A method and system for supervising dredging construction in waterway areas

Abstract

This invention discloses a method and system for supervising dredging operations in waterway areas, specifically relating to the field of waterway dredging construction supervision technology. It utilizes multi-source sensors and a digital twin platform to collect real-time data on cutting head rotation speed, propulsion force, and cutting depth, and performs time alignment and interpolation resampling with reference parameters to generate a synchronous comparison vector. Statistical features are extracted within a fixed-duration sliding window and combined with principal component analysis or kernel function mapping and time-series coding to generate a deviation feature set. Spectral coupling analysis is performed on the principal component sequence and the water flow disturbance sequence to calculate the average coupling degree and obtain a compliance decision coefficient through logarithmic mapping. When the compliance decision coefficient exceeds the limit, an automatic digital signature is generated and a compliance report is pushed. The sampling and reporting cycle is dynamically adjusted based on the rate of change of the compliance decision coefficient. This improves the real-time performance, accuracy, and traceability of dredging operation supervision.

Description

Technical Field

[0001] This invention relates to the field of waterway dredging construction supervision technology, and more specifically, to a waterway dredging construction supervision method and system. Background Technology

[0002] During dredging operations in waterways, vessels are typically equipped with multi-source sensors and digital twin platforms to collect key operating parameters of dredging equipment, including the cutting head, in real time. As the component that directly impacts underwater sediment, the cutting head's rotational speed, propulsion force, and cutting depth determine dredging efficiency, cross-sectional accuracy, and sediment disturbance; therefore, it is monitored as a core indicator. Through continuous monitoring of the cutting head's operational status, the system can synchronously collect and visualize its real-time parameters against control reference values ​​in design specifications.

[0003] The existing monitoring system can display the dynamic changes of the construction section and the working environment in real time, and automatically link and archive the cutting head operating parameters with the construction log, providing traceable data support for remote monitoring and subsequent auditing.

[0004] However, in complex aquatic environments, the cutting head is significantly affected by flow velocity disturbances and changes in seabed sediment, easily leading to abnormal fluctuations in rotation speed, propulsion force, and cutting depth, thus causing construction deviations. If the regulatory process lacks an automated comparison and compliance assessment mechanism based on real-time data, deviations cannot be identified in a timely manner, compliance reports cannot be generated, and verifiable records cannot be formed. The continuous accumulation of deviations may not only result in substandard dredging sections and difficulty in verifying sediment discharge, but will also severely restrict the effective implementation of quality acceptance and environmental audits, thereby causing navigation safety hazards and environmental compliance risks. Summary of the Invention

[0005] To overcome the aforementioned deficiencies of existing technologies, this invention provides a method and system for supervising dredging construction in waterway areas. By constructing a synchronous comparison mechanism between reference parameters and real-time monitoring data, and combining multi-source feature compression, spectral coupling analysis, and dynamic threshold judgment, a quantifiable and traceable compliance judgment process is formed, enabling accurate identification and automatic reporting of dredging deviations. This solves the problems of existing methods being unable to detect anomalies in real time and lacking effective evidence storage.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for supervising dredging construction in a waterway area, comprising:

[0007] Step 1: After aligning the actual rotational speed, propulsion force, and cutting depth of the cutting head with the reference parameters in time, a synchronization comparison vector is generated;

[0008] Step 2: Input the synchronization alignment vector into the mapping engine, and use a multi-dimensional mapping algorithm to extract and form a deviation feature set of the quantization device execution deviation;

[0009] Step 3: Combine the characteristics of cutting force fluctuations reflected in the deviation feature set with the characteristics of water flow disturbance for interactive analysis. Perform Discrete Fourier Transform (DFT) on the principal component time series of the deviation feature set to obtain the deviation spectral components; perform DFT on the environmental disturbance time series to obtain the disturbance spectral components; calculate the power spectra of the deviation spectral components, the disturbance spectral components, and their cross-power spectra respectively; in the low-frequency range, integrate and accumulate the amplitude of the cross-power spectrum to obtain the low-frequency cooperative response; integrate their respective power spectra in the same frequency band to obtain the total energy level of the deviation signal and the disturbance signal; calculate the average coupling strength factor with the cross-power spectrum integral as the numerator and the arithmetic mean of the deviation power spectrum and the disturbance power spectrum integrals as the denominator; perform logarithmic mapping on the average coupling strength factor to generate compliance decision coefficients for compliance judgment.

[0010] Step 4: When the compliance decision coefficient reaches the preset risk threshold, a compliance report with an electronic signature is automatically generated and pushed to the regulatory authority in real time;

[0011] Step 5: Adjust the monitoring granularity and reporting frequency according to the time-series changes of the compliance decision coefficient to optimize regulatory efficiency; write the compliance report into the tamper-proof evidence repository for subsequent audit traceability.

[0012] Preferably, the reference parameters are obtained in the following way:

[0013] Extract the target cross-sectional dimensions, pipe diameter specifications, and ideal rotation speed curve, thrust curve, and cutting depth curve of the cutting head under typical silt and sand physical properties from the construction design documents and equipment calibration reports to form a reference parameter set. The reference parameter set serves as a constant design specification reference for dynamic environmental adaptive correction, providing a traceable engineering benchmark for the real-time generated reference parameters, ensuring that the corrected construction parameters are always anchored to the original design requirements.

[0014] The current flow rate data is obtained using an online flow rate meter, and the bottom hardness data is obtained using a bottom hardness sensor. The flow rate correction coefficient and hardness correction coefficient are calculated using a linear interpolation method in the historical calibration dataset.

[0015] The flow velocity correction factor is multiplied with the reference speed curve point by point to generate the corrected speed curve, and the hardness correction factor is multiplied with the reference thrust curve point by point to generate the corrected thrust curve.

[0016] The calibration speed curve, calibration thrust curve and reference depth of cut curve are input into a multivariate regression model. The three are weighted by the regression coefficients to generate a reference depth of cut curve.

[0017] The generated correction speed curve, correction thrust curve, and reference cutting depth curve constitute a reference parameter set, which is stored in the comparison buffer according to the time series. Each reference parameter set is assigned a unique version identifier and archived together with the current flow rate data, substrate hardness data, and correction coefficients.

[0018] Preferably, the specific implementation of step one includes:

[0019] Collect data on the actual rotational speed, propulsion force, and depth of cut of the cutting head, including timestamps.

[0020] Extract the reference rotation speed, reference propulsion force, and reference cutting depth, along with their effective time intervals, for the corresponding working condition segment from the comparison buffer.

[0021] Determine the common comparison time range, and use linear interpolation to smooth the missing data for both the actual parameter sequence and the reference parameter sequence;

[0022] The two sequences are resampled at the same sampling frequency to generate a real parameter reconstruction sequence and a reference parameter reconstruction sequence with consistent time nodes;

[0023] The reconstruction parameters of each time point are combined into a synchronous comparison vector and written into the comparison buffer in chronological order.

[0024] The explanation is that after time alignment and interpolation resampling are completed, the generated synchronization alignment vector meets the input requirements for subsequent deviation extraction.

[0025] Preferably, the mapping engine establishes a transformation relationship based on similarity function and numerical distribution characteristics for each type of parameter difference according to the input synchronous comparison vector (containing a time-by-time difference sequence between actual rotational speed, propulsion force, and cutting depth and theoretical reference parameters), and outputs a standardized deviation metric. The mapping engine uses a calibrated input feature template set for normalization, archiving, and feature dimension consistency processing, so that the execution deviation under different times and working conditions has comparable and quantifiable characteristics.

[0026] The multidimensional mapping algorithm refers to a mapping method used to mine the coupling relationship between features in a multi-parameter input space and generate a vectorized feature set that can be used for equipment deviation identification; the multidimensional mapping algorithm is one of three methods or a combination of them: principal component analysis dimensionality reduction and reconstruction method, kernel function mapping nonlinear transformation method, and sliding window temporal pattern coding method.

[0027] Preferably, the principal component time series refers to the time variation data sequence of several main principal components extracted by principal component analysis from the deviation feature set; the environmental disturbance time series records environmental disturbance information such as the rate of change of flow velocity, the angular velocity of flow direction deviation, and the water pressure gradient within the acquisition period; the deviation spectrum component is used to reveal the energy distribution and variation law of the equipment execution deviation signal in different frequency bands, thereby reflecting the periodic anomalies and dynamic characteristics in the equipment operation status; the disturbance spectrum component is used to reflect the energy distribution of the water flow environmental disturbance signal in each frequency component, revealing the frequency characteristics and intensity of periodic or non-periodic disturbances in the environment.

[0028] Preferably, the preset risk threshold setting operation includes the following:

[0029] Data cleaning and anomaly removal: The median absolute deviation method is used to clean the time series of compliance decision coefficients; the absolute deviation of each point is calculated with the median of the whole series as the center, and the median is taken as the fluctuation scale. Points with an absolute deviation of more than three times are removed to form a stable reference sample.

[0030] Distribution Fitting and Interval Analysis: For the cleaned reference sample, a normal or log-normal distribution is used for fitting; the least squares method or maximum likelihood estimation method is used to obtain the distribution parameters including the mean and standard deviation; if the distribution does not meet the conventional assumptions, nonparametric kernel density estimation is used to model its probability density function;

[0031] Set a tolerable range: Based on the operational error limits in the construction quality inspection and evaluation standards (or other applicable industry specifications, set manually), and the typical response range of environmental disturbances, combined with the fitted distribution function, set a preset risk threshold for the compliance decision coefficient; specifically, in the cumulative distribution function of the fitted curve, select the upper limit point corresponding to the cumulative distribution probability value as the threshold point, for example, select the judgment coefficient value corresponding to the 98th percentile, ensuring that the coverage rate within the normal fluctuation range is greater than 95%;

[0032] On-site verification and dynamic adjustment: During the actual construction phase, real-time judgment coefficient data are continuously collected, and the proportion of new data falling below the threshold is statistically analyzed. If this proportion is consistently lower than expected (e.g., less than 90%), it indicates that the current threshold is too low and needs to be appropriately increased; conversely, the same applies. The adjustment range is increased or decreased by referring to the fluctuation range of the weekly collected average and standard deviation.

[0033] Preferably, the specific implementation of step four is as follows:

[0034] Compare the compliance decision coefficient with the preset risk threshold to identify events that exceed the limit;

[0035] Summarize the synchronous comparison vector segments, principal component sequences of deviation features, and environmental disturbance records corresponding to the out-of-limit deviation events, and generate a draft structured report;

[0036] Hash the report draft summary and digitally sign the summary using an RSA private key;

[0037] Add a timestamp verified by the time synchronization server and the signer's identifier to the end of the signed report;

[0038] Compliance reports with electronic signatures are pushed to the regulatory authorities via HTTPS, and transmission confirmation logs are recorded locally.

[0039] The explanation is that after completing the digital signature and secure push, the compliance report has verifiable authenticity and completeness, and is available for regulatory approval in real time.

[0040] Preferably, in step five, the fluctuation level is determined based on the rate of change of the compliance decision coefficient, and a dynamic sampling and reporting strategy is implemented to improve monitoring sensitivity and optimize resource allocation. The specific implementation method is as follows:

[0041] Outlook Summary: After the compliance decision coefficient is updated in real time, the data collection and reporting pace needs to be dynamically adjusted based on its fluctuations, taking into account both response speed and resource utilization;

[0042] A sliding window is applied to the latest compliance decision coefficient sequence to calculate the sequence of absolute values ​​of the differences between adjacent coefficients;

[0043] The average rate of change is obtained by taking the arithmetic mean of the absolute values ​​of the differences.

[0044] The average rate of change is compared with pre-determined high volatility thresholds and low volatility thresholds to classify them into high, medium, and low volatility levels.

[0045] The data sampling period and reporting period are adjusted according to the fluctuation level, and set to half, twice, and half of the original period, respectively.

[0046] Record the adjusted monitoring granularity and reporting frequency, and recall them during the next data collection and report generation.

[0047] Preferably, the method includes a dynamic resilience monitoring step:

[0048] Collect data on the actual rotational speed, propulsion force, and cutting depth of the cutting head, and simultaneously acquire the ship's position coordinates and the hardness spectrum of the seabed.

[0049] When the ship enters the preset environmentally sensitive area, a degradation coefficient for the cutting depth is generated, and the reference cutting depth is corrected based on the degradation coefficient; when the hardness spectrum of the seabed changes abruptly within the sliding window (i.e., the seabed region of the abrupt change), a speed reduction coefficient and a compensation duration coefficient are generated.

[0050] Based on the correction or generation results, generate speed reduction commands, cutting depth reduction commands, or duration compensation commands.

[0051] Based on the type of control instructions generated, the preset risk threshold of the compliance decision coefficient is dynamically lowered;

[0052] The system performs spatial and temporal dual calibration on the mutated substrate area, outputs spatial coordinates and priority labels, transforms uncontrollable anomalies into spatial tasks to be repaired, and achieves decoupled management of construction efficiency and quality.

[0053] To achieve the above objectives, the present invention provides the following technical solution: a waterway dredging construction monitoring system, comprising:

[0054] Data synchronization module: Collects three types of raw data of equipment during construction: rotation speed, propulsion force, and cutting depth. It then synchronizes these three types of data with the reference parameters defined in the design specification library in the time dimension, generates a synchronization comparison vector under the corresponding timestamp, and passes it to the feature extraction module.

[0055] Feature extraction module: Input the synchronous comparison vector into the mapping engine, use multi-dimensional mapping algorithms such as principal component analysis (PCA) to extract the principal component subset that reflects the characteristics of the device execution deviation, and calculate the statistics (mean, kurtosis, skewness, etc.) and covariance matrix to form a low-dimensional deviation feature set containing the main information, which is then passed to the spectrum coupling analysis module.

[0056] Spectrum Coupling Analysis Module: Extracts principal component time series from deviation feature set and simultaneously acquires environmental disturbance time series; performs discrete Fourier transform on both to obtain corresponding spectral components; calculates spectral energy (power spectrum) and interaction quantity (mutual power spectrum), integrates within a specified low-frequency range to obtain low-frequency cooperative response quantity as numerator, and uses the arithmetic mean of the integral values ​​of deviation power spectrum and disturbance power spectrum as denominator to calculate average coupling strength factor, and generates compliance decision coefficients for compliance judgment through logarithmic mapping, which are then transmitted to the compliance reporting module;

[0057] Compliance reporting module: Compares the compliance decision coefficient with the preset risk threshold. If the threshold is reached or exceeded, the automatic generation process of the compliance report is triggered. The report content includes the comparison vector, deviation analysis results, decision coefficient, and timestamp information. After adding electronic signature, it is pushed to the regulatory end in real time and transmitted to the evidence storage and control module.

[0058] Evidence Preservation and Control Module: Writes electronic compliance reports into an tamper-proof evidence repository to enable data traceability and audit support; at the same time, it dynamically adjusts the data monitoring granularity and reporting frequency based on the time-series changes in compliance decision coefficients, optimizes the allocation of computing and communication resources while ensuring regulatory integrity, and feeds back to the data synchronization module for adjusting the collection configuration.

[0059] The technical effects and advantages of this invention are as follows:

[0060] This method first precisely aligns and quantifies the operating parameters of the cutting head with design specifications, achieving accurate quantitative perception of the equipment's execution status. Then, through multi-dimensional mapping and in-depth interactive analysis of hydrodynamic disturbances, it obtains compliance decision coefficients with risk indicators, ensuring that the root causes of execution deviations are fully identified. Based on these coefficients, electronic signature reports are automatically triggered and pushed in real time, establishing an instant verification and archiving mechanism. Simultaneously, the monitoring granularity and reporting frequency are optimized using the temporal changes of the coefficients, achieving a dynamic balance between resources and efficiency. Through the organic collaboration of full-process coverage of data collection, feature extraction, decision-making, report generation, and tamper-proof evidence storage, an efficient, reliable, and traceable dredging construction supervision system is constructed, significantly improving the real-time performance, accuracy, construction quality, and environmental protection levels of supervision. Attached Figure Description

[0061] Figure 1 This is a flowchart of the waterway area dredging construction supervision method of the present invention.

[0062] Figure 2 This is a flowchart for obtaining the compliance decision coefficient of the present invention.

[0063] Figure 3 This is a flowchart of the dynamic resilience monitoring method of the present invention. Detailed Implementation

[0064] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0065] At the same time, it should be understood that, for ease of description, the dimensions of the various parts shown in the accompanying drawings are not drawn according to actual scale.

[0066] The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit the scope of this application and its application or use.

[0067] Techniques, methods, and equipment known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and equipment should be considered part of the specification.

[0068] Example 1, see Figure 1 The present invention provides a flowchart of a method for supervising dredging construction in waterway areas. Figure 1 The method for supervising dredging construction in a waterway area, as shown, includes:

[0069] Step 1: After aligning the actual rotational speed, propulsion force, and cutting depth of the cutting head with the reference parameters in time, a synchronization comparison vector is generated;

[0070] Step 2: Input the synchronization alignment vector into the mapping engine, and use a multi-dimensional mapping algorithm to extract and form a deviation feature set of the quantization device execution deviation;

[0071] Step 3: Combining the principal component time series reflecting the cutting force fluctuation and the water flow disturbance time series, frequency domain transformation is performed to extract their spectral components, calculate their respective power spectra and cross power spectra, and extract the cooperative response features in the low frequency band; by constructing the average coupling strength factor and performing logarithmic mapping, compliance decision coefficients characterizing the degree of correlation between deviation and disturbance are generated.

[0072] Step 4: When the compliance decision coefficient reaches the preset risk threshold, a compliance report with an electronic signature is automatically generated and pushed to the regulatory authority in real time;

[0073] Step 5: Adjust the monitoring granularity and reporting frequency according to the time-series changes of the compliance decision coefficient to optimize regulatory efficiency; write the compliance report into the tamper-proof evidence repository for subsequent audit traceability.

[0074] Logarithmic mapping is performed on the average coupling strength factor to generate compliance decision coefficients. The range of the average coupling strength factor is standardized into a finite judgment interval through logarithmic mapping. The compliance decision coefficients are used as a quantitative result to measure whether the current deviation is mainly driven by external disturbances or caused by the equipment itself deviating from the control trajectory. The compliance decision coefficients increase monotonically with the increase of coupling strength, and have clear physical interpretability and settable judgment thresholds.

[0075] The compliance decision coefficients obtained based on spectral coupling strength and logarithmic mapping can accurately reflect the deep correlation between equipment deviation and environmental disturbance, providing a quantitative basis for risk assessment.

[0076] In this embodiment of the invention, it should be explained that the reference parameters are obtained in the following way:

[0077] Extract the target cross-sectional dimensions, pipe diameter specifications, and ideal rotation speed curve, thrust curve, and cutting depth curve of the cutting head under typical silt and sand physical properties from the construction design documents and equipment calibration reports to form a reference parameter set. The reference parameter set serves as a constant design specification reference for dynamic environmental adaptive correction, providing a traceable engineering benchmark for the real-time generated reference parameters, ensuring that the corrected construction parameters are always anchored to the original design requirements.

[0078] The current flow rate data is obtained using an online flow rate meter, and the bottom hardness data is obtained using a bottom hardness sensor. The flow rate correction coefficient and hardness correction coefficient are calculated using a linear interpolation method in the historical calibration dataset.

[0079] The online flow velocity detector collects the current flow velocity in real time and calculates the corresponding flow velocity correction coefficient by linear interpolation based on the discrete calibration points of each flow velocity – ideal cutting head rotation speed in the historical calibration dataset; the substrate hardness sensor collects the current substrate hardness value in real time and calculates the corresponding hardness correction coefficient by linear interpolation based on the discrete calibration points of each hardness – required propulsion force in the historical calibration dataset.

[0080] The flow velocity correction factor is multiplied with the reference speed curve point by point to generate the corrected speed curve, and the hardness correction factor is multiplied with the reference thrust curve point by point to generate the corrected thrust curve.

[0081] The explanation is as follows: During construction, the reference rotational speed curve reflects the standard rotational speed that the cutting head should maintain over time under ideal flow conditions, while the flow rate correction coefficient describes the degree of deviation of the current water flow velocity from the ideal conditions. Multiplying the two at time points yields the corrected rotational speed target under the actual flow rate, which maintains the time change trend of the original design curve and takes into account the impact of flow rate changes on cutting efficiency and equipment stability in real time. Similarly, the reference propulsion force curve represents the propulsion force required by the equipment under typical substrate hardness, and the hardness correction coefficient reflects the difference between the current substrate hardness and the calibration conditions. By multiplying the two at time points, the generated corrected propulsion force curve can dynamically adjust the propulsion force output, ensuring accurate cutting depth while avoiding overload or shutdown caused by hard obstacles.

[0082] The calibration speed curve, calibration thrust curve and reference depth of cut curve are input into a multivariate regression model. The three are weighted by the regression coefficients to generate a reference depth of cut curve.

[0083] The explanation explains that in actual construction, the cutting depth is influenced by three factors: rotational speed, propulsion force, and designed cutting depth. The corrected rotational speed curve, corrected propulsion force curve, and reference cutting depth curve are used as inputs and fed into a multivariate regression model previously calibrated through historical experiments or simulations. This regression model contains regression coefficients for each input factor and can automatically calculate the optimal cutting depth after weighting each factor based on real-time changes in rotational speed and propulsion force, as well as the original designed cutting depth target. The reference cutting depth curve generated in this way retains the engineering design requirements while dynamically adapting to environmental and geological conditions, improving the accuracy of cutting depth prediction and construction reliability.

[0084] The generated correction speed curve, correction thrust curve, and reference cutting depth curve constitute a reference parameter set, which is stored in the comparison buffer according to the time series. Each reference parameter set is assigned a unique version identifier and archived together with the current flow rate data, substrate hardness data, and correction coefficients.

[0085] In one possible embodiment, by providing a reference parameter prediction model, the shortcomings of traditional reference parameter design specifications are compensated for, enabling real-time prediction and adaptive adjustment of equipment operating parameters, thereby improving construction efficiency and reducing environmental impact. The reference parameter prediction model is constructed using machine learning or deep learning methods, combining historical construction data and real-time environmental data to dynamically adjust equipment operating parameters during construction. For ease of understanding, this invention provides several feasible model selection schemes, including linear regression models, support vector machine regression (SVR), and long short-term memory networks (LSTM). Linear regression is suitable for scenarios where there is a linear relationship between equipment parameters and environmental factors; SVR is used to handle nonlinear problems; and LSTM... This approach is suitable for capturing dependencies in time-series data; a suitable model is selected based on the data characteristics; training data includes real-time collected data such as the cutting head rotation speed, propulsion force, and cutting depth, as well as environmental data such as water flow velocity and seabed hardness; data preprocessing steps include normalization and missing value imputation to ensure the consistency and completeness of the model input data; supervised learning is used during training, and model parameters are adjusted by optimizing loss functions, such as mean squared error (MSE) or Huber loss function; to prevent overfitting, cross-validation and early stopping are used to evaluate the model's generalization ability; environmental data is collected through real-time sensors, including flow velocity, seabed hardness, and ship position, ensuring that the model can be dynamically adjusted according to real-time data. Rated equipment parameters such as rotation speed, propulsion force, and cutting depth are provided by the equipment manual or calibration report, while correction coefficients are dynamically generated based on real-time environmental data.

[0086] In this embodiment of the invention, it should be explained that the specific implementation of step one includes:

[0087] Collect data on the actual rotational speed, propulsion force, and depth of cut of the cutting head, including timestamps.

[0088] Extract the reference rotation speed, reference propulsion force, and reference cutting depth, along with their effective time intervals, for the corresponding working condition segment from the comparison buffer.

[0089] Determine the common comparison time range, and use linear interpolation to smooth the missing data for both the actual parameter sequence and the reference parameter sequence;

[0090] The two sequences are resampled at the same sampling frequency to generate a real parameter reconstruction sequence and a reference parameter reconstruction sequence with consistent time nodes;

[0091] The reconstruction parameters of each time point are combined into a synchronous comparison vector and written into the comparison buffer in chronological order.

[0092] The explanation is that after time alignment and interpolation resampling are completed, the generated synchronization alignment vector meets the input requirements for subsequent deviation extraction.

[0093] In this embodiment of the invention, it should be explained that the mapping engine, based on the input synchronous comparison vector (containing a time-by-time difference sequence between actual rotational speed, propulsion force, and cutting depth and theoretical reference parameters), establishes a transformation relationship based on similarity function and numerical distribution characteristics for each type of parameter difference, and outputs a standardized deviation metric. The mapping engine uses a calibrated input feature template set for normalization, archiving, and feature dimension consistency processing, so that the execution deviations under different times and operating conditions have comparable and quantifiable characteristics.

[0094] The multidimensional mapping algorithm refers to a mapping method used to mine the coupling relationship between features in a multi-parameter input space and generate a vectorized feature set that can be used for equipment deviation identification; the multidimensional mapping algorithm is one of three methods or a combination of them: principal component analysis dimensionality reduction and reconstruction method, kernel function mapping nonlinear transformation method, and sliding window temporal pattern coding method.

[0095] Method 1: The implementation method of principal component analysis dimensionality reduction and reconstruction is as follows:

[0096] The synchronous comparison vector is segmented using a fixed-duration sliding window method, with each segment containing continuous time-series data on rotational speed, propulsion force, and cutting depth.

[0097] Calculate the mean, kurtosis, and skewness of the data within each window to form preliminary statistical characteristics;

[0098] Based on statistical characteristics, a covariance matrix is ​​constructed, and principal component analysis is performed to extract principal components with a cumulative contribution rate of not less than 95%.

[0099] The extracted principal components and their contribution rates are encapsulated into a bias feature set and a principal component identifier is attached.

[0100] The explanation is that, through sliding window and principal component analysis, the deviation feature set is compressed into a low-dimensional vector that reflects the main deviation information, providing a reliable feature basis for interactive analysis.

[0101] Furthermore, in addition to principal component analysis, to improve the ability of the deviation feature set to identify execution anomalies under complex disturbance conditions, the multidimensional mapping algorithm further includes the following two mapping methods, which can be flexibly combined and invoked according to real-time aquatic environmental conditions:

[0102] Method 2: The implementation of the kernel function mapping nonlinear transformation method is as follows:

[0103] The three types of working condition data (i.e., speed difference sequence, propulsion force difference sequence, and cutting depth difference sequence) in each sliding window are normalized and feature combination vectors are constructed in time order. The Gaussian kernel function is selected as the kernel mapping method to map the sample points in the original linear feature space to the high-dimensional feature space. This mapping process can reveal nonlinear coupling features that are not easy to observe in the original space.

[0104] In the high-dimensional feature space, the nonlinear coupling features between hydrodynamic disturbances and equipment response are extracted by calculating the cosine similarity between each feature vector and the rate of change of distance between adjacent samples; all extracted coupling feature indices are packaged into a subset of the deviation feature set in a standard format.

[0105] Method 3: The implementation method of the sliding window timing pattern encoding method is as follows:

[0106] Within a fixed-length sliding window, first-order difference calculations are performed on the three types of difference data to extract the rate of change vectors of each physical quantity. Extreme points are detected, and their frequency, maximum jump amplitude, and jump duration are recorded to form a time-series abrupt change feature set. Furthermore, for equipment behavior exhibiting periodic disturbance characteristics, Fourier transform is used to extract the dominant frequency component and amplitude information within the window to determine whether the deviation response has periodic synchronization characteristics. All abrupt change indicators and frequency domain features are uniformly encoded and then incorporated into the deviation feature set.

[0107] In summary, all three methods described above perform feature mining on three types of difference data constructed from synchronous comparison vectors. They can capture linear trends, nonlinear coupling (kernel functions), and abrupt response (temporal coding) information, respectively. The combined output deviation feature set is more diverse and complete. The mapping engine dynamically selects or executes the above mapping methods in parallel according to preset working condition classification rules, ensuring that stable and reliable deviation information can be extracted under various construction disturbance environments, thus providing a solid feature foundation for interactive analysis and compliance judgment in subsequent steps.

[0108] In this embodiment of the invention, it should be explained that the principal component time series refers to the time variation data sequence of several main principal components extracted by principal component analysis from the deviation feature set; the environmental disturbance time series records environmental disturbance information such as the rate of change of flow velocity, the angular velocity of flow direction deviation, and the water pressure gradient within the acquisition period; the deviation spectrum component is used to reveal the energy distribution and variation law of the equipment execution deviation signal in different frequency bands, thereby reflecting the periodic anomalies and dynamic characteristics in the equipment operation state; the disturbance spectrum component is used to reflect the energy distribution of the water flow environmental disturbance signal in each frequency component, revealing the frequency characteristics and intensity of periodic or non-periodic disturbances in the environment.

[0109] In the embodiments of the present invention, it is necessary to explain that, see the following: Figure 2 The flowchart for obtaining compliance decision coefficients, and the specific implementation method for step three are as follows:

[0110] Perform a discrete Fourier transform on the principal component time series in the deviation feature set to obtain the deviation spectral components;

[0111] To explain, the principal component time series refers to the time variation data sequence of several main principal components extracted by principal component analysis from the set of deviation features. Specifically, step two constructs a covariance matrix for the statistical characteristics (including mean, kurtosis, skewness, etc.) within the sliding window of the execution parameters. Subsequently, principal component analysis is used to extract principal components with a cumulative contribution rate of not less than 95% from the multidimensional statistical characteristics. These principal components represent the main variation directions and important signal components in the original multidimensional data, and their time series is the value change of the principal components at the corresponding time points in different sliding windows. Therefore, the principal component time series is the direct output result of step two after multidimensional mapping and dimensionality reduction of the execution parameter deviation, and it is an effective compression and expression of the original signal anomaly information. The deviation spectral components are used to reveal the energy distribution and variation law of the equipment execution deviation signal in different frequency bands, thereby reflecting the periodic anomalies and dynamic characteristics in the equipment operation status, and helping to identify the key frequency bands and potential anomaly patterns of execution parameter fluctuations.

[0112] Perform a discrete Fourier transform on the time series of environmental disturbances to obtain the spectral components of the disturbances;

[0113] The explanation is that the environmental disturbance time series records environmental disturbance information such as the rate of change of flow velocity, the angular velocity of flow direction offset, and the water pressure gradient within the acquisition period; the disturbance spectrum component is used to reflect the energy distribution of the water flow environmental disturbance signal on each frequency component, revealing the frequency characteristics and intensity of periodic or non-periodic disturbances in the environment, which is convenient for analyzing the degree of influence of environmental changes on the equipment's operating status and its dynamic response law.

[0114] The power spectra of the deviation spectral component, the perturbation spectral component, and their cross-power spectra are calculated separately to quantify the energy of a single signal and the interaction between two signals.

[0115] The explanation is as follows: based on the above-mentioned deviation spectrum component and disturbance spectrum component, the power spectrum is calculated to measure the energy concentration of a single signal, and the cross power spectrum is calculated to measure the cooperative fluctuation characteristics of the two in the frequency domain; among them, the power spectrum reflects the frequency domain energy distribution of a single signal, and the cross power spectrum is used to reveal the coherence strength of the two in the same frequency band.

[0116] Within the low-frequency range related to the equipment control rhythm, the amplitude of the cross-power spectrum is integrated and accumulated to obtain the low-frequency cooperative response; and the power spectra of each signal are integrated within the same frequency band to obtain the total energy level of the deviation signal and the disturbance signal.

[0117] The average coupling strength factor is calculated by using the integral value of the mutual power spectrum as the numerator and the arithmetic mean of the integral values ​​of the deviation power spectrum and the disturbance power spectrum as the denominator. The average coupling strength factor is used to measure the energy synchronization between the equipment deviation and the disturbance source.

[0118] Logarithmic mapping is performed on the average coupling strength factor to generate compliance decision coefficients for compliance assessment.

[0119] The explanation is that engineering disturbances typically exhibit low-frequency characteristics, and the structural response of equipment is more likely to show significant feedback in the low-frequency band. Selecting a preset low-frequency band for cross-power spectrum amplitude integration helps eliminate spurious correlation interference caused by high-frequency noise or control noise, retaining only the true physical interaction signal. Therefore, the low-frequency cross-power spectrum integral value is defined as the low-frequency cooperative response quantity, serving as the basis for subsequent coupling strength calculations. The design basis of the average coupling strength factor is that by comparing the energy of the deviation power spectrum and the disturbance power spectrum in a normalized manner, it can objectively reflect the correlation strength between equipment execution deviation and environmental disturbance. Using the arithmetic mean of the two integral values ​​as the denominator avoids the misleading amplification of coupling strength due to unilateral energy differences, ensuring the stability and fairness of the coupling index, thereby quantifying the synchronous change characteristics of the deviation signal and environmental disturbance in the frequency domain, providing a scientific and quantitative basis for subsequent compliance judgments.

[0120] In this embodiment of the invention, it should be explained that the preset risk threshold setting operation includes the following:

[0121] Data cleaning and anomaly removal: The median absolute deviation method is used to clean the time series of compliance decision coefficients; the absolute deviation of each point is calculated with the median of the whole series as the center, and the median is taken as the fluctuation scale. Points with an absolute deviation of more than three times are removed to form a stable reference sample.

[0122] Distribution Fitting and Interval Analysis: For the cleaned reference sample, a normal or log-normal distribution is used for fitting; the least squares method or maximum likelihood estimation method is used to obtain the distribution parameters including the mean and standard deviation; if the distribution does not meet the conventional assumptions, nonparametric kernel density estimation is used to model its probability density function;

[0123] Set a tolerable range: Based on the operational error limits in the construction quality inspection and evaluation standards (or other applicable industry specifications, set manually), and the typical response range of environmental disturbances, combined with the fitted distribution function, set a preset risk threshold for the compliance decision coefficient; specifically, in the cumulative distribution function of the fitted curve, select the upper limit point corresponding to the cumulative distribution probability value as the threshold point, for example, select the judgment coefficient value corresponding to the 98th percentile, ensuring that the coverage rate within the normal fluctuation range is greater than 95%;

[0124] On-site verification and dynamic adjustment: During the actual construction phase, real-time judgment coefficient data are continuously collected, and the proportion of new data falling below the threshold is statistically analyzed. If this proportion is consistently lower than expected (e.g., less than 90%), it indicates that the current threshold is too low and needs to be appropriately increased; conversely, the same applies. The adjustment range is increased or decreased by referring to the fluctuation range of the weekly collected average and standard deviation.

[0125] In this embodiment of the invention, it should be explained that the specific implementation of step four is as follows:

[0126] Compare the compliance decision coefficient with the preset risk threshold to identify events that exceed the limit;

[0127] Summarize the synchronous comparison vector segments, principal component sequences of deviation features, and environmental disturbance records corresponding to the out-of-limit deviation events, and generate a draft structured report;

[0128] Hash the report draft summary and digitally sign the summary using an RSA private key;

[0129] Add a timestamp verified by the time synchronization server and the signer's identifier to the end of the signed report;

[0130] Compliance reports with electronic signatures are pushed to the regulatory authorities via HTTPS, and transmission confirmation logs are recorded locally.

[0131] The explanation is that after completing the digital signature and secure push, the compliance report has verifiable authenticity and completeness, and is available for regulatory approval in real time.

[0132] In this embodiment of the invention, it needs to be explained that in step five, the fluctuation level is determined based on the rate of change of the compliance decision coefficient, a dynamic sampling and reporting strategy is implemented to improve monitoring sensitivity and optimize resource allocation. The specific implementation method is as follows:

[0133] Outlook Summary: After the compliance decision coefficient is updated in real time, the data collection and reporting pace needs to be dynamically adjusted based on its fluctuations, taking into account both response speed and resource utilization;

[0134] A sliding window is applied to the latest compliance decision coefficient sequence to calculate the sequence of absolute values ​​of the differences between adjacent coefficients;

[0135] The average rate of change is obtained by taking the arithmetic mean of the absolute values ​​of the differences.

[0136] The average rate of change is compared with pre-determined high volatility thresholds and low volatility thresholds to classify them into high, medium, and low volatility levels.

[0137] The data sampling period and reporting period are adjusted according to the fluctuation level, and set to half, twice, and half of the original period, respectively.

[0138] Record the adjusted monitoring granularity and reporting frequency, and recall them during the next data collection and report generation.

[0139] Summary: Example 1 provides a method for supervising dredging construction in waterway areas. The key is to achieve precise supervision and compliance determination of the construction process through real-time analysis and dynamic adjustment of multi-dimensional data. By synchronously comparing the actual and reference parameters of the cutting head through time alignment and interpolation resampling, deviation feature sets are extracted and combined with environmental disturbance analysis for interactive analysis, quantifying the coupling degree between deviation and disturbance. By constructing a compliance decision coefficient, the environmental adaptability and compliance of construction behavior can be determined, ensuring that equipment operation is within the prescribed standard range. Furthermore, the accuracy and security of supervision are guaranteed through dynamic adjustment of risk thresholds, real-time report push, and digital signature of reports. Dynamic adjustment of fluctuation levels optimizes monitoring frequency and reporting strategies, thereby improving supervision efficiency and responsiveness. The innovation of this method lies in the deep correlation quantification between equipment execution deviation and environmental disturbance, and the dynamic adaptive adjustment under actual construction conditions, which greatly improves the controllability and safety of the construction process.

[0140] Background Technology: During waterway dredging operations, sensors on the vessel continuously collect data on operating conditions such as cutting head rotation speed, propulsion force, and cutting depth. A digital twin platform simultaneously generates a visualized construction section and environmental model. Supervisory personnel view the construction status through a real-time monitoring interface, and the system automatically archives logs for review. However, when operations enter areas with rapid currents or ecologically sensitive areas, deviations from the preset design specifications begin to appear in the equipment operating parameters. Sudden increases in flow velocity cause the cutting head rotation speed to become uncontrollable, and hard rock layers in the seabed cause abnormal fluctuations in propulsion force. If the operation happens to cross a large benthic habitat reserve, even if the equipment operates according to the original design parameters, it is inevitable that it will cause irreversible disturbance to the underwater ecosystem. Under traditional regulatory logic, these problems are simply categorized as equipment execution deviations or environmental impacts, lacking targeted response mechanisms, resulting in damage to both construction efficiency and ecological protection.

[0141] The technical problem is that existing technologies are unable to dynamically adapt to sudden environmental changes and ecological protection needs during waterway dredging, leading to two core contradictions:

[0142] Environmental response lag: When ships enter ecologically sensitive areas, the monitoring system can record the location coordinates, but it cannot automatically reduce the intensity of equipment operation (such as reducing the cutting depth); the default construction parameters in the design specifications cause excessive disturbance in water-sensitive environments, and pollution can only be retrospectively discovered during the audit stage, lacking real-time intervention capabilities.

[0143] Rigid handling of sudden changes in physical properties: When encountering hard obstacles in the substrate (such as residual concrete blocks or rock layers), the system still rigidly adheres to the original design parameters for compliance comparison. If the equipment forces its way through, it will exacerbate wear and even cause malfunctions and shutdowns; if an emergency shutdown is carried out, it will disrupt the entire construction schedule. This binary logic of complete shutdown or forced passage cannot guarantee the continuity of construction, nor can it accurately correct local quality defects, ultimately lengthening the construction period and increasing costs.

[0144] Example 2 differs from Example 1 in that, see [reference] Figure 3 The flowchart of the dynamic resilience monitoring method is provided. The method includes dynamic resilience monitoring steps, including the following sub-steps:

[0145] Step S11: Collect the actual rotational speed, propulsion force, and cutting depth data of the cutting head, and simultaneously obtain the ship's position coordinates and the hardness spectrum of the seabed.

[0146] The explanation is as follows: the seabed hardness spectrum refers to the frequency domain energy distribution data generated by the fast Fourier transform of the vibration signal periodically collected by the seabed sensor. The ship's position coordinates are provided in real time by the Global Positioning System. This step outputs five types of synchronous data streams: actual rotational speed sequence, propulsion force sequence, cutting depth sequence, ship position coordinate sequence, and seabed hardness spectrum sequence.

[0147] Step S12: When the ship's position enters the preset environmentally sensitive area, a degradation coefficient for the cutting depth is generated, and the reference cutting depth is corrected based on the degradation coefficient; when the bottom hardness spectrum changes abruptly within the sliding window (i.e., the abrupt bottom area), a speed reduction coefficient and a compensation duration coefficient are generated.

[0148] The explanation is as follows: the environmentally sensitive area is a closed polygon area defined by the geofence database. If the ship's position coordinates fall into this polygon, the downgrade will be triggered.

[0149] The downgrade coefficient is generated according to the following rules: a fixed proportional value is mapped to the ecological protection level of the sensitive area (e.g., core protection zone coefficient = 0.7, buffer zone coefficient = 0.9).

[0150] The method for determining a sudden change in the substrate is as follows: within a fixed-duration sliding window, the rate of change of the energy gradient of the main frequency band of the substrate hardness spectrum is calculated. When the rate of change exceeds a set multiple of the historical baseline value, it is determined to be a sudden event.

[0151] Speed ​​reduction factor generation rule: linearly map the percentage of the device's maximum rotational speed to the energy gradient exceeding the multiple (e.g., exceeding 3 times → speed reduction to 70%).

[0152] Compensation duration coefficient generation rule: The single-point cutting time is extended proportionally based on the duration of the mutation (e.g., mutation lasts 10 seconds → compensation duration is 1.2 times the design duration).

[0153] Step S13: Based on the correction or generation results, generate a speed reduction command, a cutting depth reduction command, or a duration compensation command;

[0154] Explanation: If a speed reduction coefficient is generated, then the speed reduction command = speed reduction coefficient × rated speed of the equipment;

[0155] If a degradation factor is generated, then the cutting depth degradation command = degradation factor × designed cutting depth;

[0156] If a compensation duration coefficient is generated, then the duration compensation command = compensation duration coefficient × standard cutting duration;

[0157] Step S14: Based on the type of control instruction generated, dynamically lower the preset risk threshold of the compliance decision coefficient;

[0158] The risk threshold adjustment rule is as follows: when a speed reduction command or duration compensation command is output, the preset risk threshold of the compliance decision coefficient is lowered to the initial value by a set ratio (e.g., threshold = T0 × 0.8); when only a cutting depth reduction command exists, the risk threshold remains unchanged.

[0159] Step S15: Perform spatial and temporal dual calibration on the mutated substrate area, output spatial coordinates and priority labels, transform uncontrollable anomalies into spatial tasks to be repaired, and achieve decoupled management of construction efficiency and quality.

[0160] The explanation is as follows: Spatial calibration is used to pinpoint the location of anomalies by recording the geographical coordinates of the occurrence of sediment mutations (such as latitude and longitude boundaries) through the ship's positioning system; it also defines the physical location of the problem (e.g., hard rock layers exist within 50 meters north of the channel); temporal calibration is used to quantify the intensity of anomalies by extracting the peak value of the spectral energy gradient during the period of the mutation (i.e., time-series intensity); based on the peak value, the urgency level of repair is determined (e.g., strong mutation → red emergency, weak mutation → yellow observation); through spatial anchoring and damage quantification, the time urgency contradiction (whether to handle it immediately or not) is transformed into a spatial resource scheduling problem, turning uncontrollable sudden disturbances into planned, traceable, and verifiable engineering tasks.

[0161] Summary: Example 2 proposes a dynamic resilience monitoring method aimed at addressing sudden environmental changes and ecological protection needs during channel dredging operations. By collecting real-time data on the cutting head, vessel position, and seabed hardness spectrum, the system can automatically identify and respond to environmentally sensitive areas and sudden seabed changes. When a vessel enters an ecologically sensitive area, the system automatically generates a degradation coefficient to correct the cutting depth; when seabed hardness changes abruptly, it generates speed reduction and duration compensation commands. The system achieves real-time intervention and adjustment of the construction process by dynamically adjusting the risk threshold of the compliance decision coefficient. Furthermore, areas with abrupt seabed changes are marked using both spatial and temporal calibration, transforming uncontrollable anomalies into repairable spatial tasks, thus optimizing construction efficiency and quality management.

[0162] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for supervising dredging construction in a waterway area, characterized in that, include: Step 1: After aligning the actual rotational speed, propulsion force, and cutting depth of the cutting head with the reference parameters in time, a synchronization comparison vector is generated; Step 2: Input the synchronization alignment vector into the mapping engine, and use a multi-dimensional mapping algorithm to extract and form a deviation feature set of the quantization device execution deviation; Step 3: Combining the principal component time series reflecting the cutting force fluctuation and the water flow disturbance time series, frequency domain transformation is performed to extract their spectral components, calculate their respective power spectra and cross power spectra, and extract the cooperative response features in the low frequency band; By constructing an average coupling strength factor and performing a logarithmic mapping, compliance decision coefficients representing the degree of correlation between deviation and disturbance are generated. It also includes dynamic resilience regulatory steps: The system collects data on the actual rotational speed, propulsion force, and cutting depth of the cutting head, and simultaneously acquires the ship's position coordinates and the seabed hardness spectrum. When the ship enters a preset environmentally sensitive area, a degradation coefficient for the cutting depth is generated, and the reference cutting depth is corrected based on the degradation coefficient. When the seabed hardness spectrum changes abruptly within a sliding window, a speed reduction coefficient and a compensation duration coefficient are generated. Based on the correction or generation results, a speed reduction command, a cutting depth degradation command, or a duration compensation command is generated. According to the type of control command generated, the preset risk threshold of the compliance decision coefficient is dynamically lowered. The system performs spatial and temporal dual calibration on the abruptly changed seabed area, outputs spatial coordinates and priority labels, and transforms uncontrollable anomalies into spatial tasks to be repaired, achieving decoupled management of construction efficiency and quality. Step 4: When the compliance decision coefficient reaches the preset risk threshold, a compliance report with an electronic signature is automatically generated and pushed to the regulatory authority in real time; Step 5: Adjust the monitoring granularity and reporting frequency based on the time-series changes of the compliance decision coefficient to optimize regulatory efficiency.

2. The method for supervising dredging construction in a waterway area according to claim 1, characterized in that, The reference parameters are obtained as follows: Extract the ideal rotation speed curve, propulsion force curve, and cutting depth curve of the cutting head under the target cross-sectional dimensions, pipe diameter specifications, and typical mud and sand physical properties to form a set of reference parameters; The current flow rate data is obtained using an online flow rate meter, and the bottom hardness data is obtained using a bottom hardness sensor. The flow rate correction coefficient and hardness correction coefficient are calculated using a linear interpolation method in the historical calibration dataset. The flow velocity correction factor is multiplied with the reference speed curve point by point to generate the corrected speed curve, and the hardness correction factor is multiplied with the reference thrust curve point by point to generate the corrected thrust curve. The calibration speed curve, calibration thrust curve and reference depth of cut curve are input into a multivariate regression model. The three are weighted by the regression coefficients to generate a reference depth of cut curve. The generated corrected rotational speed curve, corrected propulsion force curve, and reference cutting depth curve form a reference parameter set, which is then stored in a comparison buffer in time sequence.

3. The method for supervising dredging construction in a waterway area according to claim 1, characterized in that, The mapping engine establishes a transformation relationship based on similarity function and numerical distribution characteristics for each type of parameter difference according to the input synchronous comparison vector, and outputs a standardized deviation metric. The mapping engine uses a calibrated input feature template set for normalization, archiving and feature dimension consistency processing, so that the execution deviation under different times and working conditions has comparable and quantifiable characteristics.

4. The method for supervising dredging construction in a waterway area according to claim 3, characterized in that, The multidimensional mapping algorithm refers to a mapping method used to mine the coupling relationship between features in a multi-parameter input space and generate a vectorized feature set that can be used for equipment deviation identification; the multidimensional mapping algorithm is one or a combination of three methods: principal component analysis dimensionality reduction and reconstruction method, kernel function mapping nonlinear transformation method, and sliding window temporal pattern coding method.

5. The method for supervising dredging construction in a waterway area according to claim 4, characterized in that, Principal component time series refers to the time variation data sequence of several main principal components extracted by principal component analysis from the deviation feature set; environmental disturbance time series records environmental disturbance information including flow velocity change rate, flow direction offset angular velocity and water pressure gradient; deviation spectral components are used to reveal the energy distribution and variation law of equipment execution deviation signal in different frequency bands, thereby reflecting the periodic anomalies and dynamic characteristics in the equipment operation status; The disturbance spectrum component is used to reflect the energy distribution of the water flow environment disturbance signal at each frequency component.

6. The method for supervising dredging construction in a waterway area according to claim 1, characterized in that, The preset risk threshold setting operation includes the following: The median absolute deviation method is used to clean the time series of compliance decision coefficients. The absolute deviation of each point is calculated with the median of the whole series as the center, and the median is taken as the fluctuation scale. Points with an absolute deviation of more than three times are removed to form a stable reference sample. For the cleaned reference sample, a normal or log-normal distribution is used for fitting; the distribution parameters, including the mean and standard deviation, are obtained; if the distribution does not meet the conventional assumptions, nonparametric kernel density estimation is used to model its probability density function. Based on the fitted distribution, the engineering allowable error limit, and the typical environmental disturbance range, the high percentage points of the distribution are selected as the preset risk threshold.

7. The method for supervising dredging construction in a waterway area according to claim 1, characterized in that, The specific implementation method of step four is as follows: Compare the compliance decision coefficient with the preset risk threshold to identify events that exceed the limit; Summarize the synchronous comparison vector segments, principal component sequences of deviation features, and environmental disturbance records corresponding to the out-of-limit deviation events, and generate a draft structured report; Hash the report draft summary and digitally sign the summary using an RSA private key; Add a timestamp verified by the time synchronization server and the signer's identifier to the end of the signed report; Compliance reports with electronic signatures are pushed to the regulatory authorities via HTTPS, and transmission confirmation logs are recorded locally.

8. The method for supervising dredging construction in a waterway area according to claim 1, characterized in that, The specific implementation method for step five is as follows: A sliding window is applied to the latest compliance decision coefficient sequence to calculate the sequence of absolute values ​​of the differences between adjacent coefficients; The average rate of change is obtained by taking the arithmetic mean of the absolute values ​​of the differences. The average rate of change is compared with pre-determined high volatility thresholds and low volatility thresholds to classify them into high, medium, and low volatility levels. The data sampling period and reporting period are adjusted according to the fluctuation level, and set to half, twice, and half of the original period, respectively. Record the adjusted monitoring granularity and reporting frequency, and recall them during the next data collection and report generation.

9. A channel area dredging construction monitoring system, used to implement the method described in claim 1, characterized in that, include: Data synchronization module: Collects three types of raw data of equipment during construction: rotation speed, thrust, and cutting depth, and synchronizes these three types of data with the reference parameters defined in the design specification library in the time dimension, generating a synchronization comparison vector under the corresponding timestamp; Feature extraction module: Input the synchronization comparison vector into the mapping engine, extract the principal component subset reflecting the device execution deviation characteristics through the multi-dimensional mapping algorithm, calculate the statistics and covariance matrix to form the deviation feature set, and pass it to the spectrum coupling analysis module; Spectrum Coupling Analysis Module: Extracts principal component time series from deviation feature set and simultaneously acquires environmental disturbance time series; performs discrete Fourier transform on both to obtain corresponding spectral components; calculates power spectrum and cross power spectrum; calculates integral in low frequency range to obtain low frequency cooperative response as numerator, and the arithmetic mean of integral values ​​of deviation power spectrum and disturbance power spectrum as denominator to calculate average coupling strength factor, and generates compliance decision coefficients for compliance judgment through logarithmic mapping; Compliance reporting module: Compares the compliance decision coefficient with the preset risk threshold. If the threshold is reached or exceeded, the automatic generation process of the compliance report is triggered. The report includes comparison vectors, deviation analysis results, decision coefficients, and timestamp information, and is pushed to the regulatory authorities in real time after being signed with an electronic signature. Evidence Preservation and Control Module: Writes electronic compliance reports into an tamper-proof evidence preservation database to enable data traceability and audit support; Based on the time-series changes in compliance decision coefficients, the data monitoring granularity and reporting frequency are dynamically adjusted to optimize the allocation of computing and communication resources while ensuring regulatory integrity.

Images