Multi-source heterogeneous ocean data fusion system and real-time processing method thereof

Through the multi-source heterogeneous marine data fusion system, the speed curve is evaluated and adjusted in real time, and the problem of personalized modeling and response to emergencies in traditional navigation decisions is solved, and effective avoidance of navigation risks and shipping is achieved. Improved efficiency.

CN119004363AActive Publication Date: 2024-11-22GUANGDONG OCEAN UNIVERSITY
View PDF 3 Cites 0 Cited by

Patent Information

Application Number
CN202411046003.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-01
Publication Date
2024-11-22
Estimated Expiration
2044-08-01

AI Technical Summary

Technical Problem

The traditional route planning and navigation decision-making process cannot be personalized for different sea areas, the prediction accuracy is limited, and the emergencies cannot be responded to emergencies in a timely manner, and the lack of real-time optimization and adjustments makes it difficult to effectively avoid navigation risks.

Method used

A multi-source heterogeneous marine data fusion system is proposed. By dividing the sea area into sub-sea areas, collecting historical risk characteristic data and labels, building a risk prediction model, evaluating navigation risks in real time, and dynamically adjusting the speed curve to avoid risks.

Benefits of technology

Significantly improve the intelligence and forward-looking nature of navigation decisions, effectively avoid navigation risks, and improve shipping efficiency and safety.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119004363A_ABST
    Figure CN119004363A_ABST
Patent Text Reader

Abstract

The invention discloses a multi-source heterogeneous ocean data fusion system and a real-time processing method thereof, and relates to the technical field of multi-source heterogeneous data processing. Historical risk feature data and historical risk labels of each sub-sea area are collected, a risk prediction model is trained, and the navigation duration required for an aircraft crew to reach each passing sub-sea area is calculated; for each passing sub-sea area, outputting a navigation risk value, and based on the navigation risk values of all the passing sub-sea areas, judging whether navigation is kept or not; if sailing is kept, sailing continues; otherwise, based on the sailing time length of each sub-sea area, the risk prediction model and the real-time risk characteristics, calculating the sailing average speed between every two sub-sea areas, and if the sailing average speed does not have a feasible solution, returning the sailing; otherwise, adjusting the sailing speed curve according to the sailing average speed; the intelligence and foresight of navigation decision making are greatly improved, the navigation risk is effectively avoided, and the shipping efficiency and safety are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of multi-source heterogeneous data processing, and in particular to a multi-source heterogeneous ocean data fusion system and a real-time processing method thereof. Background Art

[0002] Maritime navigation safety is crucial to ensure the safe operation of ships, personnel and cargo, and is a core issue that the shipping industry has always paid attention to. With global warming, extreme weather events such as tsunamis and storm surges are becoming more frequent, bringing unprecedented challenges to maritime navigation. The sea conditions are complex and changeable, and the risks of navigation are increasing day by day. Once a safety accident occurs, it will cause incalculable casualties and economic losses. Therefore, how to minimize the risks of maritime navigation and ensure that ships arrive at their destinations in a timely and safe manner has become a major issue that needs to be solved urgently.

[0003] The traditional route planning and navigation decision-making process has obvious shortcomings. First, most risk assessment models are trained based on overall sea area data, and cannot perform personalized modeling and risk quantification based on the characteristics of different sea areas, and the prediction accuracy is limited. Secondly, these models mainly rely on historical data and have a lag in responding to emergencies such as bad weather. More importantly, the traditional route planning process lacks real-time optimization and adjustment links. Once an unexpected situation is encountered during the voyage, the existing technology cannot dynamically adjust the speed curve according to the risks along the way in a timely manner, thereby reducing the navigation risk. It can only respond passively, or even have to interrupt the voyage, which has a serious impact on operational efficiency and economic benefits.

[0004] To this end, the present invention proposes a multi-source heterogeneous ocean data fusion system and a real-time processing method thereof. Summary of the invention

[0005] The present invention aims to solve at least one of the technical problems existing in the prior art. To this end, the present invention proposes a multi-source heterogeneous ocean data fusion system and a real-time processing method thereof, which greatly improves the intelligence and foresight of navigation decision-making, effectively avoids navigation risks, and improves shipping efficiency and safety.

[0006] To achieve the above objectives, a real-time processing method for multi-source heterogeneous ocean data is proposed, which includes the following steps:

[0007] Step 1: Divide the ocean into several sub-sea areas and collect historical risk characteristic data and historical risk labels for each sub-sea area;

[0008] Step 2: Preset the number of monitoring cycles T. For each monitoring cycle, construct a number of navigation samples based on the historical risk characteristic data and historical risk labels of each sub-sea area;

[0009] Step 3: For each monitoring period, the navigation samples are used as training samples to train the risk prediction model;

[0010] Step 4: collect the current position of the ship, the subsequent sailing route of the ship and the established sailing speed curve in real time, and count each sub-sea area reached by the ship as a passing sub-sea area, and calculate the sailing time required to reach each passing sub-sea area;

[0011] Step 5: For each passing sub-sea area, read the risk prediction model corresponding to the corresponding sailing time as the target model, and output the sailing risk value through the collected real-time risk characteristics and the target model;

[0012] Step 6: Based on the navigation risk values ​​of all sub-sea areas, determine whether to continue sailing; if so, continue sailing; otherwise, execute step 7;

[0013] Step 7: Based on the sailing time, risk prediction model and real-time risk characteristics of each sub-sea area, the average speed of the ship between each two sub-sea areas is calculated. If there is no feasible solution for the average speed of the ship, the ship returns; otherwise, the sailing speed curve is adjusted according to the average speed of the ship, and steps 4 to 7 are repeated;

[0014] The method of collecting the historical risk characteristic data and historical risk labels of each sub-sea area is:

[0015] For each sub-sea area:

[0016] Collect risk characteristic records of ships passing through the sub-sea area at every moment since the navigation record of the sub-sea area was available; the risk characteristic records include the timestamp, sea condition data, and ship status at each moment, and the timestamps, sea condition data, and ship status at all moments constitute the historical risk characteristic data;

[0017] A risk label is added to each timestamp in the navigation record of the sub-sea area; the risk label is 0 or 1; wherein 0 indicates that the ship sailing at the timestamp passes through the sub-sea area, and 1 indicates that the ship sailing at the timestamp fails to pass through the sub-sea area;

[0018] The risk label of each timestamp constitutes the historical risk label;

[0019] For each monitoring period, based on the historical risk characteristic data and historical risk labels of each sub-sea area, a number of navigation samples are constructed as follows:

[0020] The total duration corresponding to each monitoring cycle is marked as t;

[0021] For any total duration t:

[0022] The time step b of the preset time series forecasting model;

[0023] Traverse the navigation records of each sub-sea area, connect each sea condition data at each moment in the sub-sea area in chronological order as a first parameter sequence, and connect each ship status at each moment in the sub-sea area in chronological order as a second parameter sequence; mark each timestamp in the navigation record as c;

[0024] For timestamp c, a subsequence from timestamp ctb to timestamp ct is intercepted from the first parameter sequence as the first parameter subsequence;

[0025] For timestamp c, a subsequence from timestamp ctb to timestamp ct is intercepted from the second parameter sequence as the second parameter subsequence;

[0026] The first parameter subsequence and the second parameter subsequence are used as a training parameter sequence set corresponding to the timestamp c;

[0027] The risk label corresponding to timestamp c is used as the training risk label corresponding to timestamp c;

[0028] The set of training parameter sequences and training risk labels at timestamp c constitute a set of navigation samples;

[0029] For each monitoring period, the navigation samples are used as training samples, and the method of training the risk prediction model is as follows:

[0030] For the monitoring period t:

[0031] The training parameter sequence set of each timestamp of the monitoring period is used as the input of the risk prediction model, and the risk prediction model takes the predicted value of the navigation risk value after the monitoring period t as the output, and the navigation risk value is 0 or 1; the risk prediction model takes the training risk label of the corresponding timestamp as the prediction target, takes the difference between the predicted value of the navigation risk value and the training risk label as the prediction error, and takes minimizing the sum of squares of the prediction errors of all timestamps as the training target; the risk prediction model is trained until the sum of squares of the prediction errors reaches convergence, and the training is stopped; the risk prediction model is a time series prediction model;

[0032] The method for determining whether to continue sailing based on the navigation risk values ​​of all sub-sea areas is as follows:

[0033] If the navigation risk values ​​of all sub-sea areas are 0, it is judged to maintain navigation;

[0034] The method for calculating the average speed of the ship between each two sub-sea areas based on the sailing time of each sub-sea area, the risk prediction model and the real-time risk characteristics is as follows:

[0035] The sub-sea area whose navigation risk value is 1 is indicated as a dangerous sub-sea area;

[0036] Obtain the order of passing through each dangerous sub-sea area from the navigation route, and number the dangerous sub-sea areas as j according to the order of passing through;

[0037] For each dangerous sub-sea area, traverse the risk prediction models corresponding to all monitoring periods, input the parameter application set of the dangerous sub-sea area into each risk prediction model respectively, and obtain the navigation risk value output by each risk prediction model in turn; mark the total duration set consisting of the total duration of the monitoring period corresponding to the last traversed risk prediction model as Zj;

[0038] Traverse and combine the total duration of all dangerous sub-sea areas to form Π j Nj passage time series; Nj is the total duration of the jth dangerous sub-sea area; π is the cumulative symbol;

[0039] For each passing time series, the expected speed of passing through each two dangerous sub-sea areas is calculated based on the distance between each two adjacent dangerous sub-sea areas and the corresponding time difference in the passing time series;

[0040] The way to determine whether there is a feasible solution for the average speed of the ship is:

[0041] For each passing time sequence, if the expected speed of any two adjacent dangerous sub-sea areas is less than or equal to 0 or +∞ or greater than the preset speed threshold, the passing time sequence is judged to be unnavigable; otherwise, the passing time sequence is judged to be navigable;

[0042] All navigable passing time series are screened out as optional time series, and one optional time series is selected from all optional time series to adjust the navigation speed curve.

[0043] A multi-source heterogeneous ocean data fusion system is proposed, including a navigation sample construction module, a model training module and a navigation suggestion module; wherein each module is electrically connected to each other;

[0044] The navigation sample construction module divides the ocean into several sub-sea areas, collects historical risk characteristic data and historical risk labels of each sub-sea area, presets the number of monitoring cycles T, and for each monitoring cycle, constructs several navigation samples based on the historical risk characteristic data and historical risk labels of each sub-sea area, and sends the navigation samples to the model training module;

[0045] The model training module uses the navigation samples as training samples to train the risk prediction model for each monitoring period, and sends the risk prediction model to the navigation suggestion module;

[0046] The navigation suggestion module collects the current position of the ship, the subsequent navigation route of the ship and the established navigation speed curve in real time, and counts each sub-sea area reached by the ship as a passing sub-sea area, and calculates the navigation time required to reach each passing sub-sea area. For each passing sub-sea area, the risk prediction model corresponding to the corresponding navigation time is read as the target model, and the navigation risk value is output through the collected real-time risk characteristics and the target model. Based on the navigation risk values ​​of all passing sub-sea areas, it is judged whether to maintain navigation; if so, continue to sail; otherwise, based on the navigation time, risk prediction model and real-time risk characteristics of each passing sub-sea area, the average speed of the ship between each two sub-sea areas is calculated. If there is no feasible solution for the average speed of the ship, the ship returns; otherwise, the navigation speed curve is adjusted according to the average speed of the ship.

[0047] An electronic device comprises: a processor and a memory, wherein the memory stores a computer program that can be called by the processor; the processor executes the above-mentioned real-time processing method of multi-source heterogeneous ocean data by calling the computer program stored in the memory.

[0048] A computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the above-mentioned method for real-time processing of multi-source heterogeneous ocean data.

[0049] Compared with the prior art, the present invention has the following beneficial effects:

[0050] The present invention first divides the sea area into several sub-areas, collects historical risk characteristic data of each area, and constructs a risk label data set. Then, multiple time series risk prediction models are trained for different sailing durations. During the navigation process, the ship's position, route and speed information are obtained in real time, and the passing risk of each sub-area along the way is dynamically evaluated in combination with the risk prediction model. If there is a high-risk area, the system will intelligently plan a new feasible speed curve based on the time constraints of the sub-area, thereby minimizing the overall navigation risk. It greatly improves the intelligence and foresight of navigation decisions, effectively avoids navigation risks, improves shipping efficiency and safety, and provides strong support for the development of smart navigation and related fields. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 This is a flow chart of a method for real-time processing of multi-source heterogeneous ocean data in Example 1 of the present invention;

[0052] Figure 2 This is a module connection relationship diagram of the multi-source heterogeneous ocean data fusion system in Example 2 of the present invention;

[0053] Figure 3 This is a schematic diagram of the structure of an electronic device in Embodiment 3 of the present invention;

[0054] Figure 4This is a schematic diagram of the computer-readable storage medium structure in Example 4 of the present invention. DETAILED DESCRIPTION

[0055] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0056] Example 1

[0057] like Figure 1 As shown, the real-time processing method of multi-source heterogeneous ocean data includes the following steps:

[0058] Step 1: Divide the ocean into several sub-sea areas and collect historical risk characteristic data and historical risk labels for each sub-sea area;

[0059] Step 2: Preset the number of monitoring cycles T. For each monitoring cycle, construct a number of navigation samples based on the historical risk characteristic data and historical risk labels of each sub-sea area;

[0060] Step 3: For each monitoring period, the navigation samples are used as training samples to train the risk prediction model;

[0061] Step 4: collect the current position of the ship, the subsequent sailing route of the ship and the established sailing speed curve in real time, and count each sub-sea area reached by the ship as a passing sub-sea area, and calculate the sailing time required to reach each passing sub-sea area;

[0062] Step 5: For each passing sub-sea area, read the risk prediction model corresponding to the corresponding sailing time as the target model, and output the sailing risk value through the collected real-time risk characteristics and the target model;

[0063] Step 6: Based on the navigation risk values ​​of all sub-sea areas, determine whether to continue sailing; if so, continue sailing; otherwise, execute step 7;

[0064] Step 7: Based on the sailing time, risk prediction model and real-time risk characteristics of each sub-sea area, the average speed of the ship between each two sub-sea areas is calculated. If there is no feasible solution for the average speed of the ship, the ship returns; otherwise, the sailing speed curve is adjusted according to the average speed of the ship, and steps 4 to 7 are repeated;

[0065] The method of dividing the ocean into several sub-sea areas is as follows:

[0066] Divide the ocean into several large blocks according to its actual geographical location and geographical scope, such as dividing the monitoring target ocean into 100*100 blocks according to longitude and latitude;

[0067] Furthermore, each large block can be further divided into multiple sub-sea areas according to the similarity of sea conditions; some special areas, such as ports and straits, can be treated as a sub-sea area separately. Therefore, the specific sub-sea area location division can be obtained by navigation experience.

[0068] The method of collecting the historical risk characteristic data and historical risk labels of each sub-sea area is as follows:

[0069] For each sub-sea area:

[0070] Collect risk characteristic records of ships passing through the sub-sea area at every moment since the navigation record of the sub-sea area was available; the risk characteristic records include the timestamp, sea condition data, and ship status at each moment, and the timestamps, sea condition data, and ship status at all moments constitute the historical risk characteristic data;

[0071] Specifically, the sea condition data include but are not limited to wind field data (wind speed, wind direction, gust level), ocean current data (current speed, current direction), wave condition data (significant wave height, maximum wave height, average period, wave direction), visibility data (visibility value, fog area range), precipitation (rainfall, whether there is thunderstorm), sea ice distribution (sea ice coverage, sea ice density), seabed topography data (water depth, tidal conditions) and remaining time of hurricane attack, etc.; it is understandable that each parameter in the sea condition data can be collected and obtained in real time by installing specific sensors in each sub-sea area in advance;

[0072] The ship status includes but is not limited to ship draft, displacement, ship type, cargo load, main engine status label (0 or 1, indicating whether it is faulty), steering gear status label (0 or 1, indicating whether it is faulty), main engine maintenance times, steering gear maintenance times, etc.;

[0073] A risk label is added to each timestamp in the navigation record of the sub-sea area; the risk label is 0 or 1; wherein 0 indicates that the ship sailing at the timestamp passes through the sub-sea area, and 1 indicates that the ship sailing at the timestamp fails to pass through the sub-sea area;

[0074] The risk label of each timestamp constitutes the historical risk label;

[0075] It is understandable that the sea conditions in various sea areas in the ocean change rapidly. As time goes by, the prediction of changes in sea conditions becomes less accurate. Therefore, the number of monitoring cycles T is set so that only the sea conditions after T monitoring cycles are considered to be predicted, so as to avoid misleading the navigation strategy of the ship.

[0076] The duration of each monitoring cycle and the number of monitoring cycles T are set based on the experience of actual navigation and sea condition prediction. For example, the number of monitoring cycles T can be 5, and the duration of each monitoring cycle is 1 hour;

[0077] Furthermore, for each monitoring period, based on the historical risk characteristic data and historical risk labels of each sub-sea area, a number of navigation samples are constructed in the following manner:

[0078] The total duration of each monitoring cycle is marked as t; for example, the total duration of the third monitoring cycle is 3 hours;

[0079] For any total duration t:

[0080] A time step b of a preset time series prediction model; the time series prediction model is any one of an RNN model or an LSTM model;

[0081] Traverse the navigation records of each sub-sea area, connect each sea condition data at each moment in the sub-sea area in chronological order as a first parameter sequence, and connect each ship status at each moment in the sub-sea area in chronological order as a second parameter sequence; mark each timestamp in the navigation record as c;

[0082] For timestamp c, a subsequence from timestamp ctb to timestamp ct is intercepted from the first parameter sequence as the first parameter subsequence;

[0083] For timestamp c, a subsequence from timestamp ctb to timestamp ct is intercepted from the second parameter sequence as the second parameter subsequence;

[0084] The first parameter subsequence and the second parameter subsequence are used as a training parameter sequence set corresponding to the timestamp c;

[0085] The risk label corresponding to timestamp c is used as the training risk label corresponding to timestamp c;

[0086] The set of training parameter sequences and training risk labels at timestamp c constitute a set of navigation samples;

[0087] It can be understood that the training parameter sequence set corresponding to the timestamp c is the historical navigation data of the sub-sea area at time t before, so that the risk at time t after can be predicted based on the historical navigation data;

[0088] Furthermore, for each monitoring period, the navigation samples are used as training samples, and the method of training the risk prediction model is as follows:

[0089] For the monitoring period t:

[0090] The training parameter sequence set of each timestamp of the monitoring period is used as the input of the risk prediction model, and the risk prediction model takes the predicted value of the navigation risk value after the monitoring period t as the output, and the navigation risk value is 0 or 1; the risk prediction model takes the training risk label of the corresponding timestamp as the prediction target, takes the difference between the predicted value of the navigation risk value and the training risk label as the prediction error, and takes minimizing the sum of squares of the prediction errors of all timestamps as the training target; the risk prediction model is trained until the sum of squares of the prediction errors reaches convergence, and the training is stopped; the risk prediction model is a time series prediction model;

[0091] It is understandable that in actual navigation, ships usually set a desired speed curve in advance as part of the navigation plan. Determining a reasonable speed curve is very important for ensuring ship safety, saving fuel costs and arriving at the destination on time.

[0092] Furthermore, the predetermined sailing speed curve is a predetermined sailing speed at each moment determined before the ship sails; and the current position can be determined by satellite navigation, and the sailing route is predetermined, therefore, the collection process of the current position, the sailing route and the sailing speed curve is not described in detail;

[0093] Furthermore, it is easy to understand that the sub-sea areas that the route passes through can be determined according to the navigation route and the positions of each sub-sea area, so that the sub-sea areas that will be passed can be determined in advance; and if the positions of the sub-sea areas that will be passed and the current position of the ship are known, the distance between the current position and each sub-sea area that will be passed can be determined, so that according to the known navigation speed curve, the time required for the ship to reach each sub-sea area (distance divided by speed) can be calculated;

[0094] Therefore, after knowing the time required to reach each sub-sea area, it is possible to predict whether the ship can pass through the sub-sea area safely when it arrives at the sub-sea area;

[0095] For each passing sub-sea area, the risk prediction model corresponding to the corresponding sailing time is read as the target model, and the risk warning level is outputted through the collected real-time risk characteristics and the target model as follows:

[0096] The number of the sub-sea area passed is marked as i;

[0097] The sailing time corresponding to the i-th passing sub-sea area is marked as ti;

[0098] For the i-th passing sub-sea area, the risk prediction model of the monitoring period corresponding to the navigation duration [ti]+1 is read as the target model corresponding to the i-th passing sub-sea area; where [ti] is the integer of ti; for example, if the navigation duration is 2.5 hours, the risk prediction model corresponding to the monitoring period of 3 hours is read;

[0099] Each sea condition data and each ship status in the real-time risk characteristics of each passing sub-sea area are respectively formed into a first real-time parameter sequence and a second real-time parameter sequence in chronological order;

[0100] Separate the first real-time parameter sequence and the second real-time parameter sequence into a first application sequence and a second application sequence consisting of a timestamp from a current timestamp minus a timestamp to a current timestamp;

[0101] The first application sequence and the second application sequence are combined into a parameter application set, which is input into the target model corresponding to the sub-sea area, and the navigation risk value output by the target model is obtained.

[0102] Furthermore, the method of judging whether to maintain navigation based on the navigation risk values ​​of all sub-sea areas is:

[0103] If the navigation risk values ​​of all sub-sea areas are 0, it is judged to maintain navigation;

[0104] It can be understood that if there is a navigation risk value of 1 in any passing sub-sea area, it means that when the ship reaches the passing sub-sea area, there may be risks, so it is inappropriate to keep sailing at the current speed;

[0105] Furthermore, the method for calculating the average speed of the ship between each two sub-sea areas based on the sailing time of each sub-sea area, the risk prediction model and the real-time risk characteristics is as follows:

[0106] The sub-sea area whose navigation risk value is 1 is indicated as a dangerous sub-sea area;

[0107] Obtain the order of passing through each dangerous sub-sea area from the navigation route, and number the dangerous sub-sea areas as j according to the order of passing through;

[0108] For each dangerous sub-sea area, traverse the risk prediction models corresponding to all monitoring periods, input the parameter application set of the dangerous sub-sea area into each risk prediction model respectively, and obtain the navigation risk value output by each risk prediction model in turn; mark the total duration set consisting of the total duration of the monitoring period corresponding to the last traversed risk prediction model as Zj; it can be understood that the total duration of each monitoring period in Zj represents the duration that the ship can safely pass through the dangerous sub-sea area from the current position;

[0109] Traverse and combine the total duration of all dangerous sub-sea areas to form Π jNj passing time series; Nj is the size of the total time set of the jth dangerous sub-sea area; π is the cumulative symbol; a specific example is that the total time set of the first dangerous sub-sea area is [1, 2], and the total time set of the second dangerous sub-sea area is [2, 4], then the passing time series includes (1, 2), (2, 2), (1, 4), (2, 4);

[0110] For each passing time series, the expected speed of passing through each two dangerous sub-sea areas is calculated based on the distance between each two adjacent dangerous sub-sea areas and the corresponding time difference in the passing time series. For example, in the above example, if the distance between the first dangerous sub-sea area and the second dangerous sub-sea area is 50 nautical miles, then when the passing time series is (1, 2), the expected speed of passing through these two dangerous sub-sea areas is 50 / (2-1) = 50 knots; when the passing time series is (2, 2), the expected speed of passing through these two dangerous sub-sea areas is +∞, that is, there is no solution.

[0111] Furthermore, the method for determining whether there is a feasible solution for the average speed of the ship is:

[0112] For each passing time sequence, if the expected speed of any two adjacent dangerous sub-sea areas is less than or equal to 0 or +∞ or greater than the preset speed threshold, the passing time sequence is judged to be unnavigable; otherwise, the passing time sequence is judged to be navigable;

[0113] All navigable passing time series are screened out as optional time series, and one optional time series is selected from all optional time series to adjust the navigation speed curve;

[0114] Preferably, the optional time series may be selected by selecting the optional time series with the smallest variance of the expected speed between every two adjacent dangerous sub-sea areas. The navigation speed of the optional time series varies little, so that the navigation is more stable.

[0115] It is understandable that, due to the adjustment of the navigation speed curve, the expected time to reach each sub-sea area may change, so that the predicted navigation risk value may have errors. Therefore, the adjustment of the navigation speed curve is real-time to adapt to the ever-changing navigation time.

[0116] Example 2

[0117] like Figure 2 As shown, the multi-source heterogeneous ocean data fusion system includes a navigation sample construction module, a model training module and a navigation suggestion module; wherein each module is electrically connected;

[0118] The navigation sample construction module divides the ocean into several sub-sea areas, collects historical risk characteristic data and historical risk labels of each sub-sea area, presets the number of monitoring cycles T, and for each monitoring cycle, constructs several navigation samples based on the historical risk characteristic data and historical risk labels of each sub-sea area, and sends the navigation samples to the model training module;

[0119] The model training module uses the navigation samples as training samples to train the risk prediction model for each monitoring period, and sends the risk prediction model to the navigation suggestion module;

[0120] The navigation suggestion module collects the current position of the ship, the subsequent navigation route of the ship and the established navigation speed curve in real time, and counts each sub-sea area reached by the ship as a passing sub-sea area, and calculates the navigation time required to reach each passing sub-sea area. For each passing sub-sea area, the risk prediction model corresponding to the corresponding navigation time is read as the target model, and the navigation risk value is output through the collected real-time risk characteristics and the target model. Based on the navigation risk values ​​of all passing sub-sea areas, it is judged whether to maintain navigation; if so, continue to sail; otherwise, based on the navigation time, risk prediction model and real-time risk characteristics of each passing sub-sea area, the average speed of the ship between each two sub-sea areas is calculated. If there is no feasible solution for the average speed of the ship, the ship returns; otherwise, the navigation speed curve is adjusted according to the average speed of the ship.

[0121] Example 3

[0122] See also Figure 3 As shown, according to another aspect of the present application, an electronic device 500 is also provided. The electronic device 500 may include one or more processors and one or more memories. The memories store computer readable codes, and when the computer readable codes are executed by one or more processors, the real-time processing method of multi-source heterogeneous ocean data as described above may be executed.

[0123] The method or system according to the embodiment of the present application can also be used by Figure 3 The electronic device architecture shown in FIG. Figure 3 As shown, the electronic device 500 may include a bus 501, one or more CPUs 502, a read-only memory (ROM) 503, a random access memory (RAM) 504, a communication port 505 connected to a network, an input / output component 506, a hard disk 507, etc. The storage device in the electronic device 500, such as the ROM 503 or the hard disk 507, may store the multi-source heterogeneous ocean data real-time processing method provided by the present application. Furthermore, the electronic device 500 may also include a user interface 508. Of course, Figure 3 The architecture shown is only exemplary and can be omitted according to actual needs when implementing different devices. Figure 3One or more components of an electronic device are shown.

[0124] Example 4

[0125] See also Figure 4 As shown, a computer-readable storage medium 600 according to an embodiment of the present application is shown. Computer-readable instructions are stored on the computer-readable storage medium 600. When the computer-readable instructions are executed by the processor, the multi-source heterogeneous ocean data real-time processing method according to the embodiment of the present application described with reference to the above figures can be executed. The storage medium 600 includes, but is not limited to, for example, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and cache memory (cache), etc. Non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc.

[0126] In addition, according to the embodiments of the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the present application provides a non-transitory machine-readable storage medium, which stores machine-readable instructions, and the machine-readable instructions can be executed by a processor to execute instructions corresponding to the method steps provided by the present application;

[0127] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center through a wired network or a wireless network. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a tape), an optical medium (for example, a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0128] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in the present invention can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0129] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0130] In the several embodiments provided by the present invention, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a multi-source heterogeneous ocean data fusion system and its real-time processing method. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0131] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0132] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0133] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A real-time processing method for multi-source heterogeneous ocean data, characterized in that: The following steps are involved: Step 1: Divide the ocean into several sub-sea areas and collect historical risk characteristic data and historical risk labels for each sub-sea area; Step 2: Preset the number of monitoring cycles T. For each monitoring cycle, construct a number of navigation samples based on the historical risk characteristic data and historical risk labels of each sub-sea area; Step 3: For each monitoring period, the navigation samples are used as training samples to train the risk prediction model; Step 4: collect the current position of the ship, the subsequent sailing route of the ship and the established sailing speed curve in real time, and count each sub-sea area reached by the ship as a passing sub-sea area, and calculate the sailing time required to reach each passing sub-sea area; Step 5: For each passing sub-sea area, read the risk prediction model corresponding to the corresponding sailing time as the target model, and output the sailing risk value through the collected real-time risk characteristics and the target model; Step 6: Based on the navigation risk values ​​of all sub-sea areas, determine whether to continue sailing; if necessary, continue sailing; Otherwise, go to step 7; Step 7: Based on the sailing time of each sub-sea area, the risk prediction model and the real-time risk characteristics, the average speed of the ship between every two sub-sea areas is calculated. If there is no feasible solution for the average speed of the ship, the ship returns; otherwise, the sailing speed curve is adjusted according to the average speed of the ship, and steps 4 to 7 are repeated.

2. The method for real-time processing of multi-source heterogeneous ocean data according to claim 1, characterized in that: The method of collecting the historical risk characteristic data and historical risk labels of each sub-sea area is: For each sub-sea area: Collect risk characteristic records of ships passing through the sub-sea area at every moment since the navigation record of the sub-sea area was available; the risk characteristic records include the timestamp, sea condition data, and ship status at each moment, and the timestamps, sea condition data, and ship status at all moments constitute the historical risk characteristic data; A risk label is added to each timestamp in the navigation record of the sub-sea area; the risk label is 0 or 1; wherein 0 indicates that the ship sailing at the timestamp passes through the sub-sea area, and 1 indicates that the ship sailing at the timestamp fails to pass through the sub-sea area; The risk labels at each timestamp constitute the historical risk labels.

3. The method for real-time processing of multi-source heterogeneous ocean data according to claim 2 is characterized in that: For each monitoring period, based on the historical risk characteristic data and historical risk labels of each sub-sea area, a number of navigation samples are constructed as follows: The total duration corresponding to each monitoring cycle is marked as t; For any total duration t: The time step b of the preset time series forecasting model; Traverse the navigation records of each sub-sea area, connect each sea condition data at each moment in the sub-sea area in chronological order as a first parameter sequence, and connect each ship status at each moment in the sub-sea area in chronological order as a second parameter sequence; mark each timestamp in the navigation record as c; For timestamp c, a subsequence from timestamp ctb to timestamp ct is intercepted from the first parameter sequence as the first parameter subsequence; For timestamp c, a subsequence from timestamp ctb to timestamp ct is intercepted from the second parameter sequence as the second parameter subsequence; The first parameter subsequence and the second parameter subsequence are used as a training parameter sequence set corresponding to the timestamp c; The risk label corresponding to timestamp c is used as the training risk label corresponding to timestamp c; The set of training parameter sequences and training risk labels at timestamp c constitute a set of navigation samples.

4. The method for real-time processing of multi-source heterogeneous ocean data according to claim 3 is characterized in that: For each monitoring period, the navigation samples are used as training samples, and the method of training the risk prediction model is as follows: For the monitoring period t: The training parameter sequence set of each timestamp of the monitoring period is used as the input of the risk prediction model, and the risk prediction model takes the predicted value of the navigation risk value after the monitoring period t as the output, and the navigation risk value is 0 or 1; the risk prediction model takes the training risk label of the corresponding timestamp as the prediction target, takes the difference between the predicted value of the navigation risk value and the training risk label as the prediction error, and takes minimizing the sum of squares of the prediction errors of all timestamps as the training target; The risk prediction model is trained until the sum of squares of the prediction errors reaches convergence and then the training is stopped; the risk prediction model is a time series prediction model.

5. The method for real-time processing of multi-source heterogeneous ocean data according to claim 4 is characterized in that: The method for determining whether to continue sailing based on the navigation risk values ​​of all sub-sea areas is as follows: If the navigation risk values ​​of all sub-sea areas are 0, it is judged that the navigation is maintained.

6. The method for real-time processing of multi-source heterogeneous ocean data according to claim 5, characterized in that: The method for calculating the average speed of the ship between each two sub-sea areas based on the sailing time of each sub-sea area, the risk prediction model and the real-time risk characteristics is as follows: The sub-sea area whose navigation risk value is 1 is indicated as a dangerous sub-sea area; Obtain the order of passing through each dangerous sub-sea area from the navigation route, and number the dangerous sub-sea areas as j according to the order of passing through; For each dangerous sub-sea area, traverse the risk prediction models corresponding to all monitoring periods, input the parameter application set of the dangerous sub-sea area into each risk prediction model respectively, and obtain the navigation risk value output by each risk prediction model in turn; mark the total duration set consisting of the total duration of the monitoring period corresponding to the last traversed risk prediction model as Zj; Traverse and combine the total duration of all dangerous sub-sea areas to form Π j Nj passage time series; Nj is the total duration of the jth dangerous sub-sea area; π is the cumulative symbol; For each passing time series, the expected speed for passing through each two dangerous sub-sea areas is calculated based on the distance between each two adjacent dangerous sub-sea areas and the corresponding time difference in the passing time series.

7. The method for real-time processing of multi-source heterogeneous ocean data according to claim 6, characterized in that: The way to determine whether there is a feasible solution for the average speed of the ship is: For each passing time sequence, if the expected speed of any two adjacent dangerous sub-sea areas is less than or equal to 0 or +∞ or greater than the preset speed threshold, the passing time sequence is judged to be unnavigable; otherwise, the passing time sequence is judged to be navigable; All navigable passing time series are screened out as optional time series, and one optional time series is selected from all optional time series to adjust the navigation speed curve.

8. A multi-source heterogeneous ocean data fusion system, which is used to implement the multi-source heterogeneous ocean data real-time processing method according to any one of claims 1 to 7, characterized in that: It includes a navigation sample construction module, a model training module and a navigation suggestion module; wherein each module is electrically connected to each other; The navigation sample construction module divides the ocean into several sub-sea areas, collects historical risk characteristic data and historical risk labels of each sub-sea area, presets the number of monitoring cycles T, and for each monitoring cycle, constructs several navigation samples based on the historical risk characteristic data and historical risk labels of each sub-sea area, and sends the navigation samples to the model training module; The model training module uses the navigation samples as training samples to train the risk prediction model for each monitoring period, and sends the risk prediction model to the navigation suggestion module; The navigation suggestion module collects the current position of the ship, the subsequent navigation route of the ship and the established navigation speed curve in real time, and counts each sub-sea area reached by the ship as a passing sub-sea area, and calculates the navigation time required to reach each passing sub-sea area. For each passing sub-sea area, the risk prediction model corresponding to the corresponding navigation time is read as the target model, and the navigation risk value is output through the collected real-time risk characteristics and the target model. Based on the navigation risk values ​​of all passing sub-sea areas, it is judged whether to maintain navigation; if so, continue to sail; otherwise, based on the navigation time, risk prediction model and real-time risk characteristics of each passing sub-sea area, the average speed of the ship between each two sub-sea areas is calculated. If there is no feasible solution for the average speed of the ship, the ship returns; otherwise, the navigation speed curve is adjusted according to the average speed of the ship.

9. An electronic device, characterized in that: include: A processor and a memory, wherein the memory stores a computer program that can be called by the processor; The processor executes the real-time processing method for multi-source heterogeneous ocean data according to any one of claims 1 to 7 by calling the computer program stored in the memory.

10. A computer-readable storage medium, characterized in that: Instructions are stored, and when the instructions are executed on a computer, the computer executes the real-time processing method for multi-source heterogeneous ocean data as claimed in any one of claims 1 to 7.

Citation Information

Patent Citations

  • Ship navigation environment risk prediction method and device, electronic equipment and medium

    CN115983627A

  • Global shipping information monitoring system based on big data

    CN116911468A

  • Method and system for assessing and early warning ship collision risk

    US20220036738A1