A method for carbon assessment and fishing strategy optimization of euphausiids fishery based on carbon reduction target

Abstract

The application discloses a kind of carbon evaluation and fishing strategy optimization method of euphausiacea fishery based on carbon reduction target, comprising the following steps: (1) first, collect and organize the ship name of all fishing vessels of Antarctic euphausiacea fishery, total tonnage, main engine power and fishing vessel position and annual output data, establish a set of euphausiacea fishing boat fishery production database;(2) take Chinese euphausiacea fishing boat as case to carry out the daily oil consumption estimation model of different tonnage ship;(3) further calculate the oil consumption change and difference of the whole euphausiacea fleet each year in combination with daily or annual catch;(4) realize the visualization thematic map of the overall distribution of carbon emissions of euphausiacea fishing ground area;(5) the fishing ground forecast model result based on deep learning is combined with current environmental state real-time reinforcement learning (deep Q network) to decide optimal path.The present application constructs intelligent planning fishing route, as euphausiacea fishery carbon footprint characteristics and potential carbon reduction method, realizes the purpose of low-carbon fishing path planning strategy and carbon reduction.

Description

Technical Field

[0001] This invention relates to the field of integrated application technologies such as polar fisheries, ship energy consumption and carbon footprint assessment, and fishing route optimization strategies, and in particular to a method for carbon assessment and fishing strategy optimization for krill fisheries based on carbon reduction targets. Background Technology

[0002] Against the backdrop of the global "dual-carbon" strategy and the escalating demands for polar ecological protection, carbon emission assessment and carbon reduction pathways for polar ocean-going fishing vessels have become crucial objectives that urgently need to be addressed. Krill, a precious fishery resource yet to be fully exploited, is a veritable blue food treasure trove of high-quality animal protein and a carbon sequestration repository for polar marine ecosystems. Studies have shown that krill populations sequester up to 20 million tons of carbon annually through fecal pellets. Meanwhile, krill fishing activities generate approximately 180,000 to 200,000 tons of carbon dioxide emissions annually, making it a significant source of carbon emissions in the Antarctic Ocean. However, considering that the annual yield of krill fisheries is as high as 450,000 to 620,000 tons, the carbon dioxide emissions per unit of krill yield are very low. Compared to cattle and sheep farming and shrimp farming, the carbon dioxide emissions per ton of yield are more than 90% and 60% lower, respectively, making it a typical low-carbon fishery. However, due to significant differences in fishing days, vessel tonnage, fuel consumption, and catch yield among different krill vessels, and the annual impact of various factors on krill resources, the overall krill fishery's catch and carbon emissions vary each year. There is currently no literature on how to estimate the carbon emissions of the entire krill fleet; therefore, developing a smart fishing approach and carbon reduction strategies is of great practical and ecological significance for achieving low-carbon goals in the krill fishery and for green ecological transformation. Thus, it is necessary to expedite the development of a theoretical model for calculating carbon emissions from the krill fishery and to propose specific feasible carbon reduction pathways, methods, and optimization strategies. Summary of the Invention

[0003] To overcome the shortcomings of the existing technologies, this invention provides a method for carbon assessment and fishing strategy optimization in krill fisheries based on carbon reduction targets. This method addresses the problems of existing technologies being unable to calculate the carbon emissions of the entire krill fleet and lacking carbon reduction fishing strategies. On the one hand, it achieves overall carbon emission assessment and spatiotemporal distribution analysis of the krill fisheries. On the other hand, it uses reinforcement learning to intelligently plan fishing routes, reducing fuel consumption and carbon emissions while ensuring catch yield, thus supporting the green and low-carbon transformation of the krill fisheries.

[0004] The technical solution adopted by this invention to solve its technical problem is:

[0005] A method for carbon assessment and fishing strategy optimization in krill fisheries based on carbon reduction targets, specifically including the following steps:

[0006] (1) First, collect and organize the names, total tonnage, main engine power, and fishing vessel positions and annual production data of all fishing vessels in the Antarctic krill fishery. At the same time, download the annual catch data of each fishing area and each fishing party from the CCAMLR website to establish a database of krill fishing vessel fishery production.

[0007] (2) Taking Chinese krill fishing vessels as a case study, a daily fuel consumption calculation model for vessels of different tonnages was developed. At the same time, the differences in daily fuel consumption caused by the differences in krill processed products were considered, and empirical algorithms for different vessel types, different vessel states and daily fuel consumption were fitted.

[0008] (3) Combine the daily or annual catch and the krill product situation to calculate the fuel consumption of each krill vessel and further calculate the annual changes and differences of the fuel consumption of the entire krill fleet, and optimize the empirical algorithm model; at the same time, combine the status duration of each vessel to calculate the daily fuel consumption and daily carbon dioxide emissions of each vessel.

[0009] (4) Draw monthly variation maps of carbon dioxide emissions from krill fisheries in each fishing season and spatial variation maps of carbon dioxide emissions in different sub-regions to realize a visual thematic map of the overall distribution of carbon emissions in the entire krill fishing area.

[0010] (5) Based on Deep Q Network (DQN), the results of deep learning-based fishing ground forecast model are combined with real-time reinforcement learning of the current environmental state to make decisions on the optimal path, thereby realizing intelligent planning of fishing routes.

[0011] As a preferred embodiment, the names, gross tonnage, main engine power, and vessel positions and annual production data of all fishing vessels in the Antarctic krill fishery in step (1) are obtained through various materials, documents, CCAMLR website downloads and Doubao AI, etc.; the vessel position data comes from positioning data such as automatic identification system (AIS), with a vessel position resolution of 1 position every 5 to 10 minutes.

[0012] As a preferred embodiment, the vessel parameters for Chinese krill vessels in step (2) cover all vessel types of krill fleets worldwide, and can be obtained through... or A comparative fitting estimation of daily fuel consumption (Yi) of foreign ships under navigation conditions ( Figure 2 The estimation formula is as follows: , The number of krill fishing vessels worldwide is 8 to 13 per year, with a total tonnage ranging from 2,455 to 16,000 tons.

[0013] As a preferred embodiment, the daily catch of different krill vessels in step (3) is different. The higher the catch, the higher the fuel consumption. According to empirical calculation, for every 100 tons increase in daily krill production, the daily fuel consumption increases by about 2.2 to 2.6 tons.

[0014] As a preferred embodiment, in step (3), under the same output, the daily processing fuel consumption of a krill vessel processing frozen shrimp is about 5-15 tons lower than that of a vessel processing krill meal. Taking into account different tonnages, catch volumes, and processing methods of krill vessels, the daily fuel consumption of a certain vessel is... , where a is any constant from 2.2 to 2.6.

[0015] In a preferred embodiment, the formula for calculating the total daily carbon dioxide emissions of the krill vessel in step (4) is as follows: 3.1605. Using ArcGIS 10.0 plotting software, a map of the total monthly carbon dioxide emissions and spatial emissions for the entire krill fishery can be generated.

[0016] As a preferred embodiment, the state (s, marine environmental data features such as sea temperature, salinity, and ocean currents), action (a, fishing ground location set, longitude and latitude), and reward function (x, with catch volume and prediction accuracy as the core) of the custom deep Q-network algorithm in step (5) are designed with a multi-objective reward R(s,a), the formula of which is:

[0017] ,

[0018] In the formula, λ1 and λ2 are coefficients, and R catch R represents the catch per unit state (tons). pred This represents the error range between the predicted and the actual values.

[0019] In a preferred embodiment, step (5) involves learning the optimal strategy and forming a multi-objective reward R(s,a) value through interaction with suitable marine environmental characteristic fields, avoiding sea ice concentration and high SST areas, and fishing electronic log data.

[0020] In a preferred embodiment, in step (5), by continuously interacting with the background environment, the deep Q network optimizes the R(s,a) value to maximize the cumulative reward and realize intelligent planning of fishing routes, thereby reducing the blind exploration of fishing grounds, saving fuel, and achieving the goal of reducing carbon emissions.

[0021] As a preferred embodiment, in step (5), a multi-objective reward function with high yield and prediction accuracy, and two dimensions of time and space as the core is established. Real-time sea state information (extreme water temperature, sea ice, etc.), dynamic migration patterns of fishing grounds and ship energy consumption characteristics (carbon emissions are calculated based on sailing distance, power and fishing time) are integrated to establish a deep Q network function to further optimize the fishing ground locations to 1-2, and a low-carbon fishing path planning strategy is constructed to achieve the carbon reduction target.

[0022] Beneficial Effects: This invention provides a method for carbon assessment and fishing strategy optimization in krill fisheries based on carbon reduction targets. It fills a gap in the industry regarding methods for estimating the overall carbon emissions of krill fisheries, their carbon footprint, and research on carbon reduction pathways. Compared with existing technologies, it has the following advantages and positive effects: This invention can analyze the distribution differences and trends of carbon emissions from Antarctic krill fisheries, and proposes carbon reduction targets and pathways based on intelligent fishing path planning strategies. This is beneficial for a deeper understanding of the carbon footprint characteristics of krill fisheries and potential carbon reduction methods. Attached Figure Description

[0023] Figure 1 This is a flowchart of a krill fishery carbon assessment and fishing strategy optimization based on carbon reduction targets, according to an embodiment of the present invention.

[0024] Figure 2 This is a graph showing the fitting relationship between the main engine power, total tonnage, and daily fuel consumption of different krill boats in one embodiment of the present invention.

[0025] Figure 3 This is a monthly distribution map of carbon emissions from krill fisheries during the 2023 / 2024 fishing season, as shown in one embodiment of the present invention.

[0026] Figure 4 This is a spatial distribution map of carbon emissions from krill fisheries during the 2023 / 2024 fishing season, as shown in one embodiment of the present invention.

[0027] Figure 5 This is a schematic diagram of intelligent planning and carbon reduction of krill fishing strategies based on deep Q-networks in one embodiment of the present invention. Detailed Implementation

[0028] The embodiments of the present invention will be described in detail below with reference to the accompanying drawings: These embodiments are implemented based on the technical solution of the present invention, and provide detailed implementation methods and specific operation processes, but the protection scope of the present invention is not limited to the following embodiments.

[0029] Example:

[0030] like Figure 1 As shown, a method for carbon assessment and fishing strategy optimization in krill fisheries based on carbon reduction targets includes the following steps:

[0031] (1) First, collect and organize the names, gross tonnage, main engine power, and fishing vessel positions and annual production data of all fishing vessels in the Antarctic krill fishery. At the same time, download the annual catch of each fishing area and each fishing party from the CCAMLR website of the Antarctic Marine Living Resources Conservation Committee to establish a database of all krill fishing vessels and their parameters in the world, as shown in Table 1.

[0032] (2) Taking Chinese krill fishing vessels as a case study, a daily fuel consumption calculation model for vessels of different tonnages was developed. Simultaneously, considering the differences in daily fuel consumption caused by variations in krill processed products, empirical algorithms were fitted for different vessel types, vessel conditions, and daily fuel consumption, such as... Figure 2 As shown in Tables 2 and 3;

[0033] (3) Combine the daily or annual catch and the krill product situation to calculate the fuel consumption of each krill vessel and further calculate the annual changes and differences of the fuel consumption of the entire krill fleet, and optimize the empirical algorithm model; at the same time, combine the status duration of each vessel to calculate the daily fuel consumption and daily carbon dioxide emissions of each vessel.

[0034] (4) Plot the monthly changes in carbon dioxide emissions from krill fisheries for each fishing season (see Figure 3 This includes spatial variation maps of carbon dioxide emissions in different sub-regions, creating a thematic map visualizing the overall distribution of carbon emissions across the entire krill fishing grounds (see...). Figure 4 );

[0035] (5) Based on the Deep Q Network (DQN), the optimal path is determined by combining historical judgment experience of fishing grounds with real-time reinforcement learning of the current environmental state, which can realize intelligent planning of fishing routes. With high yield and prediction accuracy (two dimensions of time and space) as the core, a multi-objective reward function is established, which integrates real-time sea state information (extreme water temperature, sea ice, etc.), dynamic migration law of fishing grounds and energy consumption characteristics of ships (calculating carbon emissions based on sailing distance, power and fishing time). The Deep Q Network function is established to further optimize the location of fishing grounds to 1-2, and a low-carbon fishing route planning strategy is constructed to achieve the carbon reduction target.

[0036] As a preferred embodiment, the names, gross tonnage, main engine power, and vessel positions of all fishing vessels in the Antarctic krill fishery in step (1), as well as the annual yield data for the 2023-2024 fishing season (as shown in Tables 1 and 2), are obtained through various sources, documents, CCAMLR website downloads, and Doubao AI, etc.; the vessel position data comes from positioning data such as the automatic identification system (AIS), with a vessel position resolution of 1 position every 5 to 10 minutes.

[0037] Table 1. Ship parameters of krill fishing vessels worldwide

[0038]

[0039] Table 2. Krill catch by fishing party in the 2023-2024 fishing season

[0040]

[0041] The vessel parameters for Chinese krill vessels in step (2) cover all vessel types in the world's krill fleets and can be obtained through... or A comparative fitting estimation of daily fuel consumption (Yi) for foreign vessels under navigation conditions (see...) Figure 2 The estimation formula is as follows: The number of krill fishing vessels worldwide each year is estimated at 8 to 13, with a total tonnage ranging from 2,455 to 16,000 tons.

[0042] The daily catch of different krill vessels in step (3) varies. Higher catches correspond to higher fuel consumption. Empirical calculations show that for every 100 tons increase in daily krill production, daily fuel consumption increases by approximately 2.2 to 2.6 tons. Under the same production volume, krill vessels processing frozen krill consume approximately 5 to 15 tons less fuel per day than those processing krill meal. In the 2023-2024 fishing season, there were 12 krill vessels (Table 3). Considering the different tonnages, catch volumes, and processing methods of the krill vessels, the daily fuel consumption of a particular vessel was... Where a is any constant from 2.2 to 2.6, See Table 3 for quick calculation values.

[0043] Table 3. Quick Calculation Table of Daily Fuel Consumption Coefficient (tons / day) for Various Fishing Vessels in the Krill Fishery under Different Operating Conditions during the 2023-2024 Fishing Season

[0044]

[0045] The formula for calculating the daily total carbon dioxide emissions of the krill vessel in step (4) is as follows: 3.1605, where n is the number of fishing vessels in that fishing season. The total monthly and spatial emissions of carbon dioxide for the entire krill fishery can be plotted using ArcGIS 10.0 plotting software.

[0046] In step (5), the custom deep Q-network algorithm defines the state (s, marine environmental data features such as sea temperature, salinity, and ocean currents), action (a, the set of fishing ground locations, longitude and latitude), and reward function (x, with catch volume and prediction accuracy as the core). A multi-objective reward R(s,a) is designed, with the formula: In the formula, λ1 and λ2 are coefficients, and R0 catch R represents the catch per unit state (tons). predThe error range between the predicted and actual values ​​is defined. The optimal strategy is learned through interaction with suitable marine environmental characteristic fields, avoiding sea ice concentration and high SST areas, and fishery electronic log data to form a multi-objective reward R(s,a) value. Through continuous interaction with the background environment, the deep Q-network optimizes the R(s,a) value to maximize the cumulative reward, thereby intelligently planning fishing routes, reducing the distance of blindly exploring fishing grounds, saving fuel, and achieving the goal of reducing carbon emissions. Figure 5 As shown.

[0047] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A method for carbon assessment and fishing strategy optimization in krill fisheries based on carbon reduction targets, characterized in that, Specifically, the steps include the following: (1) First, collect and organize the names, total tonnage, main engine power, and fishing vessel positions and annual production data of all fishing vessels in the Antarctic krill fishery. At the same time, download the annual catch data of each fishing area and each fishing party from the CCAMLR website to establish a database of krill fishing vessel fishery production. (2) Taking Chinese krill fishing vessels as a case study, a daily fuel consumption calculation model for vessels of different tonnages was developed. At the same time, the differences in daily fuel consumption caused by the differences in krill processed products were considered, and empirical algorithms for different vessel types, different vessel states and daily fuel consumption were fitted. (3) Combine the daily or annual catch and the krill product situation to calculate the fuel consumption of each krill vessel and further calculate the annual changes and differences of the fuel consumption of the entire krill fleet, and optimize the empirical algorithm model; at the same time, combine the status duration of each vessel to calculate the daily fuel consumption and daily carbon dioxide emissions of each vessel. (4) Draw monthly variation maps of carbon dioxide emissions from krill fisheries in each fishing season and spatial variation maps of carbon dioxide emissions in different sub-regions to realize a visual thematic map of the overall distribution of carbon emissions in the entire krill fishing area. (5) Based on Deep Q Network (DQN), the results of deep learning-based fishing ground forecast model are combined with real-time reinforcement learning of the current environmental state to make decisions on the optimal path, thereby realizing intelligent planning of fishing routes.

2. The method for carbon assessment and fishing strategy optimization in krill fisheries based on carbon reduction targets, as described in claim 1, is characterized in that... The names, gross tonnage, main engine power, and vessel positions and annual production data of all fishing vessels in the Antarctic krill fishery in step (1) were obtained through various materials, documents, CCAMLR website downloads, and Doubao AI, etc.; the vessel position data came from positioning data such as the automatic identification system (AIS), with a vessel position resolution of 1 position every 5 to 10 minutes.

3. The method for carbon assessment and fishing strategy optimization in krill fisheries based on carbon reduction targets, as described in claim 1, is characterized in that... The vessel parameters for Chinese krill vessels in step (2) cover all vessel types of krill fleets worldwide, and can be obtained through... or A comparative fitting estimation of daily fuel consumption (Yi) under foreign ship navigation conditions was performed (Figure 2). The estimation formula is as follows: ， The number of krill fishing vessels worldwide is 8 to 13 per year, with a total tonnage ranging from 2,455 to 16,000 tons.

4. The method for carbon assessment and fishing strategy optimization in krill fisheries based on carbon reduction targets, as described in claim 1, is characterized in that... The daily catch of different krill vessels in step (3) is different. The higher the catch, the higher the fuel consumption. According to empirical calculation, for every 100 tons increase in daily krill production, the daily fuel consumption increases by about 2.2 to 2.6 tons.

5. The method for carbon assessment and fishing strategy optimization in krill fisheries based on carbon reduction targets, as described in claim 1, is characterized in that... In step (3), under the same output, the daily fuel consumption of a krill vessel processing frozen shrimp is about 5-15 tons lower than that of a vessel processing krill meal. Taking into account different tonnages, catch volumes, and processing methods of krill vessels, the daily fuel consumption of a certain vessel is... , where a is any constant from 2.2 to 2.

6.

6. The method for carbon assessment and fishing strategy optimization in krill fisheries based on carbon reduction targets, as described in claim 1, is characterized in that... The formula for calculating the total daily carbon dioxide emissions of the krill boat in step (4) is as follows: 3.1605. Using ArcGIS 10.0 plotting software, a map of the total monthly carbon dioxide emissions and spatial emissions for the entire krill fishery can be generated.

7. The method for carbon assessment and fishing strategy optimization in krill fisheries based on carbon reduction targets, as described in claim 1, is characterized in that... In step (5), a multi-objective reward R(s,a) is designed using the state, action, and reward function of a custom deep Q-network algorithm. The state of the deep Q-network algorithm is denoted as 's', and the marine environmental data features include sea temperature, salinity, and ocean currents. The action is denoted as 'a', which includes the set of fishing ground locations, longitude, and latitude. The reward function is denoted as 'r', with catch volume and prediction accuracy as the core parameters. The formula for R(s,a) is: ， In the formula, λ1 and λ2 are coefficients, and R catch R represents the catch per unit state (tons). pred This represents the error range between the predicted and the actual values.

8. The method for carbon assessment and fishing strategy optimization in krill fisheries based on carbon reduction targets, as described in claim 1, is characterized in that... In step (5), the optimal strategy is learned through interaction with suitable marine environmental characteristic fields, avoiding sea ice concentration and high SST areas, and fishing electronic log data to form a multi-objective reward R(s,a) value.

9. The method for carbon assessment and fishing strategy optimization in krill fisheries based on carbon reduction targets, as described in claim 1, is characterized in that... In step (5), by continuously interacting with the background environment, the deep Q network optimizes the R(s,a) value to maximize the cumulative reward and realize intelligent planning of fishing routes, thereby reducing the blind exploration of fishing grounds, saving fuel, and achieving the goal of reducing carbon emissions.

10. The method for carbon assessment and fishing strategy optimization in krill fisheries based on carbon reduction targets, as described in claim 1, is characterized in that... In step (5), a multi-objective reward function with high yield and prediction accuracy, and time and space dimensions as the core is used. Real-time sea condition information, dynamic migration patterns of fishing grounds and ship energy consumption characteristics are integrated. Carbon emissions are calculated based on navigation distance, power and fishing time. A deep Q network function is established to further optimize the fishing ground location to 1-2. A low-carbon fishing path planning strategy is constructed to achieve the carbon reduction target.

Images