Large disastrous wave occurrence probability and wave height combined prediction method, device, equipment, medium and product

Abstract

The invention discloses a disastrous big wave occurrence probability and wave height combined prediction method, device and equipment, a medium and a product, and relates to the technical field of meteorological ocean forecasting. The method comprises the following steps: acquiring an effective wave height measured value and environmental factor data of a to-be-predicted region in a historical time period; key environment factors are screened through spatial correlation analysis and linear fitting verification, a first training data set and a second training data set are constructed, and a multiple linear regression model and a logic regression model are trained respectively to obtain a wave height prediction model and a big wave occurrence probability prediction model; and inputting the key environmental factors of the current time period of the to-be-predicted region into the model to obtain the predicted occurrence probability of the disastrous waves and the predicted value of the significant wave height in the predicted time period. According to the method, the key environmental factors are screened and the two models are used for respective prediction, the influence of the various environmental factors on the big waves is considered, the occurrence probability and the wave height of the disastrous big waves can be predicted more accurately, and the ocean safety is effectively guaranteed.

Description

Technical Field

[0001] This application relates to the field of meteorological and oceanographic forecasting technology, and in particular to a method, device, equipment, medium and product for jointly predicting the probability of occurrence of disastrous large waves and wave height. Background Technology

[0002] Ocean wave disasters play a very important role in the marine field. They not only cause deaths or disappearances, but also pose significant challenges to the design, construction, and operation and maintenance of various marine engineering projects, and constitute a major threat to marine safety and security. They are one of the natural disasters that cannot be ignored in marine work.

[0003] With the continuous advancement of science and technology, machine learning technology has been widely applied in various fields. In the field of forecasting catastrophic high waves, the forecasting methods of related technologies mainly rely on physical models and empirical formulas. Although these methods can provide forecast results to a certain extent, they suffer from problems such as low forecast accuracy and poor adaptability to complex marine environments.

[0004] Therefore, there is an urgent need for a method that combines the probability of occurrence of disastrous large waves with wave height to solve the problems of low forecast accuracy and poor adaptability to complex marine environments, thereby improving the accuracy and reliability of disastrous large wave forecasts. Summary of the Invention

[0005] The purpose of this application is to provide a method, device, equipment, medium, and product for jointly predicting the probability of occurrence and wave height of disastrous large waves. This method can comprehensively utilize historical and current environmental factor data of the area to be predicted, and through scientifically selecting key factors and constructing a targeted training dataset, train accurate wave height prediction models and large wave occurrence probability prediction models respectively. This enables accurate joint prediction of the probability of occurrence and effective wave height of disastrous large waves, providing a reliable basis for marine activity planning and safety assurance of marine engineering projects, and effectively reducing the risks and losses caused by wave disasters.

[0006] To achieve the above objectives, this application provides the following solution: Firstly, this application provides a method for jointly predicting the probability of catastrophic large waves and wave height, including: Obtain the effective wave height measured values ​​and corresponding environmental factor data for the historical time period of the area to be predicted; the environmental factor data includes sea temperature, salinity, zonal current velocity, meridional current velocity, sea surface height anomaly, mean sea level pressure, zonal wind speed, meridional wind speed, and atmospheric boundary layer height. Based on historical environmental factor data of the region to be predicted, key environmental factors related to the measured values ​​of significant wave height are screened through spatial correlation analysis and linear fitting verification methods. The key environmental factors include mean sea level pressure, zonal wind speed and meridional wind speed. A first training dataset is constructed based on key environmental factors and corresponding measured values ​​of significant wave height; a second training dataset is constructed based on key environmental factors and corresponding actual probability of large waves occurring; the actual probability of large waves occurring is determined based on the measured values ​​of significant wave height. A multivariate linear regression model was trained using the first training dataset to obtain a well-trained wave height prediction model; a logistic regression model was trained using the second training dataset to obtain a well-trained large wave probability prediction model. By inputting the key environmental factors of the current time period of the area to be predicted into the trained wave height prediction model and the wave occurrence probability prediction model, the predicted probability of catastrophic large waves and the effective wave height prediction value of the predicted time period of the area to be predicted are obtained.

[0007] Secondly, this application provides a device for jointly predicting the probability of catastrophic high waves and wave height, comprising: The data acquisition module is used to acquire the effective wave height measured values ​​and corresponding environmental factor data of the area to be predicted over a historical period. The environmental factor data includes sea temperature, salinity, zonal current velocity, meridional current velocity, sea surface height anomaly, mean sea level pressure, zonal wind speed, meridional wind speed, and atmospheric boundary layer height. The key environmental factor screening module is used to screen key environmental factors related to the measured values ​​of significant wave height based on historical environmental factor data of the area to be predicted, through spatial correlation analysis and linear fitting verification methods. The key environmental factors include mean sea level pressure, zonal wind speed and meridional wind speed. The training set construction module is used to construct a first training dataset based on key environmental factors and corresponding measured values ​​of effective wave height; and to construct a second training dataset based on key environmental factors and corresponding actual probability of large waves occurring; wherein the actual probability of large waves occurring is determined based on the measured values ​​of effective wave height. The model training module is used to train a multiple linear regression model using the first training dataset to obtain a trained wave height prediction model; and to train a logistic regression model using the second training dataset to obtain a trained large wave probability prediction model. The prediction output module is used to input the key environmental factors of the area to be predicted for the current time period into the trained wave height prediction model and the wave occurrence probability prediction model, respectively, to obtain the predicted probability of catastrophic waves and the effective wave height prediction value for the predicted time period of the area to be predicted.

[0008] Thirdly, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the method for jointly predicting the probability of occurrence of catastrophic large waves and wave height as described above.

[0009] Fourthly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method for jointly predicting the probability of occurrence of catastrophic large waves and wave height as described above.

[0010] Fifthly, this application provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the method for jointly predicting the probability of occurrence of catastrophic large waves and wave height as described above.

[0011] According to the specific embodiments provided in this application, this application has the following technical effects: This application provides a method, device, equipment, medium, and product for jointly predicting the probability of hazardous large waves and wave height. By acquiring historical measured values ​​of effective wave height and various environmental factors, including sea level pressure and sea surface temperature, for the area to be predicted, it solves the problem of traditional prediction methods having limited data acquisition and lacking comprehensive marine environmental information support, thus providing a rich and fundamental data reserve for subsequent accurate predictions. Based on historical environmental factor data, spatial correlation analysis and linear fitting verification methods are used to screen out key environmental factors such as wind speed that are closely related to the measured values ​​of effective wave height. This solves the problem of traditional methods failing to effectively identify key influencing factors, leading to complex and inaccurate prediction models. It achieves a focus on key factors, simplifies the prediction model, and improves its relevance. By constructing first and second training datasets for the corresponding measured values ​​of effective wave height and the actual probability of large waves based on key environmental factors, and using these datasets to train multiple linear regression and logistic regression models, a wave height prediction model and a large wave probability prediction model are obtained. This solves the problems of low accuracy and poor adaptability to complex marine environments in traditional forecasting methods, achieving accurate joint prediction of the probability of hazardous large waves and effective wave height. Attached Figure Description

[0012] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 This is an application environment diagram of a method for jointly predicting the probability of catastrophic large waves and wave height according to an embodiment of this application; Figure 2 A flowchart illustrating a method for jointly predicting the probability of catastrophic high waves and wave height according to an embodiment of this application; Figure 3A schematic diagram of the functional modules of a device for jointly predicting the probability of occurrence of catastrophic large waves and wave height, provided in an embodiment of this application; Figure 4 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0014] First, some technical terms involved in the embodiments of this application will be introduced.

[0015] The impacts of wave disasters mainly include the following aspects: 1) Personnel safety. Maritime accidents and other marine disasters caused by ocean waves can result in injuries, fatalities, and disappearances. This is especially true for maritime workers, tourists, and ocean transport operators, as sudden changes in wave patterns can pose serious safety hazards.

[0016] 2) Marine engineering. Ocean waves have a significant impact on the design, construction, and maintenance of marine engineering projects. For example, the stability and safety of marine facilities such as ports, wharves, and coastal protection projects need to take into account the effects of ocean waves.

[0017] 3) Marine resource development. Ocean waves also affect activities such as offshore energy development and seabed resource extraction. The impact of ocean waves needs to be considered when planning and implementing projects such as offshore wind power and offshore oil and gas extraction.

[0018] 4) Marine environmental protection. Strong waves can cause the spread and deposition of marine pollutants, which in turn affects the stability and health of the marine ecosystem.

[0019] With the continuous advancement of technology, machine learning has been widely applied in various fields. Machine learning can discover potential patterns in large amounts of data through learning and analysis, and then make predictions accordingly. In the field of forecasting catastrophic high waves, traditional forecasting methods mainly rely on physical models and empirical formulas. While these methods can provide forecast results to some extent, they suffer from problems such as low forecast accuracy and poor adaptability to complex marine environments. Machine learning technology, on the other hand, can fully utilize the richness and diversity of marine observation data to improve the accuracy and reliability of catastrophic high wave forecasts by establishing data-driven models.

[0020] Therefore, this application provides a method for jointly predicting the probability of catastrophic large waves and wave height.

[0021] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0022] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0023] The method for jointly predicting the probability of catastrophic high waves and wave height provided in this application can be applied to, for example... Figure 1 The application environment shown is illustrated. Terminal 101 communicates with server 102 via a network. A data storage system can store the data that server 102 needs to process. The data storage system can be set up independently, integrated into server 102, or placed in the cloud or on another server. Terminal 101 can send the measured values ​​of significant wave height and corresponding environmental factor data of the historical time period of the area to be predicted to server 102. After receiving the measured values ​​of significant wave height and corresponding environmental factor data of the historical time period of the area to be predicted, server 102 uses spatial correlation analysis and linear fitting verification methods to screen key environmental factors related to the measured values ​​of significant wave height based on the environmental factor data of the historical time period of the area to be predicted. The key environmental factors include mean sea level pressure, zonal wind speed, and meridional wind speed. A first training dataset is constructed based on the key environmental factors and the corresponding measured values ​​of significant wave height. A second training dataset is constructed based on the key environmental factors and the corresponding actual probability of large waves. The actual probability of large waves is determined based on the measured values ​​of significant wave height. A multiple linear regression model is trained using the first training dataset to obtain a trained wave height prediction model. A logistic regression model is trained using the second training dataset to obtain a trained large wave probability prediction model. The key environmental factors of the area to be predicted for the current time period are input into the trained wave height prediction model and the large wave probability prediction model to obtain the predicted probability of disastrous large waves and the predicted value of significant wave height for the predicted time period of the area to be predicted. Server 102 can feed back the predicted probability of catastrophic waves and the predicted significant wave height for the predicted time period of the area to be predicted to terminal 101. Furthermore, in some embodiments, the joint prediction method of catastrophic wave probability and wave height can also be implemented separately by server 102 or terminal 101. For example, terminal 101 can directly perform prediction processing based on the measured significant wave height and corresponding environmental factor data for the historical time period of the area to be predicted; alternatively, server 102 can obtain the measured significant wave height and corresponding environmental factor data for the historical time period of the area to be predicted from the data storage system and perform prediction processing based on these data.

[0024] The terminal 101 can be, but is not limited to, various desktop computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. The server 102 can be implemented using a standalone server or a server cluster consisting of multiple servers, or it can be a cloud server.

[0025] In one exemplary embodiment, such as Figure 2 As shown, a method for jointly predicting the probability of catastrophic high waves and wave height is provided. This method is executed by computer equipment, specifically by a terminal or server alone, or by both a terminal and a server. In this embodiment, the method is applied to... Figure 1 Taking server 102 as an example, the explanation includes the following steps 201 to 205. Wherein: Step 201: Obtain the effective wave height measured values ​​and corresponding environmental factor data for the historical time period of the area to be predicted; the environmental factor data includes sea temperature, salinity, zonal current velocity, meridional current velocity, sea surface height anomaly, mean sea level pressure, zonal wind speed, meridional wind speed, and atmospheric boundary layer height.

[0026] Step 202: Based on historical environmental factor data of the area to be predicted, key environmental factors related to the measured values ​​of significant wave height are screened through spatial correlation analysis and linear fitting verification methods; the key environmental factors include mean sea level pressure, zonal wind speed and meridional wind speed.

[0027] Step 203: Construct a first training dataset based on key environmental factors and corresponding measured values ​​of effective wave height; construct a second training dataset based on key environmental factors and corresponding actual probability of large waves occurring; the actual probability of large waves occurring is determined based on the measured values ​​of effective wave height.

[0028] Step 204: Train a multiple linear regression model using the first training dataset to obtain a trained wave height prediction model; train a logistic regression model using the second training dataset to obtain a trained large wave probability prediction model.

[0029] Step 205: Input the key environmental factors of the current time period of the area to be predicted into the trained wave height prediction model and the large wave occurrence probability prediction model respectively to obtain the predicted probability of catastrophic large waves and the effective wave height prediction value of the predicted time period of the area to be predicted.

[0030] By implementing steps 201 to 205 above, this application can improve the accuracy and timeliness of large wave prediction. By defining the process and number of days of disastrous large waves and using machine learning models to accurately predict the probability of the occurrence of disastrous large waves, the accuracy and timeliness of large wave disaster prediction can be improved.

[0031] In another exemplary embodiment of this application, based on environmental factor data of the area to be predicted over a historical period, key environmental factors related to the measured values ​​of significant wave height are screened through spatial correlation analysis and linear fitting verification methods. This process also includes: Preprocessing is performed on the environmental factor data of the historical time period of the region to be predicted; the preprocessing operation specifically includes: labeling the values ​​of missing or abnormal data in the environmental factor data as "NAN" to obtain the environmental factor preprocessing dataset.

[0032] The region to be predicted is divided into a regular grid structure according to a preset spatial resolution, and the environmental factor data in the environmental factor preprocessing dataset is mapped to the corresponding grid points to obtain the environmental factor grid dataset.

[0033] In another exemplary embodiment of this application, based on historical environmental factor data of the area to be predicted, key environmental factors related to the measured values ​​of significant wave height are screened through spatial correlation analysis and linear fitting verification methods, specifically including: Spatial correlation analysis was performed using the following formula to obtain the Pearson correlation coefficient between each environmental factor and the measured significant wave height in the environmental factor grid dataset: .

[0034] in, Indicates the first The Pearson correlation coefficient between an environmental factor and the measured value of significant wave height, with a value range of [−1, 1], represents the degree of linear correlation between the two variables. The total time is the number of time steps (days) involved in the calculation. express time; express Measured effective wave height at time [time]. This represents the average of all measured significant wave heights within a range T. express The first moment The value of an environmental factor is the value of its environmental variable at a certain moment (such as sea temperature, salinity, current velocity component, air pressure, wind speed component, etc.). For T, the first The average of all values ​​of an environmental factor represents the average value of that environmental variable over the entire time series.

[0035] The effective wave height measured value for each day in the environmental factor grid dataset is concatenated with all corresponding environmental factors to form a one-dimensional vector, resulting in multiple one-dimensional vectors.

[0036] The linear regression method is used to fit multiple one-dimensional vectors to obtain the fitting slope, fitting intercept, and coefficient of determination.

[0037] Environmental factors whose absolute value of the Pearson correlation coefficient is greater than a preset correlation coefficient threshold and whose coefficient of determination is greater than a preset coefficient of determination threshold are considered as key environmental factors.

[0038] In another exemplary embodiment of this application, the wave height prediction model obtains the effective wave height prediction value through the following formula: .

[0039] in, This represents the predicted effective wave height. Indicates the reference value for effective wave height; Indicates the first Linear regression coefficients of key environmental factors; Indicates the first One key environmental factor; This represents the error term between the predicted and measured values ​​of the significant wave height.

[0040] In another exemplary embodiment of this application, the probability prediction model for large wave occurrence calculates the predicted probability of catastrophic large wave occurrence using the following formula: .

[0041] in, This indicates the predicted probability of a disastrous large wave event. This indicates the occurrence of a catastrophic wave. This represents the baseline probability of a catastrophic wave event. Indicates the first Logistic regression coefficients of key environmental factors; Indicates the first Key environmental factors.

[0042] In another exemplary embodiment of this application, after obtaining the predicted probability of catastrophic large waves and the predicted significant wave height for the predicted time period of the area to be predicted, the method further includes: Based on the predicted probability of catastrophic large waves and the predicted effective wave height for the predicted time period in the area to be predicted, a predicted probability map and an initial predicted wave height map for the predicted time period in the area to be predicted are generated.

[0043] Based on the predicted probability map of the predicted time period of the area to be predicted, the contour detection algorithm is used to spatially cluster cells with a predicted probability of catastrophic waves greater than or equal to a preset probability threshold, and these cells are marked as high-risk areas, thus obtaining a contour map of the high-risk areas.

[0044] As an optional implementation, based on the model output data (including predicted wave height predicted_hs and probability of big waves prob_bigwave) for the predicted area over a given time period, the maximum value of the predicted hourly effective wave height (Hs) is taken for each grid point on a daily basis. A Boolean threshold is used to determine whether Hs ≥ 4m, and a spatial mask of hazardous big waves (indicating whether a grid point is a hazardous wave height) and the daily maximum wave height value of the corresponding grid point are output. The probability map of big waves uses cmap='Oranges' color code to display the probability distribution, and high-risk areas are marked with a probability ≥ 0.7. The daily maximum wave height map uses cmap='viridis' color code to display the predicted wave height, focusing on the wave height range of 4m–8m to highlight areas with severe sea conditions. A visualization function (such as make_png) is called to generate an image, overlaying the coastline and contour lines, adding a scale bar and time label, and finally outputting a high-resolution PNG file. At the same time, the maximum wave height, probability data, and mask results are stored synchronously in a NetCDF format file.

[0045] The high-risk area contour map is overlaid onto the initial wave height prediction map to obtain the risk wave height fusion prediction map.

[0046] The following example illustrates this application using the joint prediction process of the probability of occurrence of a catastrophic large wave and its wave height.

[0047] Step S1: Collect numerical wave forecast data and meteorological oceanographic reanalysis data for the target area over a historical time period (e.g., 20 years), and obtain data such as significant wave height, sea level pressure, sea surface temperature, and sea surface salinity for the target area based on the collected data.

[0048] Step S2: Based on the above collected data, perform multivariate matching and spatiotemporal correlation methods to extract information on catastrophic wave events in the target area during historical time periods.

[0049] Step S3: Use spatial correlation analysis and linear fitting to verify the correlation coefficients between the height of the disaster wave and eight environmental factors.

[0050] Step S4: Construct wave height prediction model and large wave occurrence probability prediction model based on multiple linear regression model and logistic regression model.

[0051] Step S5: Input historical data into the machine learning model to obtain wave height prediction and large wave probability prediction.

[0052] Step S1 involves collecting historical numerical wave forecast data and meteorological-oceanographic reanalysis data for the target area. During extraction, a unified time format and data integrity checks are required to ensure complete coverage of data time periods within each natural day, meeting the statistical accuracy requirements for the number of days with severe high waves. Hourly effective wave height data for each grid point is traversed, and each time period is assessed to determine if it meets the threshold criteria. The results are grouped by natural day. If the threshold condition is met in any time period, that grid point is considered a day with severe high waves. Continuous wave heights or sustained events are not required here; only whether a severe wave height occurred once per day is counted, i.e., the "maximum wave height trigger" standard is used.

[0053] The extracted valid wave height data is decoded, and the scale_factor and add_offset variables in the variable attributes are applied to perform data transformation. Missing values ​​are then handled, and missing data are marked as NaN.

[0054] Step S2 specifically includes: Based on the collected data, performing multivariate matching and spatiotemporal correlation methods to extract information on disastrous large waves within the target area over historical time periods, marking the occurrence time, and storing the large wave events in the form of a two-dimensional logical array and a wave height value array. Converting the effective wave height data to a daily scale, recording the maximum effective wave height value each day. Converting the time variable to the standard datetime64[D] type, extracting a unique set of dates. Extracting the maximum wave height from the daily data (taking the maximum hourly), determining whether each grid point meets the threshold condition, and generating a Boolean mask (waveMask) to identify the occurrence of disastrous large waves.

[0055] The timing of the marker event serves three main purposes: Achieve spatiotemporal data matching: Ensure that the large wave event and the corresponding environmental factors are accurately aligned in the time dimension, providing a spatiotemporal correlation basis for the key environmental factor screening module.

[0056] To ensure statistical accuracy: Large wave events are grouped by natural day, and the statistical standard of "whether a disastrous wave height has occurred once a day" is achieved through time stamping, which meets the "maximum effective wave height trigger" judgment logic.

[0057] To verify the time correspondence of the prediction results: mark the occurrence time of historical events, and verify the accuracy of the prediction results in the time dimension by comparing them with the probability graph output by the model.

[0058] Step S3 specifically includes: using spatial correlation analysis and linear fitting to verify the calculation of Pearson correlation coefficients between the height of hazardous high waves and eight environmental factors (including sea surface temperature, sea surface salinity, sea surface meridional current, sea surface zonal current, sea surface height anomaly, sea level pressure, the U component of the 10m wind field, and the V component of the 10m height wind field). First, the Pearson correlation coefficient is used to measure the strength of the linear association between each factor and the significant wave height. Then, linear regression analysis is used to quantitatively assess the explanatory power (R²). The closer R² is to 1, the stronger the explanatory power. Finally, factors with high correlation coefficients, significant regression effects, and good fit to subsequent forecast models are selected as key variables.

[0059] Spatial correlation analysis: Raster data is read daily by variable and aligned with the temporal and spatial dimensions of hsValues ​​(significant wave height). The Pearson correlation coefficient is used to calculate the correlation between the variable and hsValues ​​at each grid point. The formula is shown below: .

[0060] in, Indicates the first Correlation coefficients between environmental factors and measured values ​​of significant wave height; Total time, total length of the time series (total number of times). express time; express Measured effective wave height at time [time]. This represents the average of all measured significant wave heights within a range T. express The first moment The values ​​of environmental factors, For T, the first The average of all values ​​of the environmental factors.

[0061] The calculated correlation coefficients are encapsulated in the `safe_corr()` function to avoid the influence of missing values. The `LinearRegression` function from sklearn is used to fit the model, and the fit slope, intercept, and R-squared are evaluated. 2 Output visualized related analysis results.

[0062] The linear fit validation involved reading raster data daily by variable, extracting the effective wave height field (hsValues) and corresponding eight environmental factors for each date, concatenating the pixel values ​​of all dates into a one-dimensional vector, and performing linear fitting on non-NaN samples. A multiple linear regression model was used for fitting, and the fit slope, intercept, and R-squared were evaluated. 2 The value is output, and the relevant analysis results are visualized. Finally, the wind speed is selected. The U-component of the 10m wind field, the V-component of the 10m wind field, sea level pressure, and sea surface temperature were used as key factors to train the model.

[0063] Step S4, the machine learning model building and training module, constructs a wave height prediction model and a large wave occurrence probability prediction model based on a multiple linear regression model and a logistic regression model: Linear Regression Model Training: Create a linear regression model object, fit the model using the training set, calculate the R² score on the test set, and print the R² value of the augmented linear regression. Multiple linear regression can be used to predict daily wave heights, as shown in the following formula: .

[0064] in, This represents the predicted effective wave height. Indicates the reference value for effective wave height; Indicates the first Linear regression coefficients of key environmental factors; Indicates the first One key environmental factor; This represents the error term between the predicted and measured values ​​of the significant wave height.

[0065] Logistic Regression Model Training: Create a logistic regression model object, set the maximum number of iterations to 1000, fit the model using the training set, calculate the classification accuracy on the test set, and print the accuracy of the augmented logistic regression. Logistic regression is used to predict the probability of large waves occurring, with a probability ≥ 0.7 used as the threshold for high-risk areas. The formula is as follows: .

[0066] in, This indicates the predicted probability of a disastrous large wave event. This indicates the occurrence of a catastrophic wave. This represents the baseline probability of a catastrophic wave event. Indicates the first Logistic regression coefficients of key environmental factors; Indicates the first A key environmental factor. Z indicates whether a large wave will occur (1 indicates occurrence, 0 indicates no occurrence).

[0067] Step S5 specifically includes: inputting historical data into a machine learning model, using a regression model to predict specific wave height values, using a classification model to predict the probability of large waves occurring, performing spatial clustering on grid cells with a probability ≥ 0.7, automatically marking high-risk areas, and overlaying the marking results onto the wave height prediction map, outputting both a probability map and a wave height map of disastrous large waves. Only the marking results (high-risk areas) are overlaid onto the wave height prediction map; the core function of the wave height prediction map is to "highlight disastrous large wave areas," aiming to visually identify risk areas while presenting wave height values. The output probability map does not overlay the marking results, aiming to independently reflect the probability distribution characteristics; therefore, the probability map is without overlayed marking results.

[0068] This application also provides an application scenario in which the aforementioned method for jointly predicting the probability of occurrence of disastrous large waves and wave height is applied. Specifically, the method for jointly predicting the probability of occurrence of disastrous large waves and wave height provided in this embodiment can be applied to offshore oil extraction operations. Offshore oil extraction operations include a data acquisition stage by offshore monitoring equipment, a data transmission stage, and a decision-making stage. Data enters the data transmission stage from the offshore monitoring equipment acquisition stage, is processed and analyzed by the server to obtain disastrous large wave prediction information, and then enters the decision-making stage. The method for jointly predicting the probability of occurrence of disastrous large waves and wave height provided in this embodiment belongs to the early warning stage of the decision-making stage. Specifically, during offshore oil extraction operations, offshore monitoring equipment continuously collects relevant data of the area to be predicted, including historical and current measured values ​​of effective wave height and various environmental factor data. The early warning stage of the decision-making stage, according to the prediction method of this application, first screens key environmental factors, then constructs a training dataset, and trains a wave height prediction model and a large wave occurrence probability prediction model. Based on the key environmental factors input into the model for the current time period, the predicted probability of occurrence of disastrous large waves and the predicted value of effective wave height are obtained. Based on this forecast information, staff can develop response strategies in advance before extreme weather arrives, such as suspending offshore operations, reinforcing and protecting offshore platforms and equipment, and arranging personnel evacuation, thereby effectively ensuring the safety of offshore oil extraction operations, reducing the losses caused by wave disasters to personnel, equipment and the environment, and ensuring the smooth progress of offshore oil extraction activities.

[0069] Based on the same inventive concept, this application also provides a device for jointly predicting the probability of catastrophic high waves and wave height to implement the aforementioned method for jointly predicting the probability of catastrophic high waves and wave height. The solution provided by this device is similar to the solution described in the above method. Therefore, the specific limitations of one or more embodiments of the device for jointly predicting the probability of catastrophic high waves and wave height provided below can be found in the limitations of the method for jointly predicting the probability of catastrophic high waves and wave height described above, and will not be repeated here.

[0070] In one exemplary embodiment, such as Figure 3 As shown, a device for jointly predicting the probability of catastrophic large waves and wave height is provided, comprising: The data acquisition module 301 is used to acquire the effective wave height measured values ​​and corresponding environmental factor data of the historical time period of the area to be predicted; the environmental factor data includes sea level pressure, sea surface temperature, sea surface salinity, sea surface meridional current, sea surface zonal current, sea surface height anomaly, U component of preset height wind field and V component of preset height wind field.

[0071] The key environmental factor screening module 302 is used to screen key environmental factors related to the measured values ​​of significant wave height based on historical environmental factor data of the area to be predicted, through spatial correlation analysis and linear fitting verification methods; the key environmental factors include mean sea level pressure, zonal wind speed and meridional wind speed.

[0072] The training set construction module 303 is used to construct a first training dataset based on key environmental factors and corresponding measured values ​​of effective wave height; and to construct a second training dataset based on key environmental factors and corresponding actual probability of large waves occurring; wherein the actual probability of large waves occurring is determined based on the measured values ​​of effective wave height.

[0073] The model training module 304 is used to train a multiple linear regression model using the first training dataset to obtain a trained wave height prediction model; and to train a logistic regression model using the second training dataset to obtain a trained large wave probability prediction model.

[0074] The prediction output module 305 is used to input the key environmental factors of the area to be predicted for the current time period into the trained wave height prediction model and the wave occurrence probability prediction model, respectively, to obtain the predicted probability of the occurrence of disastrous waves and the effective wave height prediction value of the area to be predicted for the predicted time period.

[0075] In summary, this application first collects data to obtain historical wave numerical forecast data and meteorological-oceanic reanalysis data for the target area. Second, it extracts information on hazardous large waves within the target area during historical time periods using multivariate matching and spatiotemporal correlation methods. Then, it verifies and selects key environmental factors that significantly influence large wave occurrence through spatial correlation analysis and linear fitting. Next, it constructs a large wave occurrence probability forecasting model using machine learning algorithms and trains and optimizes the model using the extracted historical sample set. Finally, it inputs the real-time collected data into the trained model for calculation and outputs the forecast result of the large wave occurrence probability.

[0076] Therefore, this application solves the problem of high computational cost in traditional numerical simulation and achieves efficient and quantitative prediction of the spatial distribution of large wave probability. It overcomes the limitations of traditional methods for forecasting disastrous large waves, which rely on physical models and empirical formulas, resulting in low forecast accuracy and poor adaptability to complex marine environments. By comprehensively acquiring data on multiple environmental factors and accurately selecting key environmental factors, a targeted training dataset is constructed. Multiple linear regression and logistic regression models are used to accurately predict significant wave height and the probability of disastrous large waves, respectively. This provides more reliable and accurate forecast information for marine operations, marine engineering, marine resource development, and marine environmental protection, effectively reducing the risks and losses caused by wave disasters and improving the level of marine safety.

[0077] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal, and its internal structure diagram may be as follows. Figure 4 As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs in the non-volatile storage media to run. The database is used for processing data related to the joint prediction of catastrophic wave probability and wave height. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When the computer program is executed by the processor, it implements a method for the joint prediction of catastrophic wave probability and wave height.

[0078] Those skilled in the art will understand that Figure 4 The structures shown are merely block diagrams of some structures related to the present application and do not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than shown in the figures, or combine certain components, or have different component arrangements. In an exemplary embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0079] In one exemplary embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0080] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0081] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0082] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0083] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0084] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0085] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for jointly predicting the occurrence probability and wave height of a disastrous sea, characterized in that, The disaster wave occurrence probability and wave height joint prediction method comprises: Obtain effective wave height measured values and corresponding environmental factor data of a historical time period of a region to be predicted; the environmental factor data comprises sea temperature, salinity, zonal flow velocity, meridional flow velocity, sea surface height anomaly, average sea level pressure, zonal wind speed, meridional wind speed and atmospheric boundary layer height; Based on the environmental factor data of the historical time period of the region to be predicted, key environmental factors related to the effective wave height measured values are screened through spatial correlation analysis and linear fitting verification methods; the key environmental factors comprise average sea level pressure, zonal wind speed and meridional wind speed; Based on the key environmental factors and corresponding effective wave height measured values, a first training data set is constructed; based on the key environmental factors and corresponding actual occurrence probability of a big wave, a second training data set is constructed; the actual occurrence probability of the big wave is determined based on the effective wave height measured values; A multivariate linear regression model is trained using the first training data set to obtain a trained wave height prediction model; a logic regression model is trained using the second training data set to obtain a trained big wave occurrence probability prediction model; The key environmental factors of a current time period of the region to be predicted are input into the trained wave height prediction model and the trained big wave occurrence probability prediction model respectively to obtain a predicted occurrence probability of a disaster big wave and a predicted effective wave height of a predicted time period of the region to be predicted.

2. The method according to claim 1, wherein, Based on the environmental factor data of the historical time period of the region to be predicted, key environmental factors related to the effective wave height measured values are screened through spatial correlation analysis and linear fitting verification methods, which further comprises: A pretreatment operation is performed on the environmental factor data of the historical time period of the region to be predicted; the pretreatment operation specifically comprises: marking the values of missing data or abnormal data in the environmental factor data as "NAN" to obtain an environmental factor pretreatment data set; The region to be predicted is divided into a regular grid structure according to a preset spatial resolution, and the environmental factor data in the environmental factor pretreatment data set is mapped to corresponding grid points to obtain an environmental factor grid data set.

3. The method according to claim 2, wherein, Based on the environmental factor data of the historical time period of the region to be predicted, key environmental factors related to the effective wave height measured values are screened through spatial correlation analysis and linear fitting verification methods, which specifically comprises: Spatial correlation analysis is performed through the following formula to obtain the Pearson correlation coefficient of each environmental factor in the environmental factor grid data set and the effective wave height measured values: ； in, Indicates the first Pearson correlation coefficients between environmental factors and measured values ​​of significant wave height; Total time; express time; express Measured effective wave height at time [time]. This represents the average of all measured significant wave heights within a range T. express The first moment The values ​​of environmental factors, For T, the first The average of all values ​​of a given environmental factor; The effective wave height measured values and corresponding all environmental factors of each day in the environmental factor grid data set are spliced into one-dimensional vectors to obtain a plurality of one-dimensional vectors; The plurality of one-dimensional vectors are fitted through a linear regression method to obtain a fitting slope, a fitting intercept and a determination coefficient; The environmental factors with an absolute value of the Pearson correlation coefficient greater than a preset correlation coefficient threshold value and a determination coefficient greater than a preset determination coefficient threshold value are taken as the key environmental factors.

4. The method according to claim 1, wherein, The wave height prediction model obtains the effective wave height prediction value through the following formula: ； wherein, represents the predicted value of significant wave height; represents the reference value of significant wave height; represents the linear regression coefficient of the th key environmental factor; represents the th key environmental factor; represents the error term of the predicted value of significant wave height and the measured value of significant wave height.

5. The method according to claim 1, wherein, The big wave occurrence probability prediction model calculates the predicted occurrence probability of a disaster big wave through the following formula: ； wherein, represents the probability of occurrence of a disastrous large wave; represents the occurrence of a disastrous large wave; represents the reference probability of occurrence of a disastrous large wave; represents the logistic regression coefficient of the th key environmental factor; represents the th key environmental factor.

6. The method according to claim 1, wherein, The predicted occurrence probability of a disaster big wave and the predicted effective wave height of a predicted time period of the region to be predicted are obtained, which further comprises: generate a prediction probability map and an initial wave height prediction map for the prediction time period of the to-be-predicted region based on the disastrous wave occurrence probability and the effective wave height prediction value of the to-be-predicted region prediction time period; based on the prediction probability map of the to-be-predicted region prediction time period, the contour detection algorithm is used to perform spatial clustering on the cells with a disastrous wave occurrence probability greater than or equal to a preset probability threshold, and mark them as high-risk areas to obtain a high-risk area contour map; and the high-risk area contour map is superimposed on the initial wave height prediction map to obtain a risk wave height fusion prediction map.

7. A device for jointly predicting the probability of catastrophic large waves and wave height, characterized in that, The disastrous wave occurrence probability and wave height joint prediction device applies the disastrous wave occurrence probability and wave height joint prediction method of any one of claims 1-6, and the disastrous wave occurrence probability and wave height joint prediction method device comprises: a data acquisition module configured to acquire effective wave height measured values and corresponding environmental factor data of a historical time period of a to-be-predicted region; the environmental factor data includes sea temperature, salinity, zonal flow rate, meridional flow rate, sea surface height anomaly, mean sea level pressure, zonal wind speed, meridional wind speed, and atmospheric boundary layer height; a key environmental factor screening module configured to screen key environmental factors related to the effective wave height measured values based on the environmental factor data of the historical time period of the to-be-predicted region through spatial correlation analysis and linear fitting verification methods; the key environmental factors include mean sea level pressure, zonal wind speed, and meridional wind speed; a training set construction module configured to construct a first training data set based on the key environmental factors and corresponding effective wave height measured values; and construct a second training data set based on the key environmental factors and corresponding disastrous wave actual occurrence probabilities; the disastrous wave actual occurrence probability is determined based on the effective wave height measured values; a model training module configured to train a multivariate linear regression model using the first training data set to obtain a trained wave height prediction model; and train a logistic regression model using the second training data set to obtain a trained disastrous wave occurrence probability prediction model; a prediction output module configured to input the key environmental factors of a current time period of a to-be-predicted region into the trained wave height prediction model and the trained disastrous wave occurrence probability prediction model, respectively, to obtain a disastrous wave occurrence probability and an effective wave height prediction value of a prediction time period of the to-be-predicted region.

8. A computer device comprising: A memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that the processor executes the computer program to implement the disastrous wave occurrence probability and wave height joint prediction method of any one of claims 1-6.

9. A computer readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the disastrous wave occurrence probability and wave height joint prediction method of any one of claims 1-6.

10. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to implement the disastrous wave occurrence probability and wave height joint prediction method of any one of claims 1-6.

Images