Marine ranching site selection evaluation method based on machine learning

Abstract

The invention discloses a marine ranch site selection evaluation method based on machine learning, and relates to the technical field of marine ranches, and the method comprises the following steps: determining a marine ranch site selection target, collecting the regional data of the site selection target, dividing the regional data into historical data and real-time data, carrying out the preprocessing of the collected historical data and real-time data, and carrying out the evaluation of the site selection of the marine ranch. The site selection influence factors comprise climate influence factors and ecological influence factors, and a comprehensive suitability evaluation model is constructed on the basis of historical data in combination with specific requirements of marine ranch site selection. According to the method, massive real-time data including information of climate, ecology, landform, social economy and the like are processed and analyzed through a machine learning technology, so that site selection evaluation is more accurate, errors caused by manual judgment are reduced, and meanwhile, by analyzing historical data and real-time data, the site selection evaluation accuracy is improved. And the key climate and ecological factors influencing the marine ranching site selection can be accurately identified.

Description

Technical Field

[0001] The present invention relates to the technical field of marine ranching, and in particular to a method for evaluating the site selection of marine ranching based on machine learning. Background Art

[0002] With the increasing shortage of fishery resources and people's emphasis on sustainable fishery development, marine ranching, as a new model of fishery development, has gradually received widespread attention and promotion. Through technical means such as artificial reefs, larval release, and fish behavior control, marine ranching has achieved the proliferation and protection of fishery resources, thereby improving the comprehensive output and sustainable benefits of fisheries. However, the construction of marine ranches is not a random site selection. Its site selection and construction involves many factors, including marine environment, ecological conditions, distribution of fishery resources, economic benefits, social impact, etc. These factors are interrelated and influence each other, making the site selection and evaluation work complicated and cumbersome.

[0003] In the existing technology, the construction of marine ranches is easily affected by extreme weather and climate change. It is also disturbed by human activities such as pollution and fishing operations, which affects the safety of marine ranch site selection and the quality of subsequent aquaculture resources. Therefore, how to comprehensively consider weather, climate change and human activity interference factors to quantitatively evaluate the site selection plan of marine ranches is the problem we need to solve. To this end, a marine ranch site selection evaluation method based on machine learning is proposed. Summary of the invention

[0004] The purpose of the present invention is to provide a marine ranch site selection assessment method based on machine learning to solve the problems raised in the above-mentioned background technology.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is: A method for assessing the site selection of marine ranches based on machine learning, step 1, determining the site selection objectives of marine ranches, such as fishery resource proliferation, ecological protection, economic benefits, etc., collecting regional data of the site selection objectives, and dividing the regional data into historical data and real-time data, the historical data includes marine environmental data of the past years, such as water temperature, salinity, ocean currents, seabed topography, climate data, ecological data, etc., the real-time data includes current or recent marine environmental data, meteorological data, etc., data are collected through satellite remote sensing, ocean observation stations, meteorological stations, scientific research institutions and other channels, wherein the regional data includes marine environmental data, such as water temperature, salinity, ocean currents, seabed topography, etc., marine ecological data, such as biodiversity, distribution of key species, etc., fishery resource data, such as historical catch, fish migration paths, etc., climate change data, such as frequency of extreme weather events, sea level rise, etc., human activity data, such as fishing operation intensity, shipping trajectory, pollution emissions, etc., economic benefit and social impact data, such as potential economic benefits, community participation, etc.; Step 2: Preprocess the collected historical data and real-time data, including data cleaning and normalization, and screen the factors affecting the site selection, including climate factors and ecological factors. For example, statistical methods such as variance selection method, correlation coefficient method, and mutual information method are used to evaluate the importance of the factors affecting the site selection, and the factors affecting the site selection with strong correlation with the target variable are selected. Climate factors include temperature, precipitation, wind speed, wind direction, etc., and ecological factors include biodiversity, marine pollution status, and seabed topography. Step 3: Based on historical data and in combination with the specific requirements for marine ranch site selection, a comprehensive suitability assessment model is constructed to obtain the climate impact index and ecological impact index to quantitatively assess the site selection objectives, including predicting the trend of climate change and assessing the scope of ecological change. If the goal is not only to maximize fishery output, but also to protect the ecology, maximize economic benefits, and improve social acceptance, the climate impact index reflects the climate suitability of the site selection area, including predicting the trend of climate change, and the ecological impact index reflects the ecological suitability of the site selection area, including assessing the scope of ecological change. Step 4: Based on the climate impact index and ecological impact index and combined with the comprehensive suitability assessment model, obtain the suitability assessment coefficient, determine the site selection plan, and analyze the impact of climate impact factors and ecological impact factors on site suitability; Step 5: Input the preprocessed real-time data into the comprehensive suitability assessment model to output the suitability assessment coefficient, determine the site selection plan, and evaluate and rank the site selection plans, considering factors such as the feasibility, cost, and environmental impact of the plans. Output an assessment report, including the advantages and disadvantages of the site selection, potential risks, and recommended measures, etc., formulate a specific marine ranch construction plan, and after the marine ranch is built, continuously monitor the marine environmental data, and adjust the management strategy in a timely manner to ensure the sustainable development of the marine ranch.

[0006] A further improvement of the technical solution of the present invention is that in step 1, the process of collecting data of the site selection target area is: Step 101, clarify the construction purpose. According to the construction purpose of the marine ranch, it is divided into three types: conservation type, proliferation type and leisure type. These three types of marine ranches are comprehensive marine ranches with different degrees of resource conservation, biological proliferation and leisure fishery functions. Among them, the conservation type is mainly aimed at ecological restoration and protection of marine biological resources. The key indicators are water quality, bottom type, biodiversity, ecologically sensitive areas, pollution level, etc. The proliferation type is mainly aimed at increasing the amount of specific marine biological resources and increasing fishery production. The key indicators are bait resource abundance, fish breeding habits, water flow conditions, sea productivity, etc.

[0007] The main purpose of the leisure type is to provide recreational fishing, sightseeing tourism, etc. The key indicators are transportation convenience, landscape value, tourist carrying capacity, sea area safety, etc. The key indicators of site selection are determined according to the purpose, including environmental conditions, economic benefits, ecological impact, etc.; Step 102: deploy drones, AI cameras, radars and optoelectronic equipment, and water quality sensors in the target site selection area to monitor the ecological environment, marine biological activities, and water quality in the target site selection area in real time, and obtain historical data from existing public databases and marine environmental monitoring stations, including basic geographic information, such as coastline, water depth, topography, marine functional zoning, and other basic geographic information, marine environmental data, such as water temperature, salinity, ocean currents, and seabed topography, marine ecological data, such as biodiversity and distribution of key species, fishery resource data, such as historical catches and fish migration routes, climate change data, such as the frequency of extreme weather events and sea level rise, human activity data, such as fishing operation intensity, shipping trajectories, and pollution emissions, and economic benefits and social impact data, such as potential economic benefits and community participation; Step 103, the collected historical data and real-time data are transmitted to the data management center via wireless communication, and are sorted, classified and stored to establish a marine ranch regional ecological database, and use big data, cloud computing and other technologies to conduct in-depth data mining and analysis to provide a basis for scientific management and decision-making of marine ranches.

[0008] A further improvement of the technical solution of the present invention is that in step 2, the process of obtaining the factors affecting site selection including climate factors and ecological factors is as follows: Step 201, performing pre-processing operations of data cleaning, data integration and data standardization on the collected historical data and real-time data; Step 202, based on the pre-processed historical data, analyze the climate characteristics of the marine ranch area, extract features for temperature, humidity, salinity, wind speed, wind direction, ocean current and rainfall, and obtain climate influencing factors. At the same time, analyze the ecological characteristics of the marine ranch area, obtain ecological influencing factors for biodiversity, water quality, bottom sediment, dissolved oxygen and chlorophyll, and verify the site selection influencing factors by comparing real-time data with historical data, among which temperature: analyze the water temperature change trend and seasonal pattern of the marine ranch area, humidity: evaluate the impact of humidity on marine life, salinity: measure salinity changes to understand its impact on marine habitats, wind speed and wind direction: analyze the impact of wind speed and wind direction on ocean current and surface water temperature, ocean current: study the impact of ocean current on nutrient transport and biological distribution, rainfall: evaluate the impact of rainfall on surface salinity and nutrient input, biodiversity: evaluate the richness and uniformity of species in the area, water quality: monitor water quality parameters such as dissolved oxygen and nutrient concentration, Sediment: Analyze the impact of sediment type on benthic organisms, Dissolved oxygen: Measure dissolved oxygen levels to understand its impact on the survival of marine organisms, Chlorophyll: Assess primary productivity and phytoplankton biomass; Step 203, obtaining associated data of climate influencing factors according to the pre-processed historical data, which are seasonal change data and extreme climate data; Step 204, by analyzing the correlation between ecological impact factors and climate impact factors, the associated data of ecological impact factors are obtained, namely biodiversity data and water pollution data, and then the potential impact of ecological impact factors on the growth and reproduction of marine organisms is evaluated, real-time monitoring and data analysis are performed, and the changes and trends of climate impact factors and ecological impact factors in real time are observed.

[0009] A further improvement of the technical solution of the present invention is that in step 3, the process of obtaining the climate impact index and the ecological impact index is as follows: Step 301, integrating the associated data of climate impact factors and ecological impact factors in the pre-processed historical data and real-time data to form a unified data set, and determining evaluation indicators of climate impact and ecological impact, wherein climate indicators may include temperature suitability, humidity suitability, salinity suitability, wind speed and direction suitability, ocean current suitability, etc., and ecological indicators may include biodiversity index, water cleanliness, bottom stability, dissolved oxygen content, chlorophyll content, etc.; Step 302, based on the evaluation indicators, a comprehensive suitability evaluation system is established, which should cover both climate and ecology, and consider the interaction and impact between the indicators, and build a comprehensive suitability evaluation model based on the screened climate impact factor data and ecological impact factor data; Step 303, dividing the integrated data set into a training set, a validation set and a test set, using the training set data to train the comprehensive suitability assessment model, and using the validation set and the test set data to validate the trained comprehensive suitability assessment model, and evaluating the performance of the comprehensive suitability assessment model; Step 304, based on the comprehensive suitability assessment model and the comprehensive suitability assessment system, combined with the pre-processed historical data and real-time data, obtain the climate impact index and ecological impact index of the site selection target, analyze and predict the climate change trend in the future, such as temperature increase, rainfall change, etc., and analyze the scope and trend of ecological changes, such as increase or decrease in biodiversity, improvement or deterioration of water quality, etc., and combine the fragility and resilience of the ecosystem to evaluate the possible scope of ecological changes in the future. Among them, the climate impact index reflects the suitability of the climate conditions in the site selection target area for the construction of marine ranches, and the ecological impact index reflects the suitability of the ecological environment in the site selection target area for the construction of marine ranches.

[0010] A further improvement of the technical solution of the present invention is that the calculation formula of the climate impact index is: ; ; Among them, CI is the climate impact index, T is the temperature, is the standard value of temperature, is the standard deviation of temperature, H is the humidity, is the standard value of humidity, is the humidity standard deviation, S is the salinity, is the standard value of salinity, is the salinity standard deviation, is the wind speed and direction influence function, W is the wind speed, is the minimum suitable wind speed, is the standard deviation of wind speed, V is the wind direction, The optimal wind direction is is the standard deviation of wind direction, C is the intensity of ocean current, is the standard value of ocean current strength, is the standard deviation of ocean current intensity, R is the rainfall, is the average rainfall, is the standard deviation of rainfall; ; ; Among them, EI is the ecological impact index, B is the biodiversity index, is the minimum value of biodiversity. is the maximum value of biodiversity, Q is the water quality index, is the water quality standard value, is the water quality standard deviation, is the ecological monitoring function, D is the bottom suitability index, is the optimal value of the substrate suitability index, is the standard deviation of the bottom suitability index, O is the dissolved oxygen content, is the standard value of dissolved oxygen, is the standard deviation of dissolved oxygen, L is the chlorophyll concentration, is the average chlorophyll concentration, is the standard deviation of chlorophyll.

[0011] A further improvement of the technical solution of the present invention is that in step 4, the process of obtaining the suitability evaluation coefficient is: Step 401, traverse the pre-processed historical data, combine the climate impact index and the ecological impact index, analyze the correlation between the impact of site suitability and the climate impact index and the ecological impact index, and display the correlation through visualization tools such as scatter plots and box plots to further understand the interaction between the indexes; Step 402, assigning different weights to the climate impact index and the ecological impact index, performing weighted calculation on the two indexes, and obtaining a suitability assessment coefficient; Step 403, sorting the site selection target areas according to the size of the suitability evaluation coefficient, generating a suitability evaluation sequence table, and analyzing the impact of climate and ecological factors on site selection suitability; Step 404, according to the pre-processed historical data matching suitability evaluation sequence table, the site selection target area is divided into a first-level regional level, a second-level regional level, a third-level regional level and a fourth-level regional level, wherein the site selection suitability influence of the regional level increases from the first-level regional level to the fourth-level regional level, and the first-level regional level is the most suitable site selection target, and the fourth-level regional level is the least suitable site selection target; Step 405, matching the determined different regional levels with the results of the suitability assessment coefficient, setting corresponding regional assessment thresholds for different regional levels, and determining the regional level to which they belong. Step 406, based on the regional grade and suitability evaluation coefficient, comprehensively analyze the impact of climate and ecological factors on site suitability to determine the best site selection plan.

[0012] A further improvement of the technical solution of the present invention is that the calculation formula of the suitability evaluation coefficient is: ; Among them, SA is the suitability assessment coefficient, CI is the climate impact index, is the standard value of the climate impact index, is the standard deviation of the climate impact index, is a very small positive number used to avoid the denominator being zero. is the weight of the climate impact index, EI is the ecological impact index, is the optimal value of the ecological impact index, is the standard deviation of the ecological impact index, is the weight of the ecological impact index, and .

[0013] A further improvement of the technical solution of the present invention is that: the multiple area levels correspond to the multiple area evaluation thresholds, wherein the area evaluation thresholds include an upper threshold and a lower threshold; The plurality of area levels and the plurality of area evaluation thresholds satisfy the following relationship: First-level regional level ; Secondary regional level ; Three-level regional level ; Four-level regional level ; Among them, SA is the suitability assessment coefficient, X1 is the upper threshold corresponding to the first-level regional level, X2 is the lower threshold corresponding to the second-level regional level and the upper threshold corresponding to the third-level regional level, and X3 is the lower threshold corresponding to the third-level regional level and the upper threshold corresponding to the fourth-level regional level.

[0014] A further improvement of the technical solution of the present invention is that in step 5, the process of evaluating and ranking the site selection schemes is as follows: Step 501, inputting the pre-processed real-time data into the comprehensive suitability assessment model to calculate the climate impact index and the ecological impact index respectively; Step 502, combining the calculated climate impact index and ecological impact index to obtain a suitability assessment coefficient, and determining the regional grade of the current site selection target area according to a preset regional assessment threshold; Step 503, based on the suitability evaluation coefficient and the regional level of the current site selection target area, preliminarily determine multiple site selection plans, and evaluate each site selection plan, including feasibility analysis, cost-benefit analysis, environmental impact assessment, etc., wherein the feasibility analysis should consider factors such as technical feasibility, construction difficulty, and resource acquisition, the cost-benefit analysis should calculate the construction cost, operating cost, and expected benefits, and the environmental impact assessment should evaluate the impact of the construction on the marine ecosystem, water quality, biodiversity, etc., and output an evaluation report. According to the evaluation report, the site selection plans are ranked and the optimal site selection plan is determined. The evaluation report should include the advantages and disadvantages, potential risks, and recommended measures of the site selection plan. The advantages and disadvantages should be analyzed based on the evaluation results and detailed planning, the potential risks should include technical risks, market risks, environmental risks, etc., and the recommended measures should propose specific solutions to the potential risks; Step 504, formulate a specific marine ranch construction plan based on the assessment report and site selection plan. The plan should include ranch layout design, aquaculture species selection, feed supply plan, disease prevention and control measures, environmental monitoring plan, etc. At the same time, formulate a detailed timetable and budget plan to ensure the smooth progress of the construction project. The timetable should include time nodes for the preliminary preparation, construction, commissioning and formal operation stages, and the budget plan should include the budget for construction cost, operation cost and expected income.

[0015] Due to the adoption of the above technical solution, the present invention has the following technical advances compared with the prior art: The present invention provides a method for evaluating the site selection of marine ranches based on machine learning. The machine learning technology can process and analyze massive amounts of real-time data, including information on climate, ecology, topography, socio-economy, etc., making the site selection evaluation more accurate and reducing errors in human judgment. At the same time, by analyzing historical data and real-time data, it can accurately identify key climate and ecological factors that affect the site selection of marine ranches, shorten the evaluation cycle, and provide decision makers with timely and reliable site selection recommendations.

[0016] The present invention provides a method for evaluating the site selection of marine ranches based on machine learning. The machine learning technology enables the machine learning model to have the ability to predict and evaluate different site selection plans by training the machine learning model. It can not only collect and process various types of data, but also analyze and process the monitoring system to achieve multi-faceted prediction, evaluation and guidance. It can also collect and process various types of data, analyze and process the monitoring system to achieve multi-faceted prediction, evaluation and guidance, thereby improving the scientificity and objectivity of decision-making.

[0017] The present invention provides a method for assessing the site selection of marine ranches based on machine learning. By analyzing a large amount of historical data and real-time data and utilizing a comprehensive suitability assessment model, the method predicts future trends of climate change and ecological change, and assesses the impact of potential climate change on the site selection of marine ranches, thereby formulating response strategies in advance. The method can also predict and assess the risks and benefits of different site selection options, thereby helping to ensure the long-term stable operation of marine ranches and reduce losses caused by environmental changes. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0019] Figure 1 is a flow chart of the method of the present invention; Figure 2 The flowchart of obtaining the climate impact factor and the ecological impact factor of the present invention; Figure 3 A flowchart for obtaining the climate impact index and the ecological impact index of the present invention; Figure 4 The figure is a flow chart for obtaining the suitability evaluation coefficient of the present invention. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0021] Embodiment 1, as Figure 1-Figure 4 As shown, the present invention provides a method for evaluating the site selection of a marine ranch based on machine learning, comprising the following steps: Step 1: Determine the site selection objectives of the marine ranch, such as fishery resource proliferation, ecological protection, and economic benefits. Collect regional data on the site selection objectives and divide the regional data into historical data and real-time data. Historical data includes marine environmental data, climate data, and ecological data from the past years. Real-time data includes current or recent marine environmental data and meteorological data. Collect data through satellite remote sensing, ocean observation stations, meteorological stations, scientific research institutions and other channels. Regional data include marine environmental data, such as water temperature, salinity, ocean currents, and seabed topography, and marine ecological data, such as biodiversity and distribution of key species. Collect fishery resource data, such as historical catches and fish migration routes. Collect climate change data, such as the frequency of extreme weather events, Sea level rise, etc., collect human activity data, such as fishing operation intensity, shipping trajectory, pollution emissions, etc., collect economic benefit and social impact data, such as potential economic benefits, community participation, etc.; the process of collecting data on the target site selection area is: clarify the construction purpose, and divide the marine ranch into three types according to its construction purpose: conservation type, proliferation type and leisure type. These three types of marine ranches are comprehensive marine ranches with different degrees of resource conservation, biological proliferation and leisure fishery functions. Among them, the conservation type is mainly aimed at ecological restoration and protection of marine biological resources. The key indicators are water quality, bottom type, biodiversity, ecologically sensitive areas, pollution level, etc. The proliferation type is mainly aimed at increasing the amount of specific marine biological resources and increasing fishery production. The key indicators are abundance of bait resources, fish breeding habits, water flow conditions, sea productivity, etc. The leisure type is mainly for the purpose of providing leisure fisheries, sightseeing tourism, etc. The key indicators are transportation convenience, landscape value, tourist carrying capacity, sea safety, etc. The key indicators of site selection are determined according to the purpose, including environmental conditions, economic benefits, ecological impact, etc. Drones, AI cameras, radars and optoelectronic equipment, and water quality sensors are deployed in the target site selection area to monitor the ecological environment, marine biological activities and water quality conditions in the target site selection area in real time, and historical data are obtained from existing public databases and marine environmental monitoring stations, including basic geographic information, such as coastline, water depth, topography, marine functional zoning and other basic geographic information, sea Ocean environment data, such as water temperature, salinity, ocean currents, seabed topography, etc., marine ecological data, such as biodiversity, distribution of key species, etc., fishery resource data, such as historical catch, fish migration paths, etc., climate change data, such as frequency of extreme weather events, sea level rise, etc., human activity data, such as fishing operation intensity, shipping trajectories, pollution emissions, etc., economic benefits and social impact data, such as potential economic benefits, community participation, etc. The collected historical data and real-time data will be transmitted to the data management center through wireless communication, and will be sorted, classified and stored, and a marine ranch regional ecological database will be established. The data will be deeply mined and analyzed using big data, cloud computing and other technologies to provide a basis for scientific management and decision-making of marine ranches; Step 2: Preprocess the collected historical data and real-time data, including data cleaning and normalization, and screen the factors affecting the site selection, including climate factors and ecological factors. For example, statistical methods such as variance selection method, correlation coefficient method, and mutual information method are used to evaluate the importance of the factors affecting the site selection, and select the factors affecting the site selection that have a strong correlation with the target variable. Climate factors include temperature, precipitation, wind speed, wind direction, etc., and ecological factors include biodiversity, marine pollution status, and seabed topography. The acquisition process of the factors is as follows: pre-processing operations such as data cleaning, data integration and data standardization are performed on the collected historical data and real-time data. Based on the pre-processed historical data, the climate characteristics of the marine ranch area are analyzed, and the characteristics of temperature, humidity, salinity, wind speed, wind direction, ocean current and rainfall are extracted to obtain climate influencing factors. At the same time, the ecological characteristics of the marine ranch area are analyzed, and the ecological influencing factors are obtained for biodiversity, water quality, bottom quality, dissolved oxygen and chlorophyll. The site selection influencing factors are verified by comparing the real-time data with the historical data. Among them, temperature: Water temperature trends and seasonal patterns in the field area, humidity: evaluate the impact of humidity on marine life, salinity: measure salinity changes to understand its impact on marine habitats, wind speed and direction: analyze the impact of wind speed and direction on ocean currents and surface water temperature, ocean currents: study the impact of ocean currents on nutrient transport and biological distribution, rainfall: and then evaluate the impact of rainfall on surface salinity and nutrient input, biodiversity: evaluate the richness and uniformity of species in the area, water quality: monitor water quality parameters such as dissolved oxygen and nutrient concentration, bottom sediment: analyze the impact of bottom sediment type on benthic organisms, dissolved oxygen: measure dissolved Deoxygenation level, understand its impact on the survival of marine life, chlorophyll: evaluate primary productivity and phytoplankton biomass, obtain the associated data of climate influencing factors based on pre-processed historical data, which are seasonal change data and extreme climate data, analyze the correlation between ecological influencing factors and climate influencing factors, obtain the associated data of ecological influencing factors, which are biodiversity data and water pollution data, and then evaluate the potential impact of ecological influencing factors on the growth and reproduction of marine life, real-time monitoring and data analysis, observe the changes and trends of climate influencing factors and ecological influencing factors in real time; Step 3: Based on historical data and in combination with the specific requirements for marine ranch site selection, a comprehensive suitability assessment model is constructed to obtain the climate impact index and ecological impact index to quantitatively evaluate the site selection objectives, including predicting the trend of climate change and assessing the scope of ecological change. If the goal is not only to maximize fishery output, but also to protect the environment, maximize economic benefits, and enhance social acceptance, the climate impact index reflects the climate suitability of the site selection area, including predicting the trend of climate change; the ecological impact index reflects the ecological suitability of the site selection area, including assessing the scope of ecological change; the acquisition process of the climate impact index and the ecological impact index is as follows: the associated data of the climate impact factors and ecological impact factors in the pre-processed historical data and real-time data are integrated to form a unified data set, and the evaluation indicators of climate impact and ecological impact are determined. The climate indicators include temperature suitability, humidity suitability, salinity suitability, wind speed and direction suitability, and ocean current suitability; the ecological indicators include biodiversity index, water cleanliness, bottom stability, dissolved oxygen content, and chlorophyll content; and based on the evaluation indicators, a comprehensive suitability assessment system is established, which should cover Covering both climate and ecology, and considering the interactions and impacts between various indicators, a comprehensive suitability assessment model is constructed based on the selected climate impact factor data and ecological impact factor data. The integrated data set is divided into a training set, a validation set, and a test set. The training set data is used to train the comprehensive suitability assessment model, and the validation set and test set data are used to verify the trained comprehensive suitability assessment model. The performance of the comprehensive suitability assessment model is evaluated. According to the comprehensive suitability assessment model and the comprehensive suitability assessment system, combined with the pre-processed historical data and real-time data, the climate impact index and ecological impact index of the site selection target are obtained, and the climate change trend in the future, such as temperature increase, rainfall change, etc., is analyzed and predicted. At the same time, the scope and trend of ecological changes, such as increase or decrease in biodiversity, improvement or deterioration of water quality, etc., are analyzed. Combined with the vulnerability and resilience of the ecosystem, the possible scope of ecological changes in the future is evaluated. Among them, the climate impact index reflects the suitability of the climate conditions in the site selection target area for the construction of marine ranches, and the ecological impact index reflects the suitability of the ecological environment in the site selection target area for the construction of marine ranches. Step 4: Based on the climate impact index and ecological impact index and combined with the comprehensive suitability assessment model, obtain the suitability assessment coefficient, determine the site selection plan, and analyze the impact of climate impact factors and ecological impact factors on site suitability; the process of obtaining the suitability assessment coefficient is as follows: traverse the preprocessed historical data, combine the climate impact index and the ecological impact index, analyze the correlation between the impact of site suitability and the climate impact index and the ecological impact index, display the correlation through visualization tools such as scatter plots and box plots, further understand the interaction between the indexes, assign different weights to the climate impact index and the ecological impact index, perform weighted calculation on the two indexes, obtain the suitability assessment coefficient, sort the site selection target areas according to the size of the suitability assessment coefficient, generate a suitability assessment sequence table, and analyze the suitability assessment coefficient. The influence of climate and ecological factors on site suitability is matched with the suitability evaluation sequence table based on the pre-processed historical data, and the site target area is divided into regional levels, namely, the first-level regional level, the second-level regional level, the third-level regional level and the fourth-level regional level. Among them, the site suitability influence of the regional level increases from the first-level regional level to the fourth-level regional level, and the first-level regional level is the most suitable site target, and the fourth-level regional level is the least suitable site target. The determined different regional levels are matched with the results of the suitability evaluation coefficient, and the corresponding regional evaluation thresholds are set for different regional levels to determine the regional level to which they belong. According to the regional level and the suitability evaluation coefficient, the influence of climate and ecological factors on site suitability is comprehensively analyzed to determine the best site selection plan; Step 5, input the preprocessed real-time data into the comprehensive suitability assessment model to output the suitability assessment coefficient, determine the site selection plan, and evaluate and rank the site selection plans at the same time, considering the feasibility, cost, environmental impact and other factors of the plan, output an evaluation report, including the advantages and disadvantages of the site selection, potential risks and recommended measures, etc., formulate a specific marine ranch construction plan, and after the marine ranch is built, continuously monitor the marine environmental data, and adjust the management strategy in time to ensure the sustainable development of the marine ranch; the process of evaluating and ranking the site selection plans is as follows: input the preprocessed real-time data into the comprehensive suitability assessment model, calculate the climate impact index and the ecological impact index respectively, combine the calculated climate impact index and ecological impact index to obtain the suitability assessment coefficient, and determine the regional level of the current site selection target area according to the preset regional assessment threshold, based on the suitability assessment coefficient and the regional level of the current site selection target area, preliminarily determine multiple site selection plans, and evaluate each site selection plan, including feasibility analysis, cost-benefit analysis, environmental impact assessment, etc., among which the feasibility analysis should consider Technical feasibility, construction difficulty, resource acquisition and other factors. The cost-benefit analysis should calculate the construction cost, operating cost and expected benefits. The environmental impact assessment should evaluate the impact of the construction on the marine ecosystem, water quality, biodiversity, etc., and output an assessment report. According to the assessment report, the site selection options are ranked according to their advantages and disadvantages to determine the optimal site selection option. The assessment report should include the advantages and disadvantages of the site selection option, potential risks, and recommended measures. The advantages and disadvantages should be analyzed based on the assessment results and detailed planning. Potential risks should include technical risks, market risks, environmental risks, etc. Recommended measures should propose specific solutions to potential risks. According to the assessment report and site selection plan, a specific marine ranch construction plan should be formulated. The plan should include ranch layout design, aquaculture variety selection, feed supply plan, disease prevention and control measures, environmental monitoring plan, etc. At the same time, a detailed timetable and budget plan should be formulated to ensure the smooth progress of the construction project. The timetable should include time nodes for the preliminary preparation, construction, commissioning and formal operation stages, and the budget plan should include the budget for construction costs, operating costs and expected benefits.

[0022] Embodiment 2, as Figure 1-Figure 4 As shown, based on Example 1, the present invention provides a technical solution: Preferably, the calculation formula of the climate impact index is: ; ; Where CI is the climate impact index, T is the temperature, is the standard value of temperature, is the standard deviation of temperature, H is the humidity, is the standard value of humidity, is the standard deviation of humidity, S is the salinity, is the standard value of salinity, is the salinity standard deviation, is the wind speed and direction influence function, W is the wind speed, is the minimum suitable wind speed, is the standard deviation of wind speed, V is the wind direction, The optimal wind direction is is the standard deviation of wind direction, C is the intensity of ocean current, is the standard value of ocean current strength, is the standard deviation of ocean current intensity, R is the rainfall, is the average rainfall, is the standard deviation of rainfall; ; ; Among them, EI is the ecological impact index, B is the biodiversity index, is the minimum value of biodiversity. is the maximum value of biodiversity, Q is the water quality index, is the water quality standard value, is the water quality standard deviation, is the ecological monitoring function, D is the bottom suitability index, is the optimal value of the substrate suitability index, is the standard deviation of the bottom suitability index, O is the dissolved oxygen content, is the standard value of dissolved oxygen, is the standard deviation of dissolved oxygen, L is the chlorophyll concentration, is the average chlorophyll concentration, is the standard deviation of chlorophyll; The calculation formula of suitability assessment coefficient is: ; Among them, SA is the suitability assessment coefficient, CI is the climate impact index, is the standard value of the climate impact index, is the standard deviation of the climate impact index, is a very small positive number used to avoid the denominator being zero. is the weight of the climate impact index, EI is the ecological impact index, is the optimal value of the ecological impact index, is the standard deviation of the ecological impact index, is the weight of the ecological impact index, and ; Multiple regional levels correspond to multiple regional assessment thresholds, wherein the regional assessment threshold includes an upper threshold and a lower threshold; Multiple regional levels and multiple regional assessment thresholds satisfy the following relationship: First-level regional level ; Secondary regional level ; Three-level regional level ; Four-level regional level ; Among them, SA is the suitability assessment coefficient, X1 is the upper threshold corresponding to the first-level regional level, X2 is the lower threshold corresponding to the second-level regional level and the upper threshold corresponding to the third-level regional level, and X3 is the lower threshold corresponding to the third-level regional level and the upper threshold corresponding to the fourth-level regional level.

[0023] The above are only specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for evaluating the site selection of marine ranches based on machine learning, characterized in that: The following steps are involved: Step 1: Determine the site selection target of the marine ranch, collect regional data of the site selection target, and divide the regional data into historical data and real-time data; Step 2: pre-process the collected historical data and real-time data to screen the factors affecting the site selection, where the factors affecting the site selection include climate factors and ecological factors; Step 3: Based on historical data and the specific requirements for marine ranching site selection, a comprehensive suitability assessment model is constructed to obtain climate impact index and ecological impact index to quantitatively assess the site selection target, including predicting the trend of climate change and assessing the scope of ecological change; Step 4: Based on the climate impact index and ecological impact index and combined with the comprehensive suitability assessment model, obtain the suitability assessment coefficient, determine the site selection plan, and analyze the impact of climate impact factors and ecological impact factors on site suitability; Step 5: Input the preprocessed real-time data into the comprehensive suitability assessment model to output the suitability assessment coefficient, determine the site selection plan, evaluate and rank the site selection plans, output the assessment report, and formulate a specific marine ranch construction plan.

2. The method for evaluating the site selection of marine ranches based on machine learning according to claim 1, characterized in that: In step 1, the process of collecting data on the site selection target area is as follows: Step 101, clarify the purpose of construction, and divide the marine ranches into three types according to their construction purpose: conservation type, proliferation type and leisure type; Step 102, deploy drones, AI cameras, radars and optoelectronic equipment, and water quality sensors in the target site selection area to monitor the ecological environment, marine biological activities, and water quality in the target site selection area in real time, and obtain historical data from existing public databases and marine environmental monitoring stations, including basic geographic information, marine environmental data, marine ecological data, fishery resource data, climate change data, human activity data, and economic benefits and social impact data; Step 103, the collected historical data and real-time data are transmitted to the data management center via wireless communication, and are sorted, classified and stored to establish a marine ranch regional ecological database.

3. The method for evaluating the site selection of marine ranches based on machine learning according to claim 2, characterized in that: In step 2, the process of obtaining the factors affecting site selection, including climate factors and ecological factors, is as follows: Step 201, performing pre-processing operations of data cleaning, data integration and data standardization on the collected historical data and real-time data; Step 202, based on the pre-processed historical data, analyze the climate characteristics of the marine ranch area, extract features for temperature, humidity, salinity, wind speed, wind direction, ocean current and rainfall, and obtain climate influencing factors. At the same time, analyze the ecological characteristics of the marine ranch area, extract ecological influencing factors for biodiversity, water quality, bottom quality, dissolved oxygen and chlorophyll, and compare the real-time data with the historical data to verify the site selection influencing factors. Step 203, obtaining associated data of climate influencing factors according to the pre-processed historical data, which are seasonal change data and extreme climate data; Step 204, by analyzing the correlation between ecological impact factors and climate impact factors, the associated data of ecological impact factors are obtained, which are biodiversity data and water pollution data, respectively, and then the potential impact of ecological impact factors on the growth and reproduction of marine organisms is evaluated.

4. The method for site selection and evaluation of marine ranches based on machine learning according to claim 3 is characterized in that: In step 3, the process of obtaining the climate impact index and the ecological impact index is as follows: Step 301, integrating the associated data of climate impact factors and ecological impact factors in the pre-processed historical data and real-time data to form a unified data set, and determining evaluation indicators for climate impact and ecological impact; Step 302, and based on the evaluation indicators, establish a comprehensive suitability evaluation system, and construct a comprehensive suitability evaluation model based on the screened climate impact factor data and ecological impact factor data; Step 303, dividing the integrated data set into a training set, a validation set and a test set, using the training set data to train the comprehensive suitability assessment model, and using the validation set and the test set data to validate the trained comprehensive suitability assessment model, and evaluating the performance of the comprehensive suitability assessment model; Step 304, based on the comprehensive suitability assessment model and the comprehensive suitability assessment system, combined with the pre-processed historical data and real-time data, obtain the climate impact index and ecological impact index of the site selection target, analyze the climate change trend, and analyze the scope and trend of ecological changes.

5. The method for evaluating the site selection of marine ranches based on machine learning according to claim 4, characterized in that: The calculation formula of the climate impact index is: ； ； Where CI is the climate impact index, T is the temperature, is the standard value of temperature, is the standard deviation of temperature, H is the humidity, is the standard value of humidity, is the standard deviation of humidity, S is the salinity, is the standard value of salinity, is the salinity standard deviation, is the wind speed and direction influence function, W is the wind speed, is the minimum suitable wind speed, is the standard deviation of wind speed, V is the wind direction, The optimal wind direction is is the standard deviation of wind direction, C is the intensity of ocean current, is the standard value of ocean current strength, is the standard deviation of ocean current intensity, R is the rainfall, is the average rainfall, is the standard deviation of rainfall; The calculation formula of the ecological impact index is: ； ； Among them, EI is the ecological impact index, B is the biodiversity index, is the minimum value of biodiversity. is the maximum value of biodiversity, Q is the water quality index, is the water quality standard value, is the water quality standard deviation, is the ecological monitoring function, D is the bottom suitability index, is the optimal value of the substrate suitability index, is the standard deviation of the bottom suitability index, O is the dissolved oxygen content, is the standard value of dissolved oxygen, is the standard deviation of dissolved oxygen, L is the chlorophyll concentration, is the average chlorophyll concentration, is the standard deviation of chlorophyll.

6. The method for evaluating the site selection of marine ranches based on machine learning according to claim 5, characterized in that: In step 4, the process of obtaining the suitability assessment coefficient is as follows: Step 401, traversing the pre-processed historical data, combining the climate impact index and the ecological impact index, and analyzing the correlation between the impact of site suitability and the climate impact index and the ecological impact index; Step 402, assigning different weights to the climate impact index and the ecological impact index, performing weighted calculation on the two indexes, and obtaining a suitability assessment coefficient; Step 403, sorting the site selection target areas according to the size of the suitability evaluation coefficient, generating a suitability evaluation sequence table, and analyzing the impact of climate and ecological factors on site selection suitability; Step 404, according to the pre-processed historical data matching suitability evaluation sequence table, the site selection target area is divided into a first-level regional level, a second-level regional level, a third-level regional level and a fourth-level regional level, wherein the site selection suitability influence of the regional level increases from the first-level regional level to the fourth-level regional level; Step 405, matching the determined different regional levels with the results of the suitability assessment coefficient, setting corresponding regional assessment thresholds for different regional levels, and determining the regional level to which they belong; Step 406, based on the regional grade and suitability evaluation coefficient, comprehensively analyze the impact of climate and ecological factors on site suitability to determine the best site selection plan.

7. The method for evaluating the site selection of marine ranches based on machine learning according to claim 6, characterized in that: The calculation formula of the suitability assessment coefficient is: ； Among them, SA is the suitability assessment coefficient, CI is the climate impact index, is the standard value of the climate impact index, is the standard deviation of the climate impact index, is a positive number, is the weight of the climate impact index, EI is the ecological impact index, is the optimal value of the ecological impact index, is the standard deviation of the ecological impact index.

8. The method for evaluating the site selection of marine ranches based on machine learning according to claim 7, characterized in that: The plurality of area levels correspond to the plurality of area assessment thresholds, wherein the area assessment thresholds include an upper threshold and a lower threshold; The plurality of area levels and the plurality of area evaluation thresholds satisfy the following relationship: First-level regional level ; Secondary regional level ; Three-level regional level ; Four-level regional level ; Among them, SA is the suitability assessment coefficient, X1 is the upper threshold corresponding to the first-level regional level, X2 is the lower threshold corresponding to the second-level regional level and the upper threshold corresponding to the third-level regional level, and X3 is the lower threshold corresponding to the third-level regional level and the upper threshold corresponding to the fourth-level regional level.

9. The method for evaluating the site selection of marine ranches based on machine learning according to claim 8, characterized in that: In step 5, the process of evaluating and ranking the site selection options is as follows: Step 501, inputting the pre-processed real-time data into the comprehensive suitability assessment model to calculate the climate impact index and the ecological impact index respectively; Step 502, combining the calculated climate impact index and ecological impact index to obtain a suitability assessment coefficient, and determining the regional grade of the current site selection target area according to a preset regional assessment threshold; Step 503, based on the suitability evaluation coefficient and the regional level of the current site selection target area, multiple site selection schemes are preliminarily determined, each site selection scheme is evaluated, and an evaluation report is output. According to the evaluation report, the site selection schemes are ranked according to their merits and demerits to determine the optimal site selection scheme; Step 504, formulate a specific marine ranch construction plan based on the assessment report and site selection plan, and formulate a detailed timetable and budget plan.

Images