Multi-layer nutrition-level comprehensive breeding species combination optimization method for marine ranching

Abstract

The invention provides a marine ranch multi-layer nutrition-level comprehensive breeding species combination optimization method. The method comprises the steps of determining an optimization target, determining a mathematical expression of a target function, constructing a multi-target optimization model, selecting an optimization algorithm, calculating a crowding degree, selecting an optimal solution, and implementing and verifying. Based on multi-objective optimization, the optimal species configuration combination is found, so that high productivity, good economic benefit and minimum environmental influence of the breeding system are realized at the same time.

Description

Technical Field

[0001] The present invention belongs to the technical field of aquaculture, and particularly relates to a method for optimizing the species combination of multi-trophic integrated aquaculture in a marine ranch. Background Art

[0002] With the growth of the global population and the decreasing of marine resources, marine ranching, as an intensive aquaculture model, has gradually become an important way to meet the human demand for aquatic products. By setting up artificial facilities in a specific sea area, providing a suitable aquaculture environment, and cultivating a variety of aquatic products, marine ranching can effectively increase the yield and achieve the sustainable utilization of resources. In recent years, the integrated multi-trophic aquaculture system, abbreviated as IMTA, as a new aquaculture model, has received extensive attention. The IMTA system raises organisms at different trophic levels, such as fish, shellfish, and algae, in the same ecosystem, and uses their symbiotic relationship to achieve the maximum utilization of resources and environmental friendliness.

[0003] In the multi-trophic ecological aquaculture model, the selection of species is crucial. Species selection is a necessary means to ensure the productivity and environmental sustainability of the system. Common species combinations include fish, shellfish, algae, and other low-trophic-level organisms. These species form an efficient aquaculture system through complementary nutritional requirements and waste utilization. Research shows that in the combination of shrimp (such as Litopenaeus vannamei), fish (such as Nile tilapia), and seaweed (such as Gracilaria), seaweed can effectively absorb inorganic nutrients in the water body, thereby promoting the growth of fish and shrimp. In addition, the combination of herbivorous or filter-feeding fish, such as grass carp and tilapia, and algae can maximize the nutrient recovery rate of the aquaculture system and reduce feed consumption. Another important consideration in the IMTA system is the species combination that can adapt to different environmental conditions, which can optimize the productivity and environmental benefits of the whole system. For example, the combination of highly adaptable fish, such as sea bass and barramundi, and other species. In addition, the combination of shellfish (mussels or oysters) and seaweed also shows the potential in removing oil-based waste and increasing the economic benefits of the system.

[0004] However, although the IMTA system has many theoretical advantages, the current research and practice still have significant deficiencies, which are mainly reflected in the following aspects: (1). The selection of organisms at different trophic levels is based on experience and lacks scientific pairing optimization. In the design and application of existing IMTA systems, the selection of organisms at different trophic levels, such as fish, shellfish, and algae, is usually based on experience and has not undergone scientific pairing optimization. This practice relying on experience often ignores the interaction between different organisms and its impact on the overall performance of the aquaculture system, resulting in great randomness and uncertainty in the selection of species combinations, and it is difficult to ensure the optimal performance of the system in actual operation.

[0005] (2). Currently, most research and practical applications still focus on the optimization of a single objective, such as maximizing the yield of a single species or enhancing economic benefits. This single-objective optimization approach ignores the interactions between different species in the system, as well as the conflicts and balances among multiple objectives such as productivity, economic benefits, and environmental impacts. This leads to problems of resource waste and environmental pollution while pursuing high yields or high economic benefits, making it difficult to achieve true sustainable development.

[0006] (3). Although the IMTA system itself emphasizes the synergistic effects of multiple species and multiple trophic levels, most existing optimization methods rely on linear programming or empirical rules and lack a systematic multi-objective optimization method. This makes it difficult to effectively integrate the aquaculture strategies of various species and achieve the best balance among different objectives such as productivity, economic benefits, and environmental impacts when facing complex aquaculture environments and changing market demands.

[0007] (4). Environmental impacts have not been fully emphasized. Although the IMTA system theoretically has the advantage of being environmentally friendly, in actual operation, environmental impacts are often not fully considered or are ignored. Existing optimization methods usually have limited pursuit of economic benefits and yields, while ignoring environmental protection objectives such as reducing nitrogen and phosphorus emissions and controlling water eutrophication. This situation gives priority to the pursuit of economic benefits and yields while ignoring environmental protection objectives such as reducing nitrogen and phosphorus emissions and controlling water eutrophication. This situation may lead to long-term environmental deterioration and is not conducive to the sustainable development of marine pastures.

[0008] Invention Name: A multi-trophic level integrated aquaculture system based on water quality regulation, with an application number of CN202210960932.5. This invention includes: a species release module for releasing aquaculture species into a freshwater pond; an environmental regulation module connected to the species release module for intelligently adjusting aquaculture environment parameters in real time according to aquaculture conditions to obtain a regulation result; a monitoring and early warning module connected to the environmental regulation module for comparing the regulation result with a preset threshold to obtain an early warning result; and a release adjustment module connected to the monitoring and early warning module for adjusting the proportion and density of aquaculture species according to the early warning result. The integrated aquaculture system of this invention increases the total aquaculture output without expanding the aquaculture area, stabilizes the aquaculture environment, and reduces aquaculture waste emissions. It not only improves the utilization rate of materials and space but also effectively improves the aquaculture ecological environment.

[0009] Disadvantages and deficiencies of a multi-trophic level integrated aquaculture system based on water quality regulation: (1). Limited scope of application: This patent mainly focuses on freshwater pond aquaculture systems. Although it improves the concept of multi-trophic levels, it does not fully consider the ecological niche complementarity of cultured species in the marine environment and the unique ecological conditions in the ocean, such as tides, ocean currents, salinity changes, etc. This may limit the application effect of this system in mariculture. Especially in the marine or large-scale water environment, the speed and direction of water flow are crucial for the location selection of species placement. Failing to consider this may result in the system being unable to fully utilize the advantages of multi-trophic level ecological aquaculture and unable to maximize ecological and economic benefits.

[0010] (2). Lack of application of the principle of ecological niche complementarity: The multi-trophic level aquaculture system in this patent does not fully consider the ecological niche complementarity between species, that is, the complementarity of different species in resource utilization, habitat space, and ecological functions. Failure to select and place species based on the principle of ecological niche complementarity may lead to low resource utilization efficiency, weak overall stability and stress resistance of the system, and may perform worse than systems based on ecological niche complementarity in terms of aquaculture benefits and environmental sustainability.

[0011] (3). High cost of environmental regulation: The environmental regulation of this patent relies on complex hardware systems, such as filtration, disinfection, water replenishment, water change, etc., which require high energy and resource inputs.

[0012] (4). Insufficient adaptability and flexibility: The species placement and environmental regulation involved in this patent are based on historical ecological data and fixed parameter settings, lacking sufficient flexibility and being difficult to cope with the dynamic changes in the marine environment.

[0013] The invention name is a three-dimensional multi-trophic level hybrid aquaculture workboat, and the application number is 202110544332.6, which relates to the technical field of aquaculture workboats. The three-dimensional multi-trophic level hybrid aquaculture workboat includes: an aquaculture workboat hull, a multi-layer aquaculture tank mechanism, and the multi-layer aquaculture tank mechanism is connected to the aquaculture workboat hull; a water inlet mechanism, and the water inlet mechanism is connected to the water inlet above the multi-layer aquaculture tank mechanism; a water outlet mechanism, and the water outlet mechanism is connected to the water outlet below the multi-layer aquaculture tank mechanism.

[0014] The disadvantages of a three-dimensional multi-trophic level hybrid aquaculture workboat are: (1). The mechanical design and water flow regulation of the three-dimensional multi-layer aquaculture system do not consider the principle of ecological niche complementarity, nor do they involve how to scientifically place species at different trophic levels in different aquaculture layers within the system to maximize ecological benefits. The system design focuses more on the mechanical structure and preventing pipeline blockages, while ignoring species selection in multi-trophic level aquaculture and their placement strategies under different water flow conditions.

[0015] (2). The system design is complex, including structures such as multi - layer aquaculture tanks, connecting pipes, and movable components, which may lead to high manufacturing and maintenance costs. Especially in the marine environment, complex mechanical structures may face higher failure risks and maintenance difficulties.

[0016] (3). It focuses more on short - term mechanical reliability and aquaculture efficiency, while paying less attention to the long - term ecological sustainability of the system. For example, there is no in - depth exploration of how to maintain ecological balance and prevent environmental degradation during the aquaculture process.

[0017] In summary, the following technical problems exist in the existing technologies: (1). In the IMTA system, different aquaculture goals, such as productivity, economic benefits, and environmental impact, are usually conflicting. For example, increasing the yield of a certain species may increase the nitrogen and phosphorus emissions of the system, leading to an exacerbation of environmental impact. Traditional optimization methods cannot take into account these conflicts and goals simultaneously.

[0018] (2). Identify the optimal species combination. For the IMTA system, there is no single optimal species combination, but there are multiple combinations that perform well in different goals. These combinations form the Pareto optimal solution set, and each combination cannot be further optimized in some goals without compromising other goals. The technical problem to be solved is how to effectively identify these optimal combinations. Summary of the Invention

[0019] The purpose of the present invention is to solve the defects existing in the above - mentioned existing technologies, and provide an optimization method for the species combination of multi - trophic integrated aquaculture in a marine ranch, aiming to solve the multi - objective optimization problem in the IMTA system. By introducing a multi - objective optimization algorithm, it can achieve the best balance among productivity, economic benefits, and environmental impact.

[0020] The present invention provides a set of scientific species combination optimization methods. Through evolutionary algorithm means such as non - dominated sorting and crowding degree calculation, it effectively identifies and retains diverse optimal solutions, ensuring the adaptability and flexibility of the system.

[0021] The present invention adopts the following technical solutions: An optimization method for the species combination of multi - trophic integrated aquaculture in a marine ranch, comprising the following steps: Step 1. Determine the optimization goals. The multi - objective optimization of the IMTA system usually includes three main goals: (1). Maximize productivity, (2). Maximize economic benefits, (3). Minimize environmental impact. Productivity: The sum of the total biomass of all organisms in the system, denoted as Z1. Economic benefits: The economic return of the system, denoted as Z2. Environmental impact: The adverse impact of the system on the environment, such as nitrogen and phosphorus emissions, denoted as Z3.

[0022] Step 2. Determine the mathematical expression of the objective function. Z1: The sum of the total biomass of all organisms in the system.

[0023] Where, is the weight of the i-th species, is the biomass of the i-th species, i = 1, 2, …, n.

[0024] Z2: The economic return of the system, that is, the difference between the market value of each species and its production cost.

[0025] Where, is the market price of the i-th species, is the production cost per unit area of the i-th species, is the cultivation area of the i-th species.

[0026] Z3: The adverse impact of the system on the environment, that is, based on the total nitrogen and phosphorus emissions of all species in the system.

[0027] Where, is the nitrogen emission of the i-th species, is the phosphorus emission of the i-th species.

[0028] Step 3. Construct a multi-objective optimization model. Combine the objective functions in Step 2 to form a multi-objective optimization model: Maximize Z1( B 1 , B 2 , …, B n ), Maximize Z2( , …, ), Minimize Z3((N 1 , P 1 ), …, (N n , P n )); The constraint conditions include: The biomass B i of each species must be positive. The limit of the total cultivation area A: The sum of the cultivation areas of all species cannot exceed the total cultivation area A, and A = A 1 + A 2 + … + A n . Environmental condition limitations: Environmental conditions such as water temperature and salinity must be within the tolerance range of the species.

[0029] Step 4. Select an optimization algorithm. Use the Non-dominated Sorting Genetic Algorithm II to process the multi-objective optimization model in Step 3 and generate the Pareto front. The specific steps include: generating an initial population according to the aquaculture plan, where multiple individuals represent a combination, that is, generating the proportion of different species or the aquaculture area allocation plan, calculating the Z1, Z2, and Z3 values of each individual, performing non-dominated sorting on the individuals in the population, calculating the dominance relationship, and generating different levels of Pareto fronts.

[0030] Step 5. Calculate the crowding degree. Prioritize selecting individuals located on the first Pareto front F 1 as the optimal solutions. Among the individuals within the same Pareto front, calculate the crowding degree to measure the distribution of individuals. Sort these combinations according to productivity, economic benefits, and environmental impact. For each objective Z1, Z2, Z3, the crowding degrees of the first and last individuals in the sorting are set to infinity. These individuals are located at the boundaries to ensure that they will not be preferentially eliminated. For each non-boundary individual, the crowding degree is calculated as follows: For the objective Zk, where k takes the value of 1 or 2 or 3, the crowding degree contribution of the i-th individual is:

[0031] where, is the value of the next objective after the i-th individual is sorted on the objective Zk, is the value of the previous objective after the i-th individual is sorted on the objective Zk, is the maximum value of this objective, is the minimum value of this objective.

[0032] Step 6. Select the optimal solutions. Based on the results of the crowding degree, preferentially retain individuals with a large crowding degree to ensure the diversity and quality of the solutions.

[0033] Step 7. Implementation and verification. Apply the selected species combination model to the actual IMTA system for experimental aquaculture, and verify the effectiveness of the optimization model through data feedback.

[0034] Furthermore, the specific steps of generating the Pareto front in Step 4 include: For each combination p, initialize the conditional dominance count n p . n p represents the number of times the combination p is dominated by other combinations, and the initial value is 0. The dominance set S p . S p represents the set of other combinations dominated by the combination p, and is initially empty. For each pair of combinations p i and p j , according to the dominance relationship results, check the following conditions: If p i is not inferior to p j in all objectives, and is superior to p in at least one objectivej , then p i dominates p j , add p j to S p and increment n p . Find all combinations with domination count n p equal to 0, which form the first Pareto front F 1 . Construct the next Pareto front, remove all combinations in F 1 , and recalculate the domination relationships for the remaining combinations.

[0035] Furthermore, step 5 also includes: If further screening is required, select individuals in F 1 based on crowding distance. Individuals with a larger crowding distance have greater diversity and exploration ability, so they are preferentially retained, and finally one or more solutions are selected as the optimal solutions.

[0036] Furthermore, step 5 also includes: A larger crowding distance of an individual means that it is farther from other individuals in the objective space.

[0037] Advantages of the present invention: (1). Optimize the productivity and economic benefits of the aquaculture system. By optimizing the productivity Z1 and economic benefits Z2, it is possible to select the optimal configuration among various species combinations in terms of these two objectives. This optimization can not only maximize the total biomass of the system but also improve the economic return of the farm. Through the non-dominated sorting algorithm, the system can effectively identify the combinations that achieve the best balance between productivity and economic benefits, avoiding other negative impacts that may be brought about by simply pursuing high yield or high income.

[0038] (2). Reduce environmental impact and support sustainable development. The present invention particularly focuses on the optimization of the environmental impact Z3. Using the multi-objective optimization algorithm, the system can find the optimal solutions among numerous species configuration combinations in terms of minimizing nitrogen and phosphorus emissions. Through crowding distance calculation, the system can retain diverse configuration combinations to ensure that feasible solutions with low environmental impact can be found under different environmental conditions, which provides technical support for the sustainable development of the IMTA system.

[0039] (3). Achieve efficient solution of multi-objective optimization. Through crowding distance calculation, the present invention can effectively balance multiple objectives and quickly determine the optimal configuration combination. This optimization method has significant advantages over traditional single-objective optimization schemes, and can consider multiple conflicting objectives such as productivity, economic benefits, and environmental impact simultaneously, thus providing more comprehensive and integrated optimization results.

[0040] (4). Provide personalized aquaculture solutions. The present invention can flexibly adjust and optimize strategies according to the different weight requirements of different farms and decision-makers for various objectives, and generate an optimal species combination configuration plan that meets specific requirements. Whether it focuses on high yield, pursues high returns, or pays attention to environmental protection, the present invention can provide personalized solutions to meet the needs of different users. Brief Description of the Drawings

[0041] Figure 1 It is a flowchart of the steps of the present invention. Detailed Embodiments

[0042] To make the objectives, technical solutions and advantages of the present invention clearer, the technical solutions in the present invention will be clearly and completely described below. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present invention without creative efforts shall fall within the protection scope of the present invention.

[0043] Embodiment As Figure 1 shown, a method for optimizing the species combination of multi-trophic integrated aquaculture in a marine ranch of the present invention includes the following steps: Step 1. Determine the optimization objectives. The multi-objective optimization of the IMTA system usually includes three main objectives: maximizing productivity, maximizing economic benefits, and minimizing environmental impact. Productivity: The sum of the total biomass of all organisms in the system, denoted by Z1. Economic benefits: The economic return of the system, denoted by Z2. Environmental impact: The adverse impact of the system on the environment, such as nitrogen and phosphorus emissions, denoted by Z3.

[0044] Step 2. Determine the mathematical expressions of the objective functions. Z1: The sum of the total biomass of all organisms in the system.

[0045] Among them, is the weight of the i-th species, is the biomass of the i-th species, i = 1, 2,..., n. Z2: The economic return of the system, that is, the difference between the market value of each species and its production cost.

[0046] Among them, is the market price of the i-th species, is the production cost per unit area of the i-th species, is the aquaculture area of the i-th species. Z3: The adverse impact of the system on the environment, that is, the total nitrogen and phosphorus emissions based on all species in the system.

[0047] Among them, is the nitrogen emission of the i-th species, is the phosphorus emission of the i-th species.

[0048] Step 3. Construct a multi-objective optimization model. Combine the objective functions in Step 2 to form a multi-objective optimization model: MaximizeZ1( , , …, ), MaximizeZ2( , , …, ), MinimizeZ3((N 1 , P 1 ), …, (N n , P n )); The constraint conditions include: The biomass B i of each species must be positive. The limit of the total aquaculture area A: The sum of the aquaculture areas of all species cannot exceed the total aquaculture area A, and A = A 1 + A 2 + … + A n . Environmental condition limit: Environmental conditions such as water temperature and salinity must be within the tolerance range of the species.

[0049] Step 4. Select an optimization algorithm. Use the Non-dominated Sorting Genetic Algorithm II, abbreviated as NSGA-II. This algorithm is suitable for dealing with complex multi-objective optimization models and generating the Pareto front. Use the Non-dominated Sorting Genetic Algorithm II to process the multi-objective optimization model in Step 3 and generate the Pareto front. The specific steps are as follows:

[0050] S401. Population initialization: Generate an initial population according to the aquaculture plan. Multiple individuals represent a possible combination (such as laver, mussel, and golden pompano representing a combination), that is, generate the proportion of different species or the aquaculture area allocation plan. S402. Objective function calculation: Calculate the values of Z1, Z2, and Z3 for different combinations. S403. Non-dominated sorting: Perform non-dominated sorting on different combinations to generate the Pareto front.

[0051] As in Example 1: In the IMTA system, assume there are the following six species: laver, kelp, mussel, oyster, golden pompano, and large yellow croaker. The performance of different species combinations on the three objective functions is as follows:

[0052] Objective 1: Productivity, unit: ton. Objective 2: Economic benefit, unit: 10,000 yuan. Objective 3: Environmental impact, unit: kg.

[0053] Assume there are five different combinations, as shown in Table 1: Table 1 Example of Aquaculture Biological Combinations in the IMTA System S4031. Specific steps of non - dominated sorting: Calculate the domination relationship among these combinations: For combination A and combination B: Combination B is superior to combination A in productivity Z1 (12 > 10), also superior to combination A in economic benefit Z2 (55 > 50), and also superior to combination A in environmental impact Z3 (180 < 200). Therefore, combination B completely dominates combination A.

[0054] For combination B and combination C: Combination C is superior to combination B in productivity Z1 (15 > 12), also superior to combination B in economic benefit Z2 (60 > 55), but inferior to combination B in environmental impact Z3 (220 > 180). Combination B and combination C do not dominate each other.

[0055] For combination B and combination D: Combination B is superior to combination D in productivity Z1 (12 > 8), also superior to combination D in economic benefit Z2 (55 > 45), but inferior to combination D in environmental impact Z3 (180 > 150). Therefore, combination B and combination D do not dominate each other.

[0056] For combination B and combination E: Combination E is superior to combination B in productivity Z1 (20 > 12), also superior to combination B in economic benefit Z2 (70 > 55), but inferior to combination B in environmental impact Z3 (250 > 180). Therefore, combination B and combination E do not dominate each other.

[0057] For combination C and combination D: Combination C is superior to combination D in productivity Z1 (15 > 8), also superior to combination D in economic benefit Z2 (60 > 45), but inferior to combination D in environmental impact Z3 (220 > 150). Therefore, combination C and combination D do not dominate each other.

[0058] For combination C and combination E: Combination E is superior to combination C in productivity Z1 (20 > 15), also superior to combination C in economic benefit Z2 (70 > 60), but inferior to combination C in environmental impact Z3 (250 > 220). Therefore, combination C and combination E do not dominate each other.

[0059] For combination D and combination E: Combination E is superior to combination D in productivity Z1 (20 > 8), also superior to combination D in economic benefit Z2 (70 > 45), but inferior to combination D in environmental impact Z3 (250 > 150). Therefore, combination D and combination E do not dominate each other.

[0060] S404. Generate the Pareto front. S4041. For each combination p, initialize the conditional domination count n p n p represents the number of times combination p is dominated by other combinations, with an initial value of 0. The domination set S p : represents the set of other combinations dominated by combination p, initially empty.

[0061] For each pair of combinations p i and p j , according to the domination relation result, check the following condition: If p i is not inferior to p j on all objectives, and is superior to p j on at least one objective, then p i dominates p j . Add p j to S p , and increment the value of n p .

[0062] For Example 1: Combinations A and B: Combination B dominates Combination A. Increment the domination count n A of Combination A by 1. Combinations B and C, Combinations B and D, Combinations B and E, Combinations C and E, and Combinations D and E do not dominate each other. Summary of the domination relation: Combination B dominates Combination A, so

[0063] S4042. Determine the Pareto front F 1 . Find all combinations with a domination count n p of 0, and they form the first Pareto front F 1 . In Example 1, n A = 1, so Combination A is not on the first Pareto front. Combinations B, C, D, and E are not dominated by other combinations, so they form the first Pareto front F 1 : .

[0064] S4043. Construct the next Pareto front, remove all combinations in F 1 , and recalculate the domination relations of the remaining combinations. In Example 1, Combination A forms the next Pareto front F 2 : . Combinations B, C, D, and E are part of the Pareto optimal solutions because they do not dominate each other. Combination A is dominated: Combination A is worse than Combination B on all objectives, so Combination A is a suboptimal solution.

[0065] Step 5. Calculate the crowding degree. Prefer individuals located on the first Pareto front F 1 as the optimal solutions. Prefer individuals located on the first Pareto front F 1 as the optimal solutions because they are not dominated by other individuals on all objectives. If further screening is needed, it can be done on F 1Individuals are selected based on crowding degree. Individuals with a larger crowding degree usually have greater diversity and exploration ability, so they are preferentially retained. Through this process, one or more solutions are finally selected as the optimal solutions.

[0066] Among the individuals within the same Pareto front, the crowding degree is calculated to measure the distribution among individuals. A larger crowding degree of an individual means that it is farther away from other individuals in the objective space, and retaining such individuals helps to maintain the diversity of the population.

[0067] In Example 1, in the first front Pareto solutions, there are Combinations B, C, D, and E, as shown in Table 2.

[0068] Table 2 First Front Pareto Solutions S501. Sort these combinations according to productivity, economic benefits, and environmental impact. In Example 1, for productivity Z1: D < B < C < E. For economic benefits Z2: D < B < C < E. For environmental impact Z3: E > C > B > D.

[0069] S502. Calculate the crowding degree of boundary individuals. For each objective, the crowding degrees of the first and last individuals in the sorting are set to infinity ∞ because these individuals are at the boundary to ensure that they will not be preferentially eliminated. In Example 1, for the productivity Z1 objective, the crowding degrees of Combinations D and E are set to infinity. For the economic benefits Z2 objective, the crowding degrees of Combinations D and E are set to infinity. For the Z3 objective, the crowding degrees of Combinations D and E are set to infinity.

[0070] S5013. Calculate the crowding degree of internal individuals. For each non-boundary individual, the crowding degree is calculated as follows: For the objective Zk, where k takes the value of 1 or 2 or 3, the crowding degree contribution of the i-th individual is:

[0071] where is the value of the next objective after the i-th individual is sorted on the objective Zk, is the value of the previous objective after the i-th individual is sorted on the objective Zk, is the maximum value of this objective, is the minimum value of this objective.

[0072] Specific calculation in Example 1: For Combination B: Crowding degree Z1 = (f c (Z1) - f D (Z1)) / (f E (Z1) - f D (Z2)) = (15 - 8) / (20 - 8) = 0.583. Crowding degree Z2 = (fc (Z2)-f D (Z2)) / (f E (Z2)-f D (Z2))=(60 - 45) / (70 - 45)=0.6. The crowding degree Z3=(f c (Z3)-f D (Z3)) / (f E (Z3)-f D (Z3))=(220 - 150) / (250 - 150)=0.7. Therefore, the total crowding degree of combination B is: Crowding degree B = 0.583 + 0.6 + 0.7 = 1.883.

[0073] For combination C: Crowding degree Z1=(f E (Z1)-f B (Z1)) / (f E (Z1)-f D (Z1))=(20 - 12) / (20 - 8)=0.667. Crowding degree Z2=(f E (Z2)-f B (Z2)) / (f E (Z2)-f D (Z2))=(70 - 55) / (70 - 45)=0.6. Crowding degree Z3=(f E (Z3)-f B (Z3)) / (f E (Z3)-f D (Z3)=(250 - 180) / (250 - 150)=0.7. Therefore, the total crowding degree of combination C is: Crowding degree C = 0.667 + 0.6 + 0.7 = 1.967. Finally, the results of the crowding degree are as follows: Crowding degree of combination B: 1.883. Crowding degree of combination C: 1.967. Crowding degree of combination D: ∞. Crowding degree of combination E: ∞.

[0074] Step 6. Select the optimal solution. Based on the results of the crowding degree, give priority to retaining individuals with a large crowding degree to ensure the diversity and quality of the solutions. Example 1 The boundary individuals D and boundary individuals E will definitely be retained because their crowding degree is ∞. Among the non-boundary individuals, the crowding degree of combination C is the largest at 1.967. Therefore, combination C is a relatively better solution in this set of solutions.

[0075] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for optimizing the combination of species in a multi-trophic level integrated aquaculture system in a marine ranch, characterized in that , including the following steps: Step 1. Determine the optimization goal: The multi-objective optimization of the IMTA system includes three goals: maximizing productivity, maximizing economic benefits, and minimizing environmental impact; productivity: the sum of the total biomass of all organisms in the system, represented by Z1; economic benefits: the difference between the market value of each species and its production cost, represented by Z2; environmental impact: the total nitrogen and phosphorus emissions based on all species in the system, represented by Z3; Step 2. Determine the mathematical expression of the objective function: Z1: The productivity formula is: in, is the weight of the ith species, is the biomass of the i-th species, i=1, 2, …, n; Z2: The economic benefit formula is: in, is the market price of the ith species, is the production cost per unit area of ​​the ith species, is the breeding area of ​​the i-th species; Z3: The environmental impact formula is: in, is the nitrogen emission of the ith species, is the phosphorus emission of the ith species; Step 3. Build a multi-objective optimization model: Combine the objective functions in step 2 to form a multi-objective optimization model: MaximizeZ1 ( B1, B2, ..., B n )、MaximizeZ2( , …, ), MinimizeZ3((N1,P1),…,(N n , P n )); Constraints include: the biomass B of each species i Must be a positive value; Limitation of total breeding area A: The sum of breeding areas of all species cannot exceed the total breeding area A, and A=A1+A2+…+A n ; Environmental conditions: Environmental conditions must be within the tolerance range of the species; Step 4. Select optimization algorithm: Use non-dominated sorting genetic algorithm II to process the multi-objective optimization model in step 3 and generate the Pareto frontier, including: generating an initial population according to the breeding plan, multiple individuals represent a combination, calculating the Z1, Z2 and Z3 values ​​of each individual, performing non-dominated sorting on the individuals in the population, calculating the dominance relationship, and generating the Pareto frontier; Step 5. Calculate the crowding degree: Among the individuals in the same Pareto frontier, select the individual located in the first Pareto frontier F1 as the optimal solution, and measure the distribution between individuals by calculating the crowding degree. The crowding degree calculation includes: sorting the combinations according to productivity, economic benefits and environmental impact. For each target, the crowding degree of the first and last individuals in the sorting is set to infinity. The individuals at the boundary are ensured not to be eliminated first. For non-boundary individuals, the crowding degree is calculated as follows: For the target Zk, the value of k is 1, 2 or 3, and the crowding degree contribution of the i-th individual is: in is the value of the next target after the i-th individual is sorted on the target Zk, is the value of the previous target after i individuals are sorted on the target Zk, is the maximum value of the target, is the minimum value of the target; Step 6. Select the optimal solution: Based on the crowding results, individuals with high crowding are retained first; Step 7. Implementation and verification: Apply the selected species combination pattern to the actual IMTA system, conduct trial breeding, and verify the effectiveness of the optimization model through data feedback.

2. The method according to claim 1, characterized in that Step 4 generates the Pareto frontier, which specifically includes: for each combination p, initialize the conditional dominance count n p , n p It represents the number of times that combination p is dominated by other combinations. The initial value is 0. The dominating set S p : represents the set of other combinations dominated by combination p, which is initially empty; for each pair of combinations p i and p j , according to the dominance relationship results, check the following conditions: if p i Not inferior to p on all objectives j , and outperforms p in at least one objective j , then p i Dominate p j , p j Join S i and increase n j Find all the dominating counts n p The combinations with 0 constitute the first Pareto frontier F1; construct the next Pareto frontier by removing all combinations of F1, and then recalculate the dominance relationship of the remaining combinations.

3. The method according to claim 1, characterized in that Step 5 also includes: if screening is still required, individuals are selected in F1 based on crowding. Individuals with greater crowding have greater diversity and exploration ability, so they are retained first, and finally one or more solutions are screened out as the optimal solution.

4. The method according to claim 3, characterized in that Step 5 also includes: an individual with a large crowding degree means that it is farther away from other individuals in the target space.

Images