series. Process missing values ​​and outliers and perform data smoothing (such as moving average);

[0096] In a specific embodiment, the window size is set by rolling(window=3).mean() to smooth the data;

[0097] S502. Use the autocorrelation function ACF and the partial autocorrelation function PACF to determine the order p, d, q of the model and perform a stationarity test;

[0098]

[0099] p is the number of autoregressive terms, d is the number of differences, q is the number of moving average terms, t is the time point, n is the total number of samples, k is the number of lags, X is the time series data, and Φ is the AR model parameter;

[0100] Then, the ARIMA model is fitted to generate a forecasting model;

[0101] X t =μ+φ1X t-1 +...+φ p X t-p +θ 1t-1 +...+θ q ∈ t-q +∈ t

[0102] Among them, μ is the constant term, φ is the autoregressive coefficient, θ is the moving average coefficient, and ε t is the white noise term;

[0103] In a specific embodiment, the autocorrelation and partial autocorrelation analysis are performed through plot_acf(), plot_pacf() and ARIMA(data,order=(p,d,q)) functions to generate an ARIMA prediction model;

[0104] S503. Calculate AIC and BIC to select the optimal model. Use the fitted model to predict the future maximum impact depth value and generate a confidence interval;

[0105]

[0106] Where L is the log-likelihood value of the model;

[0107] In a specific embodiment, the model is fitted and optimized by the fit() function, and the future value is predicted by fitting the model by the forecast(steps=n) function. The confidence interval is generated for the prediction result by the get_forecast().conf_int() function to measure the uncertainty of the prediction. Considering the stability and confidence interval of the prediction result, the predicted value of the maximum punching depth of 20h is 104mm.

[0108] The embodiments described above provide a detailed description of the technical solutions of the present invention. It should be understood that the above descriptions are only specific embodiments of the present invention and are not intended to limit the present invention. Any modifications, supplements or similar substitutions made within the scope of the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for estimating the maximum scour depth of a marine engineering pile foundation scour experiment based on an integrated underwater imager, characterized in that: The integrated imager is equipped with a transparency sensor and a turbidity sensor. The transparency sensor is used to measure the light transmittance of the water body and output a transparency value T. The turbidity sensor is used to measure the content of suspended particles in the water body and output a turbidity value C, thereby achieving real-time monitoring of the transparency and turbidity of the water body.

2. A method for estimating the maximum scour depth of a marine engineering pile foundation scour experiment based on an integrated underwater imager, characterized in that: The integrated imager is equipped with an adaptive algorithm that can automatically optimize scanning parameters, adjust laser intensity and scanning speed according to water transparency and turbidity in real time, thereby realizing intelligent switching and use of different scanning parameters to improve the accuracy of data collection. The principle of the adaptive algorithm is as follows: In the formula, I is the laser intensity; S is the scanning speed; T is the water transparency; C is the turbidity; α, β, and γ are the preset coefficients obtained by the experiment. The preset coefficients are obtained by the least squares method. X is the input feature matrix and Y is the output result.

3. The maximum scour depth estimation method for marine engineering pile foundation scour experiment based on integrated underwater imager according to claim 2 is characterized by: The adaptive algorithm combined with multi-wavelength laser technology allows the imager to dynamically select the best wavelength for scanning under different water conditions, so as to effectively cope with changes in water turbidity and improve the stability and reliability of measurement. The principle of the wavelength selection algorithm is as follows: ∧=[λ1,λ2,λ3...λ n ] Where ∧ is the wavelength matrix, λ1,λ2,λ3...λ n are the specific wavelength values, T threshold is the water transparency threshold, C threshold is the turbidity threshold.

4. The maximum scour depth estimation method for marine engineering pile foundation scour experiment based on integrated underwater imager according to claim 2, characterized in that: The integrated imager integrates a sonar sensor, performs sonar scanning while performing laser scanning, performs data fusion processing, and performs performance evaluation and error processing on the fused data to ensure the accuracy of data fusion, so as to provide comprehensive data analysis, enhance performance in complex underwater environments, and thus achieve more accurate depth assessment. The data fusion principle is as follows: Where D 融合 is the output scan value after fusion of multiple sensors, ω i is the weight, D i Scan output values ​​for different sensors; Where E is the error, var is the variance, V max is the maximum variance. If the variance exceeds the set threshold, adjust the weight ω i , the weight of sensors with poor performance is reduced, improving the accuracy of data fusion.

5. A method for estimating the maximum scour depth of a marine engineering pile foundation scour experiment based on an integrated underwater imager, characterized in that: The following steps are involved: S1. Initial terrain data measurement and preparation: determine the test section, obtain water conditions, transparency and turbidity data through the equipped transparency sensor and turbidity sensor, and adjust the imager parameters to set the benchmark. After adjusting the parameters, formally scan to obtain the 3D point cloud data of the initial sand pool surface after data fusion processing, and save it as the benchmark for subsequent experiments; S2. Subsequent data acquisition and preprocessing: Rapidly monitor water quality changes before each scan. If significant changes are detected, automatically adjust the laser intensity, wavelength and scanning speed, and save the fused data; S3. Calculation of the maximum scouring depth around the foundation: remove the points located on the foundation, calculate the elevation changes of the remaining points, and select the maximum scouring points around the foundation; S4. Experimental equilibrium state judgment: By analyzing the dynamic changes of the maximum punching depth, it is judged whether the experiment has reached the equilibrium state, thereby deciding whether to end the experiment; S5. Maximum punching depth estimation: After the experimental judgment balance is completed, the ARIMA model is used to analyze the maximum punching depth data obtained from each scan and predict the maximum value to improve the scientificity and practicality of the experiment.

6. The maximum scour depth estimation method for marine engineering pile foundation scour experiment based on integrated underwater imager according to claim 5 is characterized in that : The step S1 comprises the following sub-steps: S101. After leveling the sand pool and allowing it to settle naturally to ensure the terrain is stable, determine the test section, including the scanning starting point, scanning range, scanning accuracy, and the location, centroid, and plane size of the foundation; S102. Move the integrated underwater imager to the designated location, obtain water condition (transparency and turbidity) data through the transparency sensor and turbidity sensor, perform the initial scan after the imager adjusts the scanning parameters, and fuse the data to obtain the initial three-dimensional point cloud data of the sand pool surface after fusion processing in the entire scanning area; S103. The initial scanning data is saved in the form of a time-named dat file, the file content includes the plane coordinates (X, Y) and the corresponding elevation value (Z) for subsequent experimental comparison.

7. The method for estimating the maximum scour depth of a marine engineering pile foundation scour experiment based on an integrated underwater imager according to claim 5 is characterized in that : The step S2 comprises the following sub-steps: S201. Scan regularly according to preset time intervals or experimental requirements, and automatically adjust the imager parameters (laser intensity, wavelength, scanning speed) according to changes in water quality, and record the point cloud data of the sand pool surface at each moment. During each scan, the position, scanning range and accuracy of the imager remain consistent with the initial scan to ensure the uniformity of the measurement area and conditions; S202. After the scanning is completed, the fused data file is named by time and transferred to the computer, and the format remains consistent (the first two lines are the coordinate name and the scanning range, and the third line onwards are the measurement point coordinates and the corresponding elevation values).

8. The method for estimating the maximum scour depth of a marine engineering pile foundation scour experiment based on an integrated underwater imager according to claim 5, characterized in that : The step S3 comprises the following sub-steps: S301. Using the center coordinates and known radius of each foundation, first filter out the data points that fall on the foundation, and only keep the measurement points around each foundation. This can greatly reduce the amount of subsequent calculations; Calculate the distance between the measuring point and the center of the foundation. The principle is the following geometric distance formula: Where L is the distance from the point to the centroid of the foundation. If the distance is less than or equal to the radius of the foundation, that is, L≤R, the point is located on the foundation and needs to be excluded; otherwise, the points around the foundation are retained. S302. Calculate the Z value difference between the screened points and the corresponding points of the initial terrain at each time point to obtain the elevation change ΔZ of each point (positive value is siltation, negative value is scour), the principle is the following formula: ΔZ=Z i -Z0 Among them, Z i is the elevation value corresponding to the point at the i-th scan, and Z0 is the elevation value of the corresponding point at the initial scan; In order to enhance the stability of the experimental results, the average elevation change of multiple scans can be further calculated using the time integration method: in, represents the average elevation change within time t, ΔZ(τ) is the elevation change value at time point τ, and t represents the total time from the start of the experiment to the current moment; S303. Arrange the ΔZ values ​​of all remaining points in ascending order (maximum scour to minimum scour) and retain the corresponding coordinates (X, Y); S304. In the data arranged in ascending order, the first point that appears is the maximum scouring point around the foundation, and the corresponding value is the maximum scouring depth around the foundation.

9. The method for estimating the maximum scour depth of a marine engineering pile foundation scour experiment based on an integrated underwater imager according to claim 5, characterized in that : The step S4 comprises the following sub-steps: S401. After multiple measurements, the ΔZ obtained from each measurement is dynamically analyzed. If the maximum scour depth around the foundation changes within 5% for two consecutive times, it can be determined that the experiment has reached the scour equilibrium state. The principle is the following formula: Among them, Z (t) is the current maximum punching depth value, Z (t-1) The maximum punching depth value of the last time; S402. According to the experimental design requirements, after reaching the equilibrium state, the measurement can be stopped and the final maximum punching depth data can be recorded; In order to enhance the description of real-time analysis, a sliding window method is used to determine the change in drawing depth in real time: Among them, ω is the window size (such as the last three scan data), ΔZ i is the elevation change of the i-th scan.

10. The method for estimating the maximum scour depth of a marine engineering pile foundation scour experiment based on an integrated underwater imager according to claim 5, characterized in that : The step S5 comprises the following sub-steps: S501. Collect the maximum depth data obtained from each scan to form a time series. Process missing values ​​and outliers and perform data smoothing (such as moving average); S502. Use the autocorrelation function ACF and the partial autocorrelation function PACF to determine the order p, d, q of the model and perform a stationarity test. p is the number of autoregressive terms, d is the number of differences, q is the number of moving average terms, t is the time point, n is the total number of samples, k is the number of lags, X is the time series data, and Φ is the AR model parameter; Then, the ARIMA model is fitted to generate a forecast model. X t =μ+φ1X t-1 +...+φ p X t-p +θ 1t-1 +...+θ q ∈ t-q +∈ t In the formula, μ is the constant term, φ is the autoregressive coefficient, θ is the moving average coefficient, and ε t is the white noise term; S503. Calculate AIC and BIC to select the optimal model. Use the fitted model to predict the future maximum impact depth value and generate a confidence interval. Where L is the log-likelihood value of the model.

Images