Sea ice thickness prediction method and device, storage medium and electronic equipment

Abstract

The invention relates to the field of offshore photovoltaic technology, in particular to a sea ice thickness prediction method, a sea ice thickness prediction device, a storage medium and electronic equipment. The sea ice thickness prediction method comprises the steps of collecting modal data; the modal data comprises pile foundation data, ice anchor data and meteorological data; performing feature coding on each piece of modal data to obtain each piece of feature data, and dynamically adjusting the weight of each piece of feature data based on the current density of sea ice so as to perform feature fusion on each piece of feature data to obtain a cross-modal space-time fusion feature; and projecting the cross-modal space-time fusion feature to a physical parameter space to obtain a physical constraint feature, and calculating a sea ice thickness prediction value based on the physical constraint feature. According to the sea ice thickness prediction method provided by the invention, the prediction precision of the sea ice thickness under the offshore photovoltaic background can be improved.

Description

Technical Field

[0001] The present disclosure relates to the field of offshore photovoltaic technology, and in particular to a sea ice thickness prediction method, a sea ice thickness prediction device, a storage medium, and an electronic device. Background Art

[0002] Driven by energy security, photovoltaic power generation construction has ushered in new developments. In recent years, as land resources with good sunlight conditions have gradually decreased, water-based photovoltaic power generation systems (referring to freshwater) such as rivers, lakes, and reservoirs have developed rapidly.

[0003] Offshore photovoltaics are becoming a research hotspot due to their abundant water resources, unobstructed access, high light reflection, reduced thermal losses, proximity to load centers, and low dust levels. However, offshore photovoltaics face complex hydrological and climatic conditions, particularly during winter when sea ice forms off the Bohai Sea coast. This increases the risk of pile foundations and makes them susceptible to damage from ice loads. Therefore, accurate prediction of sea ice thickness is crucial.

[0004] It should be noted that the information disclosed in the above background technology section is only used to enhance the understanding of the background of the present disclosure, and therefore may include information that does not constitute prior art known to ordinary technicians in the field. Summary of the Invention

[0005] The purpose of the present disclosure is to provide a sea ice thickness prediction method, aiming to solve the problem of low sea ice thickness prediction accuracy in the context of offshore photovoltaics.

[0006] Other features and advantages of the present disclosure will become apparent from the following detailed description, or may be learned in part by practice of the present disclosure.

[0007] According to one aspect of the present disclosure, a method for predicting sea ice thickness is provided, comprising:

[0008] Collecting modal data; the modal data includes pile foundation data, ice anchor data and meteorological data;

[0009] Performing feature encoding on each of the modal data to obtain each feature data, and dynamically adjusting the weight of each feature data based on the current density of sea ice to perform feature fusion on the each feature data to obtain a cross-modal spatiotemporal fusion feature;

[0010] The cross-modal spatiotemporal fusion features are projected into a physical parameter space to obtain physical constraint features, and a sea ice thickness prediction value is calculated based on the physical constraint features.

[0011] Optionally, the pile foundation data includes pile foundation strain data and inclination sensor data, and the acquisition of modal data includes:

[0012] Sampling the distributed optical fiber strain sequence of the pile foundation according to a preset sampling rate, and extracting the strain extreme value, strain gradient and bending modal frequency of the distributed optical fiber strain sequence as the pile foundation strain data; and

[0013] The inclination angles of the pile foundation in three orthogonal directions are measured in real time, and the inclination change rate and the equivalent rotation angle are calculated based on the inclination angles as the inclination sensor data.

[0014] Optionally, the ice anchor data includes sonar point cloud data and ice anchor growth parameter data, and the acquisition modal data includes:

[0015] Segmenting the ice anchor point cloud using the DBSCAN clustering algorithm, and extracting the volume, anchoring depth, distance to the pile foundation, and contact area of ​​the ice anchor point cloud as the sonar point cloud data; and

[0016] Estimation of ice anchor growth parameter data by Kalman filtering.

[0017] Optionally, the meteorological data includes basic data and enhanced data, and the acquisition modality data includes:

[0018] respectively collecting temperature data, wind speed data, and radiation data as the basic data; and

[0019] A wind chill index is calculated based on the temperature data and the wind speed data to serve as the enhanced data.

[0020] Optionally, the characteristic data includes pile foundation strain time series characteristics, ice anchor geometry time series characteristics, and meteorological time series characteristics, and the characteristic encoding of each modal data to obtain each characteristic data includes:

[0021] Extracting the pile foundation strain time series features of the pile foundation data using a pile foundation data encoder; the pile foundation data encoder includes a convolution layer, a pooling layer, and a normalization layer;

[0022] Extracting the ice-anchor geometric temporal features of the ice-anchor data using an ice-anchor data encoder; the ice-anchor data encoder is a point cloud-specific Transformer, and the point cloud-specific Transformer includes an attention mechanism;

[0023] A meteorological data encoder is used to extract meteorological time series features of the meteorological data; the meteorological data encoder is a standard Transformer.

[0024] Optionally, dynamically adjusting the weights of the feature data based on the current density of sea ice to perform feature fusion on the feature data to obtain cross-modal spatiotemporal fusion features includes:

[0025] Configuring basic weights corresponding to the respective feature data;

[0026] Obtain the current density of sea ice in real time;

[0027] When the current density exceeds a preset threshold density, dynamically adjusting the basic weight based on the adjustment coefficient to obtain an adjustment weight corresponding to each feature data;

[0028] The cross-modal spatiotemporal fusion feature is calculated based on the adjustment weight.

[0029] Optionally, the physical constraint features include: energy balance equation regularization and Hamiltonian mechanics constraints.

[0030] According to a second aspect of the present disclosure, a sea ice thickness prediction device is provided, comprising:

[0031] An acquisition module, configured to acquire modal data, including pile foundation data, ice anchor data, and meteorological data;

[0032] A fusion module is used to perform feature encoding on each of the modal data to obtain each feature data, and dynamically adjust the weight of each feature data based on the current density of sea ice to perform feature fusion on the each feature data to obtain a cross-modal spatiotemporal fusion feature;

[0033] A prediction module is used to project the cross-modal spatiotemporal fusion features into a physical parameter space to obtain physical constraint features, and calculate the sea ice thickness prediction value based on the physical constraint features.

[0034] According to a third aspect of the present disclosure, a computer-readable storage medium is provided, on which a computer program is stored. When the program is executed by a processor, the sea ice thickness prediction method in the above embodiment is implemented.

[0035] According to a fourth aspect of an embodiment of the present disclosure, an electronic device is provided, characterized in that it includes: one or more processors; and a storage device for storing one or more programs, which, when the one or more programs are executed by the one or more processors, enables the one or more processors to implement the sea ice thickness prediction method as in the above-mentioned embodiment.

[0036] The exemplary embodiments of the present disclosure may have some or all of the following beneficial effects:

[0037] In the technical solutions provided by some embodiments of the present disclosure, multiple modal data, including pile foundation data, ice anchor data, and meteorological data, are collected as input. Feature extraction and fusion are then performed on each modal data. During the fusion process, sea ice density is used as a dynamic attention mechanism, and physical constraint features are embedded to ultimately obtain a predicted sea ice thickness value. This method, when used to predict sea ice thickness, not only extracts cross-modal features, particularly capturing the degree of mechanical coupling between ice anchors and pile foundations, but also captures both temporal evolution and spatial distribution, improving prediction accuracy. Furthermore, dynamic weighting based on ice conditions, focusing on the characteristics of dense sea ice areas, can further improve prediction accuracy.

[0038] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] The accompanying drawings are incorporated into and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the specification, are used to explain the principles of the present disclosure. Obviously, the drawings described below are only some embodiments of the present disclosure, and those skilled in the art can derive other drawings based on these drawings without inventive effort. In the drawings:

[0040] Figure 1 A schematic diagram schematically illustrates a flow chart of a sea ice thickness prediction method in an exemplary embodiment of the present disclosure;

[0041] Figure 2 A schematic diagram illustrating a flow chart of a feature cross-modal spatiotemporal fusion method in an exemplary embodiment of the present disclosure is shown;

[0042] Figure 3 A schematic diagram schematically illustrates a flow chart of a sea ice thickness prediction model in an exemplary embodiment of the present disclosure;

[0043] Figure 4 A schematic diagram schematically illustrates the composition of a sea ice thickness prediction device in an exemplary embodiment of the present disclosure;

[0044] Figure 5 The following schematically shows a structural diagram of a computer system of an electronic device in an exemplary embodiment of the present disclosure. DETAILED DESCRIPTION

[0045] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that this disclosure will be thorough and complete and will fully convey the concepts of the example embodiments to those skilled in the art.

[0046] In addition, the described features, structures or characteristics may be combined in any suitable manner in one or more embodiments. In the following description, many specific details are provided to provide a full understanding of the embodiments of the present disclosure. However, those skilled in the art will appreciate that the technical solutions of the present disclosure can be practiced without one or more of the specific details, or other methods, components, devices, steps, etc. can be adopted. In other cases, well-known methods, devices, implementations or operations are not shown or described in detail to avoid blurring various aspects of the present disclosure.

[0047] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically separate entities. That is, these functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different networks and / or processor devices and / or microcontroller devices.

[0048] The flowcharts shown in the accompanying drawings are for illustrative purposes only and do not necessarily include all contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may be decomposed, while others may be combined or partially combined. Therefore, the actual execution order may vary depending on the actual situation.

[0049] The implementation details of the technical solution of the embodiment of the present disclosure are described in detail below.

[0050] Figure 1 The following schematically illustrates a flow chart of a method for predicting sea ice thickness in an exemplary embodiment of the present disclosure. Figure 1 As shown, the sea ice thickness prediction method includes steps S101 to S103:

[0051] Step S101, collecting modal data; the modal data includes pile foundation data, ice anchor data and meteorological data;

[0052] Step S102: performing feature encoding on each of the modal data to obtain feature data, and dynamically adjusting the weight of each of the feature data based on the current density of sea ice to perform feature fusion on the feature data to obtain a cross-modal spatiotemporal fusion feature;

[0053] Step S103 : Projecting the cross-modal spatiotemporal fusion features into a physical parameter space to obtain physical constraint features, and calculating a predicted sea ice thickness value based on the physical constraint features.

[0054] In the technical solutions provided by some embodiments of the present disclosure, multiple modal data, including pile foundation data, ice anchor data, and meteorological data, are collected as input. Feature extraction and fusion are then performed on each modal data. During the fusion process, sea ice density is used as a dynamic attention mechanism, and physical constraint features are embedded to ultimately obtain a predicted sea ice thickness value. This method, when used to predict sea ice thickness, not only extracts cross-modal features, particularly capturing the degree of mechanical coupling between ice anchors and pile foundations, but also captures both temporal evolution and spatial distribution, improving prediction accuracy. Furthermore, dynamic weighting based on ice conditions, focusing on the characteristics of dense sea ice areas, can further improve prediction accuracy.

[0055] Below, each step of the sea ice thickness prediction method in this example implementation will be described in more detail with reference to the accompanying drawings and examples.

[0056] In step S101 , modal data is collected; the modal data includes pile foundation data, ice anchor data, and meteorological data.

[0057] In one embodiment of the present disclosure, the pile foundation data includes pile foundation strain data and tilt sensor data, and the acquisition of modal data includes:

[0058] Sampling the distributed optical fiber strain sequence of the pile foundation according to a preset sampling rate, and extracting the strain extreme value, strain gradient and bending modal frequency of the distributed optical fiber strain sequence as the pile foundation strain data; and

[0059] The inclination angles of the pile foundation in three orthogonal directions are measured in real time, and the inclination change rate and the equivalent rotation angle are calculated based on the inclination angles as the inclination sensor data.

[0060] Specifically, for collecting pile foundation strain data:

[0061] Capture micro-strain fluctuations of pile foundations through high-frequency sampling. Set the sampling rate, and then slide the window for time series analysis to acquire the distributed optical fiber strain sequence. For example, set the sampling rate to 1Hz and the sliding window to 10 minutes. After sampling the distributed optical fiber strain sequence, perform feature extraction to obtain the pile foundation strain data. Extract the maximum strain value within the sliding window as the strain extreme value (σ max ), σ max Used to monitor the ultimate stress state of pile foundation structures; the strain gradient (dσ / dz) is obtained by calculating the strain change rate along the depth direction of the pile foundation, and dσ / dz can reflect the deformation distribution characteristics of the structure; the bending modal frequency is obtained by extracting the first three main frequencies through fast Fourier transform (FFT), and the bending modal frequency can be used to analyze the vibration characteristics and bending modes of the pile foundation.

[0062] For collecting tilt sensor data:

[0063] First, the pile foundation is measured in three orthogonal directions (θ x ,θ y ,θ z ) to obtain a three-dimensional inclination sequence. Then, the feature engineering method is used to calculate the inclination sensor data. The rate of change of the inclination angle in each direction over time is calculated to obtain the inclination change rate (dθ / dt). dθ / dt can be used to evaluate the dynamic stability of the pile foundation; the equivalent rotation angle (θ eq ) to comprehensively characterize the overall spatial tilt state of the pile foundation.

[0064]

[0065] Based on the above method, the pile foundation strain data and inclination sensor data are used as pile foundation data for sea ice thickness prediction, which can fuse the pile foundation strain and inclination data and reveal the ice-pile mechanical coupling effect.

[0066] In one embodiment of the present disclosure, the ice anchor data includes sonar point cloud data and ice anchor growth parameter data, and the acquisition modal data includes:

[0067] Segmenting the ice anchor point cloud using the DBSCAN clustering algorithm, and extracting the volume, anchoring depth, distance from the pile foundation, and contact area of ​​the ice anchor point cloud as the sonar point cloud data; and

[0068] Estimation of ice anchor growth parameter data by Kalman filtering.

[0069] Specifically, for collecting sonar point cloud data:

[0070] First, data preprocessing was performed. This involved segmenting the sonar point cloud using the DBSCAN clustering algorithm to separate the ice anchor point cloud from background noise. This method identifies the geometric structure of ice anchors through density clustering and is suitable for unevenly distributed ice anchor point clouds.

[0071] The key to sonar point cloud segmentation lies in point cloud stitching and coordinate conversion, that is, unifying the point clouds collected from multiple perspectives into the same coordinate system to ensure spatial consistency; noise separation and vegetation removal, removing abnormal points (such as floating object reflection noise) through statistical filtering and thresholding; data resampling, reducing point cloud density, improving processing efficiency, while retaining the key morphological features of ice anchors.

[0072] Then, sonar point cloud data is collected based on the segmented ice anchor point cloud. Based on 3D reconstruction of the point cloud, the volume (V) of the space occupied by the ice anchor is calculated. V reflects the overall size of the ice anchor. The vertical distance between the bottom of the ice anchor and the plane of the pile foundation is measured to obtain the anchoring depth (D). D is used to assess the mechanical strength of the ice anchor embedded in the pile foundation. The shortest spatial distance between the ice anchor's center of mass and the pile foundation surface is calculated to obtain the distance to the pile foundation (L). L is used to analyze the risk of ice anchor impact. The actual contact area between the ice anchor and the pile foundation is estimated through point cloud surface fitting (such as triangulation), to obtain the contact area (S). S is used to evaluate the local stress distribution.

[0073] For collecting ice anchor growth parameter data:

[0074] The Kalman filter algorithm is used to estimate the ice anchor growth rate (dh / dt) in real time, and the predicted value is dynamically updated by combining the physical model with the observation data. Algorithm process:

[0075] Firstly, a kinetic model of ice anchor growth is established as the equation of state;

[0076]

[0077] Where w(k) is the process noise.

[0078] Then the observation equation is obtained by measuring the current ice thickness through sonar point cloud data;

[0079] z(k)=h(k)+v(k) (3)

[0080] Where v(k) is the measurement noise.

[0081] Finally, a prediction-correction cycle is performed to estimate the current ice thickness and covariance matrix based on the state at the previous moment, and the estimated values ​​are updated using real-time observations, the Kalman gain matrix is ​​calculated, and the estimation accuracy is optimized.

[0082] In one embodiment of the present disclosure, the meteorological data includes basic data and enhanced data, and the acquisition modality data includes:

[0083] respectively collecting temperature data, wind speed data, and radiation data as the basic data; and

[0084] A wind chill index is calculated based on the temperature data and the wind speed data to serve as the enhanced data.

[0085] Meteorological data includes basic data. Basic data can include temperature (T), which can be collected by deploying digital temperature sensors or infrared thermometers at different heights on the pile foundation. T directly affects the ice growth rate and phase change process. Basic data can also include wind speed (u), which can be collected using ultrasonic anemometers, cup anemometers, and other devices. During installation, the sensor orientation should be adjusted to align with the prevailing wind direction. u reflects the intensity of air flow and is correlated with evaporative heat dissipation from the ice surface. Basic data can also include radiation (Q), which can be measured by installing a pyranometer to measure solar radiation.

[0086] In order to fully mine the data of meteorological data, feature enhancement can also be performed, that is, enhancing data such as wind chill index is calculated based on basic data.

[0087] The Wind Chill Index (WCI) can be used to comprehensively quantify the combined effects of low temperature and wind speed on human perception, indirectly reflecting the heat dissipation rate from the ice surface. Its calculation formula is:

[0088] WCI=13.12+0.6215T-11.37u 0.16 +0.3965T·u 0.16 (4)

[0089] Where T is temperature in degrees Celsius (°C) and u is wind speed in kilometers per hour (km / h).

[0090] It should be noted that meteorological data is inherently spatiotemporal dependent and constrained by physical laws. Spatiotemporal information can be explicitly injected into the model through an encoding mechanism to improve prediction accuracy and interpretability. Therefore, meteorological data can be encoded in terms of spatiotemporal position.

[0091] The spatial coordinates (x, y) are constructed based on the latitude and longitude of the pile foundation. The sine function is used for even dimensions and the cosine function is used for odd dimensions. The timestamp encoding PE is:

[0092] PE (pos,2i) =sin(pos / 10000 2i / d ) (5)

[0093] PE (pos,2i+1) =cos(pos / 10000 2i / d ) (6)

[0094] Where i is the dimension index, which controls the frequency of different dimensions; d is the total dimension of the positional encoding, which is consistent with the model input embedding dimension.

[0095] At the same time, add the spatial distance weight term:

[0096] PE spatial =exp(-γ·||x-x0||) (7)

[0097] Where x is the coordinate vector of the current pile foundation, x0 is the coordinate of the reference point, γ is the attenuation coefficient, which controls the rate at which the weight decays with distance, and ||x-x0|| is the Euclidean distance from x0.

[0098] In step S102, feature encoding is performed on each of the modal data to obtain feature data, and the weight of each feature data is dynamically adjusted based on the current density of sea ice to perform feature fusion on the feature data to obtain cross-modal spatiotemporal fusion features.

[0099] In one embodiment of the present disclosure, the characteristic data includes pile foundation strain time series characteristics, ice anchor geometry time series characteristics, and meteorological time series characteristics. The characteristic encoding of each modal data to obtain each characteristic data includes:

[0100] Extracting the pile foundation strain time series features of the pile foundation data using a pile foundation data encoder; the pile foundation data encoder includes a convolution layer, a pooling layer, and a normalization layer;

[0101] Extracting the ice-anchor geometric temporal features of the ice-anchor data using an ice-anchor data encoder; the ice-anchor data encoder is a point cloud-specific Transformer, and the point cloud-specific Transformer includes an attention mechanism;

[0102] A meteorological data encoder is used to extract meteorological time series features of the meteorological data; the meteorological data encoder is a standard Transformer.

[0103] The input to the pile data encoder is pile data of shape [B, T, 5], where B is the batch size, T is the time step, and 5 is the feature dimension, namely strain extremes, strain gradients, bending modal frequencies, inclination angle rates, and equivalent rotation angles. The output shape of the pile data encoder is [B, 64, T], preserving the time dimension to support subsequent time series modeling.

[0104] The structure of the pile foundation data encoder includes convolution layer, pooling layer and normalization layer. The processing process is to first use the one-dimensional convolution layer to extract local time series features, then use the pooling layer to downsample and compress the time series dimension, and finally perform normalization and stabilization training through the normalization layer.

[0105] The input to the ice anchor data encoder is ice anchor data, with a shape of [B, N, 4], where B is the batch size, N is the number of point clouds, and 4 is the feature dimension, namely volume, anchor depth, distance to the pile foundation, and contact area. The output shape of the ice anchor data encoder is [B, 64, N], preserving the point cloud dimension to support spatial analysis.

[0106] The ice anchor data encoder is Point Transformer, a point cloud-specific Transformer that can support feature extraction from irregular point clouds, capture spatial geometric relationships using the attention mechanism in the local neighborhood of the point cloud, and perform dynamic feature aggregation to adapt to the diversity of ice anchor morphologies.

[0107] The input to the weather data encoder is weather data of shape [B, T, 4], where B is the batch size, T is the time step, and 4 is the feature dimension, including temperature, wind speed, radiation, and wind chill index. The output shape is [B, T, 64].

[0108] The feature encoding is performed separately through a hierarchical encoder to align the pile foundation data, ice anchor data and meteorological data, so that they can support cross-modal fusion.

[0109] After obtaining the pile foundation strain time series characteristics, ice anchor geometry time series characteristics, and meteorological time series characteristics through hierarchical coding, all three feature data are input into the global spatiotemporal Transformer. The specific processing process is as follows:

[0110] First, the input feature data is dimensionally aligned. The input feature data dimensions are adjusted from [B, C, T] (batch × channel × time step) to [T, B, C] (time step × batch × channel). This is to adapt to the Transformer's input sequence format requirements, that is, the sequence length dimension comes first to facilitate processing dependencies between time steps.

[0111] Multi-source feature fusion is then performed. Features from different data sources are spliced ​​along the channel dimension to fuse the multimodal spatiotemporal data of pile foundations, ice anchors, and meteorological conditions, preserving the independent information of each feature. This dimensional adjustment and splicing unifies spatial features and time series into a spatiotemporal sequence input, conforming to the Transformer's sequence processing paradigm.

[0112] Dynamic attention gating is then performed. A gating network generates attention weights, weightedly adjusts fused features, adaptively filters important spatiotemporal features, and suppresses noise or redundant information. The gating network structure is as follows: linear layer (192→64) → GELU activation → linear layer (64→1) → Sigmoid normalization. Compared to traditional multi-head self-attention, the use of a single-layer gating network reduces computational complexity and is suitable for multi-source heterogeneous feature fusion scenarios.

[0113] Figure 2 The following is a schematic diagram showing a flow chart of a feature cross-modal spatiotemporal fusion method in an exemplary embodiment of the present disclosure. Figure 2 As shown, the weight of each feature data is dynamically adjusted based on the current density of sea ice to perform feature fusion on each feature data to obtain a cross-modal spatiotemporal fusion feature, including:

[0114] Step S201, configuring basic weights corresponding to each of the feature data;

[0115] Step S202, obtaining the current density of sea ice in real time;

[0116] Step S203: When the current density exceeds a preset threshold density, dynamically adjust the basic weight based on the adjustment coefficient to obtain an adjustment weight corresponding to each feature data;

[0117] Step S204: Calculate the cross-modal spatiotemporal fusion features based on the adjustment weights.

[0118] Specifically, first assign the corresponding basic weight w to each feature data base Then, radar, drone and other monitoring equipment are used to obtain ice thickness and density data in the ice anchor dense area in real time as C current , which can be expressed as the ratio of the number of ice anchors in the current area to the total number of ice anchors. When the density exceeds the threshold C threshold When , the system determines it as a high-risk dense area and triggers the weight adjustment mechanism. The formula is as follows:

[0119] w dense =w base +α·(C current -C threshold ) (8)

[0120] Among them, w dense is the adjusted weight, w base is the basic weight, α is the adjustment coefficient, C current is the current density, C threshold is the threshold density, C current is the current density, and the weight of the ice anchor dense area increases nonlinearly with the density.

[0121] The adjustment coefficient α can be optimized in combination with historical ice data. For example, the recent density fluctuation rate can be calculated through the sliding window method, and α can be adaptively adjusted. The larger α is, the more significant the increase in the weight of the dense area is, reflecting a strong response to high-risk areas.

[0122] Based on the above method, through this dynamic ice condition weight adjustment mechanism, the system can adaptively focus on high-risk areas, balance the importance of global and local features, and meet the engineering needs of dynamic ice condition monitoring.

[0123] Finally, Transformer feature conversion is performed. The weighted fused features are processed using the standard Transformer architecture to generate cross-modal spatiotemporal fusion features. The parameter configuration can be set to 6 encoder layers + 3 decoder layers, multi-head attention, hidden layer dimension of 192, and the default feedforward network dimension of 2048. The encoder can capture global spatiotemporal dependencies, while the decoder can be used to generate prediction sequences. Using the standard Transformer architecture to capture global spatiotemporal dependencies is suitable for sequence-to-sequence prediction tasks. Furthermore, having fewer decoder layers than the encoder reduces the risk of overfitting in downstream prediction optimization tasks.

[0124] Based on the above method, a cross-modal feature fusion is designed, especially to capture the degree of mechanical coupling between ice anchors and pile foundations, which can simultaneously capture the temporal evolution and spatial distribution and improve the accuracy of prediction.

[0125] In step S103, the cross-modal spatiotemporal fusion features are projected into the physical parameter space to obtain physical constraint features, and the sea ice thickness prediction value is calculated based on the physical constraint features.

[0126] Specifically, after the global spatiotemporal Transformer extracts multimodal spatiotemporal features and generates a high-dimensional representation, the Physics Projector maps the representation to the physical parameter space, such as ice density and thermal conductivity, to build a bridge between deep learning and physical equations. The output of the projection layer directly corresponds to the physical quantity, enhancing the credibility of the model.

[0127] Physical constraints are embedded in the model, including regularization of the energy balance equation and Hamiltonian mechanics constraints, to ensure that the model output complies with the thermodynamic law of conservation of energy as well as the laws of mass conservation and momentum transfer, balancing data-driven and physical laws.

[0128] Among them, the energy balance equation is regularized:

[0129]

[0130] Hamiltonian mechanics constraints:

[0131]

[0132] At the same time, a parameter sharing mechanism is adopted to map the 192-dimensional features output by the Transformer to physical parameters such as ice density and thermal conductivity through the Physics Projector, thereby achieving physical interpretability of the parameters.

[0133] Based on the above method, the physical equations are embedded in the Transformer inductive bias to improve the interpretability of ice thickness prediction.

[0134] Then, the sea ice thickness prediction value h is calculated based on the above physical constraint characteristics. At the same time, the mean prediction μ and variance estimation σ can also be added. 2 The dual-headed output of is used to characterize the prediction uncertainty.

[0135] Specifically, we can also classify different risk levels based on ice thickness, and then formulate corresponding response strategies. After the sea ice thickness is predicted, timely responses can be triggered according to the risk level. The risk level can be output as a probability through Softmax.

[0136] Table 1 Risk levels and response strategies

[0137]

[0138] Figure 3 The following schematically illustrates a flow chart of a sea ice thickness prediction model in an exemplary embodiment of the present disclosure. Figure 3 As shown in the figure, after obtaining the pile foundation data, ice anchor data and meteorological data respectively, the corresponding layered encoder Local Encoder is used for feature encoding. Finally, the encoded feature data is input into the global spatiotemporal GlobalTransformer to obtain the ice thickness prediction head.

[0139] Based on the above method and the powerful representation capability of the Transformer model, a physically interpretable ice thickness prediction system was constructed, providing an intelligent solution for pile-based offshore photovoltaic systems to cope with extreme ice conditions.

[0140] Figure 4 A schematic diagram of the composition of a sea ice thickness prediction device in an exemplary embodiment of the present disclosure is shown schematically. Figure 4 As shown, the sea ice thickness prediction device 400 may include a collection module 401, a fusion module 402, and a prediction module 403.

[0141] Acquisition module 401, for acquiring modal data; the modal data includes pile foundation data, ice anchor data and meteorological data;

[0142] A fusion module 402 is configured to perform feature encoding on each of the modal data to obtain feature data, and dynamically adjust the weight of each feature data based on the current density of sea ice to perform feature fusion on the feature data to obtain a cross-modal spatiotemporal fusion feature;

[0143] The prediction module 403 is configured to project the cross-modal spatiotemporal fusion features into a physical parameter space to obtain physical constraint features, and calculate a predicted value of sea ice thickness based on the physical constraint features.

[0144] According to an exemplary embodiment of the present disclosure, the pile foundation data includes pile foundation strain data and inclination sensor data. The acquisition module 401 is further used to sample the distributed optical fiber strain sequence of the pile foundation according to a preset sampling rate, and extract the strain extreme value, strain gradient and bending modal frequency of the distributed optical fiber strain sequence as the pile foundation strain data; and measure the inclination angle of the pile foundation in three orthogonal directions in real time, and calculate the inclination change rate and equivalent rotation angle based on the inclination angle as the inclination sensor data.

[0145] According to an exemplary embodiment of the present disclosure, the ice anchor data includes sonar point cloud data and ice anchor growth parameter data. The acquisition module 401 is also used to segment the ice anchor point cloud using the DBSCAN clustering algorithm, and extract the volume, anchoring depth, distance from the pile foundation and contact area of ​​the ice anchor point cloud as the sonar point cloud data; and estimate the ice anchor growth parameter data through Kalman filtering.

[0146] According to an exemplary embodiment of the present disclosure, the meteorological data includes basic data and enhanced data, and the acquisition module 401 is also used to respectively collect temperature data, wind speed data and radiation data as the basic data; and calculate the wind chill index based on the temperature data and the wind speed data as the enhanced data.

[0147] According to an exemplary embodiment of the present disclosure, the feature data includes pile foundation strain time series features, ice anchor geometric time series features and meteorological time series features, and the fusion module 402 is further used to extract the pile foundation strain time series features of the pile foundation data using a pile foundation data encoder; the pile foundation data encoder includes a convolution layer, a pooling layer and a normalization layer; the ice anchor geometric time series features of the ice anchor data are extracted using an ice anchor data encoder; the ice anchor data encoder is a point cloud-specific Transformer, and the point cloud-specific Transformer includes an attention mechanism; the meteorological time series features of the meteorological data are extracted using a meteorological data encoder; the meteorological data encoder is a standard Transformer.

[0148] According to an exemplary embodiment of the present disclosure, the fusion module 402 is also used to configure the basic weights corresponding to each of the characteristic data; obtain the current density of sea ice in real time; when the current density exceeds a preset threshold density, dynamically adjust the basic weights based on the adjustment coefficient to obtain the adjustment weights corresponding to each of the characteristic data; and calculate the cross-modal spatiotemporal fusion features based on the adjustment weights.

[0149] According to an exemplary embodiment of the present disclosure, the physical constraint features include: energy balance equation regularization and Hamiltonian mechanics constraints.

[0150] The specific details of each module in the above-mentioned sea ice thickness prediction device 400 have been described in detail in the corresponding sea ice thickness prediction method, and therefore will not be repeated here.

[0151] It should be noted that although several modules or units of the device for action execution are mentioned in the detailed description above, this division is not mandatory. In fact, according to the embodiments of the present disclosure, the features and functions of two or more modules or units described above can be concretized in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided into multiple modules or units to be concretized.

[0152] In an exemplary embodiment of the present disclosure, a storage medium capable of implementing the above method is also provided. This storage medium may be a portable compact disc read-only memory (CD-ROM) and include program code, and may be run on a terminal device, such as a mobile phone. However, the program product of the present disclosure is not limited thereto. In this document, a readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0153] In an exemplary embodiment of the present disclosure, an electronic device capable of implementing the above method is also provided. Figure 5 The following schematically shows a structural diagram of a computer system of an electronic device in an exemplary embodiment of the present disclosure.

[0154] It should be noted that Figure 5 The computer system 500 of the electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0155] like Figure 5 As shown, computer system 500 includes a central processing unit (CPU) 501, which can perform various appropriate actions and processes according to programs stored in read-only memory (ROM) 502 or programs loaded from storage unit 508 into random access memory (RAM) 503. Various programs and data required for system operation are also stored in RAM 503. CPU 501, ROM 502, and RAM 503 are connected to each other via bus 504. Input / output (I / O) interface 505 is also connected to bus 504.

[0156] The following components are connected to the I / O interface 505: an input section 506 including a keyboard, a mouse, and the like; an output section 507 including devices such as a cathode ray tube (CRT), a liquid crystal display (LCD), and a speaker; a storage section 508 including a hard disk and the like; and a communication section 509 including a network interface card such as a LAN (Local Area Network) card or a modem. The communication section 509 performs communication processing via a network such as the Internet. A drive 510 is also connected to the I / O interface 505 as needed. Removable media 511, such as a magnetic disk, an optical disk, a magneto-optical disk, or a semiconductor memory, is installed in the drive 510 as needed, so that computer programs read therefrom can be installed into the storage section 508 as needed.

[0157] In particular, according to an embodiment of the present disclosure, the process described below with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network via the communication section 509, and / or installed from a removable medium 511. When the computer program is executed by the central processing unit (CPU) 501, various functions defined in the system of the present disclosure are performed.

[0158] It should be noted that the computer-readable medium shown in the embodiments of the present disclosure may be a computer-readable signal medium or a computer-readable storage medium or any combination of the two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or component, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, device or device. In the present disclosure, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, which carries a computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a computer-readable medium may be transmitted using any suitable medium, including but not limited to wireless, wired, or any suitable combination thereof.

[0159] The flowcharts and block diagrams in the accompanying drawings illustrate the possible implementation architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each box in the flowchart or block diagram can represent a module, program segment, or a part of code, and the above-mentioned module, program segment, or a part of code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flowchart, and the combination of boxes in the block diagram or flowchart, can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0160] The units involved in the embodiments described in this disclosure may be implemented in software or hardware, and the units described may also be provided in a processor. In some cases, the names of these units do not constitute limitations on the units themselves.

[0161] As another aspect, the present disclosure further provides a computer-readable medium, which may be included in the electronic device described in the above embodiments, or may exist independently and not be incorporated into the electronic device. The computer-readable medium carries one or more programs, and when the one or more programs are executed by the electronic device, the electronic device implements the methods described in the above embodiments.

[0162] It should be noted that although several modules or units of the device for action execution are mentioned in the detailed description above, this division is not mandatory. In fact, according to the embodiments of the present disclosure, the features and functions of two or more modules or units described above can be concretized in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided into multiple modules or units to be concretized.

[0163] Through the description of the above embodiments, it is easy for those skilled in the art to understand that the example embodiments described herein can be implemented by software or by combining software with necessary hardware. Therefore, the technical solution according to the embodiments of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, and includes several instructions to enable a computing device (which can be a personal computer, a server, a touch terminal, or a network device, etc.) to execute the method according to the embodiments of the present disclosure.

[0164] Other embodiments of the present disclosure will readily occur to those skilled in the art after considering the specification and practicing the invention disclosed herein. This disclosure is intended to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary techniques in the art not disclosed herein.

[0165] It should be understood that the present disclosure is not limited to the exact structures that have been described above and shown in the drawings, and that various modifications and changes can be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. A method for predicting sea ice thickness, characterized in that: include: Collecting modal data; the modal data includes pile foundation data, ice anchor data and meteorological data; Performing feature encoding on each of the modal data to obtain each feature data, and dynamically adjusting the weight of each feature data based on the current density of sea ice to perform feature fusion on the each feature data to obtain a cross-modal spatiotemporal fusion feature; The cross-modal spatiotemporal fusion features are projected into a physical parameter space to obtain physical constraint features, and a sea ice thickness prediction value is calculated based on the physical constraint features.

2. The sea ice thickness prediction method according to claim 1, characterized in that: The pile foundation data includes pile foundation strain data and inclination sensor data, and the collected modal data includes: Sampling the distributed optical fiber strain sequence of the pile foundation according to a preset sampling rate, and extracting the strain extreme value, strain gradient and bending modal frequency of the distributed optical fiber strain sequence as the pile foundation strain data; and The inclination angles of the pile foundation in three orthogonal directions are measured in real time, and the inclination change rate and the equivalent rotation angle are calculated based on the inclination angles as the inclination sensor data.

3. The sea ice thickness prediction method according to claim 1, characterized in that: The ice anchor data includes sonar point cloud data and ice anchor growth parameter data, and the acquisition modal data includes: Segmenting the ice anchor point cloud using the DBSCAN clustering algorithm, and extracting the volume, anchoring depth, distance from the pile foundation, and contact area of ​​the ice anchor point cloud as the sonar point cloud data; and Estimation of ice anchor growth parameter data by Kalman filtering.

4. The sea ice thickness prediction method according to claim 1, characterized in that: The meteorological data includes basic data and enhanced data, and the acquisition modal data includes: respectively collecting temperature data, wind speed data, and radiation data as the basic data; and A wind chill index is calculated based on the temperature data and the wind speed data to serve as the enhanced data.

5. The sea ice thickness prediction method according to claim 1, characterized in that: The characteristic data include pile foundation strain time series characteristics, ice anchor geometry time series characteristics and meteorological time series characteristics. The characteristic data are obtained by performing feature coding on each modal data, including: Extracting the pile foundation strain time series features of the pile foundation data using a pile foundation data encoder; the pile foundation data encoder includes a convolution layer, a pooling layer, and a normalization layer; Extracting the ice-anchor geometric temporal features of the ice-anchor data using an ice-anchor data encoder; the ice-anchor data encoder is a point cloud-specific Transformer, and the point cloud-specific Transformer includes an attention mechanism; A meteorological data encoder is used to extract meteorological time series features of the meteorological data; the meteorological data encoder is a standard Transformer.

6. The sea ice thickness prediction method according to claim 1, characterized in that: The dynamically adjusting the weights of the feature data based on the current density of sea ice to perform feature fusion on the feature data to obtain cross-modal spatiotemporal fusion features includes: Configuring basic weights corresponding to the respective feature data; Obtain the current density of sea ice in real time; When the current density exceeds a preset threshold density, dynamically adjusting the basic weight based on the adjustment coefficient to obtain an adjustment weight corresponding to each feature data; The cross-modal spatiotemporal fusion feature is calculated based on the adjustment weight.

7. The sea ice thickness prediction method according to claim 1, characterized in that: The physical constraint features include: energy balance equation regularization and Hamiltonian mechanics constraints.

8. A sea ice thickness prediction device, characterized in that: include: An acquisition module, configured to acquire modal data, including pile foundation data, ice anchor data, and meteorological data; A fusion module is used to perform feature encoding on each of the modal data to obtain each feature data, and dynamically adjust the weight of each feature data based on the current density of sea ice to perform feature fusion on the each feature data to obtain a cross-modal spatiotemporal fusion feature; A prediction module is used to project the cross-modal spatiotemporal fusion features into a physical parameter space to obtain physical constraint features, and calculate the sea ice thickness prediction value based on the physical constraint features.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the sea ice thickness prediction method according to any one of claims 1 to 7 is implemented.

10. An electronic device, characterized in that: include: one or more processors; A storage device for storing one or more programs, which, when executed by the one or more computer programs, enables the one or more processors to implement the sea ice thickness prediction method according to any one of claims 1 to 7.

Images