Ocean buoy monitoring system and method based on Beidou service

Abstract

The invention provides an ocean buoy monitoring system and method based on Beidou service, and relates to the technical field of Beidou systems.The system collects Beidou signal feature data and ocean dynamic environment data of a target ocean buoy in real time; performing analysis to obtain a signal deep feature group and an environment dynamic feature group, performing standardization processing and fusion to form a multi-dimensional fusion feature vector, inputting the multi-dimensional fusion feature vector into a pre-trained wave height prediction model, and outputting an initial wave height estimated value of the current position of the target ocean buoy; collecting equipment historical data and Beidou communication state data of the target ocean buoy, calculating and generating an equipment state attenuation factor, and correcting the initial wave height estimated value by using the equipment state attenuation factor to obtain a real-time wave height estimated value; and sending the real-time wave height estimation value and the corresponding position information to a monitoring platform through a Beidou communication link. The technical problem that in the prior art, ocean buoy wave height monitoring is prone to being interfered by environment and equipment states, and consequently precision is insufficient is solved.

Description

Technical Field

[0001] This invention relates to the field of the BeiDou system, and more particularly to a marine buoy monitoring system and method based on BeiDou services. Background Technology

[0002] Ocean wave height is an important parameter for marine environmental monitoring, and its accurate acquisition is crucial for fields such as marine disaster early warning, shipping safety, marine engineering construction, and marine scientific research.

[0003] However, traditional ocean wave height monitoring relies heavily on dedicated wave measurement equipment, which has problems such as high deployment and maintenance costs, weak resistance to harsh environments, and is easily affected by the coverage of ground communication networks in offshore areas, leading to data transmission interruptions.

[0004] Meanwhile, the BeiDou system possesses global, all-weather, and interference-resistant positioning and communication capabilities, providing new technical support for marine monitoring. However, existing attempts at marine monitoring using the BeiDou system generally suffer from shortcomings such as limited data dimensions and monitoring accuracy being greatly affected by the environment and equipment status, making it difficult to meet the high-precision monitoring needs in complex marine scenarios. Therefore, there is an urgent need to develop a reliable marine buoy monitoring method based on BeiDou services. Summary of the Invention

[0005] This invention addresses the technical problem in existing technologies where ocean buoy wave height monitoring is susceptible to interference from the environment and equipment status, resulting in insufficient accuracy. It provides an ocean buoy monitoring system and method based on BeiDou services.

[0006] The technical solution of the present invention to solve the above-mentioned technical problems is as follows:

[0007] In a first aspect, the present invention provides a marine buoy monitoring system based on BeiDou services, comprising: The data acquisition module is used to collect BeiDou signal characteristic data and marine dynamic environment data of the target ocean buoy in real time. The marine dynamic environment data includes wind speed, wind direction, air pressure, air temperature and sea surface temperature data. The feature analysis module is used to analyze the BeiDou signal feature data to obtain a deep feature group of the signal, and to analyze the marine dynamic environment data to obtain an environmental dynamic feature group. The wave height prediction module is used to standardize and fuse the deep feature group of the signal and the environmental dynamic feature group to form a multi-dimensional fused feature vector, and input the multi-dimensional fused feature vector into the pre-trained wave height prediction model to calculate and output the initial wave height estimate of the current position of the target ocean buoy. The wave height correction module is used to collect the historical equipment data and BeiDou communication status data of the target ocean buoy, combine the BeiDou signal characteristic data and the ocean dynamic environment data, calculate and generate the equipment status attenuation factor, and use the equipment status attenuation factor to correct the initial wave height estimate to obtain the real-time wave height estimate. The monitoring transmission module is used to send the real-time wave height estimate and the corresponding location information to the monitoring platform through the Beidou communication link.

[0008] Secondly, the present invention provides a method for monitoring marine buoys based on BeiDou services, including: Real-time acquisition of BeiDou signal characteristic data and marine dynamic environment data of target ocean buoys, wherein the marine dynamic environment data includes wind speed, wind direction, air pressure, air temperature and sea surface temperature data; Based on the analysis of the BeiDou signal characteristic data, a deep feature group of the signal is obtained, and based on the analysis of the marine dynamic environment data, an environmental dynamic feature group is obtained; The deep feature set of the signal and the environmental dynamic feature set are standardized and fused to form a multi-dimensional fused feature vector. The multi-dimensional fused feature vector is then input into a pre-trained wave height prediction model to calculate and output the initial wave height estimate of the current position of the target ocean buoy. The system collects historical equipment data and BeiDou communication status data of the target ocean buoy, combines the BeiDou signal characteristic data and the ocean dynamic environment data, calculates and generates an equipment status attenuation factor, and uses the equipment status attenuation factor to correct the initial wave height estimate to obtain a real-time wave height estimate. The real-time wave height estimate and the corresponding location information are sent to the monitoring platform via the BeiDou communication link.

[0009] The beneficial effects of this invention are: Compared to existing technologies, this application firstly acquires real-time BeiDou signal characteristic data and marine dynamic environment data of the target ocean buoy, achieving real-time synchronous acquisition of multi-dimensional data on BeiDou signal characteristics and marine dynamic environment, providing comprehensive and time-consistent data support for subsequent accurate wave height estimation. Secondly, it analyzes the BeiDou signal characteristic data to obtain a deep signal feature group and analyzes the marine dynamic environment data to obtain an environmental dynamic feature group, providing high-quality and targeted feature support for subsequent accurate wave height prediction. Thirdly, it standardizes and fuses the deep signal feature group and the environmental dynamic feature group to form a multi-dimensional fused feature vector, which is then input into a pre-trained wave height prediction model to calculate and output the initial wave height estimate of the target ocean buoy's current position. This eliminates the dimensional differences between the deep signal feature group and the environmental dynamic feature group, providing reliable basic data support for subsequent real-time wave height correction. Furthermore, historical equipment data and BeiDou communication status data of the target ocean buoy are collected. Combined with BeiDou signal characteristic data and marine dynamic environment data, an equipment status attenuation factor is calculated and generated. This attenuation factor is then used to correct the initial wave height estimate, resulting in a real-time wave height estimate, thus improving the accuracy and reliability of the real-time wave height estimate. Finally, the real-time wave height estimate and its corresponding location information are transmitted to the monitoring platform via the BeiDou communication link. This enables real-time, all-weather, full-domain monitoring of ocean wave height, closed-loop data management, and rapid response, providing accurate and efficient data support for marine environmental governance, disaster prevention and control, and related industry applications.

[0010] Through the above technical solution, this application acquires BeiDou signal characteristic data and marine dynamic environment data in real time. After data extraction, standardization, and fusion, an initial wave height estimate is obtained through a wave height prediction model. This estimate is then combined with historical equipment data and BeiDou communication status data to generate an equipment status attenuation factor for precise correction. Finally, the real-time wave height estimate and corresponding location information are stably transmitted to the monitoring platform via the BeiDou communication link. This effectively solves the problems of traditional monitoring methods, such as limited data dimensions, high susceptibility to environmental and equipment status interference, and limited transmission in distant waters. It improves the accuracy of wave height monitoring, the stability and integrity of data transmission, and provides reliable and efficient data support for marine disaster early warning, shipping safety assurance, marine engineering construction, and marine scientific research. Attached Figure Description

[0011] Figure 1 This invention provides a schematic diagram of the structure of a marine buoy monitoring system based on BeiDou services. Figure 2 This is a flowchart illustrating a method for monitoring marine buoys based on BeiDou services provided by the present invention.

[0012] In the attached diagram, the components represented by each number are as follows: Data acquisition module 11, feature analysis module 12, wave height prediction module 13, wave height correction module 14, monitoring and transmission module 15. Detailed Implementation

[0013] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0014] In the description of this invention, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0015] In the description of this invention, the term "for example" is used to mean "used as an example, illustration, or description." Any embodiment described as "for example" in this invention is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use the invention. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that the invention can be made without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid obscuring the description of the invention with unnecessary detail. Therefore, the invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed herein.

[0016] Example 1, as Figure 1 As shown, this embodiment of the invention provides a marine buoy monitoring system based on BeiDou services, comprising: The data acquisition module 11 is used to collect BeiDou signal characteristic data and marine dynamic environment data of the target ocean buoy in real time. The marine dynamic environment data includes wind speed, wind direction, air pressure, air temperature and sea surface temperature data.

[0017] Ocean wave height refers to the vertical distance between the crests and troughs of adjacent ocean waves. Obtaining accurate ocean wave height data can provide technical support for scenarios such as marine disaster early warning, marine engineering construction, shipping safety assurance, and marine scientific research.

[0018] Precise monitoring of ocean wave height requires the support of multi-dimensional data collaboration. Traditional monitoring methods often rely on single sensors to collect data, which are easily limited by the complex marine environment and communication coverage. In contrast, the BeiDou Navigation Satellite System has global, all-weather, and interference-resistant service capabilities, and its signal characteristic data can indirectly reflect the dynamic changes in the marine environment. Meanwhile, marine dynamic environmental data such as wind speed, wind direction, air pressure, air temperature, and sea surface temperature are the core driving factors for wave formation and energy changes.

[0019] Therefore, by simultaneously collecting BeiDou signal characteristic data and the aforementioned marine dynamic environment data, a comprehensive, real-time, and time-consistent basic data source can be provided for subsequent accurate wave height estimation, effectively solving the technical problems of traditional monitoring methods having a single data dimension and insufficient adaptability to complex marine environments.

[0020] To address the aforementioned issues, this application collects real-time BeiDou signal characteristic data and marine dynamic environment data of the target ocean buoy, wherein the marine dynamic environment data includes wind speed, wind direction, air pressure, air temperature, and sea surface temperature data.

[0021] Specifically, the data acquisition module 11 is used for: The BeiDou positioning module mounted on the target ocean buoy continuously receives BeiDou satellite signals at a preset sampling frequency and analyzes them to obtain BeiDou signal characteristic data. The BeiDou signal characteristic data includes the original time sequence of received signal strength indication, the time sequence of carrier phase observation, and the time sequence of signal-to-noise ratio. The target ocean buoy's multi-parameter meteorological sensor synchronously collects wind speed, wind direction, air pressure, and temperature data at the preset sampling frequency. The sea surface temperature data is collected at the preset sampling frequency using a water temperature sensor mounted on the target ocean buoy.

[0022] In this embodiment, the BeiDou positioning module mounted on the target ocean buoy first continuously receives BeiDou satellite signals at a preset sampling frequency and analyzes them to obtain BeiDou signal characteristic data. The BeiDou signal characteristic data includes the original received signal strength index (RSSI) time sequence, the carrier phase observation time sequence, and the signal-to-noise ratio (SNR) time sequence. The original RSSI time sequence refers to the raw data of the received signal strength index (RSSI) corresponding to each sample collected in real time during the continuous reception of BeiDou satellite signals by the BeiDou positioning module, arranged continuously in the sampling time order, reflecting the power of the BeiDou signal reaching the buoy. The carrier phase observation time sequence refers to the observed values ​​of the BeiDou signal carrier phase recorded in real time during the reception of BeiDou satellite signals by the BeiDou positioning module, arranged sequentially in the sampling time order, reflecting the dynamic changes in the BeiDou signal carrier phase. The SNR time sequence refers to the ratio of effective signal power to noise power calculated in real time during the analysis of BeiDou satellite signals by the BeiDou positioning module, arranged continuously in the sampling time order, reflecting the ratio of effective signal to noise.

[0023] For example, the BeiDou positioning module carried by the target ocean buoy supports the reception of BeiDou-2 and BeiDou-3 satellite signals, and the preset sampling frequency can be dynamically set according to actual monitoring needs, such as 1Hz, 2Hz, etc.

[0024] For example, if the preset sampling frequency is set to 1Hz, the BeiDou positioning module on the target ocean buoy receives and analyzes the BeiDou satellite signal once per second, and synchronously extracts the corresponding raw data of received signal strength indication, carrier phase observation value, and signal-to-noise ratio (SNR) data from the BeiDou satellite signal. These are then continuously integrated in chronological order to form a continuous time sequence of raw received signal strength indication, carrier phase observation, and SNR. For instance, the signal strength detection unit built into the BeiDou positioning module performs power detection on the received BeiDou satellite signal, outputting the raw data of received signal strength indication for each sample in real time. This data is then spliced ​​together in chronological order to obtain the raw time sequence of received signal strength indication. The phase measurement unit of the BeiDou positioning module observes the carrier phase of the BeiDou satellite signal in real time, recording the carrier phase observation value for each sample. This value is then organized in chronological order to obtain the carrier phase observation time sequence. Finally, the built-in algorithm of the BeiDou positioning module calculates the ratio of effective signal power to noise power in each sampled BeiDou satellite signal, and this ratio data is summarized in chronological order to obtain the SNR time sequence.

[0025] Secondly, wind speed, wind direction, air pressure, and temperature data are synchronously collected at a preset sampling frequency using a multi-parameter meteorological sensor on the target ocean buoy. This multi-parameter meteorological sensor is an integrated device combining wind speed, wind direction, air pressure, and temperature sensors, possessing adaptability to salt spray and humid marine environments. Specifically, the wind speed, wind direction, air pressure, and temperature data synchronously collected at the preset sampling frequency have completely consistent timestamps with the BeiDou signal characteristic data, ensuring the temporal matching of the two types of data and providing a data synchronization foundation for subsequent feature fusion.

[0026] Finally, sea surface temperature data is collected at a preset sampling frequency using a water temperature sensor mounted on the target ocean buoy. The sampling frequency of the water temperature sensor is consistent with that of the BeiDou positioning module and the multi-parameter meteorological sensor, and it must employ a waterproof and sealed design, with the probe directly contacting the seawater to ensure the accuracy of the measurement data. Preferably, the sampling depth of the water temperature sensor can be set to 0.5 meters below the sea surface, which effectively avoids the influence of floating debris and direct sunlight on the measurement results.

[0027] In summary, compared to existing technologies, this application acquires real-time BeiDou signal characteristic data and marine dynamic environment data of the target ocean buoy. The marine dynamic environment data includes wind speed, wind direction, air pressure, air temperature, and sea surface temperature data. This achieves real-time synchronous acquisition of multi-dimensional marine dynamic environment data, providing comprehensive and time-consistent data support for subsequent accurate wave height estimation, and improving the real-time performance and data integrity of marine monitoring.

[0028] The feature analysis module 12 is used to analyze the BeiDou signal feature data to obtain a deep feature group of the signal, and to analyze the marine dynamic environment data to obtain an environmental dynamic feature group.

[0029] The BeiDou signal feature data and marine dynamic environment data collected by data acquisition module 11 may contain noise interference and redundant information. Directly using them for wave height estimation can easily lead to low model prediction accuracy and weak anti-interference ability. Moreover, the original form of the two types of data cannot directly reflect the deep correlation with ocean wave height. Therefore, it is necessary to construct two types of feature groups through targeted analysis and processing to lay the foundation for subsequent multi-dimensional feature fusion and accurate wave height prediction.

[0030] To address the aforementioned issues, this application analyzes the BeiDou signal characteristic data to obtain a deep signal characteristic group and analyzes the marine dynamic environment data to obtain an environmental dynamic characteristic group.

[0031] Specifically, the feature analysis module 12 is used for: The original timing sequence of the received signal strength indication is preprocessed by moving average filtering to obtain a smoothed timing sequence of the received signal strength indication. A short-time Fourier transform is performed on the smoothed received signal strength indication time sequence to obtain a time-frequency matrix; The signal energy within the preset wave characteristic frequency band is extracted from the time-frequency matrix, and the ratio of the total energy within the preset wave characteristic frequency band to the total energy of the entire analysis frequency band is calculated to obtain the normalized wave energy. Calculate the standard deviation and skewness of the smoothed received signal strength indication timing sequence; The normalized fluctuation energy, the standard deviation, and the skewness are combined to form a deep feature set of the signal.

[0032] In this embodiment, the original time series of received signal strength indication (RSI) is first preprocessed with a moving average filter to obtain a smoothed RSI time series. The moving average filter is performed by selecting a time window of a certain length and calculating the arithmetic mean of the original RSI data within the window to eliminate random noise interference.

[0033] For example, if the time window length is set to 5 sampling points, that is, for the nth data in the original time sequence of the received signal strength indicator, its smoothed value is the arithmetic mean of the (n-2), (n-1), n, (n+1), and (n+2)th original data, the edge data can be processed by zero padding.

[0034] Secondly, a short-time Fourier transform is performed on the smoothed received signal strength indication time sequence to obtain the time-frequency matrix. The short-time Fourier transform involves dividing the time sequence signal into multiple time windows, performing a Fourier transform on the signal within each time window to obtain the frequency distribution corresponding to each time window, ultimately forming a two-dimensional matrix of time and frequency, which serves as the time-frequency matrix.

[0035] For example, if the time window length of the short-time Fourier transform is set to 1 second and the frequency resolution is set to 1 Hz, the time-domain features of the smoothed received signal strength indication time sequence can be converted into spatiotemporal joint features through the short-time Fourier transform, which is convenient for extracting wave-related frequency features.

[0036] Next, the signal energy within the preset wave characteristic frequency band is extracted from the time-frequency matrix, and the ratio of the total energy within the preset wave characteristic frequency band to the total energy of the entire analysis frequency band is calculated to obtain the normalized wave energy. The preset wave characteristic frequency band is set based on the common frequency range of ocean waves; for example, the preset wave characteristic frequency band is 0.05-0.5Hz, which covers the frequency range of common waves such as wind waves and swells.

[0037] For example, the signal energy can be obtained by summing the squares of the amplitudes of the corresponding frequency bands in the time-frequency matrix, and then the ratio of the total energy in the preset wave characteristic frequency band to the total energy of the entire analysis frequency band is calculated and normalized to eliminate the influence of the overall signal strength on the feature, making the feature more universal.

[0038] Furthermore, the standard deviation and skewness of the smoothed received signal strength indication time series are calculated. The standard deviation is a statistic that measures the degree of data dispersion and reflects the fluctuation range of the smoothed received signal strength indication time series; the skewness is a statistic that measures the degree of asymmetry in data distribution and reflects the distribution characteristics of the received signal strength indication data.

[0039] For example, the standard deviation can be obtained by calculating the square root of the average of the sum of squares of the differences between each data point and the mean, and the skewness can be obtained by calculating the ratio of the average of the sum of cubes of the differences between each data point and the mean to the cube of the standard deviation.

[0040] Finally, the normalized wave energy, standard deviation, and skewness are combined to form a deep feature set of the signal. Specifically, the deep feature set of the signal includes normalized wave energy, standard deviation, and skewness, which reflect the correlation between BeiDou signals and ocean wave fluctuations from three dimensions: energy characteristics, wave characteristics, and distribution characteristics, respectively, providing reliable feature support for subsequent wave height prediction.

[0041] Furthermore, the feature analysis module 12 is also specifically used for: Subtracting the air temperature data from the sea surface temperature data yields the sea-air temperature difference; Air density is calculated based on temperature and air pressure data and the ideal gas law. Based on the wind speed data and the air density, the wind energy density is calculated using the wind energy density calculation formula. The sea surface temperature difference and the wind energy density are combined to form an environmental dynamic characteristic group.

[0042] In this embodiment, the sea surface temperature data is first subtracted from the air temperature data to obtain the sea-air temperature difference. Specifically, the sea-air temperature difference is calculated because it is one of the important factors affecting wave formation. When the sea surface temperature is higher than the air temperature, water vapor evaporation from the sea surface intensifies, air convection is enhanced, and wind and waves are more likely to form; conversely, wind and wave formation is suppressed. By calculating the sea-air temperature difference, the intensity of heat exchange between the ocean and the atmosphere can be quantified, providing an environmental dynamic basis for wave height prediction.

[0043] Secondly, based on temperature and air pressure data, air density is calculated using the ideal gas law. The ideal gas law is: PV = nRT, which, after transformation, yields the air density formula: ρ = P / (RT), where ρ is the air density, P is the air pressure, and R is the gas constant for air, with R taken as 287 J / (kg). K), T is the thermodynamic temperature, which can be obtained from air temperature data, T = air temperature data + 273.15. Air density reflects the mass distribution of the atmosphere and is a key parameter for calculating wind energy density.

[0044] Next, based on wind speed data and air density, the wind energy density is calculated using the wind energy density calculation formula. Wind energy density refers to the amount of wind energy passing through a unit area per unit time. The formula for calculating wind energy density is: W = 0.5 × ρ × v², where W is the wind energy density, ρ is the air density, and v is the wind speed. Wind energy density reflects the energy intensity of the wind and is an important factor affecting the energy of ocean waves; the higher the wind energy density, the higher the wave height is usually.

[0045] Finally, the sea-temperature temperature difference and wind energy density are combined to form an environmental dynamic characteristic group. Specifically, the environmental dynamic characteristic group includes sea-temperature temperature difference and wind energy density, which reflect the impact of the marine environment on wave height from the two dimensions of heat exchange and energy input, respectively. This complements the deep signal characteristic group and improves the comprehensiveness of wave height prediction.

[0046] In summary, compared to existing technologies, this application obtains a deep signal feature group based on the BeiDou signal feature data analysis and an environmental dynamic feature group based on the marine dynamic environment data analysis. This removes noise interference and redundant information from the BeiDou signal feature data and marine dynamic environment data, extracting feature information directly related to wave height. This results in a deep signal feature group and an environmental dynamic feature group that reflect the wave wave characteristics and environmental driving laws, providing high-quality and targeted feature support for subsequent accurate wave height prediction.

[0047] The wave height prediction module 13 is used to standardize and fuse the deep feature group of the signal and the environmental dynamic feature group to form a multi-dimensional fused feature vector, and input the multi-dimensional fused feature vector into the pre-trained wave height prediction model to calculate and output the initial wave height estimate of the current position of the target ocean buoy.

[0048] The deep feature set of the signal and the dynamic feature set of the environment have different dimensions, and a single feature dimension is difficult to fully characterize the complex nonlinear relationship between wave height and multiple factors. Traditional wave height estimation methods are easily affected by feature interference and have limited prediction accuracy.

[0049] Meanwhile, standardization can eliminate the influence of dimensions, feature fusion can achieve information complementarity between BeiDou signal characteristics and marine dynamic environment characteristics, and the pre-trained wave height prediction model has the ability to accurately capture complex correlations.

[0050] To address the aforementioned issues, this application standardizes and fuses the deep feature set of the signal and the environmental dynamic feature set to form a multidimensional fused feature vector. The multidimensional fused feature vector is then input into a pre-trained wave height prediction model to calculate and output the initial wave height estimate of the current position of the target ocean buoy.

[0051] Specifically, the wave height prediction module 13 is used for: The Z-score normalization method is used to standardize each feature parameter in the deep feature group of the signal and the dynamic feature group of the environment, respectively. The standardized signal deep feature set and all feature parameters in the environmental dynamic feature set are concatenated in a predetermined order to obtain a multi-dimensional fused feature vector. Historical BeiDou signal feature data and historical marine dynamic environment data of similar buoys within a historical time period were collected and preprocessed to form a historical multidimensional fusion feature vector set, which served as the sample feature dataset. The wave height data measured by wave measurement equipment, which corresponds to each historical multidimensional fused feature vector in the sample feature dataset in time and space, is obtained and labeled as sample labels to form a sample label dataset. A wave height prediction model is constructed based on machine learning algorithms; The wave height prediction model is trained under supervision using the sample feature dataset and the sample label dataset until it is verified to converge, thus obtaining the trained wave height prediction model. The multidimensional fused feature vector is input into the trained wave height prediction model to calculate and output the initial wave height estimate of the current position of the target ocean buoy.

[0052] In this embodiment, the Z-score standardization method is first used to standardize the feature parameters in the deep signal feature group and the environmental dynamic feature group, respectively. The Z-score standardization formula is: z = (x - μ) / σ, where z is the standardized feature value, x is the original feature value, μ is the historical sample average of the feature parameter, and σ is the historical sample standard deviation of the feature parameter. Specifically, the standardization process can convert feature parameters with different dimensions and numerical ranges into standardized data with a standard deviation of 1, avoiding model training bias caused by excessive differences in feature parameter values.

[0053] Secondly, all feature parameters in the standardized signal deep feature group and the environmental dynamic feature group are concatenated in a predetermined order to obtain a multi-dimensional fused feature vector. The predetermined order can be dynamically set according to actual needs. For example, it can be set to: normalized fluctuation energy, standard deviation, skewness, sea surface temperature difference, and wind energy density. Concatenating them in the predetermined order will yield a multi-dimensional fused feature vector, with each dimension corresponding to a standardized feature parameter.

[0054] Secondly, historical BeiDou signal characteristic data and historical ocean dynamic environment data of similar buoys within a historical time period are collected. After preprocessing, a historical multidimensional fusion feature vector set is constructed, which serves as the sample feature dataset. Here, "similar buoys" refers to ocean buoys of the same or similar model and equipped with the same or similar equipment as the target ocean buoy. The historical time period can be dynamically set according to actual needs, for example, the past 6 months, 1 year, etc. The sampling frequency of the collected historical BeiDou signal characteristic data and historical ocean dynamic environment data is consistent with the preset sampling frequency in the data acquisition module. The preprocessing process is consistent with the processing process in the feature analysis module, ultimately yielding a historical multidimensional fusion feature vector set as the sample feature dataset, ensuring consistency between sample features and real-time features.

[0055] Furthermore, wave height data measured by wave measurement equipment, which corresponds to each historical multidimensional fused feature vector in the sample feature dataset in terms of time and space, is obtained and labeled as sample labels, thus forming a sample label dataset. The wave measurement equipment can be deployed on similar buoys, accurately measuring actual wave height data. The sampling time of the wave height data is consistent with the historical multidimensional fused feature vector set, ensuring the spatiotemporal matching between the sample feature dataset and the sample label dataset.

[0056] Furthermore, a wave height prediction model is constructed based on machine learning algorithms. For example, considering the complex nonlinear relationship between ocean wave height and multidimensional fused feature vectors, and the advantages of the Gradient Boosting Decision Tree (GBDT) algorithm in terms of strong anti-overfitting ability and accurate capture of feature interactions, this algorithm can be preferentially used to construct the wave height prediction model, which mainly consists of an input feature layer, a weak learner layer, a gradient calculation and residual fitting layer, and an output layer.

[0057] The input feature layer serves as the data entry point for the wave height prediction model. It receives multi-dimensional fused feature vectors, performs format verification on the feature vectors, and removes outliers, such as extreme feature values ​​exceeding ±3 times the standard deviation.

[0058] The weak learner layer is the core computational unit of the wave height prediction model. It consists of multiple independent CART (Classification and Regression Tree) regression trees connected in series. Each tree is a learner with weak performance but basic fitting ability. The structural parameters of a single CART regression tree are set as follows: the tree depth is limited to 5-8 layers, the number of leaf nodes is controlled at 10-20, and the splitting criterion adopts the minimization of mean square error.

[0059] The gradient calculation and residual fitting layer is responsible for calculating the prediction residuals of the preceding ensemble model and guiding the new weak learner to fit the residuals to achieve error correction. Using mean squared error (MSE) as the loss function, the residual between the current ensemble model's predicted value and the true value of the sample wave height is calculated. This residual is then used as the new target label and input into the next CART regression tree for fitting training, allowing the new tree to focus on compensating for the shortcomings of the preceding model.

[0060] The output layer is the result output of the wave height prediction model. It receives the final integrated prediction value and, after a simple format conversion, outputs the initial wave height estimate of the current position of the target ocean buoy. The unit of the initial wave height estimate can be uniformly set to meters, and two decimal places are retained to meet the monitoring accuracy requirements.

[0061] Furthermore, using a sample feature dataset and a sample label dataset, supervised training of the wave height prediction model is performed until validation convergence, resulting in a trained wave height prediction model. For example, the wave height prediction model can be trained using the following technical path: 1. Data preparation: Divide the sample feature dataset and the sample label dataset into a training set and a validation set in a 7:3 ratio. The training set is used for model parameter learning, and the validation set is used for model performance verification; 2. Model training: Use the historical multi-dimensional fused feature vectors in the training set as input features, and the corresponding wave height data as supervision labels. Use the mean squared error as the loss function and continuously adjust the model parameters using the gradient descent algorithm. When the mean squared error of the validation set no longer decreases after 5 consecutive iterations, the model training is considered converged, training is stopped, and the trained wave height prediction model is obtained.

[0062] Finally, the multidimensional fused feature vector is input into the trained wave height prediction model to calculate and output the initial wave height estimate of the target ocean buoy's current position. Specifically, after receiving the multidimensional fused feature vector, the wave height prediction model calculates the corresponding wave height prediction value through its internal algorithm logic, which is the initial wave height estimate of the target ocean buoy's current position.

[0063] For example, by inputting the multidimensional fused feature vector into the trained wave height prediction model, the initial wave height estimate of the target ocean buoy's current position is calculated and output. The initial wave height estimate reflects the preliminary monitoring results of the target ocean buoy's wave height at its current position and in the pre-ocean environment.

[0064] In summary, compared to existing technologies, this application standardizes and fuses the deep signal feature set and the environmental dynamic feature set to form a multi-dimensional fused feature vector. This multi-dimensional fused feature vector is then input into a pre-trained wave height prediction model to calculate and output an initial wave height estimate of the target ocean buoy's current position. This eliminates the dimensional differences between the deep signal feature set and the environmental dynamic feature set, achieving information complementarity and deep fusion between the two types of features. Furthermore, the pre-trained wave height prediction model outputs a reliable initial wave height estimate of the target ocean buoy's current position, providing reliable basic data support for subsequent real-time wave height correction.

[0065] The wave height correction module 14 is used to collect the historical equipment data and BeiDou communication status data of the target ocean buoy, combine the BeiDou signal characteristic data and the ocean dynamic environment data, calculate and generate the equipment status attenuation factor, and use the equipment status attenuation factor to correct the initial wave height estimate to obtain the real-time wave height estimate.

[0066] The initial wave height estimate does not fully consider the equipment aging and attenuation of the target ocean buoy due to historical environmental stress and cumulative working time, as well as the interference of Beidou communication status fluctuations on monitoring data. Traditional correction methods lack comprehensive quantification of multi-dimensional equipment-related influencing factors. However, historical equipment data and Beidou communication status data can accurately reflect the health status of the equipment itself. Combining Beidou signal characteristic data and marine dynamic environment data can comprehensively characterize the attenuation impact mechanism.

[0067] To address the aforementioned issues, this application collects historical equipment data and BeiDou communication status data of the target ocean buoy, combines the BeiDou signal characteristic data and the ocean dynamic environment data, calculates and generates an equipment status attenuation factor, and uses the equipment status attenuation factor to correct the initial wave height estimate to obtain a real-time wave height estimate.

[0068] Specifically, the wave height correction module 14 is used for: The equipment historical data and BeiDou communication status data of the target ocean buoy are collected, wherein the equipment historical data includes historical track data and cumulative working time; Based on the historical track data, the corresponding historical environmental parameters are queried from the pre-set marine environment database, and the historical environmental stress parameters are calculated. The historical environmental stress parameters, the cumulative working time, and the historical flight track data are input into a pre-trained equipment health prediction model to calculate the basic attenuation parameters. Based on the BeiDou signal feature data and the marine dynamic environment data, the fusion feature coupling parameters are calculated; Based on the BeiDou communication status data and the BeiDou signal characteristic data, the BeiDou-specific correction parameters are calculated. Multiply the basic attenuation parameter, the fusion feature coupling parameter, and the BeiDou-specific correction parameter to obtain the device state attenuation factor; The initial wave height estimate is multiplied by the equipment state attenuation factor to obtain the corrected real-time wave height estimate.

[0069] In this embodiment, the historical equipment data and BeiDou communication status data of the target ocean buoy are first collected. The historical equipment data includes historical track data and cumulative operating time. The BeiDou communication status data includes parameters such as link connection status, signal transmission delay, and bit error rate.

[0070] Specifically, historical track data is a sequence of historical position coordinates of the target ocean buoy, recorded and stored by the BeiDou positioning module; cumulative working time is the total operating time of the target ocean buoy from the start of deployment to the present moment, accurate to the hour; BeiDou communication status data can be obtained in real time through the BeiDou communication module, reflecting the transmission quality of the communication link.

[0071] Secondly, based on historical navigation data, corresponding historical environmental parameters are retrieved from a pre-established marine environmental database, and historical environmental stress parameters are calculated. Specifically, the pre-established marine environmental database is a large database containing historical sea surface temperature, seawater salinity, and surface current velocity data for global ocean regions, sourced from authoritative sources such as ocean observation satellites and ocean observation stations. Using the location coordinates and time information in the historical navigation data, corresponding spatiotemporal historical environmental parameters can be retrieved from the marine environmental database, providing data support for the calculation of historical environmental stress parameters.

[0072] Next, historical environmental stress parameters, cumulative working time, and historical flight track data are input into a pre-trained equipment health prediction model to calculate the basic attenuation parameters. For example, considering the temporal correlation and complex coupling between the basic attenuation parameters and factors such as historical environmental stress and working time, a pre-trained equipment health prediction model can be obtained through the following technical path: 1. Model Construction: The equipment health prediction model can be constructed based on a neural network algorithm, mainly consisting of an input layer, an LSTM temporal feature extraction layer, a fully connected feature fusion layer, and an output layer. The input layer receives historical environmental stress parameters, cumulative working time, and historical flight track data; the LSTM temporal feature extraction layer processes the temporal correlation information in the historical flight track data and can be configured with two hidden layers. The first hidden layer contains 64 LSTM units, and the second hidden layer contains 32 LSTM units. A dropout layer with a dropout rate of 0.2 is used to prevent overfitting. A gating mechanism is used to capture the long-term dependency between trajectory changes and equipment wear. The fully connected feature fusion layer performs feature interaction mining on the temporal features output from the LSTM layers through a two-layer fully connected network. The activation function is ReLU to enhance the ability to identify key feature combinations. The two-layer fully connected network can be set to have 64 neurons in the first layer and 32 neurons in the second layer. The output layer adopts a single neuron structure and the activation function is Sigmoid.

[0073] 2. Data Preparation: Collect full lifecycle data for ocean buoys of the same model, including historical environmental stress parameters, cumulative operating time, and historical track data for each buoy. This data will serve as the training dataset. Obtain the actual health score corresponding to the training dataset as a supervision label, forming a supervision label set. The actual health score can be obtained by periodically conducting on-site inspections of the buoys and having a professional engineer provide a health score from 0-100, which will then be normalized to the range of 0.8-1.0. The training dataset and the supervision label set will then be divided into a training set and a validation set in a 7:3 ratio. The training set will be used for model parameter learning, and the validation set will be used for model performance verification. 3. Model Training: The historical environmental stress parameters, cumulative working time, and historical flight track data in the training set are combined as the model input features, and the corresponding normalized health labels are used as supervision labels. The mean squared error (MSE) is selected as the loss function to measure the deviation between the model's predicted values ​​and the true labels. The Adam optimizer is used, with the initial learning rate set to 0.001, which decays to 0.9 times the previous rate every 10 iterations. The batch size is set to 64. After each training round, the mean squared error (MSE) of the model is calculated using the validation set. When the MSE of the validation set no longer decreases for 5 consecutive iterations, the model training is considered to have converged, training is stopped, and the model parameters at this time are saved to obtain the trained equipment health prediction model.

[0074] Furthermore, based on BeiDou signal feature data and marine dynamic environment data, a fusion feature coupling parameter is calculated. Specifically, the fusion feature coupling parameter is used to quantify the degree of matching between BeiDou signal features and marine dynamic environment features. When the matching degree is high, the coupling parameter value is close to 1.0, indicating that the feature fusion effect is good and the initial wave height estimation value is accurate. When the matching degree is low, the coupling parameter value is lower than 1.0, indicating that there is a deviation in feature fusion, which needs to be corrected using this parameter.

[0075] Furthermore, based on BeiDou communication status data and BeiDou signal characteristic data, BeiDou-specific correction parameters are calculated. Specifically, these parameters are used to correct the impact of interference and communication status fluctuations during BeiDou signal transmission on the monitoring results. When the BeiDou communication status is good and the signal quality is high, the correction parameter value is close to 1.0. When the BeiDou communication status is poor and the signal is severely interfered with, the correction parameter value is appropriately lowered to specifically correct the initial wave height estimate.

[0076] Furthermore, the basic attenuation parameter, the fusion feature coupling parameter, and the BeiDou-specific correction parameter are multiplied to obtain the equipment status attenuation factor. The calculation formula for the equipment status attenuation factor is: Equipment Status Attenuation Factor = Basic Attenuation Parameter × Fusion Feature Coupling Parameter × BeiDou-Specific Correction Parameter. The equipment status attenuation factor comprehensively considers three major influencing factors: the equipment's own health status, the feature fusion effect, and the quality of BeiDou communication and signal, thus comprehensively quantifying the degree of attenuation in monitoring accuracy.

[0077] Finally, the initial wave height estimate is multiplied by the equipment condition attenuation factor to obtain the corrected real-time wave height estimate. Specifically, the correction process adjusts the initial wave height estimate using the equipment condition attenuation factor to eliminate errors caused by factors such as equipment attenuation, characteristic mismatch, and signal interference, ultimately yielding a high-precision real-time wave height estimate.

[0078] For example, if the initial wave height estimate is 2.5m and the equipment status attenuation factor is 0.98, then the corrected real-time wave height estimate is 2.5 × 0.98 = 2.45m.

[0079] Furthermore, the wave height correction module 14 is also specifically used for: Historical flight track data is discretely sampled at preset time intervals to obtain basic sampling points; Based on a preset spatial distance threshold and a preset environmental parameter change rate threshold, representative track points are selected from the basic sampling points; Based on the time and location information of the representative track points, corresponding historical environmental parameters are extracted from the marine environment database. The historical environmental parameters include at least historical sea surface temperature, historical seawater salinity, and historical surface ocean current velocity. The design tolerance parameters of the target ocean buoy are obtained, wherein the design tolerance parameters include at least the design operating temperature range, the design salinity tolerance range, and the design current resistance velocity. Based on the historical environmental parameters and the corresponding design tolerance parameters, the relative stress degree of each environmental parameter is calculated. Based on preset weighting coefficients, the relative stress degree of each environmental parameter is weighted and fused to obtain the instantaneous environmental stress index of each representative track point; The instantaneous environmental stress index of all representative track points is accumulated to obtain the historical environmental stress parameters.

[0080] In this embodiment, historical flight track data is first discretely sampled at preset time intervals to obtain basic sampling points. The preset time interval can be set according to the length of the historical data. For example, the preset time interval is 1 hour, that is, a position coordinate is extracted from the historical flight track data every 1 hour as a basic sampling point, ensuring that the sampling points can cover the complete historical flight track while avoiding data redundancy.

[0081] Secondly, representative track points are selected from the basic sampling points based on preset spatial distance thresholds and preset environmental parameter change rate thresholds. Specifically, the preset spatial distance threshold refers to the minimum straight-line distance between two adjacent basic sampling points. For example, if the preset spatial distance threshold is 1 kilometer, and the straight-line distance between two adjacent basic sampling points is less than 1 kilometer, the former basic sampling point is retained as the representative track point, and the latter is deleted. The preset environmental parameter change rate threshold refers to the maximum value of the change in environmental parameters corresponding to two adjacent basic sampling points. For example, if the preset environmental parameter change rate threshold is 5%, and the sea surface temperature change rate corresponding to two adjacent basic sampling points exceeds 5%, the latter basic sampling point is retained as the representative track point.

[0082] The preset spatial distance threshold and the preset environmental parameter change rate threshold can be dynamically set according to actual needs. Selecting representative track points from the basic sampling points can reduce subsequent computation and improve processing efficiency while ensuring coverage of historical environmental features.

[0083] Next, based on the time and location information of representative track points, corresponding historical environmental parameters are extracted from the marine environmental database. These historical environmental parameters include at least historical sea surface temperature, historical seawater salinity, and historical surface ocean current velocity. The data in the marine environmental database is stored using a spatiotemporal index, and the corresponding historical environmental parameters can be quickly retrieved using the longitude and latitude coordinates and timestamps of representative track points.

[0084] Furthermore, the design tolerance parameters of the target ocean buoy are obtained. These parameters include at least the design operating temperature range, the design salinity tolerance range, and the design current resistance velocity. These parameters can be obtained from the target ocean buoy's equipment technical manual.

[0085] For example, the design operating temperature range is 10°C to 50°C, the design salinity range is 30‰ to 40‰, and the design current resistance velocity is 5 m / s, obtained from the equipment technical manual of the target ocean buoy.

[0086] Furthermore, based on historical environmental parameters and corresponding design tolerance parameters, the relative stress degree of each environmental parameter is calculated. The relative stress degree is an indicator that measures the degree of stress exerted by historical environmental parameters on the buoy equipment, ranging from 0 to 1. The closer the value is to 1, the greater the stress degree. Specifically, for historical sea surface temperature, the relative stress degree is calculated as follows: if the historical sea surface temperature is within the design operating temperature range, the relative stress degree is 0; if the historical sea surface temperature exceeds the design operating temperature range, the relative stress degree = (|design operating temperature upper / lower limit - historical sea surface temperature|) / (design operating temperature upper limit - design operating temperature lower limit). Similarly, the calculation method for the relative stress degree of historical seawater salinity and historical surface current velocity is similar to that of historical sea surface temperature, both quantified based on the upper and lower limits of the design tolerance parameters.

[0087] Furthermore, based on preset weighting coefficients, the relative stress levels of each environmental parameter are weighted and fused to obtain the instantaneous environmental stress index for each representative track point. The preset weighting coefficients can be dynamically set according to the degree of influence of each environmental parameter on the buoy equipment, and the sum of the weighting coefficients is 1. For example, the weighting coefficient for historical sea surface temperature can be set to 0.3, the weighting coefficient for historical seawater salinity to 0.2, and the weighting coefficient for historical surface current velocity to 0.5, depending on the actual situation.

[0088] Specifically, the instantaneous environmental stress index is calculated as follows: I = w1 × s1 + w2 × s2 + w3 × s3, where I is the instantaneous environmental stress index, w1, w2, and w3 are the weight coefficients of each environmental parameter, and s1, s2, and s3 are the relative stress degrees of each environmental parameter.

[0089] Finally, the instantaneous environmental stress indices of all representative track points are cumulatively calculated to obtain the historical environmental stress parameters. Specifically, the cumulative calculation uses an arithmetic summation method, adding up the instantaneous environmental stress indices of all representative track points. The sum obtained is the historical environmental stress parameter, which reflects the total degree of environmental stress experienced by the target ocean buoy during its historical operation.

[0090] Furthermore, the wave height correction module 14 is also specifically used for: Based on the BeiDou signal characteristic data, the signal transmission loss rate is obtained by comparing the actual measured value of the received signal strength indication with the theoretical value of the received signal strength indication determined according to the satellite ephemeris and buoy position. Based on the BeiDou signal characteristic data, the sampling frequency attenuation coefficient is obtained by comparing the actual sampling frequency with the standard sampling frequency. The signal characteristic deviation coefficient is obtained by arithmetically averaging the signal transmission loss rate and the sampling frequency attenuation coefficient. Based on the marine dynamic environment data, the sea temperature difference deviation rate is obtained by comparing the real-time sea temperature difference with the historical average sea temperature difference for the same period, and the wind energy density deviation rate is obtained by comparing the real-time wind energy density with the regional average wind energy density. The environmental characteristic deviation coefficient is obtained by arithmetically averaging the sea surface temperature difference deviation rate and the wind energy density deviation rate. The signal feature deviation coefficient and the environmental feature deviation coefficient are weighted, summed, and normalized, and then the difference between the sum and the first value is calculated to obtain the fusion feature coupling parameter.

[0091] In this embodiment, the signal transmission loss rate is first obtained by comparing the actual measured value of the received signal strength indication with the theoretical value of the received signal strength indication determined based on satellite ephemeris and buoy position, based on BeiDou signal characteristic data. The theoretical value of the received signal strength indication is derived by calculating the distance between the BeiDou satellite and the buoy using satellite ephemeris data and combining it with a free-space propagation model.

[0092] Specifically, the formula for calculating the signal transmission loss rate is: Signal transmission loss rate = (Theoretical value of received signal strength indication - Actual measured value of received signal strength indication) / Theoretical value of received signal strength indication × 100%. The signal transmission loss rate reflects the degree of energy loss of the BeiDou signal during transmission. The lower the signal transmission loss rate, the better the signal transmission quality.

[0093] Secondly, based on the characteristic data of BeiDou signals, the sampling frequency attenuation coefficient is obtained by comparing the actual sampling frequency with the standard sampling frequency. The standard sampling frequency is the designed sampling frequency of the BeiDou positioning module of the target marine buoy, for example, 1Hz. The actual sampling frequency is the actual sampling frequency of the BeiDou positioning module in the current monitoring scenario, which can be obtained by counting the number of sampling points per unit time.

[0094] Specifically, the formula for calculating the sampling frequency attenuation coefficient is: Sampling frequency attenuation coefficient = actual sampling frequency / standard sampling frequency. The value range of the sampling frequency attenuation coefficient is 0-1. The closer the sampling frequency attenuation coefficient is to 1, the more stable the sampling frequency is and the better the data integrity is.

[0095] Next, the signal transmission loss rate and the sampling frequency attenuation coefficient are arithmetically averaged to obtain the signal characteristic deviation coefficient. Specifically, the formula for calculating the signal characteristic deviation coefficient is: Signal characteristic deviation coefficient = (1 - Signal transmission loss rate / 100% + Sampling frequency attenuation coefficient) / 2. The signal characteristic deviation coefficient comprehensively reflects the quality of BeiDou signal characteristic data, and its value ranges from 0 to 1. The closer the signal characteristic deviation coefficient is to 1, the higher the quality of the signal characteristic data.

[0096] Furthermore, based on marine dynamic environment data, the sea temperature difference deviation rate is obtained by comparing the real-time sea temperature difference with the historical average sea temperature difference for the same period, and the wind energy density deviation rate is obtained by comparing the real-time wind energy density with the regional average wind energy density. The historical average sea temperature difference for the same period refers to the average sea temperature difference in the same area during the same time period in past preset years; the regional average wind energy density refers to the multi-year average wind energy density of the marine area where the target marine buoy is located.

[0097] Specifically, the formula for calculating the sea surface temperature difference deviation rate is: Sea surface temperature difference deviation rate = (|real-time sea surface temperature difference - historical average sea surface temperature difference|) / historical average sea surface temperature difference × 100%, and the formula for calculating the wind energy density deviation rate is: Wind energy density deviation rate = (|real-time wind energy density - regional average wind energy density|) / regional average wind energy density × 100%.

[0098] Furthermore, the arithmetic mean of the sea surface temperature difference deviation rate and the wind energy density deviation rate is calculated to obtain the environmental characteristic deviation coefficient. Specifically, the formula for calculating the environmental characteristic deviation coefficient is: Environmental characteristic deviation coefficient = (1 - Sea surface temperature difference deviation rate / 100% + 1 - Wind energy density deviation rate / 100%) / 2. The value range of the environmental characteristic deviation coefficient is 0-1. The closer the value is to 1, the higher the consistency between the real-time marine dynamic environment data and the historical and regional average levels, and the better the data reliability.

[0099] Finally, the signal feature deviation coefficient and the environmental feature deviation coefficient are weighted, summed, and normalized, and then the difference between the sum and the first value is calculated to obtain the fused feature coupling parameter. The weighting coefficients of the weighted summation can be dynamically set according to the degree of influence of signal features and environmental features on wave height prediction. For example, the weight of the signal feature deviation coefficient can be set to 0.5, and the weight of the environmental feature deviation coefficient can also be set to 0.5.

[0100] Specifically, the formula for calculating the fusion feature coupling parameter is: Fusion feature coupling parameter = 1 - [(Signal feature deviation coefficient × w1 + Environmental feature deviation coefficient × w2) / (Signal feature deviation coefficient + Environmental feature deviation coefficient)], where w1 and w2 are weighting coefficients. The closer the fusion feature coupling parameter is to 1.0, the higher the coupling degree between signal features and environmental features, and the better the fusion effect.

[0101] Furthermore, the wave height correction module 14 is also specifically used for: Extract the carrier phase observation time sequence and the signal-to-noise ratio time sequence from the BeiDou signal feature data; Spectral analysis is performed on the carrier phase observation time series to calculate the ratio of noise energy to total energy in non-preset wave characteristic frequency bands, thus obtaining the phase disturbance index; The variance of the signal-to-noise ratio time sequence within a preset time window is calculated to obtain the channel stability index; The BeiDou link error rate is obtained by parsing the BeiDou communication status data. Based on the phase disturbance index, the channel stability index, and the BeiDou link bit error rate, the BeiDou-specific correction parameters are obtained through weighted calculation.

[0102] In this embodiment, the carrier phase observation time series and the signal-to-noise ratio (SNR) time series are first extracted from the BeiDou signal feature data. Specifically, the carrier phase observation time series reflects the change of the BeiDou signal carrier phase over time and is the core data for BeiDou positioning and signal analysis; the signal-to-noise ratio (SNR) time series reflects the purity of the BeiDou signal, and a higher SNR indicates better signal quality.

[0103] Secondly, spectral analysis is performed on the carrier phase observation time series to calculate the ratio of noise energy to total energy within the non-preset wave characteristic frequency band, thus obtaining the phase disturbance index. The non-preset wave characteristic frequency band refers to the frequency range outside the preset wave characteristic frequency band. For example, if the preset wave characteristic frequency band is 0.05-0.5Hz, then the non-preset wave characteristic frequency bands are 0-0.05Hz and above 0.5Hz. Spectral analysis can be performed using the Fast Fourier Transform (FFT) method to transform the carrier phase observation time series from the time domain to the frequency domain, obtaining the energy distribution corresponding to each frequency.

[0104] Specifically, the formula for calculating the phase disturbance index is: Phase disturbance index = Noise energy in non-preset wave characteristic frequency band / Total energy × 100%, where the total energy is the total energy over the entire frequency range.

[0105] Next, the variance of the signal-to-noise ratio (SNR) time series within a preset time window is calculated to obtain the channel stability index. The preset time window can be dynamically set according to signal stability requirements. For example, a preset time window of 10 seconds means calculating the variance of the SNR data every 10 seconds. A smaller variance indicates less fluctuation in the SNR within that time window, a more stable BeiDou communication channel, and a higher channel stability index.

[0106] Furthermore, the BeiDou link bit error rate (BER) is obtained by parsing the BeiDou communication status data. The BeiDou link BER refers to the ratio of the number of erroneous data bits transmitted in the BeiDou communication link to the total number of data bits; it is an indicator of the communication link transmission quality. Specifically, the BeiDou communication module continuously monitors bit errors during communication. By parsing the bit error statistics field in the BeiDou communication status data, the BeiDou link BER can be directly obtained. The BeiDou link BER ranges from 0 to 1; the closer the value is to 0, the better the communication quality.

[0107] Finally, based on the phase perturbation index, channel stability index, and BeiDou link bit error rate, BeiDou-specific correction parameters are obtained through weighted calculation. The weighting coefficients can be set according to the degree of influence of each index on the BeiDou signal quality. For example, based on actual conditions, the weight of the phase perturbation index is set to 0.3, the weight of the channel stability index to 0.4, and the weight of the BeiDou link bit error rate to 0.3.

[0108] Specifically, the formula for calculating the BeiDou-specific correction parameter is: BeiDou-specific correction parameter = 1 - (Phase disturbance index × a1 + Channel stability index × a2 + BeiDou link bit error rate × a3), where a1, a2, and a3 are the weights of the phase disturbance index, channel stability index, and BeiDou link bit error rate, respectively. The closer the value of the BeiDou-specific correction parameter is to 1.0, the better the BeiDou signal quality and communication status, and the smaller the correction magnitude for the initial wave height estimate.

[0109] In summary, compared to existing technologies, this application collects historical equipment data and BeiDou communication status data of the target ocean buoy, combines this data with BeiDou signal characteristic data and marine dynamic environment data, calculates and generates an equipment status attenuation factor, and uses this attenuation factor to correct the initial wave height estimate to obtain a real-time wave height estimate. Thus, by integrating historical equipment data and BeiDou communication status data of the target ocean buoy, the monitoring attenuation effects caused by equipment aging, environmental stress, and communication interference are accurately quantified. The equipment status attenuation factor effectively corrects the initial wave height estimate, improving the accuracy and reliability of the real-time wave height estimate.

[0110] The monitoring transmission module 15 is used to send the real-time wave height estimate and the corresponding location information to the monitoring platform through the Beidou communication link.

[0111] The aforementioned module calculates and obtains real-time wave height estimates, which can accurately reflect the true wave height status of the target ocean buoy's current position and the dynamic change trend of the marine environment. Based on this, marine disaster early warning and prediction, marine engineering construction safety management, shipping route planning, and marine environmental scientific research data analysis can be carried out.

[0112] To address the aforementioned issues, this application transmits the real-time wave height estimate and corresponding location information to the monitoring platform via a BeiDou communication link. Specifically, the BeiDou communication link possesses strong anti-interference capabilities, wide coverage, and low transmission latency, overcoming the limitations of terrestrial communication network coverage in offshore areas and ensuring stable and reliable transmission of monitoring data. Upon receiving the data, the monitoring platform uses its built-in data processing module to visualize and structure the real-time wave height data, and automatically issues warnings for abnormal data based on preset wave height safety thresholds. If the wave height exceeds the threshold, an SMS or platform pop-up notification is triggered. This enables real-time, all-weather, full-area monitoring of ocean wave height, closed-loop data management, and rapid response, providing accurate and efficient data support for marine environmental governance, disaster prevention and control, and related industry applications.

[0113] In summary, the embodiments of this application have at least the following technical effects: Compared to existing technologies, this application first collects real-time BeiDou signal characteristic data and marine dynamic environment data of the target ocean buoy. The marine dynamic environment data includes wind speed, wind direction, air pressure, air temperature, and sea surface temperature data. This achieves real-time synchronous acquisition of multi-dimensional marine dynamic environment data, providing comprehensive and time-consistent data support for subsequent accurate wave height estimation, and improving the real-time performance and data integrity of marine monitoring.

[0114] Secondly, this application analyzes the BeiDou signal characteristic data to obtain a deep signal feature group, and analyzes the marine dynamic environment data to obtain an environmental dynamic feature group. In this way, noise interference and redundant information in the BeiDou signal characteristic data and marine dynamic environment data are removed, and feature information directly related to wave height is extracted. This yields a deep signal feature group and an environmental dynamic feature group that reflect the wave wave characteristics and environmental driving laws, providing high-quality and targeted feature support for subsequent accurate wave height prediction.

[0115] Furthermore, this application standardizes and fuses the deep signal feature set and the environmental dynamic feature set to form a multi-dimensional fused feature vector. This multi-dimensional fused feature vector is then input into a pre-trained wave height prediction model to calculate and output an initial wave height estimate of the target ocean buoy's current position. This eliminates the dimensional differences between the deep signal feature set and the environmental dynamic feature set, achieving information complementarity and deep fusion between the two types of features. The pre-trained wave height prediction model outputs a reliable initial wave height estimate of the target ocean buoy's current position, providing reliable basic data support for subsequent real-time wave height correction.

[0116] Furthermore, this application collects historical equipment data and BeiDou communication status data of the target ocean buoy, combines the BeiDou signal characteristic data and the marine dynamic environment data, calculates and generates an equipment status attenuation factor, and uses the equipment status attenuation factor to correct the initial wave height estimate to obtain a real-time wave height estimate. In this way, by integrating the historical equipment data and BeiDou communication status data of the target ocean buoy, the monitoring attenuation impact caused by equipment aging, environmental stress, and communication interference is accurately quantified. The equipment status attenuation factor effectively corrects the initial wave height estimate, improving the accuracy and reliability of the real-time wave height estimate.

[0117] Finally, this application transmits the real-time wave height estimate and corresponding location information to the monitoring platform via the BeiDou communication link. This enables real-time, all-weather, full-domain monitoring of ocean wave height, closed-loop data management, and rapid response, providing accurate and efficient data support for marine environmental governance, disaster prevention and control, and related industry applications.

[0118] Through the above technical solution, this application acquires BeiDou signal characteristic data and marine dynamic environment data in real time. After data extraction, standardization, and fusion, an initial wave height estimate is obtained through a wave height prediction model. This estimate is then combined with historical equipment data and BeiDou communication status data to generate an equipment status attenuation factor for precise correction. Finally, the real-time wave height estimate and corresponding location information are stably transmitted to the monitoring platform via the BeiDou communication link. This effectively solves the problems of traditional monitoring methods, such as limited data dimensions, high susceptibility to environmental and equipment status interference, and limited transmission in distant waters. It improves the accuracy of wave height monitoring, the stability and integrity of data transmission, and provides reliable and efficient data support for marine disaster early warning, shipping safety assurance, marine engineering construction, and marine scientific research.

[0119] Example 2, as Figure 2 As shown, this embodiment of the invention also provides a method for monitoring marine buoys based on BeiDou services, including: Real-time acquisition of BeiDou signal characteristic data and marine dynamic environment data of target ocean buoys, wherein the marine dynamic environment data includes wind speed, wind direction, air pressure, air temperature and sea surface temperature data; Based on the analysis of the BeiDou signal characteristic data, a deep feature group of the signal is obtained, and based on the analysis of the marine dynamic environment data, an environmental dynamic feature group is obtained; The deep feature set of the signal and the environmental dynamic feature set are standardized and fused to form a multi-dimensional fused feature vector. The multi-dimensional fused feature vector is then input into a pre-trained wave height prediction model to calculate and output the initial wave height estimate of the current position of the target ocean buoy. The system collects historical equipment data and BeiDou communication status data of the target ocean buoy, combines the BeiDou signal characteristic data and the ocean dynamic environment data, calculates and generates an equipment status attenuation factor, and uses the equipment status attenuation factor to correct the initial wave height estimate to obtain a real-time wave height estimate. The real-time wave height estimate and the corresponding location information are sent to the monitoring platform via the BeiDou communication link.

[0120] Specifically, the "real-time acquisition of BeiDou signal characteristic data and marine dynamic environment data of target ocean buoys" includes: The BeiDou positioning module mounted on the target ocean buoy continuously receives BeiDou satellite signals at a preset sampling frequency and analyzes them to obtain BeiDou signal characteristic data. The BeiDou signal characteristic data includes the original time sequence of received signal strength indication, the time sequence of carrier phase observation, and the time sequence of signal-to-noise ratio. The target ocean buoy's multi-parameter meteorological sensor synchronously collects wind speed, wind direction, air pressure, and temperature data at the preset sampling frequency. The sea surface temperature data is collected at the preset sampling frequency using a water temperature sensor mounted on the target ocean buoy.

[0121] Specifically, the phrase "obtaining deep feature groups of the signal based on the BeiDou signal feature data" includes: The original timing sequence of the received signal strength indication is preprocessed by moving average filtering to obtain a smoothed timing sequence of the received signal strength indication. A short-time Fourier transform is performed on the smoothed received signal strength indication time sequence to obtain a time-frequency matrix; The signal energy within the preset wave characteristic frequency band is extracted from the time-frequency matrix, and the ratio of the total energy within the preset wave characteristic frequency band to the total energy of the entire analysis frequency band is calculated to obtain the normalized wave energy. Calculate the standard deviation and skewness of the smoothed received signal strength indication timing sequence; The normalized fluctuation energy, the standard deviation, and the skewness are combined to form a deep feature set of the signal.

[0122] Specifically, the phrase "obtaining an environmental dynamic characteristic group based on the marine dynamic environment data analysis" includes: Subtracting the air temperature data from the sea surface temperature data yields the sea-air temperature difference; Air density is calculated based on temperature and air pressure data and the ideal gas law. Based on the wind speed data and the air density, the wind energy density is calculated using the wind energy density calculation formula. The sea surface temperature difference and the wind energy density are combined to form an environmental dynamic characteristic group.

[0123] Specifically, the step of "standardizing and fusing the deep feature set of the signal with the environmental dynamic feature set to form a multi-dimensional fused feature vector, and inputting the multi-dimensional fused feature vector into a pre-trained wave height prediction model to calculate and output the initial wave height estimate of the current position of the target ocean buoy" includes: The Z-score normalization method is used to standardize each feature parameter in the deep feature group of the signal and the dynamic feature group of the environment, respectively. The standardized signal deep feature set and all feature parameters in the environmental dynamic feature set are concatenated in a predetermined order to obtain a multi-dimensional fused feature vector. Historical BeiDou signal feature data and historical marine dynamic environment data of similar buoys within a historical time period were collected and preprocessed to form a historical multidimensional fusion feature vector set, which served as the sample feature dataset. The wave height data measured by wave measurement equipment, which corresponds to each historical multidimensional fused feature vector in the sample feature dataset in time and space, is obtained and labeled as sample labels to form a sample label dataset. A wave height prediction model is constructed based on machine learning algorithms; The wave height prediction model is trained under supervision using the sample feature dataset and the sample label dataset until it is verified to converge, thus obtaining the trained wave height prediction model. The multidimensional fused feature vector is input into the trained wave height prediction model to calculate and output the initial wave height estimate of the current position of the target ocean buoy.

[0124] Specifically, the step of "collecting historical equipment data and BeiDou communication status data of the target ocean buoy, combining the BeiDou signal characteristic data and the ocean dynamic environment data, calculating and generating an equipment status attenuation factor, and using the equipment status attenuation factor to correct the initial wave height estimate to obtain a real-time wave height estimate" includes: The equipment historical data and BeiDou communication status data of the target ocean buoy are collected, wherein the equipment historical data includes historical track data and cumulative working time; Based on the historical track data, the corresponding historical environmental parameters are queried from the pre-set marine environment database, and the historical environmental stress parameters are calculated. The historical environmental stress parameters, the cumulative working time, and the historical flight track data are input into a pre-trained equipment health prediction model to calculate the basic attenuation parameters. Based on the BeiDou signal feature data and the marine dynamic environment data, the fusion feature coupling parameters are calculated; Based on the BeiDou communication status data and the BeiDou signal characteristic data, the BeiDou-specific correction parameters are calculated. Multiply the basic attenuation parameter, the fusion feature coupling parameter, and the BeiDou-specific correction parameter to obtain the device state attenuation factor; The initial wave height estimate is multiplied by the equipment state attenuation factor to obtain the corrected real-time wave height estimate.

[0125] Specifically, the step of "based on the historical track data, querying the corresponding historical environmental parameters from a pre-set marine environment database, and calculating the historical environmental stress parameters" includes: Historical flight track data is discretely sampled at preset time intervals to obtain basic sampling points; Based on a preset spatial distance threshold and a preset environmental parameter change rate threshold, representative track points are selected from the basic sampling points; Based on the time and location information of the representative track points, corresponding historical environmental parameters are extracted from the marine environment database. The historical environmental parameters include at least historical sea surface temperature, historical seawater salinity, and historical surface ocean current velocity. The design tolerance parameters of the target ocean buoy are obtained, wherein the design tolerance parameters include at least the design operating temperature range, the design salinity tolerance range, and the design current resistance velocity. Based on the historical environmental parameters and the corresponding design tolerance parameters, the relative stress degree of each environmental parameter is calculated. Based on preset weighting coefficients, the relative stress degree of each environmental parameter is weighted and fused to obtain the instantaneous environmental stress index of each representative track point; The instantaneous environmental stress index of all representative track points is accumulated to obtain the historical environmental stress parameters.

[0126] Specifically, the phrase "calculating the fusion feature coupling parameters based on the BeiDou signal feature data and the marine dynamic environment data" includes: Based on the BeiDou signal characteristic data, the signal transmission loss rate is obtained by comparing the actual measured value of the received signal strength indication with the theoretical value of the received signal strength indication determined according to the satellite ephemeris and buoy position. Based on the BeiDou signal characteristic data, the sampling frequency attenuation coefficient is obtained by comparing the actual sampling frequency with the standard sampling frequency. The signal characteristic deviation coefficient is obtained by arithmetically averaging the signal transmission loss rate and the sampling frequency attenuation coefficient. Based on the marine dynamic environment data, the sea temperature difference deviation rate is obtained by comparing the real-time sea temperature difference with the historical average sea temperature difference for the same period, and the wind energy density deviation rate is obtained by comparing the real-time wind energy density with the regional average wind energy density. The environmental characteristic deviation coefficient is obtained by arithmetically averaging the sea surface temperature difference deviation rate and the wind energy density deviation rate. The signal feature deviation coefficient and the environmental feature deviation coefficient are weighted, summed, and normalized, and then the difference between the sum and the first value is calculated to obtain the fusion feature coupling parameter.

[0127] Specifically, the phrase "calculating BeiDou-specific correction parameters based on the BeiDou communication status data and the BeiDou signal characteristic data" includes: Extract the carrier phase observation time sequence and the signal-to-noise ratio time sequence from the BeiDou signal feature data; Spectral analysis is performed on the carrier phase observation time series to calculate the ratio of noise energy to total energy in non-preset wave characteristic frequency bands, thus obtaining the phase disturbance index; The variance of the signal-to-noise ratio time sequence within a preset time window is calculated to obtain the channel stability index; The BeiDou link error rate is obtained by parsing the BeiDou communication status data. Based on the phase disturbance index, the channel stability index, and the BeiDou link bit error rate, the BeiDou-specific correction parameters are obtained through weighted calculation.

[0128] In summary, the embodiments of this application have at least the following technical effects: Compared to existing technologies, this application firstly acquires real-time BeiDou signal characteristic data and marine dynamic environment data of the target ocean buoy, achieving real-time synchronous acquisition of multi-dimensional data on BeiDou signal characteristics and marine dynamic environment, providing comprehensive and time-consistent data support for subsequent accurate wave height estimation. Secondly, it analyzes the BeiDou signal characteristic data to obtain a deep signal feature group and analyzes the marine dynamic environment data to obtain an environmental dynamic feature group, providing high-quality and targeted feature support for subsequent accurate wave height prediction. Thirdly, it standardizes and fuses the deep signal feature group and the environmental dynamic feature group to form a multi-dimensional fused feature vector, which is then input into a pre-trained wave height prediction model to calculate and output the initial wave height estimate of the target ocean buoy's current position. This eliminates the dimensional differences between the deep signal feature group and the environmental dynamic feature group, providing reliable basic data support for subsequent real-time wave height correction. Furthermore, historical equipment data and BeiDou communication status data of the target ocean buoy are collected. Combined with BeiDou signal characteristic data and marine dynamic environment data, an equipment status attenuation factor is calculated and generated. This attenuation factor is then used to correct the initial wave height estimate, resulting in a real-time wave height estimate, thus improving the accuracy and reliability of the real-time wave height estimate. Finally, the real-time wave height estimate and its corresponding location information are transmitted to the monitoring platform via the BeiDou communication link. This enables real-time, all-weather, full-domain monitoring of ocean wave height, closed-loop data management, and rapid response, providing precise and efficient data support for marine environmental governance, disaster prevention and control, and related industry applications. This effectively solves the problems of traditional monitoring methods, such as single data dimensions, high accuracy affected by environmental and equipment status interference, and limited transmission in distant waters. It improves the accuracy of wave height monitoring, the stability and integrity of data transmission, and provides reliable and efficient data support for marine disaster early warning, shipping safety assurance, marine engineering construction, and marine scientific research.

[0129] It should be noted that the descriptions of each embodiment in the above embodiments have different focuses. For parts that are not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0130] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0131] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0132] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0133] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0134] Although preferred embodiments of the invention have been described, those skilled in the art, once they have learned the basic inventive concept, can make other changes and modifications to these embodiments.

[0135] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of this invention and its equivalents, this invention also intends to include these modifications and variations.

Claims

1. A marine buoy monitoring system based on BeiDou services, characterized in that, The system includes: The data acquisition module is used to collect BeiDou signal characteristic data and marine dynamic environment data of the target ocean buoy in real time. The marine dynamic environment data includes wind speed, wind direction, air pressure, air temperature and sea surface temperature data. The feature analysis module is used to analyze the BeiDou signal feature data to obtain a deep feature group of the signal, and to analyze the marine dynamic environment data to obtain an environmental dynamic feature group. The wave height prediction module is used to standardize and fuse the deep feature group of the signal and the environmental dynamic feature group to form a multi-dimensional fused feature vector, and input the multi-dimensional fused feature vector into the pre-trained wave height prediction model to calculate and output the initial wave height estimate of the current position of the target ocean buoy. The wave height correction module is used to collect the historical equipment data and BeiDou communication status data of the target ocean buoy, combine the BeiDou signal characteristic data and the ocean dynamic environment data, calculate and generate the equipment status attenuation factor, and use the equipment status attenuation factor to correct the initial wave height estimate to obtain the real-time wave height estimate. The monitoring transmission module is used to send the real-time wave height estimate and the corresponding location information to the monitoring platform through the Beidou communication link.

2. The marine buoy monitoring system based on BeiDou services according to claim 1, characterized in that, The execution steps of the data acquisition module include: The BeiDou positioning module mounted on the target ocean buoy continuously receives BeiDou satellite signals at a preset sampling frequency and analyzes them to obtain BeiDou signal characteristic data. The BeiDou signal characteristic data includes the original time sequence of received signal strength indication, the time sequence of carrier phase observation, and the time sequence of signal-to-noise ratio. The target ocean buoy's multi-parameter meteorological sensor synchronously collects wind speed, wind direction, air pressure, and temperature data at the preset sampling frequency. The sea surface temperature data is collected at the preset sampling frequency using a water temperature sensor mounted on the target ocean buoy.

3. The marine buoy monitoring system based on BeiDou services according to claim 1, characterized in that, The execution steps of the feature analysis module include: The original timing sequence of the received signal strength indication is preprocessed by moving average filtering to obtain a smoothed timing sequence of the received signal strength indication. A short-time Fourier transform is performed on the smoothed received signal strength indication time sequence to obtain a time-frequency matrix; The signal energy within the preset wave characteristic frequency band is extracted from the time-frequency matrix, and the ratio of the total energy within the preset wave characteristic frequency band to the total energy of the entire analysis frequency band is calculated to obtain the normalized wave energy. Calculate the standard deviation and skewness of the smoothed received signal strength indication timing sequence; The normalized fluctuation energy, the standard deviation, and the skewness are combined to form a deep feature set of the signal.

4. The marine buoy monitoring system based on BeiDou services according to claim 1, characterized in that, The execution steps of the feature analysis module also include: Subtracting the air temperature data from the sea surface temperature data yields the sea-air temperature difference; Air density is calculated based on temperature and air pressure data and the ideal gas law. Based on the wind speed data and the air density, the wind energy density is calculated using the wind energy density calculation formula. The sea surface temperature difference and the wind energy density are combined to form an environmental dynamic characteristic group.

5. The marine buoy monitoring system based on BeiDou services according to claim 1, characterized in that, The execution steps of the wave height prediction module include: The Z-score normalization method is used to standardize each feature parameter in the deep feature group of the signal and the dynamic feature group of the environment, respectively. The standardized signal deep feature set and all feature parameters in the environmental dynamic feature set are concatenated in a predetermined order to obtain a multi-dimensional fused feature vector. Historical BeiDou signal feature data and historical marine dynamic environment data of similar buoys within a historical time period were collected and preprocessed to form a historical multidimensional fusion feature vector set, which served as the sample feature dataset. The wave height data measured by wave measurement equipment, which corresponds to each historical multidimensional fused feature vector in the sample feature dataset in time and space, is obtained and labeled as sample labels to form a sample label dataset. A wave height prediction model is constructed based on machine learning algorithms; The wave height prediction model is trained under supervision using the sample feature dataset and the sample label dataset until it is verified to converge, thus obtaining the trained wave height prediction model. The multidimensional fused feature vector is input into the trained wave height prediction model to calculate and output the initial wave height estimate of the current position of the target ocean buoy.

6. The marine buoy monitoring system based on BeiDou services according to claim 1, characterized in that, The execution steps of the wave height correction module include: The equipment historical data and BeiDou communication status data of the target ocean buoy are collected, wherein the equipment historical data includes historical track data and cumulative working time; Based on the historical track data, the corresponding historical environmental parameters are queried from the pre-set marine environment database, and the historical environmental stress parameters are calculated. The historical environmental stress parameters, the cumulative working time, and the historical flight track data are input into a pre-trained equipment health prediction model to calculate the basic attenuation parameters. Based on the BeiDou signal feature data and the marine dynamic environment data, the fusion feature coupling parameters are calculated; Based on the BeiDou communication status data and the BeiDou signal characteristic data, the BeiDou-specific correction parameters are calculated. Multiply the basic attenuation parameter, the fusion feature coupling parameter, and the BeiDou-specific correction parameter to obtain the device state attenuation factor; The initial wave height estimate is multiplied by the equipment state attenuation factor to obtain the corrected real-time wave height estimate.

7. The marine buoy monitoring system based on BeiDou services according to claim 6, characterized in that, The execution steps of the wave height correction module also include: Historical flight track data is discretely sampled at preset time intervals to obtain basic sampling points; Based on a preset spatial distance threshold and a preset environmental parameter change rate threshold, representative track points are selected from the basic sampling points; Based on the time and location information of the representative track points, corresponding historical environmental parameters are extracted from the marine environment database. The historical environmental parameters include at least historical sea surface temperature, historical seawater salinity, and historical surface ocean current velocity. The design tolerance parameters of the target ocean buoy are obtained, wherein the design tolerance parameters include at least the design operating temperature range, the design salinity tolerance range, and the design current resistance velocity. Based on the historical environmental parameters and the corresponding design tolerance parameters, the relative stress degree of each environmental parameter is calculated. Based on preset weighting coefficients, the relative stress degree of each environmental parameter is weighted and fused to obtain the instantaneous environmental stress index of each representative track point; The instantaneous environmental stress index of all representative track points is accumulated to obtain the historical environmental stress parameters.

8. The marine buoy monitoring system based on BeiDou services according to claim 6, characterized in that, The execution steps of the wave height correction module also include: Based on the BeiDou signal characteristic data, the signal transmission loss rate is obtained by comparing the actual measured value of the received signal strength indication with the theoretical value of the received signal strength indication determined according to the satellite ephemeris and buoy position. Based on the BeiDou signal characteristic data, the sampling frequency attenuation coefficient is obtained by comparing the actual sampling frequency with the standard sampling frequency. The signal characteristic deviation coefficient is obtained by arithmetically averaging the signal transmission loss rate and the sampling frequency attenuation coefficient. Based on the marine dynamic environment data, the sea temperature difference deviation rate is obtained by comparing the real-time sea temperature difference with the historical average sea temperature difference for the same period, and the wind energy density deviation rate is obtained by comparing the real-time wind energy density with the regional average wind energy density. The environmental characteristic deviation coefficient is obtained by arithmetically averaging the sea surface temperature difference deviation rate and the wind energy density deviation rate. The signal feature deviation coefficient and the environmental feature deviation coefficient are weighted, summed, and normalized, and then the difference between the sum and the first value is calculated to obtain the fusion feature coupling parameter.

9. The marine buoy monitoring system based on BeiDou services according to claim 6, characterized in that, The execution steps of the wave height correction module also include: Extract the carrier phase observation time sequence and the signal-to-noise ratio time sequence from the BeiDou signal feature data; Spectral analysis is performed on the carrier phase observation time series to calculate the ratio of noise energy to total energy in non-preset wave characteristic frequency bands, thus obtaining the phase disturbance index; The variance of the signal-to-noise ratio time sequence within a preset time window is calculated to obtain the channel stability index; The BeiDou link error rate is obtained by parsing the BeiDou communication status data. Based on the phase disturbance index, the channel stability index, and the BeiDou link bit error rate, the BeiDou-specific correction parameters are obtained through weighted calculation.

10. A method for monitoring marine buoys based on BeiDou services, characterized in that, include: Real-time acquisition of BeiDou signal characteristic data and marine dynamic environment data of target ocean buoys, wherein the marine dynamic environment data includes wind speed, wind direction, air pressure, air temperature and sea surface temperature data; Based on the analysis of the BeiDou signal characteristic data, a deep feature group of the signal is obtained, and based on the analysis of the marine dynamic environment data, an environmental dynamic feature group is obtained; The deep feature set of the signal and the environmental dynamic feature set are standardized and fused to form a multi-dimensional fused feature vector. The multi-dimensional fused feature vector is then input into a pre-trained wave height prediction model to calculate and output the initial wave height estimate of the current position of the target ocean buoy. The system collects historical equipment data and BeiDou communication status data of the target ocean buoy, combines the BeiDou signal characteristic data and the ocean dynamic environment data, calculates and generates an equipment status attenuation factor, and uses the equipment status attenuation factor to correct the initial wave height estimate to obtain a real-time wave height estimate. The real-time wave height estimate and the corresponding location information are sent to the monitoring platform via the BeiDou communication link.

Images