Construction method of krill fishery high-precision prediction model based on multivariate information data and hierarchical coupling module

Abstract

The invention discloses a construction method of a high-precision prediction model of a krill fishery based on multivariate information data and a hierarchical coupling module. The construction method comprises the following specific steps: (1) constructing a prediction model; (2) identifying an optimal habitat index factor range of the krill fishery with high temporal-spatial resolution through the model; (3) judging whether the krill boat is in a fishing state according to the boat position of the krill boat in the fishing state; (4) transmitting the output distribution data recorded by the electronic fishing log module to the krill fishery model; (5) transmitting the data to the euphausia superba center fishery forecasting model through a local area network or a satellite internet; and (6) carrying out layered screening and coupling metadata information. All the modules are organically combined, a hierarchical progressive relation and a mutual coupling relation are presented, in specific forecasting application, module coupling optimization adjustment needs to be carried out according to actual conditions, and finally, expert knowledge and experience comprehensively research and judge a center fishing ground result and carry out network publishing.

Description

Technical Field

[0001] The present invention relates to the field of fishery application technology of multivariate information data such as satellite remote sensing, ocean models, fisheries, ship positions, fishery acoustics, and electronic fishing logs, and in particular to a method for constructing a high-precision prediction model for krill fisheries based on multivariate information data and a hierarchical coupling module. Background Art

[0002] High-precision fishing ground forecast is the most critical link in the field of marine fishing and industrial applications. How to improve the accuracy of fishing ground forecast is crucial to improving the efficiency of fishery fishing and production benefits. Traditional fishing ground forecast is mainly based on fishing logs and marine environmental data based on satellite remote sensing, establishing the correlation between the two and building statistical methods, machine learning or deep learning model methods to achieve the function of fishing ground forecast. However, due to historical conditions and the limitations of remote sensing methods themselves, remote sensing data mostly use data products delayed by 2 days, and it is impossible to obtain the marine environmental data source of the same day or the next day. The fishery data based on traditional fishing logs are delayed for half a year or more than 1 year, and the delay time with the actual production situation is long, resulting in the inability to timely reflect the current situation of the fishing ground. It can be seen that the traditional fishing ground forecast model is seriously insufficient in data timeliness, and cannot reflect the spatiotemporal distribution characteristics of the abundance of fishery resources and the actual situation of the fishing ground on the day of the year, which restricts the accuracy of fishing ground forecast and makes it difficult to achieve the timeliness and high accuracy of fishing ground forecast.

[0003] In response to the industry demand for the identification and accurate prediction of krill fishery fishing grounds, no relevant high-precision forecasting models have been reported at home and abroad. Therefore, this technical bottleneck is a common problem in the field of fishery oceanography. With the improvement of the accuracy of ocean model models, the ocean environment data based on model data is getting closer and closer to the real value, and can provide ocean environment forecast products such as water temperature and ocean currents for the current day and the next 5 days. At the same time, based on the rapid processing of krill fishery acoustic data detected on-site by krill ships and the improvement of satellite network publishing technology, the rapid processing and publishing of acoustic data with a delay of no more than 2 hours on the ship-side fishery acoustics has been achieved. The relevant acoustic data is sent to the krill fishery fishing ground forecast platform system through the ship-side LAN or satellite Internet, and the dynamic distribution of the abundance of fishery acoustic resources of the ship or other ships can be viewed in time.

[0004] At the same time, the newly developed krill fishery fishing ground forecast platform system combines the position data of all krill boats in the krill fishing ground area. By real-time tracking of the fishing status of the krill boats, the instant location of the krill boats in the fishing status can be identified through the ship position status discrimination algorithm, and then the data reported by the electronic fishing log installed on the fishing boats can be used to know in quasi-real time which krill boat has a better yield.

[0005] Therefore, constructing a high-precision krill fishery forecast model based on a four-dimensional, layered coupled forecasting method including traditional logs, historical ship position and environmental forecast modules, real-time ship position status judgment module, fishery acoustic module, and electronic fishing log module, and realizing on-site production guidance in krill fisheries is a new idea and breakthrough direction in the field of krill fishery production and fishing grounds. Summary of the invention

[0006] In order to overcome the above-mentioned deficiencies of the prior art, the present invention provides a method for constructing a high-precision prediction model for krill fisheries based on multivariate information data and a hierarchical coupling module.

[0007] The technical solution adopted by the present invention to solve its technical problem is:

[0008] A method for constructing a high-precision prediction model for krill fisheries based on multivariate information data and hierarchical coupling modules, the specific steps are as follows:

[0009] (1) constructing a high-precision prediction model for krill fisheries based on multivariate information data and hierarchical coupling modules, wherein the prediction model includes four modules, namely, a fishing ground prediction module based on traditional historical logs and marine environment, a ship position status judgment module, a fishery acoustic module, and an electronic fishing log module, and using the hierarchical coupling model of each module to construct a prediction of the krill central fishing ground;

[0010] (2) Identify the range of optimal habitat indicators for krill fisheries with high temporal and spatial resolution through the model;

[0011] (3) Based on the position of the krill boat in the fishing state, when the krill boat is in the fishing state, its location is one of the optimal conditions for further narrowing the scope of the fishing ground;

[0012] (4) The production distribution data recorded by the electronic fishing log module is then transmitted to the krill fishery model, and the range and number of krill central fishing grounds are further optimized with the daily production of each vessel as the minimum threshold. After the hierarchical screening of the module results, the number of predicted fishing grounds is reduced;

[0013] (5) Sonar integral NASC value of krill resources detected by acoustic imaging fish finder of krill boats in real time (m 2 / n mile 2, square meters per square nautical mile) is processed into binary or text character data on the ship side and transmitted to the krill center fishing ground forecast model through the local area network or satellite Internet. The krill abundance data of the sea area with a radius of 2500 meters around a krill ship can be obtained in real time. The medium and high abundance values ​​(NASC value ≧1000) are used as the minimum threshold to further screen or judge the best krill fishing ground location. After the hierarchical screening of the module results, the number of predicted fishing grounds is reduced to 1 to 3, which can accurately determine the target fishing location of the krill ship in order to obtain high yields;

[0014] (6) Based on the historical fishing ground location forecast, quasi-real-time electronic log data, fishery acoustic data and real-time krill vessel position data are combined to perform hierarchical screening and coupling of metadata information, and finally determine the optimal fishing ground location based on expert experience and knowledge, thereby improving the forecast accuracy.

[0015] As a preferred embodiment, the specific implementation process of step (1) includes the following sub-steps:

[0016] (11) The forecast module based on historical fishing logs, historical ship positions, and satellite remote sensing data analyzes the suitable habitat range of krill fisheries and key environmental factors for fishery forecasts. The temporal resolution is 3 or 7 days and the spatial resolution is 0.25° longitude and latitude grid. The temporal and spatial resolution of the fishery forecasted by this module is relatively coarse, which serves as an indication of the approximate location and direction of the fishery.

[0017] (12) Analyzing the real-time location of the fishing status of the krill boat based on the real-time ship position status judgment module, and further optimizing the precise location of the krill central fishing ground according to whether the surrounding krill boats transfer the fishing ground to the krill boat;

[0018] (13) The catch recorded in the electronic fishing log module is substituted into the krill forecasting system to further confirm and narrow the scope of the fishing ground. At this time, the spatial resolution of the forecast can reach 10 nautical miles;

[0019] (14) Based on the quasi-real-time fishery acoustic module, the krill acoustic abundance data of the ship or other ships can be obtained at any time, and sent to the krill fishery fishing ground forecast platform system through satellite Internet or ship-side local area network, and the krill fishing ground level can be determined according to the acoustic abundance threshold; at this time, the forecast spatial resolution can reach 2 to 5 nautical miles;

[0020] (15) Each module is organically combined, presenting a hierarchical progressive relationship, and at the same time a mutually coupled relationship. In specific forecasting applications, it is necessary to optimize and adjust the module coupling according to actual conditions. Finally, the expert knowledge and experience are used to comprehensively analyze the central fishery results and publish them on the Internet.

[0021] As a preferred embodiment, the time resolution in step (11) is 3 or 7 days, and the spatial resolution is 0.25° longitude and latitude grid, i.e., a range of about 150 square nautical miles.

[0022] As a preferred embodiment, the fishery grades in step (14) are divided into good fishery, better fishery and general fishery.

[0023] As a preferred embodiment, the coarse-grained range of the historical krill fishery center position is determined in step (2), specifically a 7-day, 0.25 degree longitude and latitude grid, and the formula is as follows:

[0024]

[0025] In formula (1), F location is the location of the krill central fishing ground, F 48.1 、F 48.2 、F 48.3 They are the locations of the krill central fishing grounds in sub-areas 48.1, 48.2 and 48.3, respectively. rang The optimum range of surface temperature (℃) varies from sub-region to sub-region, but is more suitable within -1.6~1.6℃. rang is the surface temperature gradient (℃ / 0.5° grid sea area), and its value range is 0.1~0.2℃ / 0.5° grid sea area, C 中度拐弯或大拐弯 It is defined as 4 to 10 consecutive unit grids, one of which is a spatial sea area of ​​2.5' latitude × 5' longitude, where the flow velocity direction difference is greater than 20°. H is the distance between the fishing ground location and the island shore baseline, in nautical miles. The number of fishing grounds predicted by this module is generally 5 to 20.

[0026] As a preferred embodiment, in the step (3), the position of the krill boat is determined according to its fishing status, and the specific speed threshold is 1.6 to 3.6 knots, which is judged as the fishing status. The heading data is not referenced, and the position delay does not exceed 20 minutes. After the hierarchical screening of the module results, the number of predicted fishing grounds is reduced to 3 to 10.

[0027] As a preferred embodiment, in step (4), the range and number of krill central fishing grounds forecast are optimized again with a daily output of 120 tons per vessel as the minimum threshold, and the number of forecasted fishing grounds is reduced to 2 to 6.

[0028] As a preferred embodiment, in step (5), the krill resource sonar integral NASC value (m 2 / n mile 2, square meters per square nautical mile) is processed into binary or text character data on the ship side and transmitted to the krill central fishing ground forecast model through the local area network or satellite Internet. The krill abundance data in the sea area with a radius of 2500 meters around a krill ship can be obtained in real time, and the medium and high abundance values ​​(NASC value ≧1000) are used as the minimum threshold to further screen or judge the best krill fishing ground location. After the hierarchical screening of the results of this module, the number of predicted fishing grounds is reduced to 1 to 3, which can accurately determine the target fishing position of the krill ship in order to obtain high yields.

[0029] As a preferred embodiment, the model structure in step (6) is as follows:

[0030]

[0031] In formula (2), F 实时预报渔场 is the real-time forecast of the location of krill fishing grounds, F 历史 is the location of the fishing ground calculated using the historical fishery data element module, F 捕捞状态船位 The real-time fishing status and location of all krill boats. 电子 The NASC value is a near-real-time krill abundance data from the electronic fishing log of krill vessels, with a delay of no more than 2 hours.

[0032] As a preferred embodiment, in step (6), Log 电子 The krill boat is equipped with an electronic fishing log reporting device for reporting production data. The krill fishery acoustic NASC value is the fishery acoustic data collected by the krill abundance acoustic data ship-side processing device system installed on the krill boat.

[0033] Beneficial effects:

[0034] The present invention fills the gap in the field of high-precision prediction of krill fisheries and has the following advantages and positive effects compared with the prior art:

[0035] The real-time state discrimination algorithm of the ship position in the present invention can judge in real time whether the krill boat is in a fishing or non-fishing state, the electronic fishing log can obtain the actual fishing output of the day, and the krill fishery acoustic NASC value can collect the abundance of krill resources in the surrounding sea areas detected by the krill boat in quasi-real time. Since the present invention integrates four forecast modules, namely, a fishing ground forecast module based on traditional historical logs and marine environment, a ship position state judgment module, a fishery acoustic module, and an electronic fishing log module, and adopts a hierarchical coupling model of each module to construct a forecast of the krill central fishing ground, the timeliness and accuracy are greatly improved compared with the previous traditional fishing ground forecast model. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 It is a schematic diagram of the construction process of the four-dimensional integrated and hierarchical coupled forecasting method of multivariate information.

[0037] Figure 2 It is a fishing ground forecast rendering (May 16, 2023) of the forecast module based on historical fishing logs, historical ship positions, and satellite remote sensing data.

[0038] Figure 3 It is based on the real-time ship position status judgment module to analyze the real-time fishing status diagram of krill boats (May 16, 2024).

[0039] Figure 4 It is a schematic diagram of krill fishery optimized based on the electronic fishing log module (May 16, 2024).

[0040] Figure 5 It is a quasi-real-time fishery acoustic module for on-site decision-making and determination of krill center fishing ground analysis map (May 16, 2024).

[0041] Figure 6 It is a schematic diagram of the high-precision central fishing ground results (May 16, 2024), which is a hierarchical coupling of the forecast results of each module and a comprehensive analysis of expert knowledge and experience. DETAILED DESCRIPTION

[0042] The present invention will be further described below in conjunction with specific embodiments. It should be understood that these embodiments are only used to illustrate the present invention and are not intended for the scope of the present invention. It should be understood here that after reading the content taught by the present invention, those skilled in the art may make various changes or modifications to the present invention, and these equivalent forms also fall within the scope limited by the appended claims of the application.

[0043] like Figure 1 As shown, the embodiment of the present invention relates to a high-precision prediction model method for krill fisheries based on multivariate information data and hierarchical coupling modules, and the steps are as follows:

[0044] (1) A new high-precision prediction model method for krill fishery grounds based on multivariate information data and hierarchical coupling modules was constructed, including four prediction modules: a fishing ground prediction module based on traditional historical logs and marine environment, a ship position status judgment module, a fishery acoustic module, and an electronic fishing log module. The prediction of the krill central fishery ground was constructed using a hierarchical coupling model of each module, which is divided into the following sub-steps:

[0045] (11) The forecast module based on historical fishing logs, historical ship positions, and satellite remote sensing data analyzes the suitable habitat range of krill fisheries and key environmental factors for fishery forecasts. The temporal resolution is 3 or 7 days and the spatial resolution is 0.25° longitude and latitude grid (i.e., a range of about 150 square nautical miles). The temporal and spatial resolution of the fishery forecasted by this module is relatively coarse, which serves as an indication of the approximate location and direction of the fishery.

[0046] (12) Ship position module: The potential direction of the fishing ground is determined based on the position of the krill boat during the fishing period (the speed is 1.6 to 3.6 knots). When there is no krill boat in the fishing area, the forecast results of this module can be ignored. When there is a krill boat fishing in the fishing area, the position of the fishing platform is used to check whether the speed of the boat is in the fishing period. If it is fishing, the forecast results of the environmental module around the boat position are superimposed and analyzed to determine the refined location of the krill center fishing ground.

[0047] (13) The catch recorded in the electronic fishing log module is substituted into the krill forecasting system to further confirm and narrow the scope of the fishing ground. At this time, the spatial resolution of the forecast can reach 10 nautical miles;

[0048] (14) Based on the quasi-real-time fishery acoustic module, the krill acoustic abundance data of the ship or other ships can be obtained at any time, and sent to the krill fishery fishing ground forecast platform system through satellite Internet or ship-side local area network. The krill fishing ground level (good fishing ground, relatively good fishing ground and general fishing ground) is determined according to the acoustic abundance threshold; at this time, the forecast spatial resolution can reach 2 nautical miles;

[0049] (15) Each module is organically combined, presenting a hierarchical and progressive relationship, and at the same time a mutually coupled relationship. In specific forecasting applications, it is necessary to optimize and adjust the module coupling according to the actual situation. Finally, the expert knowledge and experience are used to comprehensively judge the central fishery results and publish them online.

[0050] (2) The model was used to identify the optimal habitat index factor range of krill fisheries with high temporal and spatial resolution, and the coarse-grained (7-day, 0.25 degree longitude and latitude grid) range of the historical krill fishery center was determined using the following formula:

[0051]

[0052] In the formula, F location is the location of the krill central fishing ground, F 48.1 、F 48.2 、F 48.3 They are the locations of the krill central fishing grounds in sub-areas 48.1, 48.2 and 48.3 respectively. rang The optimum range of surface temperature (℃) varies from sub-region to sub-region every 7 days, but is generally within -1.6~1.6℃. rang is the surface temperature gradient (℃ / 0.5° grid sea area), and its value range is 0.1~0.2℃ / 0.5° grid sea area. 中度拐弯或大拐弯 It is defined as the sea area where the flow velocity direction difference of 4 to 10 consecutive unit grids (one grid is a spatial sea area of ​​2.5' latitude × 5' longitude) is greater than 20°. H is the distance (nautical miles) between the fishing ground location and the island coast baseline. The number of fishing grounds predicted by this module is generally 5 to 20;

[0053] (3) According to the position determination of the krill boat in the fishing state (the speed range is 1.6 to 3.6 knots, and the heading data is not used), when the krill boat is in the fishing state, its location is one of the preferred conditions for further narrowing the scope of the fishing ground. The position delay does not exceed 20 minutes. After the hierarchical screening of the results of this module, the number of predicted fishing grounds is reduced to 3 to 10.

[0054] (4) The production distribution data recorded by the electronic fishing log module was then transmitted to the krill fishery model. The range and number of krill central fishing grounds were further optimized with a daily production of 120 tons per vessel as the minimum threshold. After hierarchical screening of the module results, the number of predicted fishing grounds was reduced to 2 to 6.

[0055] (5) By processing the krill resource sonar integral NASC value detected by the krill boat's acoustic imaging fish finder in real time into binary or text character data and transmitting it to the krill central fishing ground forecast model through the local area network or satellite Internet, the krill abundance data in the sea area with a radius of 2500 meters around a krill boat can be obtained in quasi-real time. The medium and high abundance values ​​(NASC value ≧1000) are used as the minimum threshold to further screen or judge the best krill fishing ground location. After the hierarchical screening of the module results, the number of predicted fishing grounds is reduced to 1 to 3, which can accurately determine the target fishing location of the krill boat in order to obtain high yields;

[0056] (6) Based on the historical fishing ground location forecast, quasi-real-time electronic log data, fishery acoustic data and real-time krill ship position data are combined to perform hierarchical screening and coupling of metadata information, and finally determine the optimal fishing ground location in combination with expert experience and knowledge, thereby improving the forecast accuracy. The model structure is as follows:

[0057]

[0058] In formula (2), F 实时预报渔场 is the real-time forecast of the location of krill fishing grounds, F 历史 is the location of the fishing ground calculated using the historical fishery data element module, F 捕捞状态船位 The real-time fishing status and location of all krill boats, log 电子 The NASC value is a near-real-time krill abundance data from the electronic fishing log of krill vessels, with a delay of no more than 2 hours.

[0059] The Log 电子 The krill boat is equipped with an electronic fishing log reporting device for reporting production data. The krill fishery acoustic NASC value is the fishery acoustic data collected by the krill abundance acoustic data ship-side processing device system installed on the krill boat.

[0060] The present invention is further illustrated by a specific embodiment below.

[0061] 1. Forecast module analysis of historical fishing logs, historical ship positions, and satellite remote sensing data

[0062] First, the historical central fishing grounds of each sub-area 48.1, 48.2 and 48.3 were identified based on the historical fishery information metadata database, with a time resolution of weeks and a spatial resolution of 30 nautical miles, and the location of the fishing grounds, the law of change and the range of suitable environmental factor indicators corresponding to each week of each month were constructed (see formula 1). For example, fishing is carried out in the northwest corner of the 48.2 sub-area from December to March of the following year when the surface temperature range is -1 to 1°C, the surface temperature gradient is 0.1 to 0.2°C / 0.5° grid sea area, the 50-meter ocean current has a large bend, and the sea area is 20 to 70 nautical miles away from the island. The central fishing grounds of each sub-area calculated based on this condition are 2 to 6 ( Figure 2 ), there are 3 to 10 fishing grounds in the entire 48 areas.

[0063] 2. In Figure 2 Based on the forecast of the module, the fishing status of each krill boat is analyzed according to the real-time ship position status discrimination module. If the real-time position of the krill boat coincides with the historical fishing ground position, the position is considered to be the central fishing ground; if the real-time position of the krill boat is in the fishing state, the position of the krill boat is considered to be the central fishing ground; if the krill boat is in the aerial survey state, the predicted historical fishing ground position is considered to be the central fishing ground, and the krill boat is recommended to fish in the nearest fishing ground ( Figure 3 ).

[0064] 3. The production distribution data recorded by the electronic fishing log module is then transmitted to the krill fishery model. The range and number of krill central fishing grounds are further optimized with a daily production of 120 tons per ship as the minimum threshold. When a ship is found to be fishing and its electronic log production is greater than 120 tons, the sea area around the ship is determined to be the central fishing ground, and other krill ships can go to the sea area for fishing operations ( Figure 4 ), saving time for blind detection and greatly improving fishing efficiency.

[0065] 4. In the above process, if the acoustic image data of the krill ship can be obtained in real time or quasi-real time, the real-time distribution of krill abundance around a certain ship can be determined more accurately. The current krill abundance level can be understood through the NASC value of the krill resource sonar integral detected by the ship in real time ( Figure 5 ), when the NASC value is ≧2000, it is considered that the abundance of krill fishery in the current sea area is high, and it is the central fishery ( Figure 6 ). The NASC values ​​in different subregions are slightly different.

[0066] Due to the adoption of multi-level krill fishery forecast module filtering and screening and knowledge judgment based on expert experience, the accuracy of the central fishery forecast is effectively improved. At the same time, the real-time nature of the model is greatly improved by using real-time ship position and ocean model environmental data as well as quasi-real-time electronic fishing logs and fishery acoustic image data to comprehensively judge the location and resource abundance of the central fishery. Therefore, the krill fishery forecast model constructed by the present invention has high accuracy (greater than 85%) and real-time performance, which changes the effect of the traditional coarse-grained, 7-day delayed fishery forecast model, and can significantly improve the production efficiency and fishing yield of krill boats. The krill central fishery forecast map of the present invention is released every three days from January to August 2024 to the krill production ship Deep Blue for fishery production reference. At the same time, the krill fishery fishery forecast platform system installed on the ship end can also check the various elements of the fishery sea conditions every day to make on-site comprehensive judgments on the location of the fishery and the production arrangements of the operation conditions, so that its production decision-making efficiency has been greatly improved.

[0067] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A method for constructing a high-precision prediction model for krill fisheries based on multivariate information data and hierarchical coupling modules, characterized in that: The specific steps are as follows: (1) constructing a high-precision prediction model for krill fisheries based on multivariate information data and hierarchical coupling modules, wherein the prediction model includes four modules, namely, a fishing ground prediction module based on traditional historical logs and marine environment, a ship position status judgment module, a fishery acoustic module, and an electronic fishing log module, and using the hierarchical coupling model of each module to construct a prediction of the krill central fishing ground; (2) Identify the range of optimal habitat indicators for krill fisheries with high temporal and spatial resolution through the model; (3) Based on the position of the krill boat in the fishing state, when the krill boat is in the fishing state, its location is one of the optimal conditions for further narrowing the scope of the fishing ground; (4) The production distribution data recorded by the electronic fishing log module is then transmitted to the krill fishery model, and the range and number of krill central fishing grounds are further optimized with the daily production of each vessel as the minimum threshold. After the hierarchical screening of the module results, the number of predicted fishing grounds is reduced; (5) Sonar integral NASC value of krill resources detected by acoustic imaging fish finder of krill boats in real time (m 2 / nmile 2 , square meters per square nautical mile) is processed into binary or text character data on the ship side and transmitted to the krill center fishing ground forecast model through the local area network or satellite Internet. The krill abundance data of the sea area with a radius of 2500 meters around a krill ship can be obtained in real time. The medium and high abundance values ​​(NASC value ≧1000) are used as the minimum threshold to further screen or judge the best krill fishing ground location. After the hierarchical screening of the module results, the number of predicted fishing grounds is reduced to 1 to 3, which can accurately determine the target fishing location of the krill ship in order to obtain high yields; (6) Based on the historical fishing ground location forecast, quasi-real-time electronic log data, fishery acoustic data and real-time krill vessel position data are combined to perform hierarchical screening and coupling of metadata information, and finally determine the optimal fishing ground location based on expert experience and knowledge, thereby improving the forecast accuracy.

2. The method for constructing a high-precision prediction model for krill fisheries based on multivariate information data and hierarchical coupling modules according to claim 1, characterized in that: The specific implementation process of step (1) includes the following sub-steps: (11) The forecast module based on historical fishing logs, historical ship positions, and satellite remote sensing data analyzes the suitable habitat range of krill fisheries and key environmental factors for fishery forecasts. The temporal resolution is 3 or 7 days and the spatial resolution is 0.25° longitude and latitude grid. The temporal and spatial resolution of the fishery forecasted by this module is relatively coarse, which serves as an indication of the approximate location and direction of the fishery. (12) Analyzing the real-time location of the fishing status of the krill boat based on the real-time ship position status judgment module, and further optimizing the precise location of the krill central fishing ground according to whether the surrounding krill boats transfer the fishing ground to the krill boat; (13) The catch recorded in the electronic fishing log module is substituted into the krill forecasting system to further confirm and narrow the scope of the fishing ground. At this time, the spatial resolution of the forecast can reach 10 nautical miles; (14) Based on the quasi-real-time fishery acoustic module, the krill acoustic abundance data of the ship or other ships can be obtained at any time, and sent to the krill fishery fishing ground forecast platform system through satellite Internet or ship-side local area network, and the krill fishing ground level can be determined according to the acoustic abundance threshold; at this time, the forecast spatial resolution can reach 2 to 5 nautical miles; (15) Each module is organically combined, presenting a hierarchical progressive relationship, and at the same time a mutually coupled relationship. In specific forecasting applications, it is necessary to optimize and adjust the module coupling according to actual conditions. Finally, the expert knowledge and experience are used to comprehensively analyze the central fishery results and publish them on the Internet.

3. The method for constructing a high-precision prediction model for krill fisheries based on multivariate information data and hierarchical coupling modules according to claim 2, characterized in that: In the step (11), the temporal resolution is 3 or 7 days, and the spatial resolution is 0.25° longitude and latitude grid, i.e., a range of about 150 square nautical miles.

4. The method for constructing a high-precision prediction model for krill fisheries based on multivariate information data and hierarchical coupling modules according to claim 2, characterized in that: The fishery grades in step (14) are divided into good fishery, better fishery and general fishery.

5. The method for constructing a high-precision prediction model for krill fisheries based on multivariate information data and hierarchical coupling modules according to claim 1, characterized in that: In step (2), the coarse-grained range of the historical krill fishery center position is determined, specifically, a 7-day, 0.25 degree longitude and latitude grid, and the formula is as follows: In formula (1), F location is the location of the krill central fishing ground, F 48.1 、F 48.2 、F 48.3 They are the locations of the krill central fishing grounds in sub-areas 48.1, 48.2 and 48.3, respectively. rang The optimum range of surface temperature (℃) varies from sub-region to sub-region, but is more suitable within -1.6~1.6℃. rang is the surface temperature gradient (℃ / 0.5° grid sea area), and its value range is 0.1~0.2℃ / 0.5° grid sea area, C 中度拐弯或大拐弯 It is defined as 4 to 10 consecutive unit grids, one of which is a spatial sea area of ​​2.5' latitude × 5' longitude, where the flow velocity direction difference is greater than 20°. H is the distance between the fishing ground location and the island shore baseline, in nautical miles. The number of fishing grounds predicted by this module is generally 5 to 20.

6. The method for constructing a high-precision prediction model for krill fisheries based on multivariate information data and hierarchical coupling modules according to claim 1, characterized in that: In the step (3), the krill boat's fishing status is determined based on its position, and the specific speed threshold is 1.6 to 3.6 knots, which is considered fishing status. The heading data is not used as a reference, and the position delay does not exceed 20 minutes. After the hierarchical screening of the module results, the number of predicted fishing grounds is reduced to 3 to 10.

7. The method for constructing a high-precision prediction model for krill fisheries based on multivariate information data and hierarchical coupling modules according to claim 1, characterized in that: In the step (4), the range and number of the krill central fishing grounds forecast are optimized again with the daily production of 120 tons per vessel as the minimum threshold, and the number of the forecasted fishing grounds is reduced to 2 to 6.

8. The method for constructing a high-precision prediction model for krill fisheries based on multivariate information data and hierarchical coupling modules according to claim 1, characterized in that: In step (5), the krill resource sonar integral NASC value (m 2 / n mile 2 , square meters per square nautical mile) is processed into binary or text character data on the ship side and transmitted to the krill central fishing ground forecast model through the local area network or satellite Internet. The krill abundance data in the sea area with a radius of 2500 meters around a krill ship can be obtained in real time, and the medium and high abundance values ​​(NASC value ≧1000) are used as the minimum threshold to further screen or judge the best krill fishing ground location. After the hierarchical screening of the results of this module, the number of predicted fishing grounds is reduced to 1 to 3, which can accurately determine the target fishing position of the krill ship in order to obtain high yields.

9. The method for constructing a high-precision prediction model for krill fisheries based on multivariate information data and hierarchical coupling modules according to claim 1, characterized in that: The model structure in step (6) is as follows: In formula (2), F 实时预报渔场 is the real-time forecast of the location of krill fishing grounds, F 历史 is the location of the fishing ground calculated using the historical fishery data element module, F 捕捞状态船位 The real-time fishing status and location of all krill boats. 电子 The NASC value is a near-real-time krill abundance data from the electronic fishing log of krill vessels, with a delay of no more than 2 hours.

10. The method for constructing a high-precision prediction model for krill fisheries based on multivariate information data and hierarchical coupling modules according to claim 1, characterized in that: In step (6), Log 电子 The krill boat is equipped with an electronic fishing log reporting device for reporting production data. The krill fishery acoustic NASC value is the fishery acoustic data collected by the krill abundance acoustic data ship-side processing device system installed on the krill boat.

Images