Track prediction method, device, equipment, storage medium and program product

Abstract

The invention discloses a track prediction method and device, equipment, a storage medium and a program product. The method comprises the steps of obtaining track data of a ship in a preset time period before the current moment and at least one meteorological data set of an area where the ship is located; performing interpolation processing on each meteorological data group according to the time sequence of the track data, and determining meteorological data corresponding to the track of the ship in a preset time period before the current moment according to at least one meteorological data group after interpolation processing; inputting the flight path data of the ship in a preset time period before the current moment and the meteorological data corresponding to the flight path into a pre-trained flight path prediction model to predict the flight path of the ship to obtain a predicted flight path of the ship; the track prediction model is obtained by training according to historical track data of the ship and meteorological data corresponding to the historical track. According to the method, the influence of meteorological factors on the flight path is considered, and the accuracy of predicting the flight path is improved.

Description

Technical Field

[0001] The present application relates to the technical field of track prediction, and in particular to a track prediction method, device, equipment, storage medium and program product. Background Art

[0002] As global maritime traffic becomes increasingly busy, the number of ships continues to increase, the risk of ship collisions and other accidents continues to increase, and the shipping industry's demand for safe navigation of ships is increasing. Therefore, predicting the track of ships has become an important means to ensure navigation safety and improve shipping efficiency.

[0003] At present, the future track of a ship is often predicted by the ship's historical track data. However, the marine environment is complex and changeable, especially under the influence of meteorological factors such as wind and waves, the actual navigation track of the ship often deviates from the planned route. Therefore, the predicted track obtained by the ship's historical track data has the problem of low accuracy. Summary of the invention

[0004] In view of the above-mentioned defects or deficiencies in the prior art, it is desired to provide a track prediction method, device, equipment, storage medium and program product, which takes into account the influence of meteorological factors on the track, predicts the ship's track based on the track data and meteorological data before the current moment, and effectively improves the accuracy of the predicted track.

[0005] In a first aspect, the present application provides a track prediction method. The method comprises: Acquire the track data of the ship within a preset period before the current time and at least one meteorological data group of the area where the ship is located; Interpolation processing is performed on each meteorological data group according to the time series of the track data, and meteorological data corresponding to the track of the ship in a preset period before the current time is determined according to at least one meteorological data group after the interpolation processing; the time series is a sequence after the data collection time is arranged in a preset order; The ship's track data in a preset period before the current moment and the meteorological data corresponding to the track are input into a pre-trained track prediction model to predict the ship's track and obtain the ship's predicted track; the track prediction model is trained based on the ship's historical track data and the meteorological data corresponding to the historical track.

[0006] In combination with the first aspect, in a possible implementation method, each meteorological data group is interpolated according to the time series of the track data, including: for each meteorological data group, searching in the time series of the meteorological data group according to the first time point in the time series of the track data; if there is no time point matching the first time point in the time series of the meteorological data group, determining the meteorological data corresponding to the first time point based on the meteorological data corresponding to at least one time point adjacent to the first time point in the meteorological data group; updating the meteorological data group according to the meteorological data corresponding to the first time point until the meteorological data group is updated according to the time points in the time series of the track data, thereby obtaining the interpolated meteorological data group.

[0007] In combination with the first aspect, in a possible implementation method, meteorological data corresponding to a first time point is determined based on meteorological data corresponding to at least one time point adjacent to the first time point in a meteorological data group, including: determining a second time point and a third time point adjacent to the first time point in a time series of the meteorological data group based on the first time point; determining meteorological change parameters in a time period between the second time point and the third time point based on the meteorological data corresponding to the second time point and the meteorological data corresponding to the third time point; and calculating the meteorological data corresponding to the first time point based on the meteorological change parameters and the meteorological data corresponding to the second time point.

[0008] In combination with the first aspect, in a possible implementation method, meteorological data corresponding to the track of the ship within a preset time period before the current moment is determined based on at least one meteorological data group after interpolation processing, including: for each coordinate in the track data, according to the position of the meteorological observation point corresponding to each meteorological data group, at least one candidate meteorological data group whose meteorological observation point position is within the preset range of the coordinates is screened out from at least one meteorological data group after interpolation processing; the meteorological data corresponding to the coordinate is determined based on at least one candidate meteorological data group, and the meteorological data corresponding to the track is obtained based on the meteorological data corresponding to each coordinate.

[0009] In combination with the first aspect, in a possible implementation method, the training process of the track prediction model includes: obtaining the historical track data of the ship and at least one historical meteorological data group of the area where the historical track of the ship is located; interpolating each historical meteorological data group according to the time series of the historical track data, and determining the meteorological data corresponding to the historical track of the ship according to at least one historical meteorological data group after the interpolation processing; grouping the historical track data and the meteorological data corresponding to the historical track according to the time series to obtain multiple historical track data groups; using at least one historical track data group that is sorted in front after being sorted from front to back in time among the multiple historical track data groups, and the meteorological data corresponding to the historical track as training samples, and using at least one historical track data group that is sorted in the back as a test sample, to train the initial track prediction model until the accuracy of the initial track prediction model reaches a preset accuracy, and a track prediction model is obtained.

[0010] In combination with the first aspect, in a possible implementation method, the track prediction method also includes: normalizing the historical track data and the meteorological data corresponding to the historical track, and training a track prediction model based on the normalized historical track data and the meteorological data corresponding to the normalized historical track.

[0011] In a second aspect, the present application also provides a track prediction device. The device comprises: An acquisition module, used to acquire the track data of the ship within a preset period before the current moment and at least one meteorological data group of the area where the ship is located; An interpolation module is used to perform interpolation processing on each meteorological data group according to the time series of the track data, and determine the meteorological data corresponding to the track of the ship in a preset period before the current time according to at least one meteorological data group after the interpolation processing; the time series is a sequence arranged in a preset order at the time of data collection; The prediction module is used to input the ship's track data in a preset period before the current moment and the meteorological data corresponding to the track into a pre-trained track prediction model to predict the ship's track and obtain the ship's predicted track; the track prediction model is trained based on the ship's historical track data and the meteorological data corresponding to the historical track.

[0012] In a third aspect, the present application further provides a computer device, which includes a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the method described in the first aspect is implemented.

[0013] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the method described in the first aspect is implemented.

[0014] In a fifth aspect, the present application further provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, the method described in the first aspect is implemented.

[0015] The embodiment of the present application provides a track prediction method, device, equipment, storage medium and program product, which first obtains the track data of the ship in the preset time period before the current moment and at least one meteorological data group in the area where the ship is located. Then, the track data and meteorological data are unified in time series through interpolation processing, and then the meteorological data on the ship's track is calculated through each meteorological data group, and the track data and meteorological data are unified in spatial position. Finally, the unified track data and meteorological data are input into a pre-trained track prediction model, and the predicted track is output. In the embodiment of the present application, considering the influence of meteorological factors on the track, the ship's track is predicted based on the track data and meteorological data before the current moment, which effectively improves the accuracy of the predicted track. Moreover, before predicting the track, the present application aligns and unifies the track data and meteorological data in time and space dimensions, so that the track data and meteorological data are more matched, further improving the accuracy of the predicted track. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Other features, objects and advantages of the present application will become more apparent by reading the detailed description of non-limiting embodiments made with reference to the following drawings: Figure 1 A schematic diagram of a flow chart of a track prediction method in one embodiment; Figure 2 Another schematic diagram of a flow chart of a track prediction method in one embodiment; Figure 3 Another schematic diagram of a flow chart of a track prediction method in one embodiment; Figure 4 Another schematic diagram of a flow chart of a track prediction method in one embodiment; Figure 5 Another schematic diagram of a flow chart of a track prediction method in one embodiment; Figure 6 is a structural block diagram of a track prediction device in one embodiment; Figure 7 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0017] The present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the relevant invention, rather than to limit the invention. It is also necessary to explain that, for ease of description, only the parts related to the invention are shown in the accompanying drawings.

[0018] It should be noted that, in the absence of conflict, the embodiments in the present application and the features in the embodiments can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments. In addition, the term "and / or" herein is merely a description of the association relationship of associated objects, indicating that three relationships may exist, for example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. The terms "first" and "second" in the description and claims of the embodiments of the present application are used to distinguish different objects, rather than to describe a specific order of objects.

[0019] As global maritime traffic becomes increasingly busy, the number of ships continues to increase, the risk of ship collisions and other accidents continues to increase, and the shipping industry's demand for safe navigation of ships is increasing. Therefore, predicting the track of ships has become an important means to ensure navigation safety and improve shipping efficiency.

[0020] Traditional track prediction methods mainly use the ship's historical track data recorded by the Automatic Identification System (AIS) to predict the ship's future track using time series models such as moving average models, Kalman filter models, and regression models. However, the marine environment is complex and changeable, especially under the influence of meteorological factors such as wind and waves, the actual navigation track of the ship often deviates from the scheduled route. Therefore, the predicted track obtained from the ship's historical track data has the problem of low accuracy.

[0021] In one embodiment, Figure 1 As shown, a track prediction method is provided. This embodiment uses the method applied to a terminal as an example for illustration. It can be understood that the method can also be applied to a server, and can also be applied to a system including a terminal and a server, and is implemented through the interaction between the terminal and the server. In this embodiment, the method includes the following steps: Step 101: Acquire the track data of a ship in a preset period before the current moment and at least one meteorological data set of the area where the ship is located.

[0022] Among them, the track data includes at least one of coordinate data (i.e. longitude and latitude), speed, heading, and coordinate changes; the meteorological data includes at least one of wind speed, wind direction, wave speed, wave direction, wave crest period, and wave height.

[0023] In the embodiments of the present application, the track data of the ship before the current moment and the meteorological data of the area where the ship is located are mainly used to predict the track of the ship after the current moment, or to predict the track of the ship within a certain period of time (for example, within one hour) after the current moment, or to predict the track of the ship within a certain distance (for example, 5 km) after the current moment.

[0024] During the voyage, the ship can record track data in real time at a certain frequency (for example, every 1 minute), such as the ship's coordinate data, speed, heading, etc., and record the time corresponding to each track data. The time corresponding to multiple track data is sorted from front to back, which is the time series corresponding to the track data, for example, it can be expressed as [ta, tb, tc...].

[0025] When a ship needs to predict its track during navigation, it can select track data within the prediction period before the current time from the real-time track data. For example, if the track data within half an hour before the current time is selected, 30 track data can be obtained in the end.

[0026] At the same time, the impact of weather on the ship's track is taken into account. For example, changes in wind speed and direction will cause the ship's speed and heading to deviate from expectations. Especially in strong winds, the ship needs to constantly adjust its heading to maintain stable navigation. Changes in wave height and wave speed will directly affect the ship's navigation resistance and attitude. In extreme environments, the ship may sway violently or its speed may drop sharply. The wave crest period and wave direction have a great impact on the ship's roll and pitch. Therefore, when predicting the track, it is necessary to obtain meteorological data in the area where the ship is located.

[0027] The area where the ship is located is an area within a certain range from the current position of the ship. This area should include the ship's track within a preset period of time before the current moment, as well as the ship's track within a certain period of time after the current moment (ie, the predicted track).

[0028] In the embodiment of the present application, the meteorological data observed by at least one meteorological observation point in the area where the ship is located can be obtained from the meteorological monitoring platform, and the meteorological data of each meteorological observation point constitutes a meteorological data group. Therefore, at least one meteorological data group in the area where the ship is located can be obtained. The meteorological data in each meteorological data group is observed by the meteorological observation point at different times. Therefore, the observation time corresponding to each meteorological data in the meteorological data group is sorted from front to back, which is the time series corresponding to the meteorological data group, for example, it can be expressed as [t1, t2, t3...].

[0029] In addition, the acquired meteorological data may be meteorological data within a preset period before the current moment.

[0030] In a possible implementation, after obtaining the track data and at least one meteorological data set, the missing values, outliers, etc. in the above data can be processed. Specifically, for missing data, linear interpolation can be used to complete the data; for outliers, outliers in the data can be detected by outlier detection method, for example, outlier detection method based on Z score: Z=(x-μ) / σ.

[0031] Here, x is the data, μ is the sample mean, and σ is the sample standard deviation.

[0032] If Z exceeds the set threshold, the data is considered an outlier. Finally, the detected outliers are removed.

[0033] Step 102: interpolate each meteorological data group according to the time series of the track data, and determine the meteorological data corresponding to the track of the ship within a preset time period before the current moment according to at least one meteorological data group after the interpolation process.

[0034] The time series is a sequence of data collection times arranged in a preset order. The time series of track data is a sequence of track data collection times arranged in a sequence from front to back or from back to front.

[0035] In the embodiment of the present application, considering that the frequency of meteorological data collection (for example, every half an hour) is often lower than the frequency of track data collection, when obtaining meteorological data within a certain time range, the amount of meteorological data is small, therefore, each meteorological data group can be interpolated to increase the amount of meteorological data. Specifically, for each meteorological data group, a meteorological data can be interpolated at each time point in the time series of the track data, so that the time series of the interpolated meteorological data group is unified with the time series of the track data.

[0036] Further considering that the location of the meteorological observation point does not match the ship track, that is, the meteorological data of the meteorological observation point does not match the meteorological conditions on the track, the meteorological data on the ship track can be further calculated based on at least one meteorological data group after interpolation processing, so that the final meteorological data matches the ship track.

[0037] That is, the acquired track data and at least one meteorological data group are aligned and unified in the time dimension and the space dimension, so that the predicted track based on the matched track data and meteorological data has a higher accuracy.

[0038] Step 103: input the track data of the ship in a preset period before the current moment and the meteorological data corresponding to the track into a pre-trained track prediction model to predict the ship's track and obtain the predicted track of the ship.

[0039] In an embodiment of the present application, after obtaining the matching track data and the meteorological data corresponding to the track, they can be input into a pre-trained track prediction model to predict the track of the ship after the current moment to obtain the predicted track of the ship.

[0040] The track prediction model is trained based on the historical track data of the ship and the meteorological data corresponding to the historical track. The meteorological data corresponding to the historical track is obtained based on the meteorological data of multiple meteorological observation points in the area where the historical track is located, and the historical track data is obtained after the time and space are uniformly processed in the above embodiment.

[0041] In one possible implementation, after obtaining the predicted track, the meteorological data on the predicted track can be calculated, and corresponding route optimization suggestions, risk avoidance strategies, etc. can be generated based on the meteorological data on the predicted track. For example, when the meteorological data on the predicted track indicates a bad weather environment, the ship's track is replanned.

[0042] Specifically, a rule set may be pre-set, which includes route optimization strategies corresponding to different weather data. After determining the weather data on the predicted track, the corresponding route optimization strategy is searched in the rule set and provided to the user. This method can achieve rapid response for simple weather conditions.

[0043] It is also possible to generate route optimization strategies corresponding to weather data through fuzzy logic control. This method provides a more accurate route optimization strategy for complex and changeable weather environments.

[0044] The method provided in the embodiment of the present application can first obtain the track data of the ship within a preset time period before the current moment and at least one meteorological data group in the area where the ship is located. Then, the track data and meteorological data are unified in time series through interpolation processing, and then the meteorological data on the ship's track is calculated through each meteorological data group, and the track data and meteorological data are unified in spatial position. Finally, the unified track data and meteorological data are input into a pre-trained track prediction model, and the predicted track is output. In the embodiment of the present application, considering the influence of meteorological factors on the track, the ship's track is predicted based on the track data and meteorological data before the current moment, which effectively improves the accuracy of the predicted track. Moreover, before predicting the track, the present application aligns and unifies the track data and meteorological data in time and space dimensions, so that the track data and meteorological data are more matched, further improving the accuracy of the predicted track.

[0045] The above-mentioned embodiment introduces a solution for time unification of track data and meteorological data. In another embodiment of the present application, the meteorological data can be interpolated based on each time point in the time series of the track data to achieve time unification of the track data and meteorological data. This embodiment includes the following: Figure 2 Steps shown: Step 201: For each meteorological data group, search in the time series of the meteorological data group according to the first time point in the time series of the track data.

[0046] Step 202: If there is no time point matching the first time point in the time series of the meteorological data group, determine the meteorological data corresponding to the first time point based on the meteorological data corresponding to at least one time point adjacent to the first time point in the meteorological data group.

[0047] Step 203: Update the meteorological data set according to the meteorological data corresponding to the first time point until the meteorological data set is updated according to the time points in the time series of the track data, thereby obtaining the meteorological data set after interpolation processing.

[0048] In the embodiment of the present application, each meteorological data group is interpolated based on the track data. In the process of interpolating any meteorological data group, each time point in the time series of the track data can be traversed first, and each time point can be used as the first time point to search in the time series of the meteorological data group. For example, the time series of the track data [ta, tb, tc...] is traversed, and the first time point ta is used as the first time point to search in the time series of the meteorological data group [t1, t2, t3...].

[0049] If there is a time point matching the first time point in [t1, t2, t3…], it indicates that the acquired data contains the track data and weather data corresponding to the time point. Therefore, the next time point in [ta, tb, tc…] is traversed, and the time point is used as the first time point to continue searching in [t1, t2, t3…]. For example, if there is a time point identical to ta in [t1, t2, t3…], the next time point tb of ta in [ta, tb, tc…] is used as the first time point to continue searching in [t1, t2, t3…].

[0050] If there is no time point matching the first time point in [t1, t2, t3…], indicating that there is no meteorological data corresponding to the time point in the acquired data, then at least one time point adjacent to the first time point is searched in [t1, t2, t3…], that is, at least one time point with the smallest time interval with the first time point, and the meteorological data corresponding to the first time point is determined based on the meteorological data corresponding to the at least one time point found, and the meteorological data corresponding to the first time point is added to the meteorological data group. For example, if there is no time point identical to tb in [t1, t2, t3…], then the time point with the smallest time interval with tb is found to be t2 in [t1, t2, t3…], and the meteorological data corresponding to t2 is used as the meteorological data corresponding to tb and added to the meteorological data group; or the time points with the smallest time interval with tb are found to be t1 and t2, and the meteorological data corresponding to tb is calculated based on the meteorological data corresponding to t1 and t2, and added to the meteorological data group.

[0051] Next, traverse [ta, tb, tc…], take the next time point of tb as the first time point, perform time point matching, interpolation, update and other processing on the meteorological data group, until all time points in [ta, tb, tc…] are traversed and the update of the meteorological data group is completed, then the meteorological data group after interpolation processing is obtained.

[0052] In one possible implementation, after obtaining the interpolated meteorological data group, each time point in the time series of the interpolated meteorological data group can be traversed, and the meteorological data corresponding to the time points that do not match the time series of the track data can be discarded, so that the track data and the meteorological data match in both time and quantity.

[0053] In a possible implementation, the meteorological data corresponding to the first time point can be calculated based on the meteorological data corresponding to two time points adjacent to the first time point in [t1, t2, t3...]. This embodiment includes the following: Figure 3 Steps shown: Step 301: Determine, according to the first time point, a second time point and a third time point in the time series of the meteorological data group that are adjacent to the first time point.

[0054] Step 302: determine the meteorological change parameters in the time period between the second time point and the third time point according to the meteorological data corresponding to the second time point and the meteorological data corresponding to the third time point.

[0055] Step 303: Calculate the meteorological data corresponding to the first time point based on the meteorological change parameter and the meteorological data corresponding to the second time point.

[0056] In the embodiment of the present application, the time points before and after the first time point and adjacent to the first time point are searched in the time series corresponding to the meteorological data group to obtain the second time point and the third time point, wherein the second time point is less than the third time point.

[0057] Then, the difference between the meteorological data corresponding to the second time point and the meteorological data corresponding to the third time point is calculated to obtain the meteorological changes between the second time point and the third time point; then, the time interval between the second time point and the third time point is calculated, and the quotient of the meteorological changes between the second time point and the third time point and the time interval is the meteorological change parameter in the time period between the second time point and the third time point.

[0058] Finally, the meteorological changes between the first time point and the second time point are calculated based on the meteorological change parameters, and the sum of the meteorological data corresponding to the second time point and the meteorological changes between the first time point and the second time point is the meteorological data corresponding to the first time point.

[0059] The above calculation process can be expressed as follows:

[0060] Among them, t is the first time point, t1 is the second time point, t2 is the third time point, f(t1) is the meteorological data corresponding to the second time point, f(t2) is the meteorological data corresponding to the third time point, and f(t) is the meteorological data corresponding to the first time point.

[0061] It is understandable that the meteorological changes between the first time point and the third time point can also be calculated based on the meteorological change parameters, and the difference between the meteorological data corresponding to the third time point and the meteorological changes between the first time point and the third time point is the meteorological data corresponding to the first time point.

[0062] The method provided in the embodiment of the present application can search for each meteorological data group in the time series of the meteorological data group according to the first time point in the time series of the track data, and complete the meteorological data group based on the first time point, so that the track data and the meteorological data are unified in the time dimension, so that the predicted track based on the unified track data and meteorological data is more accurate.

[0063] The above-mentioned embodiment introduces a solution for spatially unifying the track data and the meteorological data. In another embodiment of the present application, the meteorological data on the track can be determined based on the coordinate information of the track data and the coordinate information of the meteorological observation point corresponding to the meteorological data. This embodiment includes the following: Figure 4 Steps shown: Step 401: for each coordinate in the track data, at least one candidate meteorological data group whose meteorological observation point position is within a preset range of coordinates is selected from at least one interpolated meteorological data group according to the position of the meteorological observation point corresponding to each meteorological data group.

[0064] Step 402: determine the meteorological data corresponding to the coordinates according to at least one candidate meteorological data group, and obtain the meteorological data corresponding to the track according to the meteorological data corresponding to each coordinate.

[0065] In an embodiment of the present application, for each coordinate information in the track data, that is, each coordinate information on the track within a preset time period before the current moment, the distance between the coordinate information and each meteorological observation point is calculated based on the coordinate information and the position information of the meteorological observation point corresponding to at least one meteorological data group.

[0066] Then, at least one meteorological observation point within the preset range of the coordinate information is selected, and the interpolated meteorological data group corresponding to the selected meteorological observation point is used as a candidate meteorological data group, and the meteorological data at the coordinate information on the track is calculated based on the candidate meteorological data group. Alternatively, at least one meteorological observation point is sorted in order from near to far in terms of distance from the coordinate on the track, and a preset number of meteorological observation points ranked first are selected, and the interpolated meteorological data group corresponding to the selected meteorological observation point is used as a candidate meteorological data group, and the meteorological data at the coordinate information on the track is calculated based on the candidate meteorological data group. The meteorological data at each coordinate information on the track is calculated, that is, the meteorological data corresponding to the track within the preset period before the current moment is obtained.

[0067] When calculating the meteorological data at the coordinates on the track based on multiple candidate meteorological data groups, a bilinear interpolation method can be used. For example, when calculating the meteorological data at the coordinates on the track based on four candidate meteorological data groups, the calculation process can be expressed as follows:

[0068] Among them, (x1, y1), (x2, y2), (x3, y3), and (x4, y4) are the location information of the meteorological observation points corresponding to the four candidate meteorological data groups; f(x1, y1), f(x2, y2), f(x3, y3), and f(x4, y4) are the four candidate meteorological data groups; f(x, y) is the meteorological data at the coordinates on the track.

[0069] In a possible implementation, after screening at least one meteorological observation point within a preset range of any coordinate information on the track, the meteorological data corresponding to the at least one meteorological observation point may be weighted averaged to obtain the meteorological data at the coordinate.

[0070] In one possible implementation, after calculating the distance between any coordinate on the track and each meteorological observation point, the meteorological observation point closest to the coordinate can be selected, and the meteorological data corresponding to the meteorological observation point can be directly used as the meteorological data at the coordinate on the track.

[0071] The method provided in the embodiment of the present application can determine the candidate meteorological data group corresponding to each coordinate on the track, and determine the meteorological data corresponding to the coordinate based on the candidate meteorological data group, thereby obtaining the meteorological data corresponding to the track. The embodiment of the present application calculates the meteorological data at each coordinate on the track, unifies the track data and the meteorological data in the spatial dimension, improves the accuracy of the meteorological data used to predict the track, and thus effectively improves the accuracy of the predicted track.

[0072] In one embodiment, a training scheme for a track prediction model is provided, which includes: Figure 5 Steps shown: Step 501: Acquire historical track data of a ship and at least one historical meteorological data set of an area where the historical track of the ship is located.

[0073] In an embodiment of the present application, the historical track data of the ship can be obtained from the AIS database through the ship identification, such as the Maritime Mobile Service Identity (MMSI) code; and the historical meteorological data group observed by at least one meteorological observation point in the area where the historical track of the ship is located is obtained from the meteorological monitoring platform.

[0074] Step 502: interpolate each historical meteorological data group according to the time series of the historical track data, and determine the meteorological data corresponding to the historical track of the ship according to at least one historical meteorological data group after the interpolation process.

[0075] In the embodiment of the present application, the process of interpolating each historical meteorological data group according to the time series of the historical track data, and determining the meteorological data corresponding to the historical track of the ship according to at least one historical meteorological data group after the interpolation process, refer to the above embodiment, and the present application will not repeat them here.

[0076] Step 503: group the historical track data and the meteorological data corresponding to the historical track according to the time series to obtain a plurality of historical track data groups.

[0077] In an embodiment of the present application, the historical track data and the meteorological data corresponding to the historical track can be grouped in time series through the same sliding time window.

[0078] For example, the five historical track data corresponding to the first time point to the fifth time point in the time series are grouped into one group, and the five historical track data corresponding to the second time point to the sixth time point are grouped into one group, until the historical track data are grouped, and multiple historical track data groups are obtained. The five meteorological data corresponding to the first time point to the fifth time point in the time series are grouped into one group, and the five meteorological data corresponding to the second time point to the sixth time point are grouped into one group, until the meteorological data corresponding to the historical track are grouped, and multiple sub-meteorological data groups are obtained.

[0079] Among them, the historical track data group and the sub-meteorological data group correspond one to one.

[0080] Alternatively, the historical track data and the meteorological data corresponding to the historical track are grouped and processed according to time series based on sliding time windows of various sizes, so that the grouped data are richer.

[0081] Step 504: use at least one historical track data group that is sorted in order from front to back among multiple historical track data groups, and the meteorological data corresponding to the historical track as training samples, and use at least one historical track data group that is sorted in the back as a test sample, to train the initial track prediction model until the accuracy of the initial track prediction model reaches a preset accuracy, thereby obtaining a track prediction model.

[0082] In the embodiment of the present application, part of the grouped historical track data group is selected as a training sample, and the meteorological data corresponding to the historical track, or the sub-meteorological data group corresponding to the selected part of the historical track data is also used as a training sample, and the remaining part of the historical track data group is used as a test sample. When selecting samples, the historical track data group with the earlier time is used as a training sample, and the historical track data group with the later time is used as a test sample in the order of time.

[0083] The initial track prediction model is trained based on the training samples and the test samples until the accuracy of the initial track prediction model reaches a preset accuracy, thereby obtaining a track prediction model.

[0084] Among them, the initial training model can adopt a deep learning model based on the self-attention mechanism, such as the Transformer model, to ensure the processing capability of multi-dimensional and highly complex data, thereby ensuring the accuracy of the predicted track. In addition, the use of a parallel computing mechanism has the advantages of ensuring the processing efficiency of large-scale data, accelerating the model training process, and reducing the computational complexity, thereby achieving real-time prediction of the track.

[0085] In a possible implementation, models such as Long Short-Term Memory (LSTM), Gate Recurrent Unit (GRU), CNN-LSTM hybrid model, CNN-Transformer hybrid model, ensemble learning, and stacking models can also be used as initial track prediction models. LSTM has better processing performance for the temporal relationship between different types of data; GRU has lower computational complexity; CNN-LSTM hybrid model can effectively process the temporal relationship between different types of data while effectively extracting complex features; CNN-Transformer hybrid model can ensure the extraction effect of spatial features of data and effectively process large-scale data; ensemble learning (such as random forest, XGBoost, etc.) can combine multiple different types of prediction models, integrate the prediction results of each model, and form complementarity between different models, thereby improving the overall prediction effect; the stacking model uses the prediction results of multiple different types of models (such as LSTM, GRU, Transformer, etc.) as input, and then uses a high-level model (such as logistic regression, decision tree, etc.) for final prediction, which can effectively combine the advantages of each basic model and improve prediction accuracy.

[0086] In one possible implementation, the training effect of the track prediction model can be verified by a loss function. For example, the mean square error (MSE) is used as the loss function of the track prediction model:

[0087] in, is the actual value, is the predicted value, is the sample size.

[0088] When the MSE is less than the preset threshold, the training of the track prediction model is completed.

[0089] In one possible implementation, after obtaining the historical track data of the ship and determining the meteorological data corresponding to the historical track, the historical track data and the meteorological data corresponding to the historical track can be further feature extracted to obtain the ship's motion status, the geographical features of the sea area where the ship is located (port location, island distribution, etc.), meteorological changes and other features, and the extracted features are added to the training samples to train the initial track prediction model. This ensures the richness of the training samples and further improves the accuracy of the track prediction model.

[0090] Among them, the above features can be extracted by using recursive neural network (RNN), aggregation function, statistical model, etc. Aggregation function can smooth data fluctuations and reduce the impact of noise on prediction, thereby reducing randomness in the data and improving the stability of the model; statistical models (such as principal component analysis PCA) can reduce the dimension of data, extract key features, and reduce redundant information to ensure that the model effectively uses important information in the data and improve the training efficiency of the model.

[0091] In one possible implementation, in order to ensure the consistency of track data and meteorological data, the historical track data and the meteorological data corresponding to the historical track can be normalized, and a track prediction model can be trained based on the normalized historical track data and the meteorological data corresponding to the normalized historical track.

[0092] The normalization process can be expressed as follows:

[0093] in, is the original data, and are the maximum and minimum values ​​in the data set, respectively. is the normalized value.

[0094] Accordingly, when using the trained track prediction model to predict the ship's track in real time, the track data within the preset period before the current time of the ship and the meteorological data corresponding to the track should be normalized before prediction. After obtaining the predicted track, the coordinate data in the predicted track should be inversely normalized:

[0095] in, is the actual value after inverse normalization, is the normalized value predicted by the model, and are the minimum and maximum values ​​used in the normalization process.

[0096] The method provided in the embodiment of the present application takes into account the influence of meteorological factors on the track, and trains the track prediction model based on the historical track data and the meteorological data corresponding to the historical track, which effectively improves the accuracy of the track prediction model. Moreover, before training the track prediction model, the present application aligns and unifies the historical track data and the historical meteorological data in the time and space dimensions, so that the track data and the meteorological data are more matched, further improving the accuracy of the track prediction model.

[0097] It should be noted that although the operations of the method of the present invention are described in a particular order in the accompanying drawings, this does not require or imply that the operations must be performed in this particular order, or that all the operations shown must be performed to achieve the desired results. On the contrary, the steps depicted in the flow chart can be performed in a different order. Additionally or alternatively, some steps can be omitted, multiple steps can be combined into one step, and / or one step can be decomposed into multiple steps.

[0098] In one embodiment, as shown in the figure, a track prediction device is provided, the device includes: an acquisition module 601, an interpolation module 602 and a prediction module 603, wherein: An acquisition module 601 is used to acquire the track data of the ship in a preset period before the current moment and at least one meteorological data group of the area where the ship is located; The interpolation module 602 is used to perform interpolation processing on each meteorological data group according to the time series of the track data, and determine the meteorological data corresponding to the track of the ship in a preset period before the current time according to at least one meteorological data group after the interpolation processing; the time series is a sequence of data collection moments arranged in a preset order; The prediction module 603 is used to input the track data of the ship in a preset time period before the current moment and the meteorological data corresponding to the track into a pre-trained track prediction model to predict the ship's track and obtain the predicted track of the ship; the track prediction model is trained based on the historical track data of the ship and the meteorological data corresponding to the historical track.

[0099] In one embodiment, the interpolation module 602 is specifically used to search in the time series of the meteorological data group according to the first time point in the time series of the track data for each meteorological data group; if there is no time point matching the first time point in the time series of the meteorological data group, the meteorological data corresponding to the first time point is determined based on the meteorological data corresponding to at least one time point adjacent to the first time point in the meteorological data group; the meteorological data group is updated according to the meteorological data corresponding to the first time point until the meteorological data group is updated according to the time points in the time series of the track data, thereby obtaining the interpolated meteorological data group.

[0100] In one embodiment, the interpolation module 602 is also used to determine the second time point and the third time point adjacent to the first time point in the time series of the meteorological data group based on the first time point; determine the meteorological change parameters in the time period between the second time point and the third time point based on the meteorological data corresponding to the second time point and the meteorological data corresponding to the third time point; and calculate the meteorological data corresponding to the first time point based on the meteorological change parameters and the meteorological data corresponding to the second time point.

[0101] In one embodiment, the interpolation module 602 is also used to select, for each coordinate in the track data, at least one candidate meteorological data group whose meteorological observation point position is within a preset range of the coordinates from at least one interpolated meteorological data group according to the position of the meteorological observation point corresponding to each meteorological data group; determine the meteorological data corresponding to the coordinate according to at least one candidate meteorological data group, and obtain the meteorological data corresponding to the track according to the meteorological data corresponding to each coordinate.

[0102] In one embodiment, the track prediction device also includes a model training module 604, which is specifically used to obtain the historical track data of the ship and at least one historical meteorological data group of the area where the historical track of the ship is located; interpolate each historical meteorological data group according to the time series of the historical track data, and determine the meteorological data corresponding to the historical track of the ship according to at least one historical meteorological data group after the interpolation process; group the historical track data and the meteorological data corresponding to the historical track according to the time series to obtain multiple historical track data groups; use at least one historical track data group that is sorted in front after being sorted from front to back in time among the multiple historical track data groups, and the meteorological data corresponding to the historical track as training samples, and use at least one historical track data group that is sorted in the back as a test sample, and train the initial track prediction model until the accuracy of the initial track prediction model reaches a preset accuracy to obtain a track prediction model.

[0103] In one embodiment, the track prediction device also includes a normalization module 605, which is specifically used to normalize the historical track data and the meteorological data corresponding to the historical track, and train a track prediction model based on the normalized historical track data and the meteorological data corresponding to the normalized historical track.

[0104] It should be understood that the units or modules recorded in the track prediction device are similar to those in the reference Figure 1 The various steps in the described method correspond to each other. Therefore, the operations and features described above for the method are also applicable to the track prediction device and the units contained therein, and will not be repeated here. The track prediction device can be pre-implemented in a browser or other security application of an electronic device, or loaded into a browser or its security application of an electronic device by downloading or the like. The corresponding units in the track prediction device can cooperate with the units in the electronic device to implement the solution of the embodiment of the present application.

[0105] Reference below Figure 7 , which shows a schematic diagram of the structure of a computer system 700 of a terminal device or server suitable for implementing an embodiment of the present application.

[0106] like Figure 7As shown, the computer system 700 includes a central processing unit (CPU) 701, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 702 or a program loaded from a storage part 708 into a random access memory (RAM) 703. In the RAM 703, various programs and data required for the operation of the system 700 are also stored. The CPU 701, the ROM 702, and the RAM 703 are connected to each other through a bus 704. An input / output (I / O) interface 705 is also connected to the bus 704.

[0107] The following components are connected to the I / O interface 705: an input section 706 including a keyboard, a mouse, etc.; an output section 707 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 708 including a hard disk, etc.; and a communication section 709 including a network interface card such as a LAN card, a modem, etc. The communication section 709 performs communication processing via a network such as the Internet. A drive 710 is also connected to the I / O interface 705 as needed. A removable medium 711, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 710 as needed, so that a computer program read therefrom is installed into the storage section 708 as needed.

[0108] In particular, according to the embodiments of the present disclosure, the above reference Figure 1 The described processes may be implemented as computer software programs. For example, embodiments of the present disclosure include a computer program product comprising a computer program tangibly embodied on a machine-readable medium, the computer program comprising a computer program for executing Figure 1 In such an embodiment, the computer program may be downloaded and installed from a network via the communication section 709 and / or installed from the removable medium 711.

[0109] It should be noted that the computer-readable medium shown in the present application may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device. In the present application, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium such as a computer-readable storage medium that can send, propagate or transmit a program for use by or in conjunction with an instruction execution system, device or device. The program code contained on the computer-readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, optical cable, RF, etc., or any suitable combination of the above.

[0110] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present invention. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, program segment or a part of a code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a different order from the order marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0111] The units or modules involved in the embodiments described in the present application may be implemented by software or hardware. The units or modules described may also be arranged in a processor. The names of these units or modules do not, in some cases, constitute limitations on the units or modules themselves.

[0112] As another aspect, the present application further provides a computer-readable storage medium, which may be included in the computer device described in the above embodiment, or may exist independently without being installed in the computer device. The above computer-readable storage medium stores one or more programs, and when the above programs are used by one or more processors to execute the method described in the present application. For example, it can be executed Figure 1~Figure 5 The individual steps of the method are shown.

[0113] The present application embodiment provides a computer program product, which includes instructions. When the instructions are executed, the method described in the present application embodiment is executed. For example, Figure 1~Figure 5 The individual steps of the method are shown.

[0114] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0115] The above description is only a preferred embodiment of the present application and an explanation of the technical principles used. Those skilled in the art should understand that the scope of the invention involved in the present application is not limited to the technical solution formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the inventive concept. For example, the above features are replaced with (but not limited to) technical features with similar functions disclosed in the present application.

Claims

1. A track prediction method, characterized in that: The method comprises: Acquire the track data of the ship within a preset period before the current moment and at least one meteorological data group of the area where the ship is located; Performing interpolation processing on each meteorological data group according to the time series of the track data, and determining the meteorological data corresponding to the track of the ship within a preset period before the current moment according to at least one meteorological data group after the interpolation processing; the time series is a sequence of data collection moments arranged in a preset order; The track data of the ship in a preset time period before the current moment and the meteorological data corresponding to the track are input into a pre-trained track prediction model to predict the track of the ship and obtain the predicted track of the ship; the track prediction model is trained based on the historical track data of the ship and the meteorological data corresponding to the historical track.

2. The method according to claim 1, characterized in that The interpolation processing of each meteorological data group according to the time series of the track data comprises: For each meteorological data group, searching in the time series of the meteorological data group according to the first time point in the time series of the track data; If there is no time point matching the first time point in the time series of the meteorological data group, determining the meteorological data corresponding to the first time point according to the meteorological data corresponding to at least one time point adjacent to the first time point in the meteorological data group; The meteorological data group is updated according to the meteorological data corresponding to the first time point until the meteorological data group is updated according to the time points in the time series of the track data, thereby obtaining the meteorological data group after interpolation processing.

3. The method according to claim 2, characterized in that The determining the meteorological data corresponding to the first time point according to the meteorological data corresponding to at least one time point adjacent to the first time point in the meteorological data group includes: Determine, according to the first time point, a second time point and a third time point in the time series of the meteorological data group that are adjacent to the first time point; Determine a meteorological change parameter in a time period between the second time point and the third time point according to the meteorological data corresponding to the second time point and the meteorological data corresponding to the third time point; The meteorological data corresponding to the first time point is calculated based on the meteorological change parameter and the meteorological data corresponding to the second time point.

4. The method according to claim 1, characterized in that: The step of determining the meteorological data corresponding to the track of the ship within a preset period of time before the current moment according to at least one meteorological data group after interpolation processing includes: For each coordinate in the track data, according to the position of the meteorological observation point corresponding to each meteorological data group, at least one candidate meteorological data group whose meteorological observation point position is within a preset range of the coordinates is selected from the at least one interpolated meteorological data group; The meteorological data corresponding to the coordinates are determined according to the at least one candidate meteorological data group, and the meteorological data corresponding to the track are obtained according to the meteorological data corresponding to each of the coordinates.

5. The method according to claim 1, characterized in that: The training process of the track prediction model includes: Acquire historical track data of the ship and at least one historical meteorological data set of the area where the historical track of the ship is located; Performing interpolation processing on each historical meteorological data group according to the time series of the historical track data, and determining meteorological data corresponding to the historical track of the ship according to at least one historical meteorological data group after the interpolation processing; The historical track data and the meteorological data corresponding to the historical track are grouped and processed according to a time series to obtain a plurality of historical track data groups; At least one historical track data group that is sorted in front of multiple historical track data groups from front to back in time, and meteorological data corresponding to the historical track are used as training samples, and at least one historical track data group that is sorted in the back is used as a test sample to train the initial track prediction model until the accuracy of the initial track prediction model reaches a preset accuracy, thereby obtaining the track prediction model.

6. The method according to claim 5, characterized in that The method further comprises: The historical track data and the meteorological data corresponding to the historical track are normalized, and the track prediction model is obtained by training based on the normalized historical track data and the meteorological data corresponding to the normalized historical track.

7. A track prediction device, characterized in that: The device comprises: An acquisition module, used to acquire the track data of the ship within a preset period before the current moment and at least one meteorological data group of the area where the ship is located; An interpolation module, used to perform interpolation processing on each meteorological data group according to the time series of the track data, and determine the meteorological data corresponding to the track of the ship in a preset period before the current moment according to at least one meteorological data group after the interpolation processing; the time series is a sequence of data collection moments arranged in a preset order; The prediction module is used to input the track data of the ship in a preset time period before the current moment and the meteorological data corresponding to the track into a pre-trained track prediction model to predict the track of the ship and obtain the predicted track of the ship; the track prediction model is trained based on the historical track data of the ship and the meteorological data corresponding to the historical track.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images