Offshore wind power plant stratum disaster risk dynamic assessment method and system

Abstract

The invention discloses a dynamic assessment method and system for the disaster risk of an offshore wind plant stratum, and belongs to the technical field of ocean engineering and wind power, and the method comprises the steps: receiving and processing the multi-source monitoring data of the offshore wind plant stratum, and generating a standardized dynamic data set; calling the soil rheological constitutive database and the real-time marine environment parameters, and calculating to obtain dynamic weight parameters; generating a risk probability space projection drawing by using the standardized dynamic data set and the dynamic weight parameters; according to the generated projection drawing, generating a disaster early warning instruction in combination with a disaster chain map; correcting data in the soil body rheological constitutive database by using vibration spectrum data fed back by construction equipment; and the corrected soil body rheology constitutive parameters are transmitted back to the soil body rheology constitutive database. According to the method, real-time fusion processing of multi-source monitoring data is adopted, and disaster risk full-time-domain sensing under the coupling action of the marine environment and the stratum state can be realized in combination with a soil body rheological constitutive parameter online correction mechanism and a disaster chain space topology analysis method.

Description

Technical Field

[0001] This invention relates to the fields of marine engineering and wind power technology, and in particular to a method and system for dynamic assessment of site-specific disaster risks in offshore wind farms. Background Technology

[0002] The dynamic assessment method and system for seabed strata disaster risks in offshore wind farms is a risk monitoring technology based on multi-source data fusion and artificial intelligence. It is used to assess the stability of seabed strata and potential disaster risks in real time, so as to ensure the safety and long-term stable operation of offshore equipment foundation structures.

[0003] This system combines intelligent sensing, the Internet of Things, big data analysis, and numerical simulation to achieve closed-loop management of the entire process from data acquisition and dynamic early warning to decision support.

[0004] In existing technologies, geological risk assessment of offshore wind farms mainly relies on static geological surveys and empirical models, but these methods still have shortcomings in dynamic marine environments. Summary of the Invention

[0005] To address the aforementioned issues, this invention provides a method and system for dynamic assessment of geological disaster risks at offshore wind farm sites. This invention employs real-time fusion processing of multi-source monitoring data, combined with an online correction mechanism for soil rheological constitutive parameters and a disaster chain spatial topology analysis method, enabling full-time perception of disaster risks under the coupled effects of marine environment and geological state.

[0006] The above objectives can be achieved through the following approach:

[0007] A method and system for dynamic assessment of site hazard risks in offshore wind farms, belonging to the fields of marine engineering and wind power technology, includes receiving and processing multi-source monitoring data of offshore wind farm sites to generate a standardized dynamic dataset; calling a soil rheological constitutive database and real-time marine environmental parameters to calculate dynamic weight parameters; generating a risk probability spatial projection map using the standardized dynamic dataset and dynamic weight parameters; generating a hazard warning command based on the generated projection map and a hazard chain map; correcting the data in the soil rheological constitutive database using vibration spectrum data fed back from construction equipment; and transmitting the corrected soil rheological constitutive parameters back to the soil rheological constitutive database.

[0008] Optionally, the step of calling the preset soil rheological constitutive database and the real-time marine environmental parameters to calculate the dynamic weight parameters of each monitoring index includes: determining a dynamic monitoring time window based on the real-time marine environmental parameters; within the dynamic monitoring time window, obtaining the allocation ratio of soil pore pressure change gradient, displacement rate fluctuation amplitude, and wave load weight; calculating the correlation factor between soil rheological state and wave load based on a preset damage accumulation model; and dynamically adjusting the allocation ratio of soil pore pressure change gradient, displacement rate fluctuation amplitude, and wave load weight according to the correlation factor.

[0009] Optionally, the wave load weight includes: if the wave load weight exceeds a preset threshold, generating a disaster chain triggering probability index, and superimposing the disaster chain triggering probability index onto the risk probability space projection map.

[0010] Optionally, generating a risk probability space projection map using the standardized dynamic dataset and the dynamic weight parameters includes: mapping the standardized dynamic dataset to preset three-dimensional geological model grid nodes; iteratively solving the soil stress-seepage field coupling equation based on a preset gradient descent method to generate an excess pore water pressure distribution cloud map; identifying edge critical zones on the excess pore water pressure distribution cloud map to generate an initial risk region; and performing probability space expansion projection on the initial risk region using the dynamic weight parameters to generate a risk probability space projection.

[0011] Optionally, receiving vibration spectrum data from construction equipment and using the vibration spectrum data to correct the data in the soil rheological constitutive database to generate corrected soil rheological constitutive parameters includes: acquiring vibration spectrum data; inputting the vibration spectrum data into a preset digital twin sand liquefaction simulation module to generate a theoretical excess pore water pressure curve; comparing the deviation amplitude between the theoretical excess pore water pressure curve and preset measured excess pore water pressure data; if the deviation amplitude exceeds a preset tolerance threshold, using the vibration spectrum data to correct the data in the soil rheological constitutive database to generate corrected soil rheological constitutive parameters.

[0012] Optionally, if the deviation exceeds a preset tolerance threshold, the step of using the vibration spectrum data to correct the data in the soil rheological constitutive database and generating corrected soil rheological constitutive parameters includes: if the deviation exceeds the preset tolerance threshold, obtaining the spectral energy shift, the dominant frequency shift, and the existing rheological parameters; and correcting the data in the soil rheological constitutive database based on the spectral energy shift, the dominant frequency shift, and the existing rheological parameters, combined with the vibration spectrum data, to generate corrected soil rheological constitutive parameters.

[0013] Optionally, the step of simulating state evolution based on the risk probability space projection map and a preset disaster chain map to generate a disaster early warning command includes: obtaining the topological boundary of the high-risk area based on the risk probability space projection map; and simulating state evolution based on the topological boundary of the high-risk area and a preset disaster chain map to generate a disaster early warning command.

[0014] Optionally, the method further includes: triggering a cross-level linkage early warning when a high-risk area in the risk probability space projection map coincides with a key node in the disaster chain map; and generating a three-dimensional visualization decision map based on the cross-level linkage early warning.

[0015] Optionally, the step of receiving multi-source monitoring data and real-time marine environmental parameters of the offshore wind farm site layer, and performing waveform framing compression processing on the multi-source monitoring data to generate noise-reduced transmission data includes: acquiring multi-source monitoring data and real-time marine environmental parameters of the offshore wind farm site layer; monitoring the multi-source monitoring data and real-time marine environmental parameters of the offshore wind farm site layer and dynamically dividing the data frames; performing median filtering smoothing and discrete cosine transform compression on abnormal peak signals in the data frames to obtain multi-source monitoring data; and performing waveform framing compression processing on the multi-source monitoring data to generate noise-reduced transmission data.

[0016] Based on the same inventive concept, the present invention also provides a dynamic assessment system for disaster risks at offshore wind farm sites. The system includes: a multi-source data receiving and noise reduction compression module, used to receive multi-source monitoring data and real-time marine environmental parameters at offshore wind farm sites, and to perform waveform frame compression processing on the multi-source monitoring data to generate noise-reduced transmission data; and a data standardization processing module, used to perform time-domain filtering and time-series alignment on the noise-reduced transmission data to generate a standardized dynamic dataset.

[0017] The system includes a dynamic weight calculation module, which calls a preset soil rheological constitutive database and the real-time marine environmental parameters to calculate the dynamic weight parameters of each monitoring indicator; a risk probability spatial projection module, which uses the standardized dynamic dataset and the dynamic weight parameters to generate a risk probability spatial projection map; a disaster chain evolution simulation and early warning module, which simulates state evolution based on the risk probability spatial projection map and a preset disaster chain map to generate disaster early warning commands; a construction vibration feedback and parameter correction module, which receives vibration spectrum data from construction equipment, uses the vibration spectrum data to correct the data in the soil rheological constitutive database, and generates corrected soil rheological constitutive parameters; and a database dynamic update module, which sends the corrected soil rheological constitutive parameters back to the soil rheological constitutive database.

[0018] Compared with the prior art, the present invention has the following advantages:

[0019] 1. This invention significantly improves the accuracy of risk assessment through dynamic weight allocation and multiphysics coupling calculation. The system can automatically adjust the weight allocation ratio of pore pressure, displacement rate and wave load according to real-time marine environmental parameters. Combined with the damage accumulation model, it accurately quantifies the correlation between soil rheological state and external load, making the risk probability space projection map more realistically reflect the actual geological conditions and avoiding the assessment bias of traditional static weight models under complex sea conditions.

[0020] 2. The introduction of a digital twin reverse parameter tuning mechanism effectively solves the technical challenge of time-varying soil parameters. By comparing the differences between the vibration spectrum of construction equipment and theoretical simulation results, a particle swarm optimization algorithm is used to dynamically correct key parameters such as soil permeability coefficient and compression modulus. This establishes a two-way feedback channel between field monitoring data and numerical models, significantly improving the reliability of constitutive parameters during long-term service and forming a continuously self-optimizing closed-loop system.

[0021] 3. The deep integration of disaster chain mapping and 3D visualization decision-making has achieved a leapfrog improvement in early warning effectiveness. By establishing a topological relationship network that includes landslide triggering paths, submarine cable breakage propagation paths, and pile foundation failure chain paths, the system can automatically generate emergency instructions with spatial coordinate correlation when the risk probability projection coincides with key nodes in the map. For example, it can prioritize the reinforcement of specific pile foundations or activate local drainage systems, transforming abstract risk data into executable engineering measures.

[0022] 4. The innovative application of multi-source data preprocessing technology ensures the stable operation of the system in harsh marine environments. Addressing the vulnerability of marine monitoring signals to wave interference, a depth-adaptive waveform framing compression algorithm is employed. Median filtering eliminates abnormal spikes, and discrete cosine transform preserves effective frequency band characteristics. This reduces data transmission volume while ensuring the integrity of critical information, providing a high-quality data foundation for subsequent analysis.

[0023] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures pointed out in the description, claims and drawings. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1This is a flowchart illustrating a dynamic assessment method for site-specific disaster risks in offshore wind farms, according to an embodiment of the present invention.

[0026] Figure 2 This is a time-series diagram of multi-source monitoring data according to an embodiment of the present invention.

[0027] Figure 3 This is a graph showing the change of dynamic weight parameters in an embodiment of the present invention.

[0028] Figure 4 This is a risk probability space projection diagram according to an embodiment of the present invention.

[0029] Figure 5 This is a comparison chart of theoretical and measured pore water pressure in an embodiment of the present invention.

[0030] Figure 6 This is a disaster chain state evolution diagram according to an embodiment of the present invention.

[0031] Figure 7 This is a schematic diagram of the structure of a dynamic assessment system for site-level disaster risks in offshore wind farms, according to an embodiment of the present invention. Detailed Implementation

[0032] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0033] Reference Figure 1 One embodiment of the present invention proposes a dynamic assessment method for geological disaster risks in offshore wind farms. It adopts real-time fusion processing of multi-source monitoring data, combined with an online correction mechanism for soil rheological constitutive parameters and a disaster chain spatial topology analysis method, which can realize full-time perception of disaster risks under the coupling effect of marine environment and geological state.

[0034] The method described in this embodiment specifically includes:

[0035] Receive multi-source monitoring data of offshore wind farm site layer and real-time marine environmental parameters, perform waveform framing compression processing on the multi-source monitoring data, and generate noise-reduced transmission data;

[0036] The noise-reduced transmission data is subjected to time-domain filtering and time-series alignment to generate a standardized dynamic dataset;

[0037] The preset soil rheological constitutive database and the real-time marine environmental parameters are called to calculate the dynamic weight parameters of each monitoring indicator;

[0038] Using the standardized dynamic dataset and the dynamic weight parameters, a risk probability space projection map is generated;

[0039] Based on the risk probability space projection map, combined with the preset disaster chain map, state evolution simulation is performed to generate disaster early warning instructions;

[0040] Receive vibration spectrum data fed back by construction equipment, use the vibration spectrum data to correct the data in the soil rheological constitutive database, and generate corrected soil rheological constitutive parameters;

[0041] The corrected soil rheological constitutive parameters are then fed back to the soil rheological constitutive database.

[0042] Specifically, the time series of multi-source monitoring data is as follows: Figure 2 As shown, the noise-reduced transmission data generated in the previous steps is obtained, which has been processed by waveform framing compression. First, a time-domain filtering operation is performed on the noise-reduced transmission data, establishing a dynamic time window mechanism: the initial window length is set to 200 milliseconds and updated in 50-millisecond steps. Within each time window, a fourth-order Butterworth low-pass filter is applied to process the original signal. The filter's cutoff frequency is dynamically determined using spectral energy analysis, specifically the frequency value corresponding to when the cumulative energy of the signal power spectral density within the window reaches 90% of the total energy. After filtering, a time-domain smoothed signal is generated. Subsequently, a time alignment operation is performed: the time reference of the multi-source sensors is unified, and a precise time protocol server is used to synchronize the timestamps of all monitoring devices at the millisecond level, with time deviation controlled within 10 milliseconds. Based on the synchronized timestamps, multi-channel data registration is performed on the time-domain smoothed signal. The matching principle is that the timestamps of data from the same source sensor are completely aligned, and the timestamp difference of data from different source sensors is less than a preset registration threshold. For misaligned sample points, signal interpolation reconstruction is performed using cubic spline interpolation, with adjacent timestamp-aligned sampling points as boundary conditions to reconstruct the complete time-domain sequence. After time-domain filtering and time-series alignment, data standardization is performed: the reference amplitude range of all sensors is extracted, the maximum and minimum values ​​of each sensor's range are calculated, and the amplitude of the time-domain aligned signal is normalized to the preset standard range interval of 0 to 1 using a linear transformation algorithm. The final output is a standardized dynamic dataset containing a multi-dimensional monitoring parameter sequence with strictly timestamp-aligned parameters, where each parameter is in a dimensionless standardized state.

[0043] Optionally, the step of calling the preset soil rheological constitutive database and the real-time marine environmental parameters to calculate the dynamic weight parameters of each monitoring indicator includes:

[0044] Based on the real-time marine environmental parameters, a dynamic monitoring time window is determined;

[0045] Within the dynamic monitoring time window, the gradient of soil pore pressure change, the amplitude of displacement rate fluctuation, and the distribution ratio of wave load weight are obtained.

[0046] Based on a pre-defined damage accumulation model, the correlation factor between soil rheological state and wave load is calculated.

[0047] Based on the aforementioned correlation factors, the distribution ratio of the soil pore pressure change gradient, displacement rate fluctuation amplitude, and wave load weight is dynamically adjusted.

[0048] Specifically, the dynamic weight parameters change as follows: Figure 3 As shown, the assessment period (i.e., the dynamic monitoring time window) is first dynamically determined based on real-time acquired marine environmental parameters (including real-time wave height, wave period, and wind speed data). The formula for calculating its duration is as follows:

[0049] ,

[0050] in The duration of the time window. The maximum effective wave height within the current observation period. The peak wave period at the corresponding time. To adjust the coefficients. Within this time window. Within, the temporal variation of soil pore pressure is simultaneously extracted from the standardized dynamic dataset. and time series data of ground displacement monitoring points Calculate the pore pressure gradient respectively. :

[0051] ,

[0052] and displacement rate fluctuation amplitude :

[0053] ,

[0054] Then, the soil rheological constitutive database was accessed, and the real-time wave load spectrum data was input. The correlation factor between wave load and soil rheological state was calculated using a pre-defined damage accumulation model (emphasizing nonlinear viscoplastic constitutive equations). :

[0055]

[0056] in For the accumulation of strain in the soil, This is a built-in coupling function in the database. It is based on the correlation factors calculated in real time. The weight distribution ratio of the three is dynamically adjusted, and the initial wave load weight is set. value:

[0057] ,

[0058] Gradient weights of soil pore pressure variation :

[0059] ,

[0060] Displacement rate fluctuation amplitude weight :

[0061] ,

[0062] This ultimately forms a dynamic weight parameter set. :

[0063] .

[0064] For example, during a typhoon at an offshore wind farm, the monitoring system acquired the effective wave height. Peak wave period Based on historical calibration coefficients Calculate the dynamic monitoring time window Within this window, the pore pressure sensor measures the maximum gradient change. The displacement monitor records the amplitude of the rate fluctuation. The soil rheological constitutive database is based on real-time wave loads. Output correlation factor Based on this, the weights are dynamically adjusted as follows: wave load weights Soil pore pressure weights Displacement rate weights This weight set Immediately update the risk assessment model to significantly enhance the weighting of wave load impact assessment during typhoon surge phases.

[0065] Optionally, the wave load weights include:

[0066] If the wave load weight exceeds a preset threshold, a disaster chain triggering probability index is generated, and the disaster chain triggering probability index is superimposed on the risk probability space projection map.

[0067] Specifically, when the calculated wave load weights are obtained... Exceeding the preset threshold At that time, the disaster chain trigger probability index generation operation is executed. First, the weight transcendence degree is calculated. The formula is:

[0068] ,

[0069] in Obtained in real time through a dynamic weight parameter set. Predefined by the marine geological risk assessment manual. Simultaneously, dynamic monitoring time windows are extracted. Internal wave load spectrum Instantaneous rate of change extreme value :

[0070] ,

[0071] Obtained through wave load time-series difference calculation. x represents the maximum value obtained from the extreme value of instantaneous variability. This is combined with the disaster expansion coefficient. and geological sensitivity coefficient Generate a disaster chain trigger probability index :

[0072] ,

[0073] Finally Map the wave load spatial coordinates to the risk probability spatial projection map and compare them with the original risk probability values. Perform probability union operation:

[0074] ,

[0075] To update the risk probabilities, complete the enhanced update of the risk probabilities in the spatial projection map.

[0076] For example, during the monitoring of an offshore platform during a typhoon, the dynamic weighting parameter returns the wave load weights. (Preset threshold) ), calculate the transcendence degree .exist Extreme values ​​of wave load change rate within a monitoring window of seconds Call historical calibration parameters and Generate a disaster chain trigger probability index This is superimposed onto the seabed area southeast of the wind turbine foundation, where the original risk probability was... Updated to .

[0077] Optionally, generating a risk probability space projection map using the standardized dynamic dataset and the dynamic weight parameters includes:

[0078] The standardized dynamic dataset is mapped to preset three-dimensional geological model grid nodes;

[0079] Based on the preset gradient descent method, the soil stress-seepage field coupling equation is solved iteratively to generate a cloud map of excess pore water pressure distribution.

[0080] Edge critical zones are identified in the ultrapore water pressure distribution cloud map to generate initial risk areas;

[0081] By combining the dynamic weight parameters, the initial risk region is subjected to probability space expansion projection to generate a risk probability space projection.

[0082] Specifically, risk probability space projection, such as Figure 4 As shown, a standardized dynamic dataset (containing time-series monitoring values ​​such as pore pressure and displacement rate) is mapped to pre-defined 3D geological model grid nodes using spatial coordinates. For each grid node, the soil stress-seepage field coupling equation is iteratively solved using the gradient descent method. The equation includes:

[0083] ,

[0084] ,

[0085] in For soil stress tensor, For spatial coordinates, For soil density, The component of gravitational acceleration, Permeability coefficient, For the Laplace operator, The water head height, For water storage rate, The iteration calculation stops when the residual is less than 10^-6 kPa, and the excess pore water pressure distribution contour map is output. The Sobel operator is used to perform edge detection on this contour map.

[0086] ,

[0087] For gradient magnitude, For excess pore water pressure, The horizontal direction of the cloud map. This refers to the vertical direction of the cloud map. (Identification) The region where the value exceeds the critical threshold of 120 kPa / m is designated as the initial risk region. Finally, a dynamic weight parameter set is used. Perform probability space expansion on the initial risk region:

[0088] ,

[0089] These are the independent risk probabilities of wave load, pore pressure gradient, and displacement rate (obtained by fitting historical disaster data). This represents the final risk probability space projection value.

[0090] For example, during the foundation construction phase of a wind farm in the Bohai Sea, a standardized dynamic dataset showed a sudden increase in pore pressure in the seabed area northwest of the pile foundation. Solving the stress-seepage field equations generated an excess pore water pressure cloud map, showing a maximum water pressure of 85 kPa at this location. Edge detection identified gradient values ​​in a 5m × 8m area. (Exceeding the critical threshold). Dynamic weight parameter set }, combined with independent risk probability Calculated The result, when projected onto a three-dimensional geological model, triggered an orange alert.

[0091] Optionally, the method of receiving vibration spectrum data from construction equipment and using this vibration spectrum data to correct the data in the soil rheological constitutive database to generate corrected soil rheological constitutive parameters includes:

[0092] Acquire vibration spectrum data;

[0093] The vibration spectrum data is input into the preset digital twin sand liquefaction simulation module to generate a theoretical excess pore water pressure curve;

[0094] Compare the deviation range between the theoretical excess pore water pressure curve and the preset measured excess pore water pressure data;

[0095] If the deviation exceeds a preset tolerance threshold, the vibration spectrum data is used to correct the data in the soil rheological constitutive database to generate corrected soil rheological constitutive parameters.

[0096] Specifically, the first step is to acquire vibration spectrum data collected by vibration sensors installed on the construction equipment. This data includes acceleration signals with a frequency range of 0.1-50 Hz, and a sampling frequency of 100 Hz. This data is then input into the sand liquefaction simulation module of the digital twin. This module is based on the modified Cambridge model, and its governing equations are:

[0097] ,

[0098] in For the body to adapt to strain, The consolidation coefficient is . For the Laplace operator, For effective stress, For the density of soil, For excess pore water pressure, The simulation module uses this time as a reference to generate theoretical excess pore water pressure curves. Simultaneously, measured curves were obtained from the pore water pressure sensor. Calculation window Deviation within seconds :

[0099] ,

[0100] when At kPa, extract spectral feature parameters: spectral energy transfer. The energy change in the main frequency band (3-8 Hz), the main frequency shift. This represents the peak frequency difference of the amplitude. It retrieves the current rheological parameters from the database: shear modulus. and permeability coefficient The corrected formula is:

[0101] ,

[0102] ,

[0103] This is the corrected shear modulus. This is the corrected permeability coefficient. This is the initial clock frequency.

[0104] For example, at a jacket installation site, vibration sensors measured a shift in the dominant frequency from 5 Hz to 7 Hz, with an energy increase of 120 joules in the 3-8 Hz band. Theoretical simulations yielded... This triggers the correction process. Current database parameters. Calculated , The updated parameters are immediately applied to the new phase of risk assessment for the construction area.

[0105] Optionally, if the deviation exceeds a preset tolerance threshold, the data in the soil rheological constitutive database is corrected using the vibration spectrum data to generate corrected soil rheological constitutive parameters, including:

[0106] If the deviation exceeds a preset tolerance threshold, the spectral energy shift, the main frequency shift, and the existing rheological parameters are obtained.

[0107] The data in the soil rheological constitutive database are corrected based on the spectral energy shift, dominant frequency shift, and existing rheological parameters, combined with the vibration spectrum data, to generate corrected soil rheological constitutive parameters.

[0108] Specifically, the theory and pore water pressure are compared, for example... Figure 5As shown, when the deviation between the theoretical excess pore water pressure curve and the measured data exceeds the preset tolerance threshold of 5 kPa, the soil rheological parameter correction process is initiated. First, vibration spectrum data collected by the accelerometer of the construction equipment is acquired. This data has a frequency range of 0.1 to 50 Hz, and a sampling frequency of 100 Hz. Two key spectral feature parameters are extracted: dominant frequency shift. By identifying the peak frequency of the current vibration spectrum amplitude and subtracting the historical reference dominant frequency value. The obtained difference, spectral energy transfer The energy variation in the 3-8 Hz main frequency band is calculated using the numerical integration method. The formula is as follows:

[0109] ,

[0110] in This is the current power spectral density function (obtained by calculating the vibration spectrum using fast Fourier transform). For the historical baseline power spectral density function, the current parameters stored in the soil rheological constitutive database are called: shear modulus. With permeability coefficient Perform dual-parameter synchronous correction, generate corrected soil rheological constitutive parameters, and then overwrite the original database records to complete the dynamic update.

[0111] For example, during pile driving operations at an offshore wind farm, vibration spectrum monitoring showed that the peak amplitude frequency shifted from the historical reference value of 4.6 Hz to 5.3 Hz, and the energy integral value in the 3-8 Hz band increased by 125 joules compared to the reference. The theoretical excess pore water pressure curve output by the digital twin sand liquefaction module deviated by 8.2 kPa from the measured pore water pressure sensor data, exceeding the threshold and triggering correction. The current database parameter is retrieved: shear modulus. Permeability coefficient Calculate and update parameters: The corrected parameters are immediately applied to the real-time risk warning model for pile foundation construction in this area.

[0112] Optionally, the step of generating a disaster early warning instruction by simulating the state evolution based on the risk probability space projection map and a preset disaster chain map includes:

[0113] Based on the risk probability space projection map, the topological boundary of the high-risk area is obtained;

[0114] Based on the topological boundaries of the high-risk areas and a preset disaster chain map, state evolution simulation is performed to generate disaster early warning instructions.

[0115] Specifically, the evolution of the disaster chain state is as follows: Figure 6As shown, firstly, continuous regions with risk values ​​exceeding a preset safety threshold are extracted from the generated risk probability spatial projection map. Then, a morphological closing algorithm is used for boundary closure processing to obtain the closed topological boundary of the high-risk region. This topological boundary consists of a sequence of polygon vertex coordinates, where each vertex contains coordinates from the 3D geological model. Axis coordinate data. A pre-defined disaster chain map database is invoked, which stores the logical association rules and state transition probabilities between various disaster events such as soil liquefaction, seabed landslides, and pile foundation instability. Using the topological boundary of the high-risk area as initial input, the disaster chain evolution is simulated using the following state transition equation:

[0116] ,

[0117] in This represents the probability tensor of the disaster chain state at the current moment, which is updated through iterative calculation; The probability of the state at the previous moment is given by the initial value determined by the high-risk value of the risk projection map. The time decay coefficient matrix is ​​an exponential decay function generated by fitting historical disaster cases. The disaster diffusion function is predefined in the disaster chain map database based on soil type; Represents the topological boundary vertex to the first Euclidean distance between nodes in a disaster chain; The soil sensitivity coefficient is provided in real time by a soil rheological constitutive database. During the iterative calculation, when the probability value of any disaster node in the state probability tensor exceeds the preset warning threshold, a structured disaster warning instruction is automatically generated. This instruction includes the disaster type, radius of influence, evolution time window, and emergency response measures, completing the transformation from static risk distribution to dynamic disaster warning.

[0118] For example, during the monitoring of an offshore wind farm in the East China Sea, the risk probability spatial projection showed a high-risk zone with a risk value of 0.88 (preset safety threshold 0.6) around pile foundation No. 9. The system uses a 3×3 pixel closed-loop kernel to generate a closed topological boundary and obtain the vertex coordinate sequence. The disaster chain map shows that the area is associated with sand liquefaction (nodes). ) and pile foundation tilt (node) Disaster chain, pre-set arrive The state transition probability is 0.85. Input parameter: time decay matrix. Take 0.92 (storm duration condition), disaster diffusion function Inverse distance weighting function, soil sensitivity coefficient (Sandy soil layer). When iterating to the fifth time step ( At (minutes), the pile foundation tilt node The probability rose to 0.63 (warning threshold 0.6), triggering a red warning order: immediately stop pile driving operations at position 9 and activate the casing reinforcement plan.

[0119] Optionally, the method further includes:

[0120] When a high-risk area in the risk probability space projection map is detected to overlap with a key node in the disaster chain map, a cross-level linkage early warning is triggered.

[0121] Based on the cross-level linkage early warning, a three-dimensional visualization decision map is generated.

[0122] Specifically, it receives and parses real-time data from the risk probability spatial projection map and the preset disaster chain map. When the spatial coordinates of a high-risk area in the projection map overlap with the spatial coordinates of a predefined key node in the disaster chain map, a cross-level linkage early warning mechanism is automatically triggered. Key nodes are defined as disaster types that act as hubs in the disaster chain transmission, and their spatial coordinates are stored in the disaster chain map database. The mathematical expression for executing cross-level early warning is:

[0123] ,

[0124] in The cross-level linkage early warning intensity value is output through weighted calculation; For the first The transmission weights of hierarchical disasters are predefined by the disaster chain map; For the first The real-time risk probability of hierarchical nodes is extracted directly from the risk probability space projection map; The spatial overlap coefficient is calculated using the following formula: .based on The specific steps for generating a 3D visualization decision map are as follows: overlaying the grid topology of the 3D geological model; mapping the real-time probability values ​​of each disaster level to different color gradients; marking the warning radiation range in spatially overlapping areas; and embedding the emergency measures text from the historical emergency response plan database into the corresponding coordinate nodes.

[0125] For example, in the monitoring of an offshore wind farm in the South China Sea, the risk probability projection map shows that there is a volumetric gas on the southwest side of the No. 2 wind turbine foundation. The area is identified as a high-risk zone for excess pore water pressure (probability value 0.91). The system detects the connection between this area and the critical node "soil liquefaction - pile foundation tilting" in the disaster chain map (preset volume). ) 65% overlap, calculate the overlap coefficient. Extracting hierarchical parameters: propagation weights (Liquefaction layer) (Pile foundation); Real-time probability , Calculated The system triggered a cross-level red alert, overlaying a red cloud map of the liquefaction zone, a yellow stress zone of the pile foundation, and radial warning range lines on the 3D map. At the same time, a pop-up window pushed the "Immediately start foundation grouting + anchor cable reinforcement" plan.

[0126] Optionally, the step of receiving multi-source monitoring data and real-time marine environmental parameters at the offshore wind farm site, and performing waveform framing compression processing on the multi-source monitoring data to generate noise-reduced transmission data includes:

[0127] Acquire multi-source monitoring data and real-time marine environmental parameters at offshore wind farm sites;

[0128] The multi-source monitoring data and real-time marine environmental parameters of the offshore wind farm site are monitored, and the data frames are dynamically divided.

[0129] The abnormal spike signals in the data frame are smoothed by median filtering and compressed by discrete cosine transform to obtain multi-source monitoring data.

[0130] Waveform framing and compression processing is performed on multi-source monitoring data to generate noise-reduced transmission data.

[0131] Specifically, the first step is to acquire multi-source monitoring data of the offshore wind farm site and real-time marine environmental parameters. The multi-source monitoring data includes pore water pressure signals, soil displacement velocity signals, and vibration acceleration signals collected by a seabed sensor array. Real-time marine environmental parameters, provided by ocean buoys, include wave height, wave period, and wind speed data. A dynamic framing algorithm is used to determine the time frame based on the real-time wave height change rate. length:

[0132] ,

[0133] in The frame length is dynamic, and the calculation results are used to adjust the data processing time window in real time. The baseline frame length is set to 1 by default. The adjustment coefficient was calibrated to 0.05 through marine characteristic experiments. The real-time rate of change of wave height is obtained by differential calculation of two adjacent wave height acquisition values.

[0134] Median filtering is performed on the sensor signals within each dynamic time frame: a window of size 2m+1 sampling points (m=5) is used, and the median value of the data within the window is used to replace the current point to eliminate abnormal spike interference. Then, discrete cosine transform is applied for the first stage of compression, as shown in the following formula:

[0135] ,

[0136] in These are the discrete cosine transform coefficients; For the first frame Values ​​of each sampling point; The frame length is determined by... Multiply by the sampling frequency of 100 Hz to determine the result. Only the first 20% of the transform coefficients are retained, and the remaining coefficients are set to zero. The signal is then reconstructed through inverse transform. Finally, all multi-source data is compressed in frames of a fixed length of 1 minute: each frame overlaps by 50%, and a discrete cosine transform is performed again, retaining coefficients with 90% energy percentage. The resulting data is then merged and output as noise-reduced transmission data.

[0137] For example, during the construction phase of a wind farm in the South China Sea, a wave height variation rate of 0.4 was monitored. The dynamic frame length was calculated. =10 / (1+0.05×0.4)≈9.8 seconds. Median filtering (11-point window) was applied to the pore water pressure signal (980 sampling points) within 9.8 seconds to remove amplitude abrupt changes caused by ocean current impact. The first compression retained 196 discrete cosine transform coefficients, and the signal-to-noise ratio of the reconstructed signal was improved to 42 dB. Subsequently, the 6 dynamic frames were merged into a 58.8-second data segment, and after zero-padding and alignment according to a 1-minute fixed frame, a second compression was performed: the top 542 (90% energy) were selected from 6000 transform coefficients, and the final noise-reduced transmission data volume was only 23% of the original data.

[0138] Based on the same inventive concept, such as Figure 7 As shown, the present invention also provides a dynamic assessment system for site-specific disaster risks in offshore wind farms, the system comprising:

[0139] The multi-source data receiving and noise reduction compression module is used to receive multi-source monitoring data and real-time marine environmental parameters at the offshore wind farm site, and to perform waveform frame compression processing on the multi-source monitoring data to generate noise-reduced transmission data.

[0140] The data standardization processing module is used to perform time-domain filtering and time-series alignment on the noise-reduced transmission data to generate a standardized dynamic dataset.

[0141] The dynamic weight calculation module is used to call the preset soil rheological constitutive database and the real-time marine environmental parameters to calculate the dynamic weight parameters of each monitoring indicator.

[0142] The risk probability space projection module is used to generate a risk probability space projection map using the standardized dynamic dataset and the dynamic weight parameters.

[0143] The disaster chain evolution simulation and early warning module is used to simulate the state evolution based on the risk probability space projection map and the preset disaster chain map, and generate disaster early warning instructions.

[0144] The construction vibration feedback and parameter correction module is used to receive vibration spectrum data fed back by construction equipment, use the vibration spectrum data to correct the data in the soil rheological constitutive database, and generate corrected soil rheological constitutive parameters.

[0145] The database dynamic update module is used to transmit the corrected soil rheological constitutive parameters back to the soil rheological constitutive database.

[0146] It should be noted that the electrical connections between the various units described above do not necessarily represent direct or indirect connections. Any indirect connection method can be applied to the embodiments of the present invention as long as it achieves the purpose of the present invention. The above descriptions are merely exemplary embodiments of the present invention and should not be construed as limiting the scope of the present invention.

[0147] All equivalent changes and modifications made in accordance with the teachings of this invention are still within the scope of this invention. Those skilled in the art will readily conceive of other embodiments of this invention upon considering the specification and the disclosure of practical truth. This application is intended to cover any variations, uses, or adaptations of this invention that follow the general principles of this invention and include common knowledge or conventional techniques in the art not described herein.

Claims

1. A method for dynamic assessment of site-specific disaster risks in offshore wind farms, characterized in that, The method includes: Receive multi-source monitoring data of offshore wind farm site layer and real-time marine environmental parameters, perform waveform framing compression processing on the multi-source monitoring data, and generate noise-reduced transmission data; The noise-reduced transmission data is subjected to time-domain filtering and time-series alignment to generate a standardized dynamic dataset; The preset soil rheological constitutive database and the real-time marine environmental parameters are called to calculate the dynamic weight parameters of each monitoring indicator; Using the standardized dynamic dataset and the dynamic weight parameters, a risk probability space projection map is generated; Based on the risk probability space projection map, combined with the preset disaster chain map, state evolution simulation is performed to generate disaster early warning instructions; The system receives vibration spectrum data from construction equipment, uses the vibration spectrum data to correct the data in the soil rheological constitutive database, and generates corrected soil rheological constitutive parameters. The corrected soil rheological constitutive parameters are then fed back to the soil rheological constitutive database.

2. The method for dynamic assessment of site-specific disaster risks in offshore wind farms according to claim 1, characterized in that, The step of calling the preset soil rheological constitutive database and the real-time marine environmental parameters to calculate the dynamic weight parameters of each monitoring indicator includes: Based on the real-time marine environmental parameters, a dynamic monitoring time window is determined; Within the dynamic monitoring time window, the gradient of soil pore pressure change, the amplitude of displacement rate fluctuation, and the distribution ratio of wave load weight are obtained. Based on a pre-defined damage accumulation model, the correlation factor between soil rheological state and wave load is calculated. Based on the aforementioned correlation factors, the distribution ratio of the soil pore pressure change gradient, displacement rate fluctuation amplitude, and wave load weight is dynamically adjusted.

3. The method for dynamic assessment of site-specific disaster risks in offshore wind farms according to claim 2, characterized in that, The wave load weights include: If the wave load weight exceeds a preset threshold, a disaster chain triggering probability index is generated, and the disaster chain triggering probability index is superimposed on the risk probability space projection map.

4. The method for dynamic assessment of site-specific disaster risks in offshore wind farms according to claim 1, characterized in that, The step of generating a risk probability space projection map using the standardized dynamic dataset and the dynamic weight parameters includes: The standardized dynamic dataset is mapped to preset three-dimensional geological model grid nodes; Based on the preset gradient descent method, the soil stress-seepage field coupling equation is solved iteratively to generate a cloud map of excess pore water pressure distribution. Edge critical zones are identified in the ultrapore water pressure distribution cloud map to generate initial risk areas; By combining the dynamic weight parameters, the initial risk region is subjected to probability space expansion projection to generate a risk probability space projection.

5. The method for dynamic assessment of site-specific disaster risks in offshore wind farms according to claim 1, characterized in that, The method of receiving vibration spectrum data from construction equipment and using this vibration spectrum data to correct the data in the soil rheological constitutive database to generate corrected soil rheological constitutive parameters includes: Acquire vibration spectrum data; The vibration spectrum data is input into the preset digital twin sand liquefaction simulation module to generate a theoretical excess pore water pressure curve; Compare the deviation range between the theoretical excess pore water pressure curve and the preset measured excess pore water pressure data; If the deviation exceeds a preset tolerance threshold, the vibration spectrum data is used to correct the data in the soil rheological constitutive database to generate corrected soil rheological constitutive parameters.

6. The method for dynamic assessment of site-specific disaster risks in offshore wind farms according to claim 5, characterized in that, If the deviation exceeds a preset tolerance threshold, the vibration spectrum data is used to correct the data in the soil rheological constitutive database, generating corrected soil rheological constitutive parameters including: If the deviation exceeds a preset tolerance threshold, the spectral energy shift, the main frequency shift, and the existing rheological parameters are obtained. The data in the soil rheological constitutive database are corrected based on the spectral energy shift, dominant frequency shift, and existing rheological parameters, combined with the vibration spectrum data, to generate corrected soil rheological constitutive parameters.

7. The method for dynamic assessment of site-specific disaster risks in offshore wind farms according to claim 1, characterized in that, The step of simulating state evolution based on the risk probability space projection map and combining it with a preset disaster chain map to generate disaster early warning instructions includes: Based on the risk probability space projection map, the topological boundary of the high-risk area is obtained; Based on the topological boundaries of the high-risk areas and a preset disaster chain map, state evolution simulation is performed to generate disaster early warning instructions.

8. The method for dynamic assessment of site-specific disaster risks in offshore wind farms according to claim 7, characterized in that, The description also includes: When a high-risk area in the risk probability space projection map is detected to overlap with a key node in the disaster chain map, a cross-level linkage early warning is triggered. Based on the cross-level linkage early warning, a three-dimensional visualization decision map is generated.

9. The method for dynamic assessment of site-specific disaster risks in offshore wind farms according to claim 1, characterized in that, The process of receiving multi-source monitoring data from offshore wind farm sites and real-time marine environmental parameters, and performing waveform framing compression processing on the multi-source monitoring data to generate noise-reduced transmission data includes: Acquire multi-source monitoring data and real-time marine environmental parameters at offshore wind farm sites; The multi-source monitoring data and real-time marine environmental parameters of the offshore wind farm site are monitored, and the data frames are dynamically divided. The abnormal spike signals in the data frame are smoothed by median filtering and compressed by discrete cosine transform to obtain multi-source monitoring data. Waveform framing and compression processing is performed on multi-source monitoring data to generate noise-reduced transmission data.

10. A dynamic risk assessment system for strata disasters at offshore wind farm sites, applied to the dynamic risk assessment method for strata disasters at offshore wind farm sites as described in any one of claims 1-9, characterized in that, The system includes: The multi-source data receiving and noise reduction compression module is used to receive multi-source monitoring data and real-time marine environmental parameters at the offshore wind farm site, perform waveform frame compression processing on the multi-source monitoring data, and generate noise-reduced transmission data. The data standardization processing module is used to perform time-domain filtering and time-series alignment on the noise-reduced transmission data to generate a standardized dynamic dataset. The dynamic weight calculation module is used to call the preset soil rheological constitutive database and the real-time marine environmental parameters to calculate the dynamic weight parameters of each monitoring indicator. The risk probability space projection module is used to generate a risk probability space projection map using the standardized dynamic dataset and the dynamic weight parameters. The disaster chain evolution simulation and early warning module is used to simulate the state evolution based on the risk probability space projection map and the preset disaster chain map, and generate disaster early warning instructions. The construction vibration feedback and parameter correction module is used to receive vibration spectrum data fed back by construction equipment, use the vibration spectrum data to correct the data in the soil rheological constitutive database, and generate corrected soil rheological constitutive parameters. The database dynamic update module is used to transmit the corrected soil rheological constitutive parameters back to the soil rheological constitutive database.

Images