Track prediction method, system and equipment based on surveying vessel, medium and product

Abstract

The invention provides a track prediction method, system and equipment based on a surveying vessel, a medium and a product, and relates to the technical field of ships, and the method comprises the steps: obtaining the real-time state information of a navigation environment generated by a target vessel in a current time period based on a preset data sampling window; acquiring target measurement and control key data corresponding to the target ship according to the real-time state information of the navigation environment; the target measurement and control key data is input to a corresponding track prediction model, a track prediction result of the target ship output by the track prediction model is obtained, and the track prediction model is obtained through sample measurement and control key data and ship track position sample information corresponding to the sample measurement and control key data. And training a machine learning model, wherein the data types of the sample measurement and control key data and the target measurement and control key data are the same. According to the invention, a more accurate track prediction result can be obtained.

Description

Technical Field

[0001] The present invention relates to the field of ship technology, and in particular to a track prediction method, system, equipment, medium and product based on a survey ship. Background Art

[0002] Track prediction is a core technology in the field of marine navigation safety. This technology not only improves the safety and efficiency of ship navigation, but also provides critical data support for rapid response in emergency situations.

[0003] Currently, the field of trajectory prediction is primarily divided into two categories: statistically based and artificial intelligence (AI)-based methods. Statistical methods, such as Kalman filters, particle filters, fuzzy logic, and hidden Markov models, construct or assume a target motion model and leverage prior information and observational data to estimate and predict a ship's future position. AI-based methods, without the need for precise mathematical models, can directly learn and predict a ship's trajectory from large amounts of data.

[0004] Existing track prediction methods still have some drawbacks. For example, statistical methods rely heavily on the accuracy of the target's motion model and prior information, which are often difficult to fully obtain in real marine environments or contain errors, resulting in unstable prediction results. Neural network-based methods, while possessing powerful learning capabilities, are prone to falling into local optimal solutions rather than global optimal solutions during training. Therefore, there is an urgent need for a track prediction method, system, equipment, medium, and product based on survey vessels to address these issues. Summary of the Invention

[0005] In response to the problems existing in the prior art, the present invention provides a track prediction method, system, equipment, medium and product based on a survey ship.

[0006] The present invention provides a track prediction method based on a survey ship, comprising: Based on the preset data sampling window, obtain the real-time status information of the navigation environment generated by the target ship in the current time period; Acquiring target measurement and control key data corresponding to the target ship based on the real-time status information of the navigation environment; The target measurement and control key data is input into the corresponding track prediction model to obtain the track prediction result of the target ship output by the track prediction model, wherein the track prediction model is obtained by training a machine learning model through sample measurement and control key data and the ship track position sample information corresponding to the sample measurement and control key data, and the data types of the sample measurement and control key data and the target measurement and control key data are the same.

[0007] According to a track prediction method based on a survey ship provided by the present invention, the track prediction model is trained by the following steps: Obtain navigation environment sample status information; Based on the preset track prediction conditions, corresponding sample measurement and control key data is obtained according to the navigation environment sample state information, wherein the preset track prediction conditions are determined based on the data type of the measurement and control key data; Determining the track coordinate information corresponding to the ship in a historical time period based on the sample measurement and control key data, so as to obtain the ship track position sample information through the track coordinate information; A training sample set is constructed based on the sample measurement and control key data and the ship track position sample information, and a support vector machine is trained using the training sample set. After determining that the training results meet the preset training conditions, the track prediction model is obtained.

[0008] According to a track prediction method based on a survey vessel provided by the present invention, after inputting the target measurement and control key data into a corresponding track prediction model and obtaining the track prediction result of the target ship output by the track prediction model, the method further includes: Receive first input information, wherein the first input information includes an operation instruction to adjust a data type of key measurement and control data corresponding to a current preset track prediction condition; In response to the first input information, a target track prediction condition is obtained, and based on the data type of the key measurement and control data in the target track prediction condition, the sample key measurement and control data in the training sample set is adjusted to obtain an adjusted training sample set; wherein the target track prediction condition is the preset track prediction condition after the data type is adjusted; Adjusting the model parameters of the track prediction model using the adjusted training sample set to obtain an adjusted track prediction model; The step of obtaining target measurement and control key data corresponding to the target ship according to the real-time state information of the navigation environment includes: Based on the target track prediction conditions and the real-time status information of the navigation environment, key target measurement and control data with adjusted data types are obtained.

[0009] According to a survey ship track prediction method provided by the present invention, before acquiring the real-time state information of the navigation environment of the target ship generated in the current time period based on a preset data sampling window, the method includes: receiving second input information, wherein the second input information includes instruction information for adjusting the preset data sampling window; In response to the second input information, adjusting the window size of the current preset data sampling window to obtain an adjusted preset data sampling window; The step of obtaining the real-time state information of the navigation environment of the target ship generated in the current time period based on the preset data sampling window includes: Based on the adjusted preset data sampling window, the real-time status information of the navigation environment generated by the target ship in the current time period is obtained.

[0010] According to a track prediction method based on a survey ship provided by the present invention, the target measurement and control key data includes at least one of the speed data, heading information, position information and ocean current speed of the target ship.

[0011] According to a ship track prediction method provided by the present invention, obtaining target measurement and control key data corresponding to the target ship based on the real-time status information of the navigation environment includes: performing data preprocessing on the real-time state information of the navigation environment to obtain the real-time state information of the navigation environment after data preprocessing; Based on the current preset track prediction conditions and the real-time status information of the navigation environment after the data preprocessing, the target measurement and control key data corresponding to the target ship are obtained.

[0012] The present invention also provides a track prediction system based on a survey ship, comprising: The data sampling module is used to obtain the real-time status information of the navigation environment of the target ship generated in the current time period based on the preset data sampling window; A data processing module is used to obtain key target measurement and control data corresponding to the target ship based on the real-time status information of the navigation environment; A track prediction module is used to input the target measurement and control key data into the corresponding track prediction model to obtain the track prediction result of the target ship output by the track prediction model, wherein the track prediction model is obtained by training a machine learning model through sample measurement and control key data and the ship track position sample information corresponding to the sample measurement and control key data, and the data types of the sample measurement and control key data and the target measurement and control key data are the same.

[0013] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the method for predicting a track based on a survey vessel as described above is implemented.

[0014] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for predicting a track based on a survey vessel as described above is implemented.

[0015] The present invention also provides a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the computer program implements any of the above-mentioned track prediction methods based on a survey vessel.

[0016] The track prediction method, system, equipment, medium and product based on the survey ship provided by the present invention determine the target measurement and control key data from the real-time status information of the navigation environment generated by the target ship in the current time period, and then use the track prediction model obtained by training with the corresponding sample measurement and control key data and the ship track position sample information to predict the ship track, thereby obtaining a more accurate track prediction result. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0018] Figure 1 A schematic flow chart of a track prediction method based on a survey vessel provided by the present invention; Figure 2 A schematic structural diagram of a track prediction system based on a survey vessel provided by the present invention; Figure 3 This is a schematic structural diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0019] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0020] Track prediction is a technology that uses historical data and information such as a vessel's current position, speed, and heading to predict its movement and trajectory over future time periods. Track prediction can help prevent collisions, groundings, and other accidents. Accurate track prediction is crucial for ensuring safe and efficient navigation, particularly in complex ocean environments. It allows for early detection of potential hazards, helping ships avoid dangerous areas and ensuring safe navigation. Track prediction also helps ships develop optimal routes and sailing plans, taking into account factors such as sea conditions, currents, and wind speeds, to save fuel and improve efficiency. Track prediction also helps plan ship arrival times, reducing waiting times and improving port efficiency. In maritime rescue and emergency response scenarios, when a vessel is in distress or encounters other emergencies, track prediction can help authorities better determine search and rescue scopes and action plans, thereby improving rescue efficiency.

[0021] Existing methods for ship track prediction primarily include statistical and neural network-based approaches. Statistical methods, such as Kalman filtering, particle filtering, fuzzy logic, and hidden Markov models, rely on target motion models and prior information, which are often difficult to obtain or inaccurate in real-world situations.

[0022] Existing neural network-based track prediction methods, such as back propagation (BP) neural networks, can learn and store a large number of input-output mapping relationships without requiring prior knowledge of the mathematical expressions of these mapping relationships. However, during training, they can become stuck in local optimal solutions rather than global optimal solutions. Radial basis function (RBF) neural networks, for example, can leverage the inherent characteristics of the data for prediction without the need to establish a target motion model. However, their simple structure prevents them from fully utilizing the spatiotemporal information in track data. For example, long short-term memory (LSTM) networks for ship track prediction only consider the time series properties of the track and lack effective modeling of the track motion pattern, leaving significant room for improvement in their performance.

[0023] This invention uses a track prediction algorithm based on the measured data characteristic values ​​of a ship's measurement and control equipment. It is a method for predicting a ship's future trajectory using historical navigation data. This method primarily relies on historical measurement and control data from the survey vessel, such as the ship's position, speed, heading, current speed and direction, and other data. By extracting the characteristic information between this historical measurement and control data and the ship's historical track position during navigation, a track prediction model is established. This allows for accurate prediction of future ship track data, providing a basis for subsequent decision-making by ship-related personnel.

[0024] Figure 1 The flow chart of the track prediction method based on the survey ship provided by the present invention is as follows: Figure 1 As shown, the present invention provides a track prediction method based on a survey ship, comprising: Step 101: Based on a preset data sampling window, obtain the real-time status information of the navigation environment of the target ship generated in the current time period.

[0025] In this invention, the preset data sampling window is a predefined sampling window for collecting and analyzing information about the target vessel's navigational environment. The size of this sampling window can be flexibly adjusted based on actual needs, for example, from a few seconds to a few minutes, or even longer (e.g., hourly weather conditions). Regarding data sampling accuracy, the size of the sampling window also reflects the sampling frequency and accuracy. For example, when measuring ship speed, the sampling window can be set to ±0.5 knots, indicating that the speed measurement accuracy must be maintained within ±0.5 knots.

[0026] In the present invention, the preset data sampling window can be set based on the nature of the navigation mission, the environmental characteristics of the navigation area, the real-time requirements for data processing, and the limitations of storage and transmission capabilities. By setting such a data sampling window, the temporal continuity and consistency of the collected data can be ensured, facilitating subsequent data processing and analysis.

[0027] Furthermore, during navigation, relevant sensors installed on the ship will collect measurement and control data generated during navigation. At the same time, specific sensors will also collect relevant information about the environment in which the ship is located during navigation. In the present invention, real-time navigation environment status information refers to status data of various environmental factors closely related to the navigation of the target ship, including but not limited to the following aspects: Marine meteorological information: such as wind speed, wind direction, temperature, air pressure, humidity, etc. This information has an important impact on the navigation stability and safety of ships.

[0028] Water current information: including the speed, direction and tidal information of ocean currents, which directly affect the navigation speed and heading control of the ship.

[0029] Visibility information: Especially in foggy weather or at night, visibility is crucial to the navigation safety of ships.

[0030] Channel and obstacle information: including the location, width, water depth of the channel and the obstacles on both sides of the channel. This information helps ships avoid potential dangerous areas.

[0031] Ship measurement and control information: information such as the ship's position, speed, and heading.

[0032] Step 102: Acquire target measurement and control key data corresponding to the target ship based on the real-time status information of the navigation environment.

[0033] In the present invention, real-time status information of the navigation environment collected by sensors on various target ships is received and analyzed. This information includes but is not limited to marine meteorological data (such as wind speed and wind direction), water flow data (such as current speed and direction), channel and obstacle information, and ship operation status information.

[0034] Based on the above embodiment, the target measurement and control key data includes at least one of the speed data, heading information, position information and ocean current speed of the target ship.

[0035] In the present invention, the real-time status information of the navigation environment is analyzed, and key data important for predicting the subsequent track of the target ship is extracted from it. The key data includes at least one of the following aspects: Ship's speed information: Speed ​​information is an important indicator for assessing the ship's navigation status. It can be calculated through the ship's speed sensor or by combining satellite positioning data and navigation time. Speed ​​information reflects the ship's current navigation speed and is crucial for determining whether the ship is sailing according to the predetermined plan.

[0036] Vessel's heading information: Heading information refers to the vessel's current direction of travel. This is usually calculated using the vessel's gyroscope, magnetic compass, or satellite positioning data.

[0037] Current velocity: Current velocity is a crucial parameter in the ocean environment, directly affecting a ship's speed and course. Current velocity information can be obtained using a current meter or by combining it with other oceanographic observation data.

[0038] Position information: Position information is fundamental data in ship measurement and control. Using satellite positioning data or other positioning technologies, a ship's latitude and longitude can be obtained in real time. This information is used not only to determine the ship's current position but also to calculate parameters such as the ship's trajectory and speed.

[0039] On the basis of the above embodiment, the step of obtaining target measurement and control key data corresponding to the target ship according to the real-time status information of the navigation environment includes: performing data preprocessing on the real-time state information of the navigation environment to obtain the real-time state information of the navigation environment after data preprocessing; Based on the current preset track prediction conditions and the real-time status information of the navigation environment after the data preprocessing, the target measurement and control key data corresponding to the target ship are obtained.

[0040] In the present invention, the received raw data (i.e., the original real-time navigation environment status information) may contain errors, duplications, or invalid information. Therefore, this data needs to be verified to identify and correct errors. Furthermore, a data cleansing step is used to remove duplicate and invalid data to ensure the accuracy and efficiency of subsequent analysis. For example, real-time navigation environment status information often contains noise, which may originate from the sensor itself or interference during transmission. Denoising and filtering techniques can effectively reduce the impact of noise on data analysis and improve the signal-to-noise ratio. Since different sensors may have different measurement ranges and accuracies, data standardization and normalization are necessary to enable comparison and analysis of data from different sources on the same scale. In one embodiment, different types of pre-processed data can be fused to form a more comprehensive and accurate description of the navigation environment. For example, wind speed and direction data can be fused with ocean current data to more accurately assess the environmental impacts on the ship.

[0041] Step 103: Input the target measurement and control key data into the corresponding track prediction model to obtain the track prediction result of the target ship output by the track prediction model, wherein the track prediction model is obtained by training a machine learning model through sample measurement and control key data and the ship track position sample information corresponding to the sample measurement and control key data, and the data types of the sample measurement and control key data and the target measurement and control key data are the same.

[0042] This method requires the collection of a large amount of sample key measurement and control data. This data is derived from actual records of past ship voyages and includes information such as speed, heading, position, and current velocity. Similar to the target key measurement and control data, this data also undergoes a series of data preprocessing steps, such as verification, cleaning, denoising, and standardization, to ensure data quality and consistency.

[0043] At the same time, it is also necessary to collect the ship track position sample information corresponding to these sample measurement and control key data. This information records the position coordinates of the ship at different time points and forms the actual navigation track of the ship.

[0044] Furthermore, the machine learning model is trained using the collected sample measurement and control key data and the corresponding ship track position information. The model learns the patterns and regularities in the sample data and attempts to establish a mapping relationship between the sample measurement and control key data and the ship's track position. By continuously optimizing the model parameters, the model can more accurately predict the ship's future track.

[0045] In this invention, after model training is complete, the model needs to be evaluated to verify its predictive performance. A portion of sample data can be used as a test set to assess the model's generalization ability on unknown data. Based on the evaluation results, the model can be further optimized and adjusted to improve prediction accuracy and stability. After the model is constructed and trained, it can be applied to track prediction during actual navigation. Key target measurement and control data (also preprocessed) is input into the trained track prediction model, and the model will use this data to calculate the target ship's future track prediction.

[0046] In the present invention, more types of sample measurement and control key data can be selected for model training based on actual track prediction conditions. Accordingly, in the actual prediction process of the model, the data type involved in the target measurement and control key data input into the model is the same as the data type of the sample data selected during the model training, thereby training a track prediction model that meets different track prediction conditions.

[0047] In one embodiment, a track prediction model is trained using sample data such as speed data, heading information, and sailing time. After training, the model is fed with the current ship data (i.e., the speed data, heading information, and sailing time for the current time period). Specifically, the input data is sailing speed: speed = 10.0, sailing direction: direction = math.radians(30), and sailing time: time = 1.0 hours.

[0048] Set the starting and ending coordinates of the ship to: start_point = (0, 0); end_point = (0, 0); The x-coordinate increment delta_x of the track predicted by the model is: delta_x=speed×math.cos(direction)×time; The y-coordinate increment delta_y of the track predicted by the model is: delta_y=speed ×math.sin(direction) ×time; At this point, the final destination coordinates of the ship are updated as follows: end_point= (start_point[0] + delta_x, start_point[1]+ delta_y); At this time, during the model training process, all extracted feature values ​​will be used as factors for coordinate extraction to calculate the new end point coordinates, thereby obtaining a range model for possible future tracks.

[0049] The track prediction method based on the survey ship provided by the present invention determines the target measurement and control key data from the real-time status information of the navigation environment generated by the target ship in the current time period, and then uses the track prediction model trained by the corresponding sample measurement and control key data and the ship track position sample information to predict the ship track, thereby obtaining a more accurate track prediction result.

[0050] Based on the above embodiment, the track prediction model is trained by the following steps: Obtain navigation environment sample status information; Based on the preset track prediction conditions, corresponding sample measurement and control key data is obtained according to the navigation environment sample state information, wherein the preset track prediction conditions are determined based on the data type of the measurement and control key data; Determining the track coordinate information corresponding to the ship in a historical time period based on the sample measurement and control key data, so as to obtain the ship track position sample information through the track coordinate information; A training sample set is constructed based on the sample measurement and control key data and the ship track position sample information, and a support vector machine is trained using the training sample set. After determining that the training results meet the preset training conditions, the track prediction model is obtained.

[0051] In order to construct a track prediction model, the present invention requires collecting relevant navigation environment data. This data typically includes, but is not limited to, current velocity, wind direction and speed, seawater temperature, water depth, wave height, tidal information, and the positions and speeds of other vessels surrounding the vessel. This information is a crucial indicator of the navigation environment and has a direct impact on the vessel's trajectory. Sensors, radar, satellite navigation systems, and other devices can be used to obtain this navigation environment status information in real time or periodically.

[0052] After collecting the sample state information for the navigation environment, the next step is to filter and extract key measurement and control data based on preset track prediction conditions. These conditions are determined based on the type of key measurement and control data. For example, these conditions include changes in navigation speed over a specific time period, the direction and speed of currents in a specific sea area, and the navigation performance of a specific ship type at different water depths. By matching these preset conditions with the sample state information for the navigation environment, key data closely related to track prediction can be accurately extracted.

[0053] After determining the sample key measurement and control data, this data is used to trace the ship's trajectory over a certain period of time. This involves analyzing the ship's navigation records, including but not limited to the ship's GPS position, heading angle, and speed. This information can be used to plot the ship's track over the historical period and extract the corresponding track coordinates. This coordinate information will serve as the ship's track position sample information for subsequent training and prediction.

[0054] After obtaining sample measurement and control key data and ship track position information, they can be combined to construct a training sample set. Each sample contains a set of measurement and control key data and corresponding track coordinate information. This training sample set is then used to train a support vector machine (SVM). In this invention, track prediction is treated as a regression problem, predicting future track coordinates based on current measurement and control key data.

[0055] During training, the SVM learns the mapping between key measurement and control data and track coordinates, and continuously optimizes its internal parameters to minimize prediction error. When the training results meet preset training conditions (such as error threshold and number of iterations), the training process ends, resulting in the final track prediction model.

[0056] Specifically, the present invention is based on the two key steps of the SVM algorithm: margin maximization and kernel function selection, and its goal is to obtain the optimal solution that minimizes the structural risk under the existing information. Assuming that the training sample set is given for: ; in, for dimensional input variables, namely, input speed data, heading information, position information, current speed and other information; for The corresponding output variables, namely the track coordinate related information, are expressed as follows: ; ; Define the hyperplane for: ; in, The normal vector determines the direction of the hyperplane. The intercept determines the distance between the hyperplane and the origin. The points correspond to a type of data points, The points correspond to another type of data points. The normalized linearly separable sample set is: ; For each sample point Introducing slack variables (used to describe the size of the deviation from the sensitive area) and penalty function C (used to adjust the balance between empirical risk and expressive power), using ε -The SVM with insensitive loss function can be described as the following optimization problem: ; in, and is the solution to the above convex quadratic programming.

[0057] By constructing the Lagrange function, the optimization problem is transformed into a dual problem : ; ; in, is the Lagrange multiplier, is the solution to the above dual problem.

[0058] is the kernel function, Indicates that The feature vector after nonlinear mapping to high-dimensional feature space, where: ; The regression decision function obtained after solving the optimization problem It can be written as: ; To address the enormous number of telemetry parameters that influence the trajectory of a survey vessel, the present invention utilizes the Support Vector Machine (SVM) algorithm to maximize the data margin during predictive analysis. The SVM algorithm finds an optimal hyperplane to separate sample points of different categories. For linearly separable problems, a linear hyperplane can be found that completely separates the data points. For linearly inseparable problems, a kernel function can be introduced to map the data into a high-dimensional space, thereby finding a hyperplane that separates the data points. Furthermore, the distance from the sample points closest to the hyperplane to the hyperplane is maximized. This distance is known as the margin, and the support vectors are the sample points closest to the hyperplane. To find the optimal hyperplane, an objective function must be minimized. This function consists of two components: the inverse of the margin, which represents the distance from the sample point to the hyperplane, and the model complexity. The inverse of the margin represents the distance from the sample point to the hyperplane, and this distance is desired to be as large as possible. During training, as many training samples as possible should be used. By minimizing the error between the predicted and actual values, an optimal hyperplane can be found for regression. This present invention maximizes the sample data margin without affecting subsequent model training or results.

[0059] The model training process involves parameter tuning and model evaluation. The present invention adjusts model parameters through methods such as cross-validation to find the optimal parameter combination. Simultaneously, model performance metrics such as accuracy, precision, and recall are evaluated to verify the model's predictive power and stability. Furthermore, the present invention performs model training based on the data hyperplane and optimal algorithm function obtained in the above steps to obtain the optimal solution for track prediction. During the model training step, factors influencing the track can be set to adjust model parameters for different data types.

[0060] The present invention can also predict the data in the trained model according to different formulas and influence levels based on the influencing factors of track prediction. When conducting preliminary model evaluation, a small amount of data can be used for prediction, and the accuracy of the track prediction results can be verified before a large amount of data is used for prediction, so as to infer the track line, track direction, etc. through the prediction algorithm.

[0061] On the basis of the above embodiment, after inputting the target measurement and control key data into the corresponding track prediction model and obtaining the track prediction result of the target ship output by the track prediction model, the method further includes: Receive first input information, wherein the first input information includes an operation instruction to adjust a data type of key measurement and control data corresponding to a current preset track prediction condition; In response to the first input information, a target track prediction condition is obtained, and based on the data type of the key measurement and control data in the target track prediction condition, the sample key measurement and control data in the training sample set is adjusted to obtain an adjusted training sample set; wherein the target track prediction condition is the preset track prediction condition after the data type is adjusted; Adjusting the model parameters of the track prediction model using the adjusted training sample set to obtain an adjusted track prediction model; The step of obtaining target measurement and control key data corresponding to the target ship according to the real-time state information of the navigation environment includes: Based on the target track prediction conditions and the real-time status information of the navigation environment, key target measurement and control data with adjusted data types are obtained.

[0062] In the present invention, the extracted characteristic information changes according to different weather conditions, equipment status and other factors. The prediction model can be modified according to the actual situation, the measurement and control parameters and calculation formulas can be modified at any time, and the track prediction model that meets the current requirements can be configured at any time. It has good flexibility and can more realistically reflect the actual situation.

[0063] Specifically, target track prediction conditions are set based on the needs of actual application scenarios. These conditions can be adjusted based on the prediction time range, accuracy requirements, and specific environmental factors (such as wind speed and temperature). In this invention, the existing training sample set is processed according to the data type requirements specified in the acquired target track prediction conditions to ensure that the data in the training sample set meets the requirements of model training, especially for key measurement and control data that has a significant impact on the prediction results.

[0064] After these adjustments, the data in the training sample set will be more standardized and consistent, and better meet the requirements of the target track prediction conditions. This adjusted training sample set will become the basis for subsequent model training and adjustments. Furthermore, the adjusted training sample set will be used to train or further adjust the existing track prediction model. During this process, the model will learn or optimize its internal parameters (such as weights and biases) based on the provided sample data to more accurately fit the patterns and regularities in the data. Through iterative training, the model will gradually improve its prediction capabilities until it reaches the preset performance indicators or convergence conditions.

[0065] Ultimately, a trained and tuned trajectory prediction model can generate more accurate and reliable trajectory predictions based on new or similar input data. Such a model is crucial for improving the accuracy and reliability of trajectory predictions, especially in complex applications requiring high-precision predictions.

[0066] On the basis of the above embodiment, before acquiring the real-time state information of the navigation environment of the target ship generated in the current time period based on the preset data sampling window, the method includes: receiving second input information, wherein the second input information includes instruction information for adjusting the preset data sampling window; In response to the second input information, adjusting the window size of the current preset data sampling window to obtain an adjusted preset data sampling window; The step of obtaining the real-time state information of the navigation environment of the target ship generated in the current time period based on the preset data sampling window includes: Based on the adjusted preset data sampling window, the real-time status information of the navigation environment generated by the target ship in the current time period is obtained.

[0067] In the present invention, during model training or actual track prediction, historical or real-time navigation data is collected, including information such as the ship's position, speed, and heading, to ensure data integrity and accuracy. During data collection, if high data accuracy is required, for example, if the speed needs to be measured with an accuracy of ±0.5 knots, the data sampling window can be set to ±0.5 knots; if data needs to be sampled once per second, the data sampling window can be set to 1 second.

[0068] In the present invention, the size of the data sampling window can be calculated based on the required accuracy and sampling frequency. The window size can also be adjusted based on actual needs. For example, if a faster response speed is required, the data sampling window size can be reduced; if higher accuracy is required, the sampling frequency can be increased or the data sampling window size can be reduced. Based on the window calculation results, the present invention uses a sliding window method to process ship navigation data. Specifically, data within a certain time range is selected as a window, and the statistics of each measurement and control data within the window are calculated, which are used as the characteristic values ​​of this time period. By continuously sliding the window, a series of characteristic values ​​are obtained, which can comprehensively reflect the navigation status and behavior patterns of the survey vessel.

[0069] The present invention adaptively adjusts the sampling data window during the data sampling process, and maximizes the data interval through the SVM algorithm during feature value extraction to improve the efficiency of the ship track prediction algorithm.

[0070] In one embodiment, the track prediction method based on ship measurement and control equipment data provided by the present invention is generally described. The specific process is as follows: First, the measurement and control data of the survey ship is obtained, including key information such as the ship's position, speed, heading, current speed and direction. This data is collected in real time by ship-borne sensors and transmitted to the ship's prediction system through communication equipment.

[0071] This method uses a support vector machine (SVM) algorithm to maximize data intervals to capture valid feature data, including the ship's motion characteristics and surrounding environmental conditions. By extracting feature information from historical measurement and control data during navigation, it builds a track prediction model. This model combines historical data with current conditions to assist operators in predicting future navigation trajectories. Furthermore, the track prediction model's flexibility lies in its ability to adjust model parameters in real time based on varying circumstances, improving prediction accuracy and adaptability.

[0072] Furthermore, the real-time measurement and control data collected is input into the track prediction model for track prediction. In the present invention, the track prediction model can modify the measurement and control parameters and calculation formulas in the early stages based on factors such as weather conditions and equipment status, maintaining the model's flexibility and customization. This allows for model adjustments based on actual needs, or the construction of track prediction models with different input standards.

[0073] Optionally, in one embodiment, the size of the sampling data window during the data sampling process can be adjusted according to acquisition requirements to adapt to different prediction requirements and data conditions.

[0074] The present invention provides a method for track prediction based on the data characteristic values ​​of measurement and control equipment of a survey ship, which makes the ship track prediction more accurate and practical, and improves navigation safety and efficiency.

[0075] The track prediction system based on a survey ship provided by the present invention is described below. The track prediction system based on a survey ship described below and the track prediction method based on a survey ship described above can refer to each other.

[0076] Figure 2 The structural diagram of the track prediction system based on the survey ship provided by the present invention is as follows: Figure 2 As shown, the present invention provides a track prediction system based on a survey ship, including a data sampling module 201, a data processing module 202 and a track prediction module 203, wherein the data sampling module 201 is used to obtain the real-time status information of the navigation environment generated by the target ship in the current time period based on a preset data sampling window; the data processing module 202 is used to obtain the target measurement and control key data corresponding to the target ship according to the real-time status information of the navigation environment; the track prediction module 203 is used to input the target measurement and control key data into the corresponding track prediction model to obtain the track prediction result of the target ship output by the track prediction model, wherein the track prediction model is obtained by training a machine learning model through sample measurement and control key data and the ship track position sample information corresponding to the sample measurement and control key data, and the data types of the sample measurement and control key data and the target measurement and control key data are the same.

[0077] The track prediction system based on the survey ship provided by the present invention determines the target measurement and control key data from the real-time status information of the navigation environment generated by the target ship in the current time period, and then uses the track prediction model trained by the corresponding sample measurement and control key data and the ship track position sample information to predict the ship track, thereby obtaining a more accurate track prediction result.

[0078] The system provided by the present invention is used to execute the above-mentioned method embodiments. Please refer to the above-mentioned embodiments for the specific processes and detailed contents, which will not be repeated here.

[0079] Figure 3 A schematic diagram of the structure of the electronic device provided by the present invention, such as Figure 3 As shown, the electronic device may include: a processor 301, a communications interface 302, a memory 303, and a communication bus 304, wherein the processor 301, the communications interface 302, and the memory 303 communicate with each other via the communication bus 304. The processor 301 may call logic instructions in the memory 303 to execute a track prediction method based on a survey vessel, the method comprising: obtaining real-time navigation environment status information generated by a target vessel within a current time period based on a preset data sampling window; obtaining target measurement and control key data corresponding to the target vessel based on the real-time navigation environment status information; and inputting the target measurement and control key data into a corresponding track prediction model to obtain a track prediction result of the target vessel output by the track prediction model. The track prediction model is obtained by training a machine learning model using sample measurement and control key data and ship track position sample information corresponding to the sample measurement and control key data, and the sample measurement and control key data and the target measurement and control key data have the same data type.

[0080] Furthermore, the logic instructions in the aforementioned memory 303 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage media include various media capable of storing program code, such as USB flash drives, mobile hard drives, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical disks.

[0081] On the other hand, the present invention also provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes program instructions. When the program instructions are executed by the computer, the computer can execute the track prediction method based on the survey ship provided by the above methods, the method including: based on a preset data sampling window, obtaining the real-time status information of the navigation environment generated by the target ship in the current time period; according to the real-time status information of the navigation environment, obtaining the target measurement and control key data corresponding to the target ship; inputting the target measurement and control key data into the corresponding track prediction model to obtain the track prediction result of the target ship output by the track prediction model, wherein the track prediction model is obtained by training a machine learning model with sample measurement and control key data and the ship track position sample information corresponding to the sample measurement and control key data, and the data types of the sample measurement and control key data and the target measurement and control key data are the same.

[0082] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the track prediction method based on the survey ship provided in the above-mentioned embodiments, the method comprising: obtaining the real-time status information of the navigation environment generated by the target ship in the current time period based on a preset data sampling window; obtaining the target measurement and control key data corresponding to the target ship according to the real-time status information of the navigation environment; inputting the target measurement and control key data into the corresponding track prediction model to obtain the track prediction result of the target ship output by the track prediction model, wherein the track prediction model is obtained by training a machine learning model with sample measurement and control key data and the ship track position sample information corresponding to the sample measurement and control key data, and the data types of the sample measurement and control key data and the target measurement and control key data are the same.

[0083] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they may be located in one location or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0084] Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented using software plus a necessary general-purpose hardware platform, or of course, hardware. Based on this understanding, the essence of the above technical solution, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, or an optical disk, and includes a number of instructions for causing a computer device (such as a personal computer, server, or network device) to execute the methods described in each embodiment or certain portions of the embodiments.

[0085] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A track prediction method based on a survey ship, characterized in that: include: Based on the preset data sampling window, obtain the real-time status information of the navigation environment generated by the target ship in the current time period; Acquiring target measurement and control key data corresponding to the target ship based on the real-time status information of the navigation environment; The target measurement and control key data is input into the corresponding track prediction model to obtain the track prediction result of the target ship output by the track prediction model, wherein the track prediction model is obtained by training a machine learning model through sample measurement and control key data and the ship track position sample information corresponding to the sample measurement and control key data, and the data types of the sample measurement and control key data and the target measurement and control key data are the same.

2. The track prediction method based on a survey ship according to claim 1, characterized in that: The track prediction model is trained by the following steps: Obtain navigation environment sample status information; Based on the preset track prediction conditions, corresponding sample measurement and control key data is obtained according to the navigation environment sample state information, wherein the preset track prediction conditions are determined based on the data type of the measurement and control key data; Determining the track coordinate information corresponding to the ship in a historical time period based on the sample measurement and control key data, so as to obtain the ship track position sample information through the track coordinate information; A training sample set is constructed based on the sample measurement and control key data and the ship track position sample information, and a support vector machine is trained using the training sample set. After determining that the training results meet the preset training conditions, the track prediction model is obtained.

3. The track prediction method based on a survey ship according to claim 2, characterized in that: After inputting the target measurement and control key data into the corresponding track prediction model to obtain the track prediction result of the target ship output by the track prediction model, the method further includes: Receive first input information, wherein the first input information includes an operation instruction to adjust a data type of key measurement and control data corresponding to a current preset track prediction condition; In response to the first input information, a target track prediction condition is obtained, and based on the data type of the key measurement and control data in the target track prediction condition, the sample key measurement and control data in the training sample set is adjusted to obtain an adjusted training sample set; wherein the target track prediction condition is the preset track prediction condition after the data type is adjusted; Adjusting the model parameters of the track prediction model using the adjusted training sample set to obtain an adjusted track prediction model; The step of obtaining target measurement and control key data corresponding to the target ship according to the real-time state information of the navigation environment includes: Based on the target track prediction conditions and the real-time status information of the navigation environment, key target measurement and control data with adjusted data types are obtained.

4. The track prediction method based on a survey ship according to claim 1, characterized in that: Before acquiring the real-time state information of the navigation environment of the target ship generated in the current time period based on the preset data sampling window, the method includes: receiving second input information, wherein the second input information includes instruction information for adjusting the preset data sampling window; In response to the second input information, adjusting the window size of the current preset data sampling window to obtain an adjusted preset data sampling window; The step of obtaining the real-time state information of the navigation environment of the target ship generated in the current time period based on the preset data sampling window includes: Based on the adjusted preset data sampling window, the real-time status information of the navigation environment of the target ship generated in the current time period is obtained.

5. The track prediction method based on a survey ship according to claim 1, characterized in that: The target measurement and control key data includes at least one of the speed data, heading information, position information and ocean current speed of the target ship.

6. The track prediction method based on a survey vessel according to any one of claims 2 to 5, characterized in that: The step of obtaining target measurement and control key data corresponding to the target ship according to the real-time state information of the navigation environment includes: performing data preprocessing on the real-time state information of the navigation environment to obtain the real-time state information of the navigation environment after data preprocessing; Based on the current preset track prediction conditions and the real-time status information of the navigation environment after the data preprocessing, the target measurement and control key data corresponding to the target ship are obtained.

7. A track prediction system based on a survey vessel, characterized in that: include: The data sampling module is used to obtain the real-time status information of the navigation environment of the target ship generated in the current time period based on the preset data sampling window; A data processing module is used to obtain key target measurement and control data corresponding to the target ship based on the real-time status information of the navigation environment; A track prediction module is used to input the target measurement and control key data into the corresponding track prediction model to obtain the track prediction result of the target ship output by the track prediction model, wherein the track prediction model is obtained by training a machine learning model through sample measurement and control key data and the ship track position sample information corresponding to the sample measurement and control key data, and the data types of the sample measurement and control key data and the target measurement and control key data are the same.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the track prediction method based on the survey ship according to any one of claims 1 to 6 is implemented.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the track prediction method based on a survey ship as claimed in any one of claims 1 to 6 is implemented.

10. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the track prediction method based on a survey ship as claimed in any one of claims 1 to 6 is implemented.

Images