CN111860146A

CN111860146A - Ocean front region acquisition method and device, computer equipment and storage medium

Info

Publication number: CN111860146A
Application number: CN202010528843.4A
Authority: CN
Inventors: 任磊; 姬进财; 潘广维; 杨清书
Original assignee: Sun Yat Sen University
Current assignee: Sun Yat Sen University
Priority date: 2020-06-11
Filing date: 2020-06-11
Publication date: 2020-10-30
Anticipated expiration: 2040-06-11
Also published as: CN111860146B

Abstract

The application relates to a method and a device for acquiring a sea front area, computer equipment and a storage medium. The method comprises the following steps: determining a rough estimation ocean front region, a target path of a plurality of unmanned ships reaching the rough estimation ocean front region and a target time length of the unmanned ships reaching the rough estimation ocean front region along the target path according to the obtained satellite remote sensing observation data; determining each initial position of each unmanned ship in the rough estimation sea front area according to received arrival confirmation information sent when each unmanned ship arrives at the rough estimation sea front area; outputting an adjusting instruction according to regional change information of a roughly estimated sea front region determined by a plurality of groups of satellite remote sensing observation data received within a target time length to indicate that each unmanned naval vessel is adjusted based on the corresponding initial position of each unmanned naval vessel, so as to obtain the adjusted position of each unmanned naval vessel; and determining the target ocean front region according to the received adjusted positions. By adopting the method, the identification accuracy of the ocean front region in the sea area can be improved.

Description

Ocean front region acquisition method and device, computer equipment and storage medium

Technical Field

The present application relates to the field of marine front identification technologies, and in particular, to a method and an apparatus for acquiring a marine front region, a computer device, and a storage medium.

Background

The narrow intersection zone between two or more water bodies with obviously different properties in the ocean is called as the ocean front, which is a jump zone of ocean environmental parameters and can be described by elements such as seawater temperature, salinity, density, speed, color, chlorophyll and the like; and the scale of the ocean front can reach hundreds of kilometers, the ocean front exists on the surface layer, the middle layer and the bottom layer of the ocean, and the intensive turbulence mixing exchange, horizontal convergence (divergence) and vertical movement in the front area of the ocean front not only influence shipping and fishery, but also influence underwater communication, ship safety, maritime search and rescue and other activities, so that the accurate identification of the ocean front is one of the problems which need to be solved urgently at present.

In the conventional art, when identifying a sea front region in a sea area using a gradient threshold method, a reasonable threshold is set by different criteria of different parameters (such as temperature, salinity) in the sea area to identify the sea front region according to the reasonable threshold.

However, the conventional method at present is difficult to set a reasonable threshold for different sea areas and different seasons, which results in low recognition accuracy of the ocean front area in the sea area.

Disclosure of Invention

In view of the above, it is necessary to provide a method, an apparatus, a computer device and a storage medium for acquiring a sea front region, which can improve the accuracy of identifying the sea front region in a sea area.

A method of acquiring a marine front region, the method comprising:

acquiring satellite remote sensing observation data, and determining a rough estimation ocean front area according to the satellite remote sensing observation data; the satellite remote sensing observation data comprises at least one of a water color value, a temperature value, a salinity value, a chlorophyll concentration value and a suspended sediment concentration value of the sea area to be identified;

determining target paths of a plurality of unmanned ships to be dispatched to the rough estimation ocean front area and target time lengths of the unmanned ships to the rough estimation ocean front area along the target paths according to the satellite remote sensing observation data; wherein the target path is used to characterize a shortest path for the plurality of unmanned vessels to reach the coarse estimation sea front area;

determining each initial position of each unmanned ship in the rough estimation sea front area according to received arrival confirmation information sent when each unmanned ship arrives at the rough estimation sea front area;

Determining regional change information of the rough estimation ocean front region according to a plurality of groups of satellite remote sensing observation data received in the target time length;

outputting an adjusting instruction according to the region change information to indicate that each unmanned naval vessel is adjusted based on the corresponding initial position, so as to obtain the adjusted position of each unmanned naval vessel;

and determining a target ocean front region according to the received adjusted positions.

In one embodiment, the method further comprises the following steps:

acquiring a plurality of groups of satellite remote sensing observation data at different moments in the target time length;

determining a plurality of roughly estimated ocean front areas according to the plurality of groups of satellite remote sensing observation data;

and carrying out interpolation fitting processing on the plurality of roughly estimated ocean front regions according to the plurality of different moments to obtain the region change information.

In one embodiment, the method further comprises the following steps:

determining the area estimation value and frontal line length of the roughly estimated ocean front area according to the satellite remote sensing observation data;

determining the number of unmanned ships to be dispatched according to the area estimation value and the first ratio result of the frontal line length;

determining a plurality of linear distances between each unmanned ship and each preset vertex position of the rough estimation sea front region, and selecting the smallest one of the linear distances corresponding to each unmanned ship as a target path for the unmanned ship to reach the rough estimation sea front region;

And acquiring the maximum speed per hour of the plurality of unmanned ships, and taking a second ratio result of the target path of the unmanned ships reaching the rough estimation ocean front area and the maximum speed per hour as the target duration.

In one embodiment, further comprising:

determining a visual region distribution map of the rough estimation ocean front region according to the satellite remote sensing observation data;

according to the position coordinates of each vertex of the visual region distribution diagram, carrying out regular graphic calibration processing on the visual region distribution diagram to obtain a target rectangular frame containing the visual region distribution diagram;

determining an area estimation value of the roughly estimated sea front region corresponding to the target rectangular frame;

and dividing the contour line of the visual region distribution diagram into a plurality of contour points, and linearly superposing the distance between two adjacent contour points to obtain the frontal line length.

In one embodiment, further comprising:

carrying out inversion processing on the satellite remote sensing observation data to obtain a sea area distribution map of the sea area to be identified;

if the sea area distribution map meets a pre-stored interface determining condition, determining an interface existing in the sea area distribution map as the rough estimation sea front area;

And carrying out visualization processing on the rough estimation sea front area to obtain a visualization area distribution map of the rough estimation sea front area.

A method of predicting a marine front region, the method comprising:

outputting an instrument layout instruction according to the target ocean front region; the instrument layout instruction is used for indicating that a plurality of detection instruments are crossly laid at intervals according to a preset distance;

acquiring sea area observation elements detected by the detection instrument, and performing cleaning treatment operation on the acquired sea area observation elements to obtain cleaned observation elements; wherein the cleaning treatment operation comprises a missing value filling operation and/or an abnormal value deleting operation, and the sea area observation elements comprise the humidity, salinity, temperature, wind speed, wind direction, flow speed, wave height, wave direction, wave period, water depth and navigational speed of the sea area to be observed;

performing feature selection processing on the cleaned observation elements by using a support vector machine recursive feature elimination method to obtain target features;

inputting the target characteristics and the set time into a preset marine front prediction model to obtain a predicted marine front region of the target marine front region at the preset time; the preset ocean front prediction model comprises a model which is trained by using the target characteristics to a random forest model;

And carrying out visualization processing on the predicted ocean front region to obtain a predicted region visualization image of the predicted ocean front region.

In one embodiment, further comprising:

the target characteristics are used as input of a random forest model, and model training is carried out by combining with the initial value of the sensitive parameters of the random forest model to obtain a preliminary result after model training;

determining different groups of values of the sensitive parameters according to the preliminary result, and determining different root mean square errors between predicted values and actual observed values of the random forest model after the different groups of values and the target feature verification;

and sorting the different root mean square errors, and taking the random forest model corresponding to the smallest root mean square error as the preset ocean front prediction model.

An acquisition system for a marine front region, the system comprising: the processor and the unmanned ship formation controller are respectively arranged in a land base station of a sea area to be identified, and the processor is electrically connected with the unmanned ship formation controller;

the processor is used for acquiring satellite remote sensing observation data and determining a rough estimation ocean front area according to the satellite remote sensing observation data; determining target paths of a plurality of unmanned ships to be dispatched to the rough estimation ocean front area and target time lengths of the unmanned ships to the rough estimation ocean front area along the target paths according to the satellite remote sensing observation data; determining each initial position of each unmanned ship in the rough estimation sea front area according to received arrival confirmation information sent when each unmanned ship arrives at the rough estimation sea front area; determining regional change information of the rough estimation ocean front region according to a plurality of groups of satellite remote sensing observation data received in the target time length; outputting an adjusting instruction according to the region change information to indicate that each unmanned naval vessel is adjusted based on the corresponding initial position, so as to obtain the adjusted position of each unmanned naval vessel; determining a target ocean front region according to the received adjusted positions; the satellite remote sensing observation data comprise at least one of a water color value, a temperature value, a salinity value, a chlorophyll concentration value and a suspended sediment concentration value of a sea area to be identified, and the target paths are used for representing the shortest paths of the plurality of unmanned ships to the roughly estimated ocean front area;

The unmanned ship formation controller is used for controlling the unmanned ships with corresponding number to run according to the number of the unmanned ships to be dispatched.

An acquisition apparatus for a marine front region, the apparatus comprising:

the first determination module is used for acquiring satellite remote sensing observation data and determining a rough estimation ocean front area according to the satellite remote sensing observation data; the satellite remote sensing observation data comprises at least one of a water color value, a temperature value, a salinity value, a chlorophyll concentration value and a suspended sediment concentration value of the sea area to be identified;

the second determination module is used for determining target paths of a plurality of unmanned ships to be dispatched to the rough estimation ocean front area and target time lengths of the unmanned ships to reach the rough estimation ocean front area along the target paths according to the satellite remote sensing observation data; wherein the target path is used to characterize a shortest path for the plurality of unmanned vessels to reach the coarse estimation sea front area;

a third determining module, configured to determine, according to received arrival confirmation information sent when each unmanned ship arrives at the rough estimation sea front region, each initial position of each unmanned ship in the rough estimation sea front region;

The fourth determination module is used for determining the regional change information of the rough estimation ocean front region according to a plurality of groups of satellite remote sensing observation data received within the target time length;

the adjusting module is used for outputting an adjusting instruction according to the region change information so as to indicate each unmanned naval vessel to adjust based on the corresponding initial position of each unmanned naval vessel, and the adjusted position of each unmanned naval vessel is obtained;

and the fifth determining module is used for determining the target ocean front area according to the received adjusted positions.

An apparatus for predicting a region of a ocean front, the apparatus comprising:

the instruction output module is used for outputting an instrument layout instruction according to the target sea front region in the acquisition method of the sea front region; the instrument layout instruction is used for indicating that a plurality of detection instruments are crossly laid at intervals according to a preset distance;

the cleaning processing module is used for acquiring the sea area observation elements detected by the detection instrument and performing cleaning processing operation on the acquired sea area observation elements to obtain cleaned observation elements; wherein the cleaning treatment operation comprises a missing value filling operation and/or an abnormal value deleting operation, and the sea area observation elements comprise the humidity, salinity, temperature, wind speed, wind direction, flow speed, wave height, wave direction, wave period, water depth and navigational speed of the sea area to be observed;

The characteristic selection module is used for carrying out characteristic selection processing on the cleaned observation elements by using a support vector machine recursive characteristic elimination method to obtain target characteristics;

the prediction module is used for inputting the target characteristics and the set time into a preset marine front prediction model to obtain a predicted marine front region of the target marine front region at the preset time; the preset ocean front prediction model comprises a model which is trained by using the target characteristics to a random forest model;

and the visualization processing module is used for performing visualization processing on the predicted ocean front region to obtain a predicted region visualization image of the predicted ocean front region.

A computer device comprising a memory and a processor, the memory storing a computer program, the processor implementing the following steps when executing the computer program:

A computer-readable storage medium, on which a computer program is stored which, when executed by a processor, carries out the steps of:

The method for acquiring the ocean front region comprises the steps of firstly determining a roughly estimated ocean front region according to acquired satellite remote sensing observation data, wherein the satellite remote sensing observation data comprises at least one index data of a water color value, a temperature value, a salinity value, a chlorophyll concentration value and a suspended sediment concentration value of a sea region to be identified, so that the purpose of determining the roughly estimated ocean front region according to two or more index data in the satellite remote sensing observation data can be realized, the defects of high identification difficulty and single mode caused by the fact that the ocean front region can be identified only after threshold values are set according to different index numbers in the prior art are overcome, and the flexibility and the timeliness for determining the roughly estimated ocean front region are improved; furthermore, target paths of a plurality of unmanned ships to be dispatched to reach the rough estimation sea front area and target duration of the unmanned ships reaching the rough estimation sea front area along the target paths are determined according to the satellite remote sensing observation data, so that the problem of low identification precision of the sea front area caused by the fact that the traditional technology can only judge whether the sea front exists in the sea area is solved, and a foundation is laid for obtaining an accurate sea front area subsequently; then determining each initial position of each unmanned ship in the rough estimation sea front area according to received arrival confirmation information sent when each unmanned ship arrives at the rough estimation sea front area; determining regional change information of the roughly estimated sea front region according to a plurality of groups of satellite remote sensing observation data received in the target time length, so as to achieve the purpose of capturing development changes of the roughly estimated sea front region in real time in the target time length and provide a basis for subsequently determining an accurate sea front region; outputting an adjusting instruction according to the region change information to indicate that each unmanned naval vessel is adjusted based on the corresponding initial position, so as to obtain the adjusted position of each unmanned naval vessel; and determining a target ocean front region according to the received adjusted positions. Therefore, the purposes that when a plurality of unmanned ships reach the roughly estimated sea front region, respective initial positions can be adjusted in real time according to the region change information, and the sea front region can be accurately positioned according to the adjusted positions are achieved, the defects that in the prior art, a large amount of sea front identification prior information about the sea region is needed to be used as reference information to identify the sea front region, and therefore the identification method is not reusable and low in universality are overcome, and the reusability and universality of the sea front region are improved; further, each initial position of each unmanned ship is adjusted according to the regional change information of the roughly estimated sea front region in the target duration, so that the purpose that the roughly estimated sea front region is repositioned according to the adjusted position to obtain a precise sea front region is achieved, and therefore the accuracy and the reliability of determining the target sea front region are improved.

Drawings

FIG. 1 is a schematic diagram of a marine front;

FIG. 2 is a schematic flow chart illustrating a method for obtaining a marine front region according to an embodiment;

FIG. 3 is a schematic flow chart illustrating a method for obtaining a marine front region in another embodiment;

FIG. 4 is a schematic flow chart illustrating a method for obtaining a marine front region in yet another embodiment;

FIG. 5 is a schematic flow chart illustrating a method for obtaining a marine front region in yet another embodiment;

FIG. 6 is a flow diagram illustrating a method for predicting a marine front region, according to one embodiment;

FIG. 7 is a block diagram of an apparatus for acquiring a marine front region according to an embodiment;

FIG. 8 is a block diagram of an apparatus for predicting a region of a ocean front in one embodiment;

FIG. 9 is a diagram of the internal architecture of a system for acquiring a marine front region in one embodiment;

FIG. 10 is a diagram showing an internal structure of a computer device according to an embodiment.

Detailed Description

The narrow zone of intersection between two or more bodies of water of significantly different nature in the ocean is called the ocean front, as shown in fig. 1, and the ocean front is a jump zone of ocean environmental parameters, which can be described by elements such as seawater temperature, salinity, density, velocity, color, chlorophyll, etc. The size of the ocean front can reach hundreds of kilometers, and the ocean front exists in the surface layer, the middle layer and the bottom layer of the ocean. The intense turbulent mixing exchange, horizontal convergence (divergence) and vertical motion present in the frontal region. These ocean front motions not only affect shipping and fishery, but also affect activities such as underwater communication, ship safety, maritime search and rescue, and the like; the accurate real-time identification of the ocean front is the key for solving the problems, the traditional ocean front identification method is a gradient threshold method, and because different parameters (such as temperature, salinity and the like) have different standards, even if the parameters are the same, the standards for division and definition are different in different sea areas and different seasons, the gradient threshold method is difficult to determine a reasonable threshold; the current method can only judge whether the ocean front exists in the target sea area, and can not identify the form details of the ocean front; in addition, the existing ocean front identification method can identify the ocean front of the local sea area, but has the problem that the non-local sea area identification standard is not uniform. In the past, a large amount of experience needs to be summarized to provide reference information for sea area division, and for different sea areas, the methods have no recyclable characteristic, so that the method has no good universality; furthermore, although the existing knowledge about the ocean front is mostly limited to the identification thereof, the number of prediction methods established based on the identification features is small.

In view of the above problems in the prior art, the present application provides a method and an apparatus for acquiring a marine front region, a computer device, and a storage medium. In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.

According to the acquisition method of the ocean front region, an execution main body can be an acquisition device of the ocean front region, and the acquisition device of the ocean front region can be realized into part or all of computer equipment in a software, hardware or software and hardware combination mode. Optionally, the Computer device may be an electronic device with a processor function, such as a Personal Computer (PC), a portable device, a notebook Computer, a smart phone, a tablet Computer, a portable wearable device, and the like, for example, a tablet Computer, a mobile phone, and the like, and the specific form of the Computer device is not limited in the embodiments of the present application.

It should be noted that the execution subject of the method embodiments described below may be part or all of the computer device described above. The following method embodiments are described by taking the execution subject as the computer device as an example.

In one embodiment, as shown in fig. 2, there is provided a method for acquiring a sea front region, including the steps of:

step S11, acquiring satellite remote sensing observation data, and determining a rough estimation ocean front area according to the satellite remote sensing observation data; the satellite remote sensing observation data comprise at least one of a water color value, a temperature value, a salinity value, a chlorophyll concentration value, a suspended sediment concentration value and the like of the sea area to be identified.

The water color value of the sea area to be identified can be determined by the optical properties of the seawater, suspended substances contained in the seawater, the depth of the seawater, the characteristics of cloud layers and other factors; said salinity value may comprise the ratio of total dissolved solids in seawater to the weight of seawater, typically expressed in grams per kilogram of seawater; the chlorophyll concentration values may include chlorophyll concentration values of different layers of the sea area to be identified, such as a surface layer, a subsurface layer, a middle layer and a bottom layer, and may be determined jointly by factors such as an ocean water temperature environment, nutrient distribution, phytoplankton species, and the like; the sea area silt content that many factors such as silt source, the system of flowing through of suspension silt concentration value can the sign disappear length, velocity of flow strong and weak and stormy waves and produce, just suspension silt concentration value can change along with at least one change in season, size tide and the tide that rises, for example the silt content when the tide is little is higher than the tide, and the silt content when the tide falls is higher than the tide that rises, and the silt content is higher relatively in winter and spring, and the summer and autumn silt content is relatively low.

Specifically, when the computer device obtains the satellite remote sensing observation data, it may set a water color value, a temperature value, a salinity value, a chlorophyll concentration value, and a suspended sediment concentration value included in the satellite remote sensing observation data as different index data, then select one index data (such as a salinity value) in the remote sensing observation data to perform inversion processing and/or visualization processing, obtain a salinity visualization distribution map, compare the salinity visualization distribution map with a pre-stored interface determination condition, when a first target boundary matched with the interface determination condition exists in the salinity visualization distribution map, determine the first target interface as a rough estimation sea front region, for example, determine that a narrow band is a first target interface when a sea water salinity difference value at two sides of the narrow band in the visualization salinity distribution map is greater than or equal to a preset salinity difference threshold value, and if the salinity visualization distribution map does not have a first target interface matched with the interface determination condition, namely the salinity difference value of the seawater on the two sides of the narrow band is smaller than a preset salinity difference value threshold value, determining that no ocean front region exists in the sea area to be identified.

Then, selecting another index data (such as suspended sediment concentration value) from the satellite remote sensing observation data to perform inversion processing and/or visualization processing to obtain a suspended sediment concentration visualization distribution diagram, performing the same matching operation with the previously stored interface determination condition on the suspended sediment concentration visualization distribution diagram, determining a second target interface existing in the suspended sediment concentration visualization distribution diagram, comparing the first target interface with the second target interface, if the goodness of fit of the first target interface and the second target interface is higher than a preset goodness of fit threshold, determining that the first target interface is reasonable, and if the goodness of fit of the first target interface and the second target interface is lower than the preset goodness of fit threshold, based on a difference point or region of the two previous index data (such as salinity value and suspended sediment concentration value), selecting another index data (such as a water color value) in the satellite remote sensing observation data to perform inversion processing and/or visual processing to obtain a water color visual distribution map, and judging whether the coincidence degree of a third target interface and a second target interface in the water color visual distribution map is higher than a preset coincidence degree threshold value through the same operation; until the goodness of fit of the target interface determined based on the previous index data and the target interface determined based on the current index data is higher than a preset goodness of fit threshold, at this time, it may be determined that the first target interface is a rough estimated ocean front region. Wherein the rough estimate sea front region may comprise a region formed by at least two contour lines. Alternatively, the goodness-of-fit threshold may be 80%.

Step S12, determining target paths of a plurality of unmanned ships to be dispatched to reach the rough estimation ocean front area and target time lengths of the unmanned ships to reach the rough estimation ocean front area along the target paths according to the satellite remote sensing observation data; wherein the target path is used to characterize a shortest path for the plurality of unmanned vessels to reach the coarse estimation sea front area.

Specifically, as shown in fig. 3, the computer device may determine the target path and the target duration by the following sub-steps:

and step S121, determining the area estimation value and frontal line length of the roughly estimated ocean front area according to the satellite remote sensing observation data.

As shown in fig. 4, the process of determining the area estimation value and the frontal line length by the computer device may include:

and S1211, determining a visual region distribution diagram of the coarse estimation ocean front region according to the satellite remote sensing observation data.

Specifically, the process of the computer device determining the visualized region distribution map of the coarse estimation sea front region may comprise: firstly, carrying out inversion processing on the satellite remote sensing observation data to obtain a sea area distribution map of the sea area to be identified; if the sea area distribution map meets a pre-stored interface determining condition, determining an interface existing in the sea area distribution map as the rough estimation sea front area; and carrying out visualization processing on the rough estimation sea front area to obtain a visualization area distribution map of the rough estimation sea front area.

In the actual processing process, the process of determining the roughly estimated sea front region by the computer device according to the satellite remote sensing observation data is the same as the process of determining the roughly estimated sea front region in step S11, and details are not repeated here. And the computer device determined visualization region distribution map of the rough estimation sea front region may comprise visualization contour lines forming at least two contour lines of the rough estimation sea front region.

Step S1212, performing regular graph calibration processing on the visualization region distribution map according to each vertex position coordinate of the visualization region distribution map, to obtain a target rectangular frame including the visualization region distribution map.

Specifically, when determining the visualized area distribution map of the rough estimation sea front area, the computer device may first obtain a topographic map of the visualized area distribution map, then determine, according to the topographic map, position coordinates of each vertex of the rough estimation sea front area, where the position coordinates of each vertex may include M coordinate positions of M vertices of the rough estimation sea front area, then select the M vertices of the rough estimation sea front area, and determine, in a regular image calibration manner (for example, a rectangular frame calibration manner), a target rectangular frame including the visualized area distribution map, where the target rectangular frame may also be a rectangular frame including the rough estimation sea front area. Wherein M is a positive integer. Alternatively, M may take the value 4.

Step S1213, determining an area estimation value of the rough estimation ocean front region corresponding to the target rectangular frame.

Specifically, when the computer device determines the target rectangular box containing the visualization region distribution map, the computer device may further determine an area value of the target rectangular box, and use the area value of the target rectangular box as the estimated area value of the rough estimation sea front region.

Step S1214, dividing the contour line of the visual region distribution map into a plurality of contour points, and linearly superimposing the distances between two adjacent contour points to obtain the frontal line length.

Specifically, computer equipment is determining the regional distribution diagram of the visual region in rough estimate ocean front region, also determines formation that includes in the regional distribution diagram of the visual region in rough estimate ocean front region when the visual contour line of two at least contour lines in rough estimate ocean front region, can further evenly divide a plurality of contour points with every visual contour line according to preset distance interval to calculate the distance between every two adjacent contour points, then carry out linear stack with each distance calculated, obtain the frontal line length in rough estimate ocean front region.

And S122, determining the number of unmanned ships to be dispatched according to the area estimation value and the first ratio result of the frontal line length.

Specifically, when the area estimation value and the frontal line length of the roughly estimated sea front region are determined, the computer device may further determine a first ratio result of the area estimation value and the frontal line length, and determine the number of unmanned ships to be dispatched according to the first ratio result, that is, a first target ratio obtained by rounding up the first ratio result is used as the number of unmanned ships to be dispatched. Alternatively, when a first target ratio obtained by rounding the first ratio result is less than 2, it may be determined that 2 unmanned vessels are to be dispatched.

Step S123, determining a plurality of linear distances between each unmanned ship and each preset vertex position of the rough estimation sea front area, and selecting the smallest one of the plurality of linear distances corresponding to each unmanned ship as a target path for the unmanned ship to reach the rough estimation sea front area.

Specifically, when the number of unmanned ships to be dispatched is determined by the computer device, for example, when the number of unmanned ships to be dispatched is P, the P unmanned ships may be set to be located at the same starting point, each vertex coordinate position of the rough estimation sea front region is set to be each preset vertex position, for example, the number of the preset vertex positions of the device is T, then, the linear distances between the starting point and each preset vertex position are calculated, so that T linear distances are obtained, at this time, the T linear distances are sorted according to different distance values, the linear distance with the smallest distance value is selected from the T linear distances sorted according to the size, and the selected linear distance with the smallest distance value is used as a target path for the unmanned ships to reach the rough estimation sea front region. For example, when P unmanned vessels are located at the same starting point O and the preset vertex positions include the coordinate position of A, B, C, D, the linear distances of OA, OB, OC and OD are calculated, and the linear distance with the smallest distance value among the linear distances of OA, OB, OC and OD is used as the target path of the unmanned vessel to the rough estimation ocean front region.

Step S124, obtaining the maximum speed per hour of the plurality of unmanned ships, and taking the second ratio result of the target path of the unmanned ships reaching the rough estimation ocean front area and the maximum speed per hour as the target duration.

Specifically, when the computer device determines that the number of the unmanned ships to be dispatched is P, it may further determine, according to the number of the unmanned ships to be dispatched and the weight of each unmanned ship, target carrier devices that can carry P unmanned ships, for example, the target carrier devices may be large ships, and the carrier speed per hour of the target carrier devices may be greater than the maximum ship speed per hour of the P unmanned ships. Then, the carrier hourly speed of the target carrier device is used as the maximum hourly speed of the plurality of unmanned ships, a second ratio result of a target path of the unmanned ships reaching the rough estimation ocean front area to the maximum hourly speed is further determined, and the second ratio result is used as the target time length of the plurality of unmanned ships to be dispatched reaching the rough estimation ocean front area.

Step S13, determining each initial position of each unmanned ship in the rough estimation sea front area according to the received arrival confirmation information sent when each unmanned ship arrives at the rough estimation sea front area.

Specifically, when the number of unmanned ships to be dispatched is determined to be P, the computer device may first send a marine front area detection instruction to the unmanned ship formation controller, where the marine front area detection instruction carries the number of unmanned ships to be dispatched, the target carrier device for carrying the P unmanned ships, the maximum time speed of the target carrier device, a target path for the plurality of unmanned ships to arrive at the rough estimation marine front area, and a target time length for the plurality of unmanned ships to arrive at the rough estimation marine front area along the target path, and the unmanned ship formation controller may control the target carrier device for carrying the P unmanned ships to arrive at the rough estimation marine front area along the target path when the target time length is reached, and send confirmation arrival information to the computer device, where the confirmation arrival information may be used to represent that the P unmanned ships have arrived at a vertex coordinate position in the rough estimation marine front area.

When the computer equipment receives the arrival confirmation information sent by the unmanned ship formation controller, the initial positions of the unmanned ships in the rough estimation sea front area can be determined, namely when the number of the unmanned ships to be dispatched is P, the P unmanned ships can be used for determining the initial positions of the unmanned ships in the rough estimation sea front area according to the front line length of the rough estimation sea front area at equal intervals, and position instruction information carrying the P unmanned ships at the initial positions of the rough estimation sea front area is sent to the unmanned ship formation controller, so that the unmanned ship formation controller controls the unmanned ships to run to the corresponding initial positions in the rough estimation sea front area.

In the actual processing process, the land base station of the sea area to be identified may include a remote controller, a carrier PC of a control platform, and a communication tool, and a plurality of unmanned vessels may be stored in the land base station, and each unmanned vessel may include the unmanned vessel itself, and communication transmission devices such as a GPS, an inertial navigation unit, and other sensors, radio stations, and the like mounted thereon.

And step S14, determining the regional change information of the roughly estimated ocean front region according to the multiple groups of satellite remote sensing observation data received in the target time length.

Specifically, as shown in fig. 5, the computer device may determine the region variation information of the rough estimation sea front region by the following sub-steps:

and step S141, acquiring a plurality of groups of satellite remote sensing observation data at different moments in the target time length.

Specifically, the computer device may obtain satellite remote sensing observation data in real time during the process that the plurality of unmanned vessels travel to the rough estimation ocean front region along the target path, so as to obtain a plurality of sets of satellite remote sensing observation data at different times within a target time length when the plurality of unmanned vessels reach the rough estimation ocean front region along the target path, where each set of satellite remote sensing observation data includes at least one of a water color value, a temperature value, a salinity value, a chlorophyll concentration value, and a suspended sediment concentration value of a sea area to be identified, which are obtained at the corresponding time.

And S142, determining a plurality of roughly estimated ocean front areas according to the plurality of groups of satellite remote sensing observation data.

Specifically, when the computer device acquires the plurality of sets of satellite remote sensing observation data, one rough estimation ocean front region may be correspondingly determined for each set of satellite remote sensing observation data, so as to determine a plurality of rough estimation ocean front regions, where the plurality of rough estimation ocean front regions correspond to different acquisition times. The process of determining the rough estimation sea front region corresponding to the acquisition time according to each group of satellite remote sensing observation data is the same as the process of determining the rough estimation sea front region in step S11, and is not repeated here.

And step S143, performing interpolation fitting processing on the plurality of roughly estimated ocean front regions according to the plurality of different moments to obtain the region change information.

Specifically, when determining the plurality of coarse estimation sea front regions, the computer device may further perform interpolation fitting processing on the plurality of coarse estimation sea front regions according to different respective corresponding acquisition times of the plurality of coarse estimation sea front regions to obtain interpolation fitting processed sea front region information, where the interpolation fitting processed sea front region information may be used to represent development change paths of the coarse estimation sea front regions within the target duration, such as leftward and/or rightward movement. Therefore, the interpolation-fit processed sea front region information may be used as the region variation information of the coarse estimation sea front region.

And step S15, outputting an adjusting instruction according to the region change information to indicate that each unmanned naval vessel is adjusted based on the corresponding initial position, so as to obtain the adjusted position of each unmanned naval vessel.

Specifically, when the computer device determines the region change information of the rough estimation ocean front region, the computer device may first obtain a corresponding relationship between the rough estimation ocean front region and the region change information, and then according to the corresponding relationship, according to the region change information and P initial positions of P unmanned ships in the rough estimation ocean front region, respectively send an adjustment instruction carrying the region change information to the unmanned ships at each initial position to instruct the unmanned ships at each initial position to adjust based on the respective corresponding initial position, so as to obtain P adjusted positions of the P unmanned ships, and send the P adjusted positions to the computer device. Wherein each adjusted position may be used to characterize a development change of the initial position of the corresponding unmanned vessel over the target duration.

In an actual processing process, when the computer device determines the region change information of the rough estimation sea front region, a position adjustment instruction can be sent to the unmanned ship formation controller according to the obtained corresponding relationship between the rough estimation sea front region and the region change information, wherein the position adjustment instruction carries the region change information and the corresponding relationship, so that the unmanned ship formation controller controls and adjusts P initial positions of P unmanned ships according to the region change information and the corresponding relationship, and then P adjusted positions of the P unmanned ships are obtained and sent to the computer device.

And step S16, determining a target ocean front region according to the received adjusted positions.

Specifically, when receiving each adjusted position sent by each unmanned ship or unmanned ship formation controller, the computer device may determine the target ocean front region again according to each adjusted position. Wherein the target ocean front region may comprise a final ocean front region that is precisely identified.

According to the method for acquiring the ocean front region, the roughly estimated ocean front region is determined according to the acquired satellite remote sensing observation data, and the satellite remote sensing observation data comprises at least one index data of a water color value, a temperature value, a salinity value, a chlorophyll concentration value and a suspended sediment concentration value of the ocean to be identified, so that the purpose of determining the roughly estimated ocean front region according to two or more index data in the satellite remote sensing observation data can be realized, the defects of high identification difficulty and single mode caused by the fact that the ocean front region can be identified only after thresholds are set according to different index numbers in the prior art are overcome, and the flexibility and timeliness for determining the roughly estimated ocean front region are improved; furthermore, target paths of a plurality of unmanned ships to be dispatched to reach the rough estimation sea front area and target duration of the unmanned ships reaching the rough estimation sea front area along the target paths are determined according to the satellite remote sensing observation data, so that the problem of low identification precision of the sea front area caused by the fact that the traditional technology can only judge whether the sea front exists in the sea area is solved, and a foundation is laid for obtaining an accurate sea front area subsequently; then determining each initial position of each unmanned ship in the rough estimation sea front area according to received arrival confirmation information sent when each unmanned ship arrives at the rough estimation sea front area; determining regional change information of the roughly estimated sea front region according to a plurality of groups of satellite remote sensing observation data received in the target time length, so as to achieve the purpose of capturing development changes of the roughly estimated sea front region in real time in the target time length and provide a basis for subsequently determining an accurate sea front region; outputting an adjusting instruction according to the region change information to indicate that each unmanned naval vessel is adjusted based on the corresponding initial position, so as to obtain the adjusted position of each unmanned naval vessel; and determining a target ocean front region according to the received adjusted positions. Therefore, the purposes that when a plurality of unmanned ships reach the roughly estimated sea front region, respective initial positions can be adjusted in real time according to the region change information, and the sea front region can be accurately positioned according to the adjusted positions are achieved, the defects that in the prior art, a large amount of sea front identification prior information about the sea region is needed to be used as reference information to identify the sea front region, and therefore the identification method is not reusable and low in universality are overcome, and the reusability and universality of the sea front region are improved; further, each initial position of each unmanned ship is adjusted according to the regional change information of the roughly estimated sea front region in the target duration, so that the purpose that the roughly estimated sea front region is repositioned according to the adjusted position to obtain a precise sea front region is achieved, and therefore the accuracy and the reliability of determining the target sea front region are improved.

In one embodiment, as shown in fig. 6, a method for predicting a marine front region is provided, comprising the following steps:

step S21, outputting an instrument layout instruction according to the target ocean front region in the ocean front region prediction method; the instrument layout instruction is used for indicating that a plurality of detection instruments are arranged at intervals in a crossed mode according to a preset distance.

Specifically, the instrument layout instruction may carry attribute information of the instrument to be laid, where the attribute information includes a type of the instrument to be laid and a number of each instrument. Alternatively, the apparatus to be deployed may include a thermohaline deep section profiler (CTD), an Acoustic Doppler Current Profiler (ADCP), an optical backscatter turbidity meter (OBS), an acoustic backscatter meter (ABS), etc., the CTD is a water body detector for measuring the temperature and salinity of seawater at different depths, the ADCP is a hydroacoustic current meter for measuring the water velocity, the OBS is an optical instrument for measuring the turbidity of the water body, and the ABS is an acoustic instrument for measuring the silt particle concentration, particle size, and silt flux.

In the actual processing process, in order to improve the acquisition precision of the information of the ocean front region, the computer device may lay a plurality of detection instruments according to the following process:

(1) Firstly, a computer device preliminarily judges an initial contour line of a target ocean front region according to received satellite remote sensing observation data, and outputs first initial layout information to indicate that two unmanned ships are respectively laid on two sides of the initial contour line, and each unmanned ship carries the same instrument comprising CTD, ADCP, OBS, ABS and the like.

(2) And then the computer equipment issues the instrument of the initial layout information on each unmanned ship to the water through a control instruction so as to control the instrument to observe dynamic environment information at two sides of the initial contour line of the ocean front region at the same time interval (for example, 5-10 minutes), wherein the dynamic environment information comprises flow velocity, temperature, salinity, turbidity, sand particle concentration, particle size, sediment flux and the like, and the observed dynamic environment information is sent to a terminal monitoring centralized control system.

(3) And then the computer equipment receives the power environment information sent by the terminal monitoring centralized control system, analyzes and compares the power environment information of the two unmanned naval vessels on the two sides of the initial contour line, if the turbidity, the sand particle concentration, the particle size and the silt flux index on the two sides are remarkably different (for example, the difference of at least one data quantity value is more than 10 times), the initial contour line is judged to be correct and reasonable, and at the moment, a distance reducing instruction is sent to an unmanned naval vessel formation controller to indicate that the distance between the two unmanned naval vessels on the two sides of the initial contour line is reduced.

(4) Enabling the distance between the two unmanned ships after the distance is reduced to be a new initial contour line, repeatedly performing the steps of analyzing and comparing the dynamic environment information of the two unmanned ships at two sides of the initial contour line until the turbidity, the sand particle concentration, the particle size and the silt flux index difference at two sides of the initial contour line are not significant (for example, the difference of at least one data value is more than 10 times), stopping sending a distance reduction instruction, and taking the relative positions of the unmanned ships at two sides of the current target ocean front area as optimal positions; if the indexes of turbidity, sand particle concentration, particle size and silt flux on the two sides are not significantly different (for example, the difference of data values reaches more than 10 times) after analysis and comparison, sending a distance expansion instruction to the unmanned ship formation controller to instruct to expand the distance between the two unmanned ships on the two sides of the initial contour line; and stopping sending the distance expansion instruction until indexes of turbidity, sand particle concentration, particle size and silt flux on two sides are remarkably different, and taking the relative positions of two unmanned naval vessels on two sides of the current ocean front region as optimal positions.

(5) Setting other two positions on two sides of the initial contour line as new positions of the two unmanned naval vessels, and returning to the step (1) to execute the step of obtaining the optimal position; until the relative positions of each two unmanned naval vessels arranged at the two sides of the initial contour line are judged to be the optimal positions. And determining a plurality of detection instruments distributed in the target ocean front region.

Step S22, obtaining sea area observation elements detected by the detecting instrument, and carrying out cleaning treatment operation on the obtained sea area observation elements to obtain cleaned observation elements; wherein the cleaning treatment operation comprises a missing value filling operation and/or an abnormal value deleting operation, and the sea area observation elements comprise the humidity, salinity, temperature, wind speed, wind direction, flow speed, wave height, wave direction, wave period, water depth and navigational speed of the sea area to be observed.

Specifically, the computer device may receive, in real time, the sea area observation elements detected by the detection instrument, and may perform a missing value filling operation on the obtained sea area observation elements, for example, the observation elements obtained at time t are complete (that is, the observation elements include humidity, salinity, temperature, wind speed, wind direction, flow speed, wave height, wave direction, wave period, water depth, and navigational speed), the observation elements other than salinity are obtained at time t +1, the observation data obtained at time t +2 and time t +3 are complete, and at this time, linear difference processing may be performed on the missing salinity information at time t +1 according to the observation elements obtained at time t +2, and time t +3, so as to fill a missing value, thereby obtaining observation elements after the missing value filling operation. For example, when the wind speed in the observation elements acquired at the time t 'exceeds a wind speed threshold value, the wind speed acquired at the time t' is deleted as an abnormal value, and the wind speed threshold value is the maximum value of the wind speed in the historical observation elements; the computer device may also perform an abnormal value deletion operation on the acquired sea area observation element, for example, when the wind speed in the observation element acquired at the time t 'exceeds a wind speed threshold, the wind speed acquired at the time t' is deleted as an abnormal value, and the wind speed threshold may be a maximum wind speed in the historical observation elements, so as to obtain the observation element after the abnormal value deletion operation, and thus it may be determined that the observation element after the cleaning processing may include the observation element after the missing value filling operation and/or the observation element after the abnormal value deletion operation.

And step S23, performing feature selection processing on the cleaned observation elements by using a support vector machine recursive feature elimination method to obtain target features.

Specifically, the computer device may adopt a support vector machine recursive feature elimination algorithm (SVM-RFE) to perform feature selection on the observed features after the cleaning processing, so as to obtain the target features. Among them, SVM-RFE is a gene selection method proposed by Guyon et al in classifying cancers, and higher classification accuracy can be obtained using SVM-RFE.

The characteristic extraction process by using the SVM-RFE algorithm comprises the following steps: the core of the SVM-RFE algorithm is that all the features are modeled and sorted in a descending order according to the influence weight of each feature on a result variable, the features with smaller influence weights are deleted earliest, and the remaining features of the remaining influence weights are used for re-modeling and sorting in the descending order, so that the loop is iterated continuously until only one remaining feature of the remaining influence weights exists.

The SVM-RFE algorithm can effectively reduce the number of features and has better robustness, and is a cyclic process, and each step comprises the following sub-steps:

1) the classifier is trained with the current data set to obtain features that are related to features of the classifier. For example, in a linear kernel support vector machine, the relevant information may include a weight for each feature.

2) And calculating the ranking criterion scores of the characteristics according to a preset rule.

3) The feature corresponding to the smallest ranking criterion score is removed from the current dataset.

And circularly executing the process until the last variable in the feature set is executed, finishing the circular process, wherein the corresponding execution result is a list of obtained feature sequence numbers which are sorted according to feature importance when the circular process is finished, and each feature sequence number corresponds to a target feature.

Step S24, inputting the target characteristics and the set time into a preset marine front prediction model to obtain a predicted marine front region of the target marine front region at the preset time; and the preset ocean front prediction model comprises a model which is trained by using the target characteristics to train a random forest model.

The set time may include any time after the target time of the target ocean front region is determined.

Specifically, the process of determining the preset ocean front prediction model by the computer device comprises the following steps: firstly, the target characteristics are used as the input of a random forest model, and model training is carried out by combining the initial value of the sensitive parameters of the random forest model to obtain a preliminary result after model training; then determining different groups of values of the sensitive parameters according to the preliminary result, and determining different root mean square errors between predicted values and actual observed values of the random forest model after the different groups of values and the target feature verification; and finally, sorting the different root mean square errors, and taking the random forest model corresponding to the smallest root mean square error as the preset ocean front prediction model.

In the actual processing process, the training process of the random forest model comprises the following substeps:

(1) there are N samples randomly selected back in the plurality of target features, and the N samples selected are used to train a decision tree, and the N samples may be samples at a root node of the decision tree. Wherein each sample is a corresponding target feature, and after selecting one sample each time, the random selection with putting back may include putting the selected sample back into the plurality of target features for random selection again.

(2) And when each sample is detected to comprise M attributes and the ith node of the decision tree needs to be split, randomly selecting M attributes from the M attributes, wherein M < < M. And then selecting 1 attribute from the m attributes by adopting a preset strategy (such as an information gain strategy) as the splitting attribute of the ith node.

(3) And (2) returning to the step (2) for continuously executing the step of obtaining the splitting attribute of the ith node by taking the i +1 as a new i until the splitting attribute of the ith node is the attribute corresponding to the splitting of the parent node of the ith node, and determining that the ith node reaches the leaf node without continuously splitting.

(4) Establishing a large number of decision trees according to the steps (1) to (3) to form a random forest, wherein the random forest has 2 parameters, one parameter is the number of trees in the forest, and a larger value is usually selected; the other is the size of M, which is usually equal to the root mean square of M. And the predicted value of the target variable in the random forest model is the mean value of the decision trees.

(5) And (4) based on the established random forest, comparing the mean value output by the random forest in the step (4) with the observed value of the target variable, and calculating a correlation coefficient and a root mean square error.

(6) By considering different input variable structures (namely the number of used variables, the same variable data type and the like), repeating the steps (1) to (5), obtaining statistical values (namely the correlation coefficient and the root mean square error) during model training when the variable structures are different, carrying out comparative analysis on the statistical values, and selecting the model with the most significant correlation coefficient and the least root mean square error as the optimal model.

Step S25, performing visualization processing on the predicted ocean front region to obtain a predicted region visualization image of the predicted ocean front region.

Specifically, when the computer device determines the predicted marine front region, the computer device may further perform visualization processing on the predicted marine front region to obtain a predicted region visualization image of the predicted marine front region, where the predicted region visualization image of the predicted marine front region may include visualization contour lines forming at least two contour lines of the predicted marine front region.

In this embodiment, an instrument layout instruction is output through the target sea front region determined in the sea front region prediction method, so as to achieve the purpose of acquiring sea observation elements in real time through various detection instruments arranged, and the obtained sea observation elements are subjected to cleaning processing operation including missing value filling operation and/or abnormal value deleting operation, and the cleaned observation elements are subjected to feature selection processing by using a support vector machine recursive feature elimination method to determine target features, so that the sea front element information can be captured dynamically and finely in real time, so that the purpose of extracting feature elements influencing sea front distribution change based on multi-element observation information is achieved, and guarantee is provided for subsequently improving the prediction accuracy of a development path of the target sea front region; furthermore, the target feature and the set time are input into the preset ocean front prediction model which is trained and trained by using the target feature, the prediction ocean front region of the target ocean front region at the preset time is determined, and the prediction region visualization image of the prediction ocean front region is determined by performing visualization processing on the prediction ocean front region, so that the purpose of predicting elements influencing ocean front distribution change by adopting an aggregation type decision tree random forest machine learning method is achieved, and the accuracy and the prediction precision of the prediction ocean front region are effectively improved.

It should be understood that although the various steps in the flow charts of fig. 2-6 are shown in order as indicated by the arrows, the steps are not necessarily performed in order as indicated by the arrows. The steps are not performed in the exact order shown and described, and may be performed in other orders, unless explicitly stated otherwise. Moreover, at least some of the steps in fig. 2-6 may include multiple steps or multiple stages, which are not necessarily performed at the same time, but may be performed at different times, which are not necessarily performed in sequence, but may be performed in turn or alternately with other steps or at least some of the other steps.

In one embodiment, as shown in fig. 7, there is provided an acquisition apparatus for a marine front region, comprising: a first determination module 11, a second determination module 12, a third determination module 13, a fourth determination module 14, an adjustment module 15, and a fifth determination module 16, wherein:

the first determining module 11 is used for acquiring satellite remote sensing observation data and determining a rough estimation ocean front area according to the satellite remote sensing observation data; the satellite remote sensing observation data comprise at least one of a water color value, a temperature value, a salinity value, a chlorophyll concentration value and a suspended sediment concentration value of the sea area to be identified.

The second determining module 12 is configured to determine, according to the satellite remote sensing observation data, a target path where the plurality of unmanned ships to be dispatched reach the rough estimation ocean front area and a target duration for reaching the rough estimation ocean front area along the target path; wherein the target path is used to characterize a shortest path for the plurality of unmanned vessels to reach the coarse estimation sea front area.

A third determining module 13, configured to determine each initial position of each unmanned ship in the rough estimation sea front region according to the received arrival confirmation information sent when each unmanned ship arrives at the rough estimation sea front region.

And a fourth determining module 14, configured to determine, according to multiple groups of satellite remote sensing observation data received within the target time duration, regional change information of the rough estimation ocean front region.

And the adjusting module 15 is configured to output an adjusting instruction according to the region change information to instruct each unmanned ship to adjust based on the corresponding initial position, so as to obtain an adjusted position of each unmanned ship.

A fifth determining module 16, configured to determine a target ocean front region according to each received adjusted position.

The second determining module 12 may include: a first determination submodule, a second determination submodule, a third determination submodule, and a fourth determination submodule.

Specifically, the first determining submodule may be configured to determine an area estimation value and a frontal line length of the roughly estimated ocean front region according to the satellite remote sensing observation data; the second determining submodule can be used for determining the number of unmanned ships to be dispatched according to the area estimated value and the first ratio result of the frontal line length; a third determining submodule, configured to determine a plurality of linear distances between each unmanned ship and each preset vertex position of the rough estimation sea front region, and select a minimum linear distance from the plurality of linear distances corresponding to each unmanned ship as a target path for the unmanned ship to reach the rough estimation sea front region; the fourth determining submodule may be configured to obtain a maximum speed per hour of the plurality of unmanned vessels, and use a second ratio result of the target path of the unmanned vessel reaching the rough estimation ocean front region and the maximum speed per hour as the target duration.

A first determination submodule, which may include: the device comprises a first determination unit, a calibration processing unit, a second determination unit and a third determination unit.

Specifically, the first determining unit may be configured to determine a visual region distribution map of the rough estimation ocean front region according to the satellite remote sensing observation data; the calibration processing unit may be configured to perform calibration processing of a regular graph on the visualization region distribution map according to each vertex position coordinate of the visualization region distribution map, so as to obtain a target rectangular frame including the visualization region distribution map; a second determining unit, configured to determine an area estimation value of the rough estimation sea front region corresponding to the target rectangular frame; the third determining unit may be configured to divide the contour line of the visualization region distribution map into a plurality of contour points, and linearly superimpose distances between two adjacent contour points to obtain the frontal line length.

The first determination unit may include: a first determining subunit, a second determining subunit, and a third determining subunit.

Specifically, the first determining subunit may be configured to perform inversion processing on the satellite remote sensing observation data to obtain a sea area distribution map of the sea area to be identified; a second determining subunit, configured to determine, if the sea area distribution map meets a pre-stored interface determination condition, that an interface where the sea area distribution map exists is the rough estimation sea front region; the third determining subunit may be configured to perform visualization processing on the rough estimation sea front region, so as to obtain a visualization region distribution map of the rough estimation sea front region.

The fourth determining module 14 may include: the obtaining submodule, the fifth determining submodule and the fitting processing submodule.

Specifically, the obtaining submodule can be used for obtaining a plurality of groups of satellite remote sensing observation data at different moments in the target duration; the fifth determining submodule can be used for determining a plurality of roughly estimated ocean front areas according to the plurality of groups of satellite remote sensing observation data; and the fitting processing sub-module can be used for carrying out interpolation fitting processing on the plurality of roughly estimated ocean front regions according to the plurality of different moments to obtain the region change information.

In one embodiment, as shown in fig. 8, there is provided a prediction apparatus of a marine front region, including: instruction output module 21, washing processing module 22, characteristic selection module 23, prediction module 24 and visualization processing module 25, wherein:

the instruction output module 21 may be configured to output an instrument layout instruction according to the target sea front region in the acquisition device of the sea front region; the instrument layout instruction is used for indicating that a plurality of detection instruments are crossly laid at intervals according to a preset distance;

a cleaning processing module 22, configured to obtain the sea area observation element detected by the detection instrument, and perform a cleaning processing operation on the obtained sea area observation element to obtain an observation element after the cleaning processing; wherein the cleaning treatment operation comprises a missing value filling operation and/or an abnormal value deleting operation, and the sea area observation elements comprise the humidity, salinity, temperature, wind speed, wind direction, flow speed, wave height, wave direction, wave period, water depth and navigational speed of the sea area to be observed;

The feature selection module 23 may be configured to perform feature selection processing on the cleaned observation elements by using a support vector machine recursive feature elimination method to obtain target features;

the prediction module 24 may be configured to input the target feature and the set time into a preset marine front prediction model, so as to obtain a predicted marine front region of the target marine front region at the preset time; the preset ocean front prediction model comprises a model which is trained by using the target characteristics to a random forest model;

the visualization processing module 25 may be configured to perform visualization processing on the predicted marine front region to obtain a predicted region visualization image of the predicted marine front region.

The prediction module 24 may include: the model training sub-module, the first processing sub-module and the second processing sub-module.

Specifically, the model training submodule may be configured to use the target feature as an input of a random forest model, and perform model training in combination with an initial value of a sensitive parameter of the random forest model to obtain a preliminary result after the model training; the first processing submodule can be used for determining different groups of values of the sensitive parameters according to the preliminary result and determining different root mean square errors between predicted values and actual observed values of the random forest model after the different groups of values and the target feature verification; and the second processing submodule can be used for sorting the different root mean square errors and taking the random forest model corresponding to the smallest root mean square error as the preset ocean front prediction model.

For the specific definition of the acquisition device of the marine front region, the above definition of the acquisition method of the marine front region may be referred to, and for the specific definition of the prediction device of the marine front region, the above definition of the prediction method of the marine front region may be referred to, and details are not described herein again. The acquisition means of the ocean front region the prediction means of the ocean front region and the modules in the sum may be implemented wholly or partially by software, hardware, or a combination thereof. The modules can be embedded in a hardware form or independent from a processor in the computer device, and can also be stored in a memory in the computer device in a software form, so that the processor can call and execute operations corresponding to the modules.

In one embodiment, as shown in fig. 9, there is provided a system for acquiring a marine front region, the system comprising: the processor and the unmanned ship formation controller are respectively arranged in a land base station of a sea area to be identified, the land base station comprises a remote controller, a carrier PC (personal computer) of a control platform and a communication tool, the processor is electrically connected with the unmanned ship formation controller, and the unmanned ship formation controller comprises:

The processor is used for acquiring satellite remote sensing observation data and determining a rough estimation ocean front area according to the satellite remote sensing observation data; determining target paths of a plurality of unmanned ships to be dispatched to the rough estimation ocean front area and target time lengths of the unmanned ships to the rough estimation ocean front area along the target paths according to the satellite remote sensing observation data; determining each initial position of each unmanned ship in the rough estimation sea front area according to received arrival confirmation information sent when each unmanned ship arrives at the rough estimation sea front area; determining regional change information of the rough estimation ocean front region according to a plurality of groups of satellite remote sensing observation data received in the target time length; outputting an adjusting instruction according to the region change information to indicate that each unmanned naval vessel is adjusted based on the corresponding initial position, so as to obtain the adjusted position of each unmanned naval vessel; determining a target ocean front region according to the received adjusted positions; the satellite remote sensing observation data comprise at least one of a water color value, a temperature value, a salinity value, a chlorophyll concentration value and a suspended sediment concentration value of the sea area to be identified, and the target paths are used for representing the shortest paths of the plurality of unmanned ships to the roughly estimated ocean front area.

The processor is also used for acquiring a plurality of groups of satellite remote sensing observation data at different moments in the target time length; determining a plurality of roughly estimated ocean front areas according to the plurality of groups of satellite remote sensing observation data; and carrying out interpolation fitting processing on the plurality of roughly estimated ocean front regions according to the plurality of different moments to obtain the region change information.

The processor is further used for determining the area estimation value and the frontal line length of the roughly estimated ocean front area according to the satellite remote sensing observation data; determining the number of unmanned ships to be dispatched according to the area estimation value and the first ratio result of the frontal line length; determining a plurality of linear distances between each unmanned ship and each preset vertex position of the rough estimation sea front region, and selecting the smallest one of the linear distances corresponding to each unmanned ship as a target path for the unmanned ship to reach the rough estimation sea front region; and acquiring the maximum speed per hour of the plurality of unmanned ships, and taking a second ratio result of the target path of the unmanned ships reaching the rough estimation ocean front area and the maximum speed per hour as the target duration.

The processor is further used for determining a visual region distribution map of the coarse estimation ocean front region according to the satellite remote sensing observation data; according to the position coordinates of each vertex of the visual region distribution diagram, carrying out regular graphic calibration processing on the visual region distribution diagram to obtain a target rectangular frame containing the visual region distribution diagram; determining an area estimation value of the roughly estimated sea front region corresponding to the target rectangular frame; and dividing the contour line of the visual region distribution diagram into a plurality of contour points, and linearly superposing the distance between two adjacent contour points to obtain the frontal line length.

The processor is further used for carrying out inversion processing on the satellite remote sensing observation data to obtain a sea area distribution map of the sea area to be identified; if the sea area distribution map meets a pre-stored interface determining condition, determining an interface existing in the sea area distribution map as the rough estimation sea front area; and carrying out visualization processing on the rough estimation sea front area to obtain a visualization area distribution map of the rough estimation sea front area.

The processor is further used for outputting an instrument layout instruction according to the target ocean front region; the instrument layout instruction is used for indicating that a plurality of detection instruments are crossly laid at intervals according to a preset distance; acquiring sea area observation elements detected by the detection instrument, and performing cleaning treatment operation on the acquired sea area observation elements to obtain cleaned observation elements; wherein the cleaning treatment operation comprises a missing value filling operation and/or an abnormal value deleting operation, and the sea area observation elements comprise the humidity, salinity, temperature, wind speed, wind direction, flow speed, wave height, wave direction, wave period, water depth and navigational speed of the sea area to be observed; performing feature selection processing on the cleaned observation elements by using a support vector machine recursive feature elimination method to obtain target features; inputting the target characteristics and the set time into a preset marine front prediction model to obtain a predicted marine front region of the target marine front region at the preset time; the preset ocean front prediction model comprises a model which is trained by using the target characteristics to a random forest model; and carrying out visualization processing on the predicted ocean front region to obtain a predicted region visualization image of the predicted ocean front region.

The processor is further used for inputting the target characteristics as a random forest model and performing model training by combining initial values of sensitive parameters of the random forest model to obtain a preliminary result after model training; determining different groups of values of the sensitive parameters according to the preliminary result, and determining different root mean square errors between predicted values and actual observed values of the random forest model after the different groups of values and the target feature verification; and sorting the different root mean square errors, and taking the random forest model corresponding to the smallest root mean square error as the preset ocean front prediction model.

In the actual processing process, as shown in fig. 9, the acquisition system of the marine front region includes a water surface unmanned ship formation controller and a terminal monitoring centralized control system, where the water surface unmanned ship formation controller includes a water surface unmanned ship formation formed by a plurality of unmanned ships, an observation device (such as an ADCP \ CTD \ meteorological sensor, etc.), a control system including a basic motion controller and a sensor data processor, a power supply system (such as a normal state and abnormal state emergency standby power supply), and an information receiving and transmitting system. The terminal monitoring centralized control system can comprise: a display, a communication system, a data storage system (e.g., storing sensor data, track information, ocean front size information), a remote control system, a parameter setting system (e.g., setting formation shape, desired path, etc.).

In one embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as shown in fig. 10. The computer device includes a processor, a memory, a communication interface, a display screen, and an input device connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of an operating system and computer programs in the non-volatile storage medium. The communication interface of the computer device is used for carrying out wired or wireless communication with an external terminal, and the wireless communication can be realized through WIFI, an operator network, NFC (near field communication) or other technologies. The computer program is executed by a processor to implement a method of acquisition of a marine front region. The display screen of the computer equipment can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer equipment can be a touch layer covered on the display screen, a key, a track ball or a touch pad arranged on the shell of the computer equipment, an external keyboard, a touch pad or a mouse and the like.

Those skilled in the art will appreciate that the architecture shown in fig. 10 is merely a block diagram of some of the structures associated with the disclosed aspects and is not intended to limit the computing devices to which the disclosed aspects apply, as particular computing devices may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.

In one embodiment, a computer device is provided, comprising a memory and a processor, the memory having a computer program stored therein, the processor implementing the following steps when executing the computer program:

In one embodiment, the processor, when executing the computer program, further performs the steps of:

acquiring a plurality of groups of satellite remote sensing observation data at different moments in the target time length; determining a plurality of roughly estimated ocean front areas according to the plurality of groups of satellite remote sensing observation data; and carrying out interpolation fitting processing on the plurality of roughly estimated ocean front regions according to the plurality of different moments to obtain the region change information.

Determining the area estimation value and frontal line length of the roughly estimated ocean front area according to the satellite remote sensing observation data; determining the number of unmanned ships to be dispatched according to the area estimation value and the first ratio result of the frontal line length; determining a plurality of linear distances between each unmanned ship and each preset vertex position of the rough estimation sea front region, and selecting the smallest one of the linear distances corresponding to each unmanned ship as a target path for the unmanned ship to reach the rough estimation sea front region; and acquiring the maximum speed per hour of the plurality of unmanned ships, and taking a second ratio result of the target path of the unmanned ships reaching the rough estimation ocean front area and the maximum speed per hour as the target duration.

determining a visual region distribution map of the rough estimation ocean front region according to the satellite remote sensing observation data; according to the position coordinates of each vertex of the visual region distribution diagram, carrying out regular graphic calibration processing on the visual region distribution diagram to obtain a target rectangular frame containing the visual region distribution diagram; determining an area estimation value of the roughly estimated sea front region corresponding to the target rectangular frame; and dividing the contour line of the visual region distribution diagram into a plurality of contour points, and linearly superposing the distance between two adjacent contour points to obtain the frontal line length.

carrying out inversion processing on the satellite remote sensing observation data to obtain a sea area distribution map of the sea area to be identified; if the sea area distribution map meets a pre-stored interface determining condition, determining an interface existing in the sea area distribution map as the rough estimation sea front area; and carrying out visualization processing on the rough estimation sea front area to obtain a visualization area distribution map of the rough estimation sea front area.

outputting an instrument layout instruction according to the target ocean front region; the instrument layout instruction is used for indicating that a plurality of detection instruments are crossly laid at intervals according to a preset distance; acquiring sea area observation elements detected by the detection instrument, and performing cleaning treatment operation on the acquired sea area observation elements to obtain cleaned observation elements; wherein the cleaning treatment operation comprises a missing value filling operation and/or an abnormal value deleting operation, and the sea area observation elements comprise the humidity, salinity, temperature, wind speed, wind direction, flow speed, wave height, wave direction, wave period, water depth and navigational speed of the sea area to be observed; performing feature selection processing on the cleaned observation elements by using a support vector machine recursive feature elimination method to obtain target features; inputting the target characteristics and the set time into a preset marine front prediction model to obtain a predicted marine front region of the target marine front region at the preset time; the preset ocean front prediction model comprises a model which is trained by using the target characteristics to a random forest model; and carrying out visualization processing on the predicted ocean front region to obtain a predicted region visualization image of the predicted ocean front region.

the target characteristics are used as input of a random forest model, and model training is carried out by combining with the initial value of the sensitive parameters of the random forest model to obtain a preliminary result after model training; determining different groups of values of the sensitive parameters according to the preliminary result, and determining different root mean square errors between predicted values and actual observed values of the random forest model after the different groups of values and the target feature verification; and sorting the different root mean square errors, and taking the random forest model corresponding to the smallest root mean square error as the preset ocean front prediction model.

In one embodiment, a computer-readable storage medium is provided, having a computer program stored thereon, which when executed by a processor, performs the steps of:

In one embodiment, the computer program when executed by the processor further performs the steps of:

It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware instructions of a computer program, which can be stored in a non-volatile computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. Any reference to memory, storage, database or other medium used in the embodiments provided herein can include at least one of non-volatile and volatile memory. Non-volatile Memory may include Read-Only Memory (ROM), magnetic tape, floppy disk, flash Memory, optical storage, or the like. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM), among others.

The technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, but should be considered as the scope of the present specification as long as there is no contradiction between the combinations of the technical features.

The above-mentioned embodiments only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present patent shall be subject to the appended claims.

Claims

1. A method for acquiring a marine front region, the method comprising:

2. The method of claim 1, wherein determining regional variation information for the coarse estimation ocean front region based on the plurality of sets of satellite remote sensing observations received within the target time period comprises:

3. The method of claim 1, wherein determining a target path for a plurality of unmanned vessels to be dispatched to reach the coarse estimation sea front area and a target time duration for reaching the coarse estimation sea front area along the target path according to the satellite remote sensing observation comprises:

4. The method of claim 3, wherein determining an area estimate and a frontal line length for the coarse estimate sea front region from the satellite remote sensing observations comprises:

5. The method of claim 4, wherein determining a visual region distribution map of the coarse estimate ocean front region from the satellite telemetry observations comprises:

6. A method for predicting a marine front region, the method comprising:

the target ocean front region of any one of claims 1 to 5, outputting instrumentation deployment instructions; the instrument layout instruction is used for indicating that a plurality of detection instruments are crossly laid at intervals according to a preset distance;

7. The method of claim 6, wherein the pre-defined ocean front prediction model comprises:

8. An acquisition system for a marine front region, the system comprising: the processor and the unmanned ship formation controller are respectively arranged in a land base station of a sea area to be identified, and the processor is electrically connected with the unmanned ship formation controller;

9. An apparatus for acquiring a marine front region, the apparatus comprising:

10. An apparatus for predicting a marine front region, the apparatus comprising:

an output module for outputting instrumentation deployment instructions for the target ocean front region according to the method of any one of claims 1 to 5; the instrument layout instruction is used for indicating that a plurality of detection instruments are crossly laid at intervals according to a preset distance;

the cleaning module is used for acquiring the sea area observation elements detected by the detection instrument and cleaning the acquired sea area observation elements to obtain cleaned observation elements; wherein the cleaning treatment operation comprises a missing value filling operation and/or an abnormal value deleting operation, and the sea area observation elements comprise the humidity, salinity, temperature, wind speed, wind direction, flow speed, wave height, wave direction, wave period, water depth and navigational speed of the sea area to be observed;

The processing module is used for performing feature selection processing on the cleaned observation elements by using a support vector machine recursive feature elimination method to obtain target features;

and the visualization module is used for performing visualization processing on the predicted ocean front region to obtain a predicted region visualization image of the predicted ocean front region.

11. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor, when executing the computer program, implements the steps of the method of any of claims 1 to 7.

12. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 7.