A multi-source ocean observation data fusion system and method

Abstract

The embodiments of the present specification provide a multi-source ocean observation data fusion system and method, the method comprising: obtaining basic information of multiple ocean observation points and a sharing request for target observation data, the basic information comprising at least one of location information, attribute information and data collection information of the ocean observation points; grouping the multiple ocean observation points based on the basic information to obtain at least two observation groups; for each of the at least two observation groups, fusing the observation data uploaded by at least one ocean observation point contained in the observation group to obtain fused observation data; grouping and storing the fused observation data, and establishing an index for the fused observation data of the at least two observation groups; determining a target observation group from the at least two observation groups based on the sharing request, the target observation group being the observation group to which the source containing the target observation data belongs; and retrieving the target observation data corresponding to the sharing request based on the index.

Description

Technical Field

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

[0002] Marine observation data plays a very important role in maintaining marine rights and interests, developing marine resources, warning marine disasters, protecting the marine environment, strengthening marine construction, and seeking new marine development space. The marine data management platform can collect marine observation data from different marine observation points (for example, marine observation stations, buoys, shore-based radars, observation equipment, ships, etc.), and then provide real-time data support for disaster prevention and mitigation services. Since marine observation data from different sources are generally small file data and the number is huge, the efficiency of the marine data management platform in processing marine observation data (for example, receiving, storing, sharing, etc.) will decrease as the number of small files continues to increase. At the same time, when there are more data sharing needs, the real-time nature of data sharing may be affected. In addition, the transmission and sharing of marine observation data may also be affected by various situations (for example, network fluctuations), further affecting the real-time nature of data sharing.

[0003] Therefore, it is hoped to provide a multi-source ocean observation data fusion system and method, which can improve the processing efficiency of a large amount of ocean observation data from multiple sources and ensure the real-time nature of data sharing. Summary of the invention

[0004] One or more embodiments of the present specification provide a multi-source ocean observation data fusion method, which is executed by an ocean data management platform, including: obtaining basic information of multiple ocean observation points and a sharing request for target observation data, the basic information including at least one of location information, attribute information and data collection information of the ocean observation points; grouping the multiple ocean observation points based on the basic information to obtain at least two observation groups; for each of the at least two observation groups, fusing the observation data uploaded by at least one ocean observation point included in the observation group to obtain fused observation data; grouping and storing the fused observation data, and establishing an index for the fused observation data of the at least two observation groups; determining a target observation group from the at least two observation groups based on the sharing request, the target observation group being the observation group to which the source containing the target observation data belongs; and retrieving the target observation data corresponding to the sharing request based on the index.

[0005] One or more embodiments of the embodiments of the present specification provide a multi-source ocean observation data fusion system, the system comprising: an acquisition module, configured to acquire basic information of multiple ocean observation points and a sharing request for observation data, the basic information comprising at least one of location information, attribute information and data acquisition information of the ocean observation points; a module configured to group the multiple ocean observation points based on the basic information to obtain at least two observation groups, group and store the fused observation data, and establish an index for the fused observation data of the at least two observation groups; a target determination module, configured to determine a target observation group from the at least two observation groups based on the sharing request, the target observation group being the observation group to which the source of the target observation data belongs; and a retrieval module, configured to retrieve the observation data corresponding to the target observation group.

[0006] One or more embodiments of the present specification provide a multi-source ocean observation data fusion device, the device comprising at least one processor and at least one memory, the at least one memory being used to store computer instructions; the at least one processor being used to execute a multi-source ocean observation data fusion method as described in any embodiment of the present specification.

[0007] One or more embodiments of the present specification provide a computer-readable storage medium, wherein the storage medium stores computer instructions. When a computer reads the computer instructions in the storage medium, the computer executes a multi-source ocean observation data fusion method as described in any embodiment of the present specification. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] This specification will be further described in the form of exemplary embodiments, which will be described in detail by the accompanying drawings. These embodiments are not restrictive, and in these embodiments, the same number represents the same structure, wherein:

[0009] Figure 1 is a module diagram of a multi-source ocean observation data fusion system according to some embodiments of this specification;

[0010] Figure 2 is an exemplary flow chart of multi-source ocean observation data fusion according to some embodiments of this specification;

[0011] Figure 3 is an exemplary schematic diagram of grouping multiple ocean observation points based on a prediction model according to some embodiments of this specification;

[0012] Figure 4 is an exemplary schematic diagram of obtaining observation data according to some embodiments of this specification;

[0013] Figure 5It is an exemplary schematic diagram of the observation map shown in some embodiments of the present specification. DETAILED DESCRIPTION

[0014] In order to more clearly illustrate the technical solutions of the embodiments of this specification, the following is a brief introduction to the drawings required for the description of the embodiments. Obviously, the drawings described below are only some examples or embodiments of this specification. For ordinary technicians in this field, this specification can also be applied to other similar scenarios based on these drawings without creative work. Unless it is obvious from the language environment or otherwise explained, the same reference numerals in the figures represent the same structure or operation.

[0015] It should be understood that the "system", "device", "unit" and / or "module" used herein are a method for distinguishing different components, elements, parts, portions or assemblies at different levels. However, if other words can achieve the same purpose, the words can be replaced by other expressions.

[0016] As shown in this specification and claims, unless the context clearly indicates an exception, the words "a", "an", "an" and / or "the" do not refer to the singular and may also include the plural. Generally speaking, the terms "comprise" and "include" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list. The method or device may also include other steps or elements.

[0017] Flowcharts are used in this specification to illustrate the operations performed by the system according to the embodiments of this specification. It should be understood that the preceding or following operations are not necessarily performed precisely in order. Instead, the steps may be processed in reverse order or simultaneously. At the same time, other operations may also be added to these processes, or one or more operations may be removed from these processes.

[0018] In order to improve the processing efficiency of sharing requests for ocean observation data and ensure the real-time nature of data sharing, some embodiments of the present specification provide a multi-source ocean observation data fusion system, method, device and storage medium. When a sharing request is received, the observation data are grouped based on the basic information of multiple ocean observation points to obtain at least two observation groups, and the observation data uploaded by the ocean observation points included in the observation groups are fused to obtain fused observation data. The fused observation data are grouped and stored and indexed. Based on the sharing request, a target observation group is determined from the at least two observation groups, thereby retrieving the target observation data corresponding to the sharing request based on the index, thereby reducing the amount of data to be searched and improving data processing efficiency.

[0019] Figure 1 It is a module diagram of a multi-source ocean observation data fusion system according to some embodiments of this specification.

[0020] like Figure 1 As shown, in some embodiments, the multi-source ocean observation data fusion system 100 may include an acquisition module 110 , a grouping module 120 , a target determination module 130 and a retrieval module 140 .

[0021] In some embodiments, the acquisition module 110 is configured to acquire basic information of multiple ocean observation points and a sharing request for observation data. In some embodiments, the basic information may include at least one of location information, attribute information, and data collection information of the ocean observation points.

[0022] In some embodiments, the grouping module 120 is configured to group multiple ocean observation points based on basic information to obtain at least two observation groups; for each of the at least two observation groups, the observation data uploaded by at least one ocean observation point included in the observation group is integrated to obtain integrated observation data; the integrated observation data is grouped and stored, and an index is established for the integrated observation data of at least two observation groups. In some embodiments, the grouping module 120 can be further configured to determine the real-time requirements of multiple ocean observation points based on historical sharing requests; and to group multiple ocean observation points based on basic information and real-time requirements. In some embodiments, the grouping module 120 can be further configured to determine the future real-time requirements of multiple ocean observation points based on real-time requirements, future meteorological information, observation data of key observation points, and observation point maps; and to group multiple ocean observation points based on basic information, real-time requirements, and future real-time requirements.

[0023] In some embodiments, the target determination module 130 is configured to determine a target observation group from at least two observation groups based on the sharing request. In some embodiments, the target observation group is an observation group to which a source including target observation data belongs.

[0024] In some embodiments, the retrieval module 140 is configured to retrieve the target observation data corresponding to the sharing request based on the index.

[0025] In some embodiments, the multi-source ocean observation data fusion system 100 may further include a transmission module (not shown in the figure). In some embodiments, for one of the multiple ocean observation points, the transmission module may be configured to determine the estimated change information of the ocean observation point based on a preset acquisition rule; based on the estimated change information of the ocean observation point, determine the transmission parameters of the ocean observation point, and the transmission parameters are used to control the acquisition device to transmit the collected observation data to the ocean data management platform. In some embodiments, the transmission module may be further configured to adjust the preset acquisition rules based on the associated sharing request of the ocean observation point; based on the adjusted preset acquisition rules, determine the estimated change information of the ocean observation point.

[0026] For more information about the acquisition module 110, the grouping module 120, the target determination module 130 and the retrieval module 140, see Figure 2-Figure 4 Related description.

[0027] It should be understood that Figure 1 The system and its modules shown can be implemented in various ways. For example, in some embodiments, the multi-source ocean observation data fusion system 100 can be implemented by an ocean data management platform or a processor. In some embodiments, the acquisition module 110, the grouping module 120, the target determination module 130 and the retrieval module 140 can be integrated on the ocean data management platform. The ocean data management platform is a platform for collecting, aggregating, storing and managing the observation data of all ocean observation points. When other devices send a sharing request to the management platform, the ocean data management platform will return the corresponding observation data to the device that issued the sharing request. In some embodiments, the ocean data management platform may include a storage device to store the observation data or other information of the ocean observation point (for example, the basic information of the ocean observation point).

[0028] It should be noted that the above description of the multi-source ocean observation data fusion system 100 and its modules is only for convenience of description and does not limit the present specification to the scope of the embodiments. It is understandable that, after understanding the principle of the system, those skilled in the art may arbitrarily combine the modules or form a subsystem to connect with other modules without deviating from the principle. In some embodiments, Figure 1 The acquisition module, grouping module, target determination module and retrieval module disclosed in the specification can be different modules in a system, or a module can realize the functions of two or more modules mentioned above. For example, each module can share a storage module, or each module can have its own storage module. Such variations are all within the scope of protection of this specification.

[0029] Figure 2 is an exemplary flow chart of a multi-source ocean observation data fusion method according to some embodiments of this specification. Figure 2 As shown, the process 200 includes the following steps. In some embodiments, the process 200 may be executed by the multi-source ocean observation data fusion system 100 or the ocean data management platform (referred to as the management platform for short).

[0030] Step 210, obtaining basic information of multiple ocean observation points and a sharing request for target observation data.

[0031] An ocean observation point is a point where the ocean is observed. A variety of acquisition equipment may be set up at an ocean observation point to observe the ocean conditions at the corresponding position of the point and a certain area around it. The acquisition equipment is a device that collects observation data at an ocean observation point. The acquisition equipment may include equipment for sampling, measuring, observing, and analyzing ocean data. In some embodiments, the acquisition equipment may use various technologies such as computing technology, remote sensing technology, hydroacoustic technology, laser technology, and anti-corrosion technology to collect observation data. The relevant description of the above acquisition equipment is for illustrative purposes only and is not intended to limit the scope of this specification.

[0032] In some embodiments, depending on the observation method, the ocean observation points may include ocean observation stations, buoys, submerged buoys, shore-based radars, observation equipment, ships, etc. The multiple ocean observation points in some embodiments of this specification may be understood as all ocean observation points related to the ocean data management platform. For example, all ocean observation points related to the ocean data management platform refer to all ocean observation points that transmit the collected observation data to the ocean data management platform.

[0033] In some embodiments, the basic information of the ocean observation point may include at least one of location information, attribute information, and data collection information of the ocean observation point.

[0034] The location information of the ocean observation point is information that can reflect the location of the ocean observation point. In some embodiments, the location information of the ocean observation point may include at least one of the latitude and longitude information of the ocean observation point, the relative geographical distance to a specific location (e.g., an ocean data management platform), and the sub-area to which the ocean observation point belongs within the observation area. The observation area is a set area for observing the ocean conditions. The observation area can be pre-divided into multiple observation sub-areas.

[0035] The attribute information of the ocean observation point is information that can reflect the type of the ocean observation point. In some embodiments, the attribute information of the ocean observation point may include shore-based observation, buoy observation, submerged buoy observation, ship observation, ocean remote sensing, offshore oil platform, tower observation, seabed-based observation, underwater self-propelled ocean observation, and other types.

[0036] The data collection information of an ocean observation point refers to the number of observation file fragments generated by the ocean observation point per unit time (e.g., 1 hour). Each collection by an ocean observation point may form an observation file fragment. In some embodiments, different ocean observation points may collect ocean observation data according to different preset collection rules. For more information about preset collection rules, see Figure 4 and its related description.

[0037] In some embodiments, basic information of multiple ocean observation points can be pre-stored in a storage device, and the ocean data management platform can obtain the basic information of multiple ocean observation points from the storage device. The storage device can be a storage device that comes with the ocean data management platform, or it can be an external storage device that does not belong to the ocean data management platform, such as a hard disk, an optical disk, etc.

[0038] Observation data refers to ocean-related data obtained at an ocean observation point. In some embodiments, the observation data may include at least one of ocean hydrological observation data, ocean meteorological observation data, and other observation data.

[0039] The marine hydrological observation data refers to the observation data related to the distribution and change patterns of marine hydrological elements. In some embodiments, the marine hydrological observation data may include one or more combinations of tides, waves, currents, sea ice, seawater temperature, salinity, depth, etc.

[0040] Marine meteorological observation data refers to observation data related to marine meteorological conditions. In some embodiments, the marine meteorological observation data may include one or more combinations of wind force, air pressure, temperature, relative humidity, rainfall, effective visibility on the sea surface, clouds, fog, weather phenomena, etc.

[0041] In some embodiments, other observation data may include one or more combinations of sea luminescence, water color, noise, radiance, sea surface illumination, sea surface height, etc. Sea luminescence refers to the luminescence phenomenon of the sea surface caused by luminescent organisms.

[0042] In some embodiments, the observation data can be collected by a collection device at an ocean observation point and transmitted to an ocean data management platform. In some embodiments, the collection device can transmit the observation data to the ocean data management platform based on transmission parameters. For more information about observation data, see Figure 4 and its related description.

[0043] A sharing request refers to a request to obtain target observation data.

[0044] Target observation data refers to the observation data required by users based on work needs. For example, the target observation data may be the observation data of a specific ocean observation point in a certain time period. For example, the target observation data may be the observation data of a certain observation area or observation sub-area on the ocean in a certain time period. For another example, the target observation data may be the observation data of an ocean observation point related to a certain meteorological state (for example, an observation point on a typhoon path). For another example, the target observation data may also be a certain type of observation data (for example, tidal data) in a certain observation area or observation sub-area. Among them, users may be departments or units that need to obtain observation data based on work needs.

[0045] The sharing request may be issued by a user. For example, a user may initiate a communication connection (e.g., a network connection) with the ocean data management platform and issue a sharing request to the ocean data management platform. In some embodiments, when the ocean data management platform receives a sharing request for target observation data, the ocean data management platform may find the corresponding observation data (i.e., the target observation data) based on the sharing request and return the corresponding observation data to the device that issued the sharing request.

[0046] Step 220: grouping the multiple ocean observation points based on the basic information to obtain at least two observation groups.

[0047] An observation group is a group that contains at least one ocean observation point.

[0048] In some embodiments, the ocean data management platform may group the multiple ocean observation points in a variety of ways based on basic information of the multiple ocean observation points to determine at least two observation groups.

[0049] In some embodiments, the ocean data management platform may group multiple ocean observation points based on their location information. For example, based on the distribution of the ocean observation points, the ocean observation points belonging to the same observation area or observation sub-area may be grouped into one observation group.

[0050] In some embodiments, the ocean data management platform may group multiple ocean observation points based on the attribute information of the multiple ocean observation points, for example, dividing the same type of ocean observation points into one observation group.

[0051] In some embodiments, the ocean data management platform may group multiple ocean observation points based on data collection information of the multiple ocean observation points. For example, ocean observation points with the same or similar number of observation file fragments generated per unit time are grouped into one observation group.

[0052] In some embodiments, the ocean data management platform can group multiple ocean observation points by clustering based on the basic information of multiple ocean observation points to determine at least two observation groups. For example, the ocean data management platform can convert the basic information of all ocean observation points into vector form, and divide the ocean observation points whose vector distance is less than the distance threshold into a group of observation groups according to the clustering method (e.g., K-means). Among them, the method of calculating the vector distance may include but is not limited to Euclidean distance, cosine distance, Mahalanobis distance, Chebyshev distance and / or Manhattan distance, etc.

[0053] In some embodiments, the ocean data management platform may determine the real-time requirements of multiple ocean observation points based on historical sharing requests, and group the multiple ocean observation points based on basic information and real-time requirements.

[0054] The historical sharing requests are all sharing requests received by the ocean management platform within the historical preset time period. The historical preset time period can include multiple historical sub-time periods. There is no intersection between each historical sub-time period, the union of all historical sub-time periods is the historical preset time period, and the length of each historical sub-time period is not limited to be unified.

[0055] The real-time demand is a numerical value reflecting the degree of demand for observation data of an ocean observation point. In some embodiments, for one of the multiple ocean observation points, the ocean data management platform can determine the real-time demand of the ocean observation point by the number of sharing requests for the ocean observation point in each historical sub-time period.

[0056] As an example only, for one of the multiple ocean observation points, the ocean data management platform can determine the real-time requirement of the ocean observation point by formula (1):

[0057] d=k1×s1+k2×s2+…+k n ×s n (1)

[0058] Among them, d represents the real-time demand of the ocean observation point, s1-s n are the number of sharing requests for the ocean observation point in the historical sub-time periods 1-n, k n is the weight coefficient corresponding to historical sub-time period n. n is the total number of historical sub-time periods, and n is a natural number greater than or equal to 1. The smaller n is, the closer the corresponding historical sub-time period is to the current moment.

[0059] It can be understood that the closer the historical sub-time period is to the current moment, the more the number of sharing requests in the historical sub-time period can reflect the real-time demand of the current moment than the number of sharing requests in the earlier historical sub-time period. Therefore, the size relationship of the weight coefficient can be expressed as k1>k2>...>k n .

[0060] In some embodiments, the real-time requirement is also related to the local consistency within each historical sub-time period in the historical sharing request.

[0061] Local consistency is a parameter that reflects the consistency of the type of observation data requested by the sharing request in each historical sub-time period.

[0062] In some embodiments, for one of the historical sub-time periods within a historical preset time period, the ocean data management platform may determine the local consistency of the historical sub-time period based on the observation data type and request time corresponding to all sharing requests within the historical sub-time period.

[0063] In some embodiments, the ocean data management platform may determine the local consistency of the historical sub-time period according to consistency rules based on the observed data types and request times corresponding to all sharing requests in the historical sub-time period.

[0064] The consistency rule is a preset rule for determining local consistency. In some embodiments, the consistency rule may include that the number of sharing requests for a type of observation data (e.g., tidal data) in a portion of a historical sub-time period exceeds a first preset number. In some embodiments, the consistency rule may also include that the number of sharing requests for a type of observation data in a historical sub-time period exceeds a second preset number. The first preset number and the second preset number may be preset manually or by the system.

[0065] In some embodiments, the local consistency within a historical sub-time period may be positively correlated to the number of times that each observation data type corresponding to all sharing requests within the historical sub-time period satisfies the consistency rule within the historical sub-time period, for example, the local consistency may be equal to the number of times that the consistency rule is satisfied. As an example only, if all sharing requests within a historical sub-time period include 10 types of observation data, among which only the number of sharing requests corresponding to the two observation data types of tidal data and rainfall data exceeds the first preset number, the local consistency may be 2.

[0066] In some embodiments, the ocean data management platform can determine the weight coefficient corresponding to each historical sub-time period based on the local consistency in each historical sub-time period in the historical sharing request. In some embodiments, the ocean data management platform can determine the weight coefficient corresponding to each historical sub-time period through a vector database based on the local consistency in each historical sub-time period in the historical sharing request. The vector database is a preset database containing consistency vectors and corresponding reference weight coefficient sets. The consistency vector is the local consistency of each historical sub-time period represented in vector form. The reference weight coefficient set is the weight information corresponding to the consistency vector in the vector database.

[0067] In some embodiments, the ocean data management platform can convert the local consistency of each historical sub-time period within the historical preset time period into a consistency vector. The consistency vector C can be expressed as C = (c1, c2, ..., c n ). Among them, c n c1-c n Respectively represent the local consistency of historical sub-time periods 1-n.

[0068] In some embodiments, the ocean data management platform can determine multiple consistency vectors and corresponding reference weight coefficient sets to construct a vector database. When constructing the database, the ocean data management platform can set different candidate weight coefficient sets (k1, k2, ..., k n ), and each candidate weight coefficient set (k1, k2, …, k n ) also needs to satisfy k1>k2>...>k n For each consistency vector and each corresponding candidate weight coefficient set, the ocean data management platform can use the candidate weight coefficient set with the fastest processing speed as the reference weight coefficient set corresponding to the consistency vector through data retrieval experiments. For example, for each candidate weight coefficient set, the corresponding real-time requirement is determined based on the candidate weight coefficient set and further grouped based on the real-time requirement, and a data retrieval experiment is performed based on the grouping result to obtain the processing speed of the target observation data corresponding to the retrieval sharing request.

[0069] In some embodiments, when determining the weight coefficients corresponding to each historical sub-time period through a vector database, the ocean data management platform can select a reference weight coefficient set corresponding to a consistency vector with the highest similarity to the consistency vector in the vector database, and then determine the weight coefficients corresponding to each historical sub-time period.

[0070] In an embodiment of the present specification, through the local consistency in each historical sub-time period in the historical sharing request, the observation data type in which the number of sharing requests in each historical sub-time period exceeds the first preset number or the second preset number is taken into consideration, and then a more targeted weight coefficient can be obtained by searching the vector database, so that more accurate real-time requirements can be determined, thereby making the grouping of ocean observation points more in line with actual needs.

[0071] In some embodiments, the ocean data management platform can group multiple ocean observation points in a variety of ways based on basic information and real-time requirements. For example, the ocean data management platform can construct feature vectors based on the basic information and real-time requirements of the ocean observation points, and then divide the ocean observation points whose vector distance is less than a distance threshold into a group of observation groups according to the clustering method.

[0072] In some embodiments, the ocean data management platform may group multiple ocean observation points based on basic information, and then group them again based on the order of the real-time requirements of the ocean observation points. For example, after grouping multiple ocean observation points based on basic information (e.g., the first grouping), each initial grouping (i.e., the result obtained by the first grouping) is further split based on the real-time requirements (e.g., the second grouping). For each initial grouping, the ocean observation points in the initial grouping whose real-time requirements are higher than a preset threshold are divided into one observation group, and the remaining ocean observation points whose real-time requirements are not higher than the preset threshold are divided into another observation group. The preset threshold can be used to group ocean observation points according to real-time requirements. The preset threshold can be set manually.

[0073] In the embodiments of the present specification, multiple ocean observation points are grouped according to historical sharing requests, so that the grouping process can be more in line with actual needs, and the grouping can be adjusted according to changes in actual needs, which facilitates subsequent fusion management of observation data.

[0074] In some embodiments, the ocean data management platform can determine the future real-time requirements of multiple ocean observation points based on real-time requirements, future meteorological information, observation data of key observation points, and observation point maps; and group multiple ocean observation points based on basic information, real-time requirements, and future real-time requirements.

[0075] Future weather information refers to weather information within a period of time in the future. In some embodiments, future weather information may include weather types, coverage areas, and basic information corresponding to each weather type within a period of time in the future. For example, weather types may include typhoons, ocean currents, tsunamis, rainstorms, etc., and the basic information corresponding to the weather type may be the intensity, path, direction, range, etc. of the weather. For example, if the weather type is a typhoon, the basic information is the level and path of the typhoon.

[0076] In some embodiments, the ocean data management platform can obtain future weather information in a variety of ways, such as directly from a weather platform, which is not limited here.

[0077] In some embodiments, a key observation point may be an observation point with a higher importance in location and / or corresponding observation data among all ocean observation points. The importance may be measured artificially.

[0078] In some embodiments, the ocean data management platform can determine the key observation points based on the attribute information of the ocean observation points. For example, the offshore oil platform\tower observation point can be determined as the key observation point.

[0079] In some embodiments, the ocean data management platform can determine key observation points based on historical data. For example, ocean observation points whose number of sharing requests in historical data exceeds a preset request threshold can be used as key observation points. The preset request threshold can be set manually.

[0080] In some embodiments, the ocean data management platform may also determine key observation points based on future meteorological information. For example, the ocean observation points in the area with the most prominent meteorological performance in the future meteorological information (e.g., the area with the strongest typhoon) may be determined as key observation points. Since future meteorological information may change, key observation points may also change according to changes in future meteorological information.

[0081] The observation point map is a map that reflects the positional relationship of all ocean observation points related to the ocean data management platform. The map is a data structure composed of nodes and edges, and the edges connect the nodes. The nodes and edges can have attribute characteristics. In some embodiments, the nodes in the observation point map represent ocean observation points, and an undirected edge is connected between two nodes whose actual distance is less than a preset distance threshold. The preset distance threshold can be set manually. In some embodiments, the point features of the observation point map may include the real-time requirements and future real-time requirements of the ocean observation points, and the edge features of the observation point map may include the distance between two ocean observation points.

[0082] In some embodiments, the ocean data management platform can process future meteorological information, basic information of multiple ocean observation points, and observation point maps through embedded models to determine key observation points.

[0083] In some embodiments, the embedding model may be a machine learning model of a custom structure as described below. The prediction model may also be a machine learning model of other structures, such as a neural network model. In some embodiments, the input of the embedding model may include future meteorological information, an observation point map, and basic information of multiple ocean observation points, and the output may include key observation points.

[0084] In some embodiments, the embedding model can be trained by multiple first training samples with first labels. In some embodiments, the first training sample can include at least sample meteorological information, sample observation point maps and basic information of multiple sample ocean observation points. The first training sample can be obtained from historical data. In some embodiments, the first label can be a key observation point among the multiple sample ocean observation points included in the first training sample. The first label can be determined by artificial annotation.

[0085] In some other embodiments, the embedding model can also be obtained by joint training with the prediction model. Figure 3 and its related description.

[0086] The future real-time demand is the real-time demand for observation data of ocean observation points within a period of time in the future. The future period of time corresponding to the future meteorological information and the future period of time corresponding to the future real-time demand may be related. For example, the future meteorological information and the future real-time demand may correspond to the same future period of time.

[0087] In some embodiments, the ocean data management platform can determine future real-time requirements based on the matching degree between the observation data of key observation points and future meteorological information, real-time requirements, and observation point maps.

[0088] Most types of future weather information will have corresponding data changes reflected in the observation data before they occur, which is called observation data performance. For example, before a typhoon occurs, the waves will increase and the sea surface temperature will rise. Therefore, the ocean data management platform can build a meteorological preset table based on historical data to determine the observation data performance that the relevant observation points should have before each type of meteorological type occurs. The relevant observation points refer to the ocean observation points within the area covered by the future meteorological information.

[0089] The meteorological preset table is a preset table including meteorological information and the corresponding observation data performance before the meteorological information occurs. The observation data performance corresponding to each meteorological information can be at least one item. For the same type of meteorology, there will be differences in the observation data performance corresponding to meteorology of different levels or different intensities. For example, the observation data performance corresponding to the occurrence of a level 1 typhoon and a level 5 typhoon (for example, the size of the waves, the range of temperature rise) will also be different. Therefore, for the same type of meteorology, multiple groups of corresponding observation data performances can be set according to the basic information corresponding to the meteorological type (for example, the intensity of the meteorology).

[0090] In some embodiments, the ocean data management platform can search for the observation data performance corresponding to the future meteorological information in the meteorological preset table based on the future meteorological information, and then determine the matching degree between the observation data of the key observation point and the observation data performance corresponding to the future meteorological information. As an example only, for a certain meteorological information, there are 7 observation data performances set in the meteorological preset table (for example, the sea surface temperature increases by 1-2°C, the humidity increases by 2-5%, etc., a total of seven items), and the key observation point meets 5 of the observation data performances, then the matching degree is 5 / 7. For each of the 7 observation data performances in the meteorological preset table, the ocean data management platform can compare the observation data performance with the observation data of the same type in the observation data of the key observation point. If the observation data of the key observation point meets the range of the observation data performance, it is judged to meet the observation data performance.

[0091] Since the weather preset table is a rule obtained by summarizing historical data, the matching degree of the key observation points determined based on the weather preset table should be high, for example, close to 1. If the matching degree is low, it means that the predicted future weather information may not occur at the key observation point in the future, so the future real-time demand of the key observation point may be low. This is because the future weather information is predicted, and the sea surface weather conditions are complex, so inaccurate predictions or sudden changes in weather are possible.

[0092] In some embodiments, the ocean data management platform can determine the future real-time requirements of key observation points based on the matching degree and real-time requirements of the observation data of the key observation points and future meteorological information. In some embodiments, the future real-time requirements of key observation points can be positively correlated with the matching degree and real-time requirements of the observation data of the key observation points and future meteorological information. For example, the higher the matching degree between the observation data of the key observation points and future meteorological information, the higher the future real-time requirements of the key observation points. The higher the real-time requirements of the key observation points, the higher the future real-time requirements of the key observation points. As an example only, the ocean data management platform can determine the future real-time requirements of the key observation points by formula (2):

[0093] D= (k+m) × d (2)

[0094] Among them, d represents the real-time demand of the observation point, D represents the future real-time demand of the observation point, m represents the matching degree between the observation data of the key observation point and the future meteorological information, and k is the preset coefficient. The value of k can be 1.

[0095] In some embodiments, the ocean data management platform can determine the future real-time requirements of the remaining observation points based on the future real-time requirements of the key observation points and the observation map. In some embodiments, the ocean data management platform can determine the future real-time requirements of the first observation point for each key observation point based on the future real-time requirements of the key observation point and the distance between the key observation point and the observation point (referred to as the first observation point) connected to the key observation point by an edge. As an example only, the ocean data management platform can determine the future real-time requirements of the first observation point by formula (3):

[0096] B=e+v×D (3)

[0097] Where B represents the future real-time demand of the first observation point, e represents the real-time demand of the ocean observation point (i.e., the first observation point) that needs to solve the future real-time demand, D represents the future real-time demand of the key observation point, v represents the weight, and 0 <v<1。

[0098] In some embodiments, v may be related to the distance between the first observation point and the key observation point, as well as whether they are within the coverage area of the same future meteorological information. For example, the closer the distance between the first observation point and the key observation point, the larger v. Also, for example, v is larger when within the same coverage area than when not within the same coverage area.

[0099] In some embodiments, after determining the future real-time requirement of the first observation point, the ocean data management platform may calculate the future real-time requirement of the observation point (referred to as the second observation point) connected to the first observation point by an edge in a manner similar to calculating the future real-time requirement of the first observation point. Only as an example, the ocean data management platform may also determine the future real-time requirement of the second observation point through formula (3), where the difference is that B represents the future real-time requirement of the second observation point, D represents the future real-time requirement of the first observation point, and e represents the real-time requirement of the second observation point.

[0100] In some embodiments, after determining the future real-time requirement of the second observation point, the ocean data management platform may continue to calculate the future real-time requirement of the observation point (referred to as the third observation point) connected to the second observation point by an edge in a manner similar to calculating the future real-time requirement of the first observation point or the second observation point, and so on, until the future real-time requirements of all nodes are obtained.

[0101] Only as an example, as Figure 5 shown, assume that O is the key observation point, M, N, P, Q are the observation points (i.e., the first observation points) adjacent to O, and X and Y are the observation points (i.e., the second observation points) adjacent to P, where the rectangular box is the coverage area of a future meteorological information. The ocean data management platform may determine the future real-time requirements of points M, N, P, and Q based on the future real-time requirement of point O. Since the distance L1 between point P and point O is the smallest and they are within the same coverage area, the distance L2 between point Q and point O is the second, followed by the distance L3 between point N and point O and the distance L4 between point M and point O, which can be expressed as L1 < L2 < L3 < L4, then the weight relationship of these four nodes is v P > v Q > v N > v M . After calculating the future real-time requirement of point P, the ocean data management platform may calculate the future real-time requirements of points X and Y based on the future real-time requirement of point P.

[0102] In some embodiments, if an observation point is an adjacent point of multiple key observation points, multiple future real-time requirements of this observation point may be calculated through the above method. In this case, the maximum value among the multiple future real-time requirements may be selected as the future real-time requirement of this observation point.

[0103] In some embodiments, there may be isolated nodes in the observation graph. An isolated node is a node that has no edge connection with other nodes. The future real-time requirements of an isolated node do not need to be calculated. In some embodiments, the real-time requirements of an isolated node can be determined as the future real-time requirements of the isolated node.

[0104] In some embodiments, the ocean data management platform can group multiple ocean observation points by clustering, combining real-time requirements and future real-time requirements. For example, the ocean data management platform can construct a feature vector based on the basic information, real-time requirements and future real-time requirements of the ocean observation points, and then divide the ocean observation points whose vector distance is less than the distance threshold into a group of observation groups according to the clustering method.

[0105] In some embodiments, the ocean data management platform can determine the future real-time requirements of multiple ocean observation points in multiple future time periods based on real-time requirements, future meteorological information, observation data of key observation points, and observation point maps. The ocean data management platform can also group multiple ocean observation points based on basic information, real-time requirements, and future real-time requirements of multiple future time periods. The time lengths of multiple future time periods can be the same or different. The method for determining the future real-time requirements of each future time period is similar to the method for determining a single future real-time requirement.

[0106] In some embodiments, the ocean data management platform can determine the future real-time requirements of multiple ocean observation points through the prediction layer in the first prediction model based on real-time requirements, future meteorological information, observation data of key observation points, and observation point maps, and further group the multiple ocean observation points through the grouping layer in the first prediction model based on the basic information, real-time requirements, and future real-time requirements of the multiple ocean observation points. For more information, please refer to Figure 3 and its related description.

[0107] In the embodiments of this specification, the grouping of ocean observation points can be made more in line with actual needs through basic information, real-time requirements and future real-time requirements.

[0108] Step 230: For each of the at least two observation groups, the observation data uploaded by at least one ocean observation point included in the observation group are fused to obtain fused observation data.

[0109] The data collected at each ocean observation point may form an observation file fragment, so the observation data of an ocean observation point may be a collection of multiple observation file fragments, and the observation data may include multiple types of data.

[0110] In some embodiments, the ocean data management platform can fuse the same type of observation data from different observation points of the same observation group according to the collection time. In other words, the ocean data management platform can fuse the same type of observation data from different ocean observation points in the same time period to obtain fused observation data (also called fused data).

[0111] Step 240: group and store the fused observation data, and create indexes for the fused observation data of at least two observation groups.

[0112] Group storage means storing the fused data of different observation groups separately. In other words, the fused data of each observation group is stored together, and the fused data of different observation groups are stored separately.

[0113] An index is a data structure that stores key information of a data item and the corresponding storage location. In some embodiments, for each of at least two observation groups, the ocean data management platform can establish an index for each key information in the observation group. Key information refers to information that can characterize the characteristics of the observation data. For example, the key information may include the ocean observation point, observation item, collection time, etc. corresponding to the observation data.

[0114] By establishing an index, when searching for observation data, the ocean data management platform can directly locate the storage location of the queried observation data without traversing the entire storage space one by one, thereby greatly improving the search efficiency.

[0115] Step 250: determining a target observation group from at least two observation groups based on the sharing request.

[0116] The target observation group is the observation group to which the source of the target observation data corresponding to the sharing request belongs.

[0117] In some embodiments, the ocean data management platform may determine the target observation point corresponding to the observation data corresponding to the sharing request, and determine the observation group where the target observation point is located as the target observation group. The target observation point refers to the source of the observation data corresponding to the sharing request.

[0118] In some embodiments, the number of target observation points can be multiple, so that the number of target observation groups can also be multiple. For example, if the observation data requested to be retrieved in the sharing request is the observation data of a certain ocean observation point (for example, a designated ocean observation station), then the ocean observation point is the target observation point. For another example, if the observation data requested to be retrieved in the sharing request is the observation data of a certain ocean area, then all ocean observation points in the ocean area can be used as target observation points.

[0119] Step 260: Based on the index, retrieve the target observation data corresponding to the sharing request.

[0120] In some embodiments, the ocean data management platform can search the target observation group according to the index, determine the target observation data corresponding to the sharing request, and retrieve the target observation data. In some embodiments, the ocean data management platform can also select the index corresponding to the target observation data from the indexes of multiple key information contained in the target observation group, and directly locate the location where the target observation data is stored according to the index corresponding to the target observation data, and retrieve the target observation data.

[0121] Before grouping, a large amount of observation data is stored in the data management platform. Therefore, after obtaining a sharing request, if the conventional search method is used, the data management platform needs to traverse and compare all the data to find the required data. The entire processing process is time-consuming and inefficient. In the embodiments of the present specification, by storing the observation data in groups, the data management platform can narrow the search range according to the grouping information of the observation point and only search in specific groups, thereby reducing the amount of data that needs to be searched. When there are many data sharing requests, the processing efficiency of data sharing requests can be improved to ensure the real-time nature of data sharing. After the index is established, it can also be quickly located according to the corresponding index without traversing the entire data set one by one, thereby greatly improving the efficiency of the search.

[0122] It should be noted that the above description of the process 200 is only for example and illustration, and does not limit the scope of application of this specification. For those skilled in the art, various modifications and changes can be made to the process 200 under the guidance of this specification. However, these modifications and changes are still within the scope of this specification.

[0123] Figure 4 This is an exemplary schematic diagram of grouping multiple ocean observation points based on a prediction model according to some embodiments of this specification.

[0124] In some embodiments, the ocean data management platform may group multiple ocean observation points based on a prediction model.

[0125] The prediction model is a machine learning model. In some embodiments, the prediction model can be a machine learning model of a custom structure described below. The prediction model can also be a machine learning model of other structures, such as a neural network model. In some embodiments, the input of the prediction model can include real-time requirements, future meteorological information, observation data of key observation points, observation point maps, and basic information of multiple ocean observation points, and the output can include grouping results of multiple ocean observation points. For more information about real-time requirements, future meteorological information, observation data of key observation points, and observation point maps, please refer to Figure 2And the related description will not be repeated here.

[0126] In some embodiments, the prediction model can be obtained by training multiple second training samples with second labels. The second training samples can be obtained from historical data. In some embodiments, the second training samples may at least include sample real-time requirements, sample future meteorological information, observation data of sample key observation points, sample observation point maps, and basic information of sample ocean observation points. Among them, the time period corresponding to the sample future meteorological information is after the time period corresponding to the sample real-time requirements, and the time period corresponding to the sample future meteorological information and the time period corresponding to the sample real-time requirements are both historical time periods. The second label can be the grouping result after grouping the corresponding multiple sample ocean observation points in the second training sample. The second label can be obtained through data retrieval experiments. For example, a data retrieval experiment is performed based on multiple grouping results to obtain the processing speed of retrieving the target observation data corresponding to the shared request under each grouping result, and the grouping result with the fastest processing speed is used as the second label.

[0127] In some embodiments, the prediction model may be composed of multiple processing layers. Figure 3 As shown, the prediction model 310 may include a prediction layer 320 and a grouping layer 330 .

[0128] like Figure 3 As shown, the input of the prediction layer 320 may include the real-time requirements 321 of multiple ocean observation points, future meteorological information 322, observation data 323 of key observation points, and observation point map 324, and the output of the prediction layer 320 may include the future real-time requirements 340 of each ocean observation point in the observation point map 324. For more information about future real-time requirements, please refer to Figure 2 And the related description will not be repeated here.

[0129] like Figure 3 As shown, the input of the grouping layer 330 may include the future real-time requirements 340 of each ocean observation point output by the prediction layer 320, the real-time requirements 321 of multiple ocean observation points and the basic information 331 of multiple ocean observation points, and the output of the grouping layer 330 may include the grouping results 350 of multiple ocean observation points.

[0130] In some embodiments, the prediction layer 320 and the grouping layer 330 can be trained separately. The third training sample of the prediction layer 320 may include sample real-time requirements, sample future weather information, observation data of sample key observation points, and sample observation point maps. The third label of the prediction layer 320 may be the real-time requirements of the sample real-time requirements in the third training sample, that is, the real-time requirements subsequent to the sample real-time requirements. Among them, the time period corresponding to the sample future weather information is located after the time period corresponding to the sample real-time requirements, and the time period corresponding to the sample future weather information and the time period corresponding to the sample real-time requirements are both historical time periods.

[0131] In some embodiments, the fourth training sample of the grouping layer 330 may include sample real-time requirements, basic information of sample ocean observation points, and sample future real-time requirements, and the fourth label of the grouping layer 330 may be a grouping result after grouping the corresponding multiple sample ocean observation points in the fourth training sample. The sample real-time requirements and sample future real-time requirements in the fourth training sample are respectively the same as the sample real-time requirements and sample future real-time requirements in the third training sample.

[0132] In some embodiments, the output of the prediction layer 320 may be used as the input of the grouping layer 330 , so the prediction layer 320 and the grouping layer 330 in the prediction model 310 may be jointly trained.

[0133] An exemplary joint training process includes: inputting the sample real-time requirements, sample future meteorological information, observation data of sample key observation points, and sample observation point maps in the third training sample into the initial prediction layer, and obtaining the real-time requirements of the samples in the third training sample output by the initial prediction layer in a subsequent period of time; using the output of the initial prediction layer as the future real-time requirements of the samples in the fourth training sample, and inputting the fourth training sample into the initial grouping layer, and obtaining the grouping results of multiple ocean observation points output by the initial grouping layer. A loss function is constructed based on the fourth training label and the grouping results of multiple ocean observation points output by the initial grouping layer, and the parameters of the initial prediction layer and the initial grouping layer are updated synchronously. Through parameter updating, the trained prediction layer 320 and grouping layer 330 are obtained.

[0134] In some embodiments, the output of the prediction layer may include future real-time requirements corresponding to multiple ocean observation points in multiple future time periods. At this time, the third label of the prediction layer may be the real-time requirements of the samples in the third training sample in multiple subsequent time periods.

[0135] In some embodiments, the prediction model may further include an embedding layer configured to determine key observation points based on future meteorological information, basic information of multiple ocean observation points, and an observation point map.

[0136] In some embodiments, the input of the embedding layer may include future weather information, basic information of multiple ocean observation points, and observation point maps, and the output may include key observation points. In some embodiments, the output of the embedding layer may also include observation data corresponding to key observation points. The embedding layer can be trained separately, and its training method is the same as the embedding model described above. For more information, see Figure 2 and its related description.

[0137] In some embodiments, the embedding layer may also be a separate embedding model, and the embedding model may be combined with the prediction model. In some embodiments, the observation data corresponding to the key observation points output by the embedding layer may be used as the input of the prediction layer, so the embedding layer may be jointly trained with the prediction layer. The method of jointly training the embedding layer and the prediction layer is similar to the method of jointly training the prediction layer and the grouping layer. For more information, please refer to the relevant description above.

[0138] In the embodiments of this specification, by using a prediction model to determine the future real-time demand and the grouping results of multiple ocean observation points, the self-learning ability of the machine learning model can be used to find patterns from a large amount of historical data, obtain the relationship between real-time demand, future meteorological information, observation data of key observation points, observation point maps and basic information of ocean observation points, etc., and improve the accuracy and efficiency of determining the grouping results of multiple ocean observation points. Through the joint training of the prediction layer and the grouping layer of the prediction model, in some cases, it is helpful to solve the problem of difficulty in obtaining labels when training the prediction layer alone, and it can also enable the prediction layer to better obtain grouping results that reflect future real-time demand.

[0139] Figure 5 It is an exemplary schematic diagram of obtaining observation data according to some embodiments of this specification.

[0140] In some embodiments, the observation data can be collected by collection devices at multiple ocean observation points and transmitted to the ocean data management platform. For more information about collection devices, see Figure 2 and its related description.

[0141] In some embodiments, for one of the multiple ocean observation points, the ocean data management platform can control the collection device to collect the observation data of the ocean observation point based on preset collection rules and transmit it to the ocean data management platform.

[0142] The preset collection rules are rules for the collection device to collect observation data. In some embodiments, the preset collection rules may include collection parameters and rules for changing the collection parameters. In some embodiments, the collection parameters may include collection frequency, collection range, and collection items. The collection frequency is the number of times the collection device collects data per unit time. The collection range is the sea area range that the collection device is to collect. The collection item is the type of observation data to be collected by the collection device, such as temperature, humidity, etc. The rules for changing the collection parameters may be rules for changing the collection frequency and rules for changing the collection items within a preset time period. For example, the rules for changing the collection parameters include collecting collection item a1 at a collection frequency p1 in a first sub-period within a preset time period, collecting collection item a1 and collection item a2 at a collection frequency p2 in a second sub-period within the preset time period, and so on.

[0143] In some embodiments, the ocean data management platform may preset collection rules based on historical experience or manual settings.

[0144] In some embodiments, for one of the multiple ocean observation points, such as Figure 4 As shown, the ocean data management platform can determine the estimated change information 420 of the ocean observation point based on the preset collection rules 410; based on the estimated change information 420 of the ocean observation point, determine the transmission parameters 430 of the ocean observation point, and the transmission parameters 430 can be used to control the collection equipment to transmit the collected observation data to the ocean data management platform.

[0145] The estimated change information refers to the change in the amount of observed data in a future preset time period relative to the amount of observed data in the current time period.

[0146] In some embodiments, the ocean data management platform can determine the estimated change information of the ocean observation points in a variety of ways based on preset collection rules.

[0147] In some embodiments, in response to the fact that the collection parameters in the preset collection rules have not changed, indicating that the amount of observation data generated in the future preset time period is the same as the amount of observation data generated in the current time period, at this time, the ocean data management platform can determine the estimated change information as 0.

[0148] In some embodiments, in response to changes in the collection parameters in the preset collection rules (for example, the collection frequency is increased or decreased), the amount of observation data generated in the future preset time period will increase or decrease. At this time, the ocean data management platform can determine the estimated change information as not zero.

[0149] In some embodiments, the ocean data management platform can also determine the estimated change information through a preset algorithm based on the collection frequency, collection range and collection items in the preset collection rules. For example, the ocean data management platform can determine the numerical relationship between the four items within a fixed time length based on the collection frequency, collection range, collection items in the historical data and the observation data through a fitting algorithm. The ocean data management platform can determine the estimated change information based on the numerical relationship, the collection frequency, collection range and collection items in the preset collection rules. Among them, the fitting algorithm can include but is not limited to linear fitting, least squares fitting, polynomial fitting, exponential fitting and nonlinear curve fitting.

[0150] In some embodiments, the ocean data management platform may adjust and update preset collection rules based on future meteorological information.

[0151] Considering that the preset collection rules generally do not change or do not change in a timely manner (for example, the collection rules for the morning and afternoon generally do not change), the ocean data management platform can actively update the preset collection rules. For example, when abnormal weather is predicted in the future time period, the ocean data management platform can actively update the preset collection rules corresponding to the future time period. As an example only, if heavy rain is predicted for the next day, the preset collection rules for the next day can be updated to modify the collection parameters. For example, increase the collection frequency.

[0152] In some embodiments, the ocean data management platform may determine updated preset collection rules based on future meteorological information; and determine estimated change information based on the updated preset collection rules.

[0153] In some embodiments, the ocean data management platform can determine updated preset collection rules based on future meteorological information by collecting preset tables. The collection preset table is a preset table that includes future meteorological information and corresponding preset collection rules. In some embodiments, the collection preset table can be constructed through human experience. The ocean data management platform can search for the preset collection rules corresponding to the future meteorological information in the collection preset table as updated preset collection rules.

[0154] The method of determining the estimated change information based on the updated preset collection rule is the same as the above-mentioned method of determining the estimated change information based on the preset collection rule, and will not be repeated here.

[0155] In some embodiments, for one of the multiple ocean observation points, the ocean data management platform can adjust the preset collection rules based on the associated sharing request of the ocean observation point; and determine the estimated change information of the ocean observation point based on the adjusted preset collection rules.

[0156] The associated sharing request of an ocean observation point is the request associated with the ocean observation point among all sharing requests. For example, if a sharing request is for all observation data in area A, and observation point a is in area A, then the sharing request is the associated sharing request of observation point a. For another example, if a sharing request is for all observation data with a temperature exceeding a certain temperature value, and observation point a meets this condition, then the sharing request is also the associated sharing request of observation point a.

[0157] In some embodiments, the ocean data management platform can determine the number of associated sharing requests for ocean observation points, and determine whether the current preset collection rules meet the preset conditions based on the number of associated sharing requests; in response to the current preset collection rules meeting the preset conditions, the preset collection rules are not adjusted; in response to the current preset collection rules meeting the preset conditions, the preset collection rules are adjusted. The more the number of associated sharing requests, the more data demand there is for the ocean observation point, and more observation data needs to be collected for analysis and processing; the fewer the number of associated sharing requests, the less data demand there is for the observation point, and therefore it is not necessary to maintain high-intensity collection.

[0158] The preset condition is a condition for determining whether the preset collection rule needs to be adjusted. In some embodiments, the preset condition may include: under the current preset collection rule, the amount of observation data actually generated by a single collection (hereinafter referred to as the actual amount of observation data) is within a preset reasonable range.

[0159] In some embodiments, the preset reasonable range may be a reasonable range of the amount of observation data collected at a single time. The preset reasonable range of the amount of observation data collected at a single time corresponding to different numbers of associated sharing requests may be different. The reasonable range of the amount of observation data collected at a single time corresponding to different numbers of associated sharing requests may be set in advance based on experience.

[0160] In some embodiments, the ocean data management platform can determine a preset reasonable range of the amount of observation data collected in a single time corresponding to the number of associated sharing requests based on the number of associated sharing requests, and then determine whether the current preset collection rules meet the preset conditions.

[0161] In some embodiments, in response to the actual amount of observed data being within a preset reasonable range, the ocean data management platform may not adjust the preset collection rules.

[0162] In some embodiments, in response to the actual amount of observed data being less than a preset reasonable range, the ocean data management platform may adjust the preset collection rules to increase the actual amount of observed data, for example, by increasing the collection frequency, increasing the collection range or collection items, and so on.

[0163] In some embodiments, in response to the actual amount of observed data being greater than a preset reasonable range, the ocean data management platform may adjust the preset collection rules to reduce the actual amount of observed data, for example, by reducing the collection frequency, reducing the collection scope or collection items, and so on.

[0164] The method of determining the estimated change information based on the adjusted preset collection rule is the same as the above-mentioned method of determining the estimated change information based on the preset collection rule, which will not be repeated here.

[0165] In some embodiments, the ocean data management platform can determine the future demand degree based on the future real-time demand and the associated sharing request; and adjust the preset collection rules based on the future demand degree. For more information about future real-time demand, please refer to Figure 2 and its related description.

[0166] In some embodiments, the ocean data management platform can determine future associated sharing requests based on future real-time requirements and associated sharing requests. Future associated sharing requests are sharing requests associated with a certain observation point in a future time period among all predicted sharing requests. Both future real-time requirements and future associated sharing requests can reflect the level of demand for observation data of ocean observation points, and can be combined to determine the future demand degree. The future demand degree can intuitively reflect the numerical value of the demand degree for future observation data of ocean observation points.

[0167] In some embodiments, the ocean data management platform can determine the future demand degree through weighted fusion based on the future real-time demand and the number of future associated sharing requests. As an example only, the ocean data management platform can determine the future demand degree through formula (4):

[0168] F=t1×D+t2×H (4)

[0169] Wherein, F represents the future demand degree, D represents the future real-time demand, and H represents the number of future associated sharing requests. t1 and t2 are preset weight coefficients. For example, t1 and t2 can both be 0.5.

[0170] The method of adjusting the preset collection rules based on the future demand is the same as that of adjusting the preset collection rules based on the associated sharing request, except that the associated sharing request is replaced by the future demand. In addition, the difference between adjusting the preset collection rules based on the future demand and adjusting the preset collection rules based on the associated sharing request is that the adjustment based on the future demand is an advance adjustment, while the adjustment based on the associated sharing request is a real-time adjustment.

[0171] Transmission parameters are the parameters for transmitting observation data to the ocean data management platform.

[0172] In some embodiments, the transmission parameters are used to control the collection device to transmit the collected observation data to the ocean data management platform.

[0173] In some embodiments, the transmission parameters may include a transmission threshold and a transmission priority. The transmission threshold refers to the threshold condition reached when the upload operation of the observation data is triggered. For example, after the ocean observation point collects a certain amount of observation data, it needs to upload it, and this amount of data is the transmission threshold. The transmission priority is the priority of the ocean data management platform when receiving observation data from multiple ocean observation points. When the ocean data management platform receives transmission requests for observation data from multiple ocean observation points at the same time, the higher the transmission priority of the ocean observation point, the higher the priority of the ocean data management platform in receiving the observation data from the ocean observation point.

[0174] In some embodiments, the ocean data management platform can determine the transmission parameters in a variety of ways based on the estimated change information. For example, the larger the estimated change information, the lower the transmission threshold and the higher the transmission priority. The transmission priority can be determined based on the ranking of the estimated change information of all observation points. The estimated change information can be sorted from large to small, and the higher the ranking of the estimated change information, the larger the estimated change information, and the higher the transmission priority. The estimated change information can also be sorted from small to large, and the higher the ranking of the estimated change information, the smaller the estimated change information, and the lower the transmission priority.

[0175] In some embodiments, the ocean data management platform may control the collection device to transmit the observation data to the ocean data management platform based on the determined transmission parameters.

[0176] In the embodiments of this specification, observation data is collected and transmitted to the ocean data management platform by presetting collection rules and transmission parameters, which can ensure the real-time performance of the transmitted observation data, thereby achieving good real-time performance of data sharing.

[0177] One or more embodiments of the present specification also provide a multi-source ocean observation data fusion device, which includes at least one processor and at least one memory; the at least one memory is used to store computer instructions; the at least one processor is used to execute at least part of the computer instructions to implement a multi-source ocean observation data fusion method as described in any embodiment of the present specification.

[0178] One or more embodiments of the present specification also provide a computer-readable storage medium, wherein the storage medium stores computer instructions, and is characterized in that after the computer reads the computer instructions in the storage medium, the computer executes a multi-source ocean observation data fusion method as described in any embodiment of the present specification.

[0179] The basic concepts have been described above. Obviously, for those skilled in the art, the above detailed disclosure is only for example and does not constitute a limitation of this specification. Although not explicitly stated here, those skilled in the art may make various modifications, improvements and corrections to this specification. Such modifications, improvements and corrections are suggested in this specification, so such modifications, improvements and corrections still belong to the spirit and scope of the exemplary embodiments of this specification.

[0180] At the same time, this specification uses specific words to describe the embodiments of this specification. For example, "one embodiment", "an embodiment", and / or "some embodiments" refer to a certain feature, structure or characteristic related to at least one embodiment of this specification. Therefore, it should be emphasized and noted that "one embodiment" or "an embodiment" or "an alternative embodiment" mentioned twice or more in different positions in this specification does not necessarily refer to the same embodiment. In addition, certain features, structures or characteristics in one or more embodiments of this specification can be appropriately combined.

[0181] In addition, unless explicitly stated in the claims, the order of the processing elements and sequences described in this specification, the use of alphanumeric characters, or the use of other names are not intended to limit the order of the processes and methods of this specification. Although the above disclosure discusses some invention embodiments that are currently considered useful through various examples, it should be understood that such details are only for illustrative purposes, and the attached claims are not limited to the disclosed embodiments. On the contrary, the claims are intended to cover all modifications and equivalent combinations that are consistent with the essence and scope of the embodiments of this specification. For example, although the system components described above can be implemented by hardware devices, they can also be implemented only by software solutions, such as installing the described system on an existing server or mobile device.

[0182] Similarly, it should be noted that in order to simplify the description disclosed in this specification and thus help understand one or more embodiments of the invention, in the above description of the embodiments of this specification, multiple features are sometimes combined into one embodiment, figure or description thereof. However, this disclosure method does not mean that the features required by the subject matter of this specification are more than the features mentioned in the claims. In fact, the features of the embodiments are less than all the features of the single embodiment disclosed above.

[0183] In some embodiments, numbers describing the number of components and attributes are used. It should be understood that such numbers used in the description of the embodiments are modified by the modifiers "about", "approximately" or "substantially" in some examples. Unless otherwise specified, "about", "approximately" or "substantially" indicate that the numbers are allowed to vary by ±20%. Accordingly, in some embodiments, the numerical parameters used in the specification and claims are approximate values, which may change according to the required features of individual embodiments. In some embodiments, the numerical parameters should take into account the specified significant digits and adopt the general method of retaining digits. Although the numerical domains and parameters used to confirm the breadth of their range in some embodiments of this specification are approximate values, in specific embodiments, the setting of such numerical values ​​is as accurate as possible within the feasible range.

[0184] Each patent, patent application, patent application publication, and other materials, such as articles, books, specifications, publications, documents, etc., cited in this specification are hereby incorporated by reference in their entirety. Except for application history documents that are inconsistent with or conflicting with the contents of this specification, documents that limit the broadest scope of the claims of this specification (currently or later attached to this specification) are also excluded. It should be noted that if the descriptions, definitions, and / or use of terms in the materials attached to this specification are inconsistent or conflicting with the contents described in this specification, the descriptions, definitions, and / or use of terms in this specification shall prevail.

[0185] Finally, it should be understood that the embodiments described in this specification are only used to illustrate the principles of the embodiments of this specification. Other variations may also fall within the scope of this specification. Therefore, as an example and not a limitation, alternative configurations of the embodiments of this specification may be considered consistent with the teachings of this specification. Accordingly, the embodiments of this specification are not limited to the embodiments explicitly introduced and described in this specification.

Claims

1. A multi-source ocean observation data fusion method, the method being executed by an ocean data management platform, comprising: Obtaining basic information of a plurality of ocean observation points and a sharing request for target observation data, wherein the basic information includes at least one of location information, attribute information, and data collection information of the ocean observation points; Based on the basic information, the plurality of ocean observation points are grouped to obtain at least two observation groups, including: Determining the real-time requirements of the plurality of ocean observation points based on the historical sharing requests; The multiple ocean observation points are grouped based on the basic information and the real-time requirement, including: Determine the future real-time requirements of the plurality of ocean observation points based on the real-time requirements, future meteorological information, observation data of key observation points, and observation point maps; Based on the basic information, the real-time requirement and the future real-time requirement, the multiple ocean observation points are grouped by a prediction model, the prediction model includes an embedding layer, a prediction layer and a grouping layer, the input of the embedding layer includes the future meteorological information, the basic information and the observation point map, the output of the embedding layer includes the observation data of the key observation points, the input of the prediction layer includes the real-time requirement, the future meteorological information, the observation data of the key observation points, and the observation point map, the output of the prediction layer includes the future real-time requirement, the input of the grouping layer includes the basic information, the real-time requirement and the future real-time requirement, and the output of the grouping layer includes the grouping result; For each of the at least two observation groups, fusing the observation data uploaded by at least one ocean observation point included in the observation group to obtain fused observation data; storing the fused observation data in groups, and creating indexes for the fused observation data of the at least two observation groups; Determining a target observation group from the at least two observation groups based on the sharing request, the target observation group being the observation group to which the source of the target observation data belongs; Based on the index, the target observation data corresponding to the sharing request is retrieved.

2. The method according to claim 1, characterized in that The observation data is collected by the collection equipment at the ocean observation point and transmitted to the ocean data management platform, and the method further includes: For one of the plurality of ocean observation points: Determining estimated change information of the ocean observation point based on preset collection rules; Based on the estimated change information of the ocean observation point, a transmission parameter of the ocean observation point is determined, and the transmission parameter is used to control the acquisition device to transmit the collected observation data to the ocean data management platform.

3. The method according to claim 2, characterized in that The method further comprises: Adjusting the preset collection rule based on the associated sharing request of the ocean observation point; Based on the adjusted preset collection rules, the estimated change information of the ocean observation point is determined.

4. A multi-source ocean observation data fusion system, the system comprising: an acquisition module, configured to acquire basic information of a plurality of ocean observation points and a sharing request for target observation data, wherein the basic information includes at least one of location information, attribute information and data collection information of the ocean observation points; The grouping module is configured as follows: Based on the basic information, the plurality of ocean observation points are grouped to obtain at least two observation groups, including: Determining the real-time requirements of the plurality of ocean observation points based on the historical sharing requests; The multiple ocean observation points are grouped based on the basic information and the real-time requirement, including: Determine the future real-time requirements of the plurality of ocean observation points based on the real-time requirements, future meteorological information, observation data of key observation points, and observation point maps; Based on the basic information, the real-time requirement and the future real-time requirement, the multiple ocean observation points are grouped by a prediction model, the prediction model includes an embedding layer, a prediction layer and a grouping layer, the input of the embedding layer includes the future meteorological information, the basic information and the observation point map, the output of the embedding layer includes the observation data of the key observation points, the input of the prediction layer includes the real-time requirement, the future meteorological information, the observation data of the key observation points, and the observation point map, the output of the prediction layer includes the future real-time requirement, the input of the grouping layer includes the basic information, the real-time requirement and the future real-time requirement, and the output of the grouping layer includes the grouping result; For each of the at least two observation groups, fusing the observation data uploaded by at least one ocean observation point included in the observation group to obtain fused observation data; storing the fused observation data in groups, and creating indexes for the fused observation data of the at least two observation groups; a target determination module, configured to determine a target observation group from the at least two observation groups based on the sharing request, the target observation group being the observation group to which the source containing the target observation data belongs; The retrieving module is configured to retrieve the target observation data corresponding to the sharing request based on the index.

5. The system according to claim 4, characterized in that The observation data is collected by the collection equipment at the multiple ocean observation points and transmitted to the ocean data management platform. The system also includes a transmission module, which is configured as follows: For one of the plurality of ocean observation points: Determining estimated change information of the ocean observation point based on preset collection rules; Based on the estimated change information of the ocean observation point, a transmission parameter of the ocean observation point is determined, and the transmission parameter is used to control the acquisition device to transmit the collected observation data to the ocean data management platform.

6. A multi-source ocean observation data fusion device, characterized in that: The apparatus comprises at least one processor and at least one memory; The at least one memory is used to store computer instructions; The at least one processor is used to execute at least part of the computer instructions to implement a multi-source ocean observation data fusion method as described in any one of claims 1 to 3.

7. A computer-readable storage medium storing computer instructions, characterized in that: After the computer reads the computer instructions in the storage medium, the computer executes the multi-source ocean observation data fusion method as described in any one of claims 1 to 3.

Images