Marine organism species diversity monitoring system

Abstract

The invention relates to the technical field of ecological monitoring, in particular to a marine organism species diversity monitoring system, which comprises a species selection and classification module used for automatically selecting and classifying species in the ocean, automatically selecting and classifying representative species by using an ADLF (adaptive deep learning framework) technology, and automatically selecting and classifying the representative species; then the data collecting and processing module collects historical survey and latest monitoring data, stores the data in a basic database, then performs data statistical analysis and model simulation, constructs an evaluation index system reflecting the functional condition of the estuary ecosystem according to the biological characteristics for standardization processing, and then performs optimization by comparing with the model library module; and finally, the monitoring area module is used for displaying an evaluation result and spatial distribution. The system not only improves the ability of China in the fields of marine biological resource utilization and ecological protection, but also enhances the adaptability and coping ability to marine environment change.

Description

Technical Field

[0001] The present invention relates to the technical field of ecological monitoring and evaluation, and in particular to the technical field of marine ecology and environmental science. Background Art

[0002] Currently, environmental DNA (eDNA) technology, remote sensing and GIS integration technology combined with remote sensing image data and geographic information system analysis tools, and niche models using species ecology data are available on the market. Although these systems have achieved certain success in their respective fields, these technologies each have their limitations. Environmental DNA technology is costly and easily interfered by environmental factors, remote sensing and GIS technology are complex in data parsing and analysis, and the accuracy of niche models is subject to data quality. Therefore, in practical applications, it is usually necessary to combine multiple technologies and methods to overcome their respective shortcomings and improve the accuracy and efficiency of monitoring and evaluation.

[0003] In summary, although the existing technical solutions have made certain progress in some aspects, they still cannot meet the needs of marine species diversity monitoring systems. Since these technologies in the existing systems may face challenges such as high cost, strong data dependence, high environmental sensitivity, great operational complexity and subjective interpretation of results during implementation, in view of this, a marine species diversity monitoring system is provided. Summary of the invention

[0004] The purpose of the present invention is to provide a marine species diversity monitoring system to solve the problems that the existing systems and technologies mentioned in the above background technology may face during implementation, such as high cost, strong data dependence, high environmental sensitivity, great operational complexity, and subjective challenges in result interpretation.

[0005] To achieve the above objectives, the present invention provides a marine biological species diversity monitoring system, including: a species selection and classification module, which is used for automatic selection and classification of species in the ocean, and uses an adaptive deep learning framework ADLF technology to automatically select and classify representative species, thereby improving the efficiency and accuracy of species identification.

[0006] The specific formula of ADLF technology is: the state space is the model parameter, the action space is the direction and size of parameter update, and the reward function Defined as: .

[0007] Update formula for Q-learning: .

[0008] in, Indicates in status Take action of Value, that is, from the state Through Action estimates of expected future rewards that may be obtained; Indicates the current state, representing the current condition or position of a system or environment; The current action taken, that is, in the current state The operations performed below, For control The step size of the value update is between 0 and 1; Reward function, expressed in state Take action The instant rewards you receive, Used to adjust the current value of future rewards, the value is also between 0 and 1; Indicates that in the next state All possible actions The maximum Q value, Take Action After transferring to the new state, In the new state Any possible action.

[0009] The data collection and processing module is used to collect historical surveys and the latest monitoring data, and adopts the marine biological image intelligent recognition system and the data fusion prediction and analysis system to realize automatic and efficient data processing and species identification, and stores these data in the basic database. The basic database is used to store species data, environmental parameters, socio-economic data and literature information, and then the data is subjected to data statistical analysis and model simulation through the data analysis module.

[0010] The formula involved in the marine life image intelligent recognition system is: .

[0011] in, Represents an image, Represents the feature representation of the positive sample of this image, Represents the similarity between two features. is a temperature parameter, is the total number of images in the batch.

[0012] The specific steps of the marine organism image intelligent recognition system are as follows: S1.1. Select a feature extractor, SEResNext as the basic model, to extract useful feature representations from the images of organisms.

[0013] S1.2. Use contrastive learning to train the feature extractor. This process involves selecting a set of positive and negative samples, comparing their feature representations, and optimizing the model parameters so that the model can better distinguish different biological images.

[0014] S1.3. Determine the specific category of the image to be identified by comparing the similarity of the features of the image to be identified with those of the images in the retrieval library.

[0015] S1.4. Provide an interface that allows users to check and verify recognition results and expand the search library as needed to continuously improve and enhance recognition performance.

[0016] Among them, the formula involved in the data fusion prediction and analysis system mentioned is: ;in is the predicted value, is the number of decision trees, It is The prediction results of the tree for the input variables; the species data mentioned in the basic database module include the classification, distribution, quantity and ecological habits of various organisms. These data are the basis for identifying ecosystem changes and assessing the health of biodiversity; environmental parameters include water temperature, salinity, dissolved oxygen content, pH value, which are crucial to understanding the physical and chemical conditions of marine ecosystems; socio-economic data include fishery catch, sea area utilization, and pollution emission data; literature information is the existing knowledge and latest progress in marine biodiversity research.

[0017] Among them, the formula involved in data statistical analysis is: Shannon-Wiener diversity index: ;in, It is The relative abundance of species, is the total number of species, Express The result of natural logarithmic transformation of the proportion; the formula involved in the model simulation statistical analysis is: Logistic growth model: ;in, is the population size, is the intrinsic growth rate, is the environmental carrying capacity, It's time. Indicates population size Over time The rate of change of is the growth rate of the population.

[0018] The indicator system construction module is used to construct an evaluation indicator system reflecting the functional status of the estuarine ecosystem based on the characteristics of functional and structural indicator organisms, and select the SEI index ecological evaluation method for standardization.

[0019] The model library module includes a model building module of environmental changes, structural indicators and functional indicators.

[0020] The monitoring area module is used to optimize the monitoring area data and display the evaluation results and spatial distribution.

[0021] As a further improvement of the present technical solution, the species selection and classification module 1 is based on data preprocessing and enhancement, an adaptive deep learning framework ADLF, and a method for species identification and classification, wherein the data preprocessing and enhancement module is used to perform preliminary processing on the collected marine biological images, including cropping, scaling to a uniform size, and standardizing brightness and contrast.

[0022] Among them, the adaptive deep learning framework ADLF module is used for specific marine life identification tasks. It uses the pre-trained deep learning models ResNet and VGG as the basis, fine-tunes the last few layers of the network, defines a reward mechanism to encourage the model to identify samples with higher difficulty, and uses a reinforcement learning algorithm to fine-tune the model parameters to achieve self-optimization and adaptation in the ever-changing marine environment.

[0023] Among them, the species recognition and classification module is used to classify marine organisms into corresponding categories according to the recognition results, and use the ADLF model to perform high-precision recognition of marine organism images to determine which species they belong to. This step not only includes simple species recognition, but also involves the selection of representative high-trophic-level economic species and the classification of low-trophic-level organisms for subsequent ecosystem analysis.

[0024] As a further improvement of the present technical solution, the data collection and processing module 2 includes data collection and preprocessing, biological image recognition and data fusion, analysis storage and feedback.

[0025] The data collection and preprocessing are as follows: automatically collecting marine biological images, environmental parameters and socio-economic data, and performing preliminary processing such as cleaning, formatting and missing value processing on the collected data;

[0026] Among them, the marine life image intelligent recognition system framework processes the collected biological images, automatically identifies and classifies marine organisms through machine learning algorithms, and inputs the recognition results and other types of data into the intelligent data fusion and predictive analysis framework IDFPAF. According to the analysis results, the data collection and processing strategies are adjusted to optimize the image recognition and data fusion algorithms.

[0027] Among them, analysis, storage and feedback are: using the IDFPAF framework to conduct in-depth analysis of the fused data, including statistical analysis and trend prediction, and storing the analysis results and original data in the basic database.

[0028] As a further improvement of the present technical solution, the data in the basic database 3 include species, environmental chemical elements, correlation analysis between biology and chemistry, social and economic development, and literature information.

[0029] Among them, species data include the types, numbers, distribution, and ecological habits of marine organisms, assess biodiversity and ecosystem health, identify key species, and protect priority areas.

[0030] Among them, environmental chemical elements include water temperature, salinity, dissolved oxygen, pH value, nutrient concentration and the content of harmful substances. These parameters are monitored to assess environmental quality and identify pollution sources.

[0031] Among them, the correlation analysis data between biology and chemistry analyzes the relationship between biological data and environmental chemical elements, reveals the impact of specific environmental factors on biodiversity, and understands the functions of ecosystems and the effects of organisms on environmental changes.

[0032] Among them, social and economic development data include fishery catches, marine tourism income, pollution emissions, and information on sea area utilization, reflecting the impact of human activities on the marine ecosystem.

[0033] Among them, literature information includes scientific research articles, reports, and policy documents, which provide the latest research results on marine biodiversity, ecosystem management and protection. Designers and implementers of monitoring systems can understand current scientific trends and policy backgrounds to guide monitoring activities and protection decisions.

[0034] As a further improvement of the technical solution, the data analysis module is implemented by data statistical analysis and model simulation technology, wherein the formula involved in the data statistical analysis technology is: Shannon-Wiener diversity index: ;in, It is The relative abundance of species, is the total number of species, Express The result of natural logarithm transformation of the proportions.

[0035] Among them, the formula involved in the model simulation statistical technical analysis is: Logistic growth model: ;in, is the population size, is the intrinsic growth rate, is the environmental carrying capacity, It's time. Indicates population size Over time The rate of change of is the growth rate of the population.

[0036] As a further improvement of the technical solution, the indicator system construction module is used to construct an evaluation indicator system reflecting the functional status of the estuary ecosystem, and the SEI index of the ecological evaluation method is selected for standardization.

[0037] As a further improvement of the present technical solution, the model library module is used to analyze and process basic data. By analyzing the basic data, the model library reveals the changing trends of marine biological populations and their ecological environment. By analyzing the correlation between environmental parameters and biological data, it helps to identify the impact of different environmental factors on marine biodiversity. By considering indicators of species richness, uniformity, and interspecies relationships, it evaluates and monitors the health status and biodiversity level of marine ecosystems.

[0038] As a further improvement of this technical solution, the monitoring area module is used to verify, supplement and improve the estuary ecological monitoring and evaluation system established in the above work, and to further optimize the monitoring indicators, monitoring technology, station layout, monitoring frequency, evaluation indicators and evaluation standards.

[0039] The optimization steps are as follows: S8.1. Evaluate the effectiveness of existing monitoring indicators, determine key indicators that can accurately reflect the health status of estuarine ecosystems, introduce new monitoring indicators, and improve the comprehensiveness and accuracy of monitoring based on the latest scientific research results and technological developments.

[0040] Among them, the effectiveness of existing monitoring indicators is usually evaluated by analyzing the accuracy, sensitivity and representativeness of these indicators in reflecting ecological changes, predicting environmental trends and guiding conservation decisions.

[0041] Among them, the new monitoring indicators include biodiversity index based on the latest scientific research findings, ecosystem service function assessment, ecological pressure response model, and changes in environmental parameters monitored using advanced technology.

[0042] S8.2. Use modern technology such as remote sensing, automatic monitoring stations and drones to improve monitoring efficiency and data accuracy, and apply big data and machine learning techniques to improve the efficiency and accuracy of data analysis.

[0043] S8.3. Scientifically plan the layout of monitoring sites to ensure that monitoring points cover key and representative ecological areas, and flexibly adjust the monitoring frequency according to ecological events and environmental changes, especially during key ecological periods and in areas sensitive to changes.

[0044] S8.4. Summarize and standardize monitoring experience, form a set of standardized and systematic monitoring technical specifications, conduct standardized training and promotion, improve the professional level of monitoring personnel, and ensure the standardization and commercialization of monitoring activities.

[0045] As a further improvement of the technical solution, the model library module includes: a target layer, a criterion layer, a factor layer, an indicator layer, and assignment of related indicators.

[0046] Among them, the target layer: the degree of environmental change, the purpose is to quantify the impact of environmental change on marine biodiversity, the target layer expression is: ;in Represents the degree of environmental change; Represents structural indicators; Represents functional indicators.

[0047] Criteria layer: Construction of structural indicators and functional indicators, structural indicators Including species richness, evenness, functional indicators Covering primary productivity and nutrient cycle, the expression of the criterion layer is ; ;in are the values ​​of various indicators of structural indicators; are the values ​​of various indicators of functional indicators; for Weight, for Weight; represents the total number of structural indicators; Represents the total number of functional indicators.

[0048] Factor layer: The purpose is to identify the species that are most sensitive or representative to environmental changes. The factor layer function is: ;in, represents the factor layer function; Representative species dataset; is the screening threshold or criterion.

[0049] Indicator layer: Specific monitoring indicators for selected species, expressed as: ;in, is the monitoring indicator value; Datasets for selected species; is the environmental parameter; express The functional relationship between them.

[0050] Among them, the mathematical expression of the data analysis process is: ;in, The weight of the representative structural indicators in the environmental change assessment, The weight of the representative functional indicators in the environmental change assessment, is the weighted sum of structural indicators, is the weighted sum of the functional indicators, The total number of structural indicators, Representative The weight of each structural indicator, Representative The value of the structural indicator, The total number of functional indicators, Representative The weight of each functional indicator, Representative The value of a functional indicator.

[0051] As a further improvement of the technical solution, the formula involved in the data fusion prediction and analysis system is: ;in is the predicted value, is the number of decision trees, It is The prediction results of the tree for the input variables.

[0052] Among them, the formula involved in data statistical analysis is: Shannon-Wiener diversity index: ;in, It is The relative abundance of species, is the total number of species, Express The result of natural logarithm transformation of the proportions.

[0053] Among them, the formula involved in the model simulation statistical analysis is: Logistic growth model: ;in, is the population size, is the intrinsic growth rate, is the environmental carrying capacity, It's time. Indicates population size Over time The rate of change of is the growth rate of the population.

[0054] Compared with the prior art, the invention has the beneficial effects.

[0055] 1. By implementing the marine biological species diversity monitoring system, the needs for systematic collection, analysis and evaluation of typical marine habitats and their biological resources in my country have been met, and the purpose of technical guarantee and data support has been achieved. The system not only effectively captures the changes in species diversity and accurately reflects the health status of the marine ecosystem, but also successfully discovers marine organisms with potential medicinal and industrial value, providing an important basis for drug development, especially the research and development of anti-tumor drugs.

[0056] 2. By implementing a marine biological species diversity monitoring system, the challenge of accurately grasping the evolutionary trends, characteristics and potential problems of the marine ecosystem has been solved. By analyzing the role of marine ecological monitoring in ecological early warning, the project provides a basis for predicting and warning of the future direction of the marine ecological environment, and provides important support for curbing the deterioration of the marine ecological environment and ensuring the sustainable use of natural resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 It is a flowchart of the overall process of the present invention.

[0058] The meanings of the numbers in the figure are: 1. Species selection and classification module; 2. Data collection and processing module; 3. Basic database; 4. Data analysis module; 5. Indicator system construction module; 6. Model library module; 7. Monitoring area module. DETAILED DESCRIPTION

[0059] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only 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.

[0060] Example:

[0061] See also Figure 1 As shown, a marine biological species diversity monitoring system is provided, including a species selection and classification module 1, wherein the species selection and classification module 1 is used for automatic selection and classification of species in the ocean, and uses an adaptive deep learning framework ADLF technology to automatically select and classify representative species, thereby improving the efficiency and accuracy of species identification.

[0062] The specific formula of ADLF technology is: the state space is the model parameter, the action space is the direction and size of parameter update, and the reward function Defined as: .

[0063] Q-learning update formula ;in, Indicates in status Take action of Value, that is, from the state Through Action estimates of expected future rewards that may be obtained; Indicates the current state, representing the current condition or position of a system or environment; The current action taken, that is, in the current state The operations performed below, For control The value update step size is between 0 and 1.

[0064] in, Reward function, expressed in state Take action The instant rewards you receive, Used to adjust the current value of future rewards, also between 0 and 1.

[0065] in, Indicates that in the next state All possible actions The maximum Q value, Take Action After transferring to the new state, In the new state Any possible action.

[0066] The data preprocessing and enhancement module is used to perform preliminary processing on the collected marine life images, including cropping, scaling to a uniform size, and standardizing brightness and contrast.

[0067] Among them, the adaptive deep learning framework ADLF module is used for specific marine life identification tasks.

[0068] Using the pre-trained deep learning models ResNet and VGG as the basis, the last few layers of the network are fine-tuned, a reward mechanism is defined to encourage the model to recognize samples with higher difficulty, and a reinforcement learning algorithm is used to fine-tune the model parameters to achieve self-optimization and adaptation in the ever-changing ocean environment.

[0069] Among them, the species recognition and classification module is used to classify marine organisms into corresponding categories according to the recognition results, and use the ADLF model to perform high-precision recognition of marine organism images to determine which species they belong to. This step not only includes simple species recognition, but also involves the selection of representative high-trophic-level economic species and the classification of low-trophic-level organisms for subsequent ecosystem analysis.

[0070] The marine biological species diversity monitoring system also includes a data collection and processing module 2, which is used to collect historical surveys and the latest monitoring data, and uses a marine biological image intelligent recognition system and a data fusion prediction and analysis system to achieve automatic and efficient data processing and species identification, and stores these data in a basic database 3. The basic database 3 is used to store species data, environmental parameters, socio-economic data and literature information, and then the data is subjected to data statistical analysis and model simulation through a data analysis module 4.

[0071] The data collection and preprocessing are as follows: automatically collecting marine biological images, environmental parameters and socio-economic data, and performing preliminary processing such as cleaning, formatting and missing value processing on the collected data.

[0072] Among them, the marine life image intelligent recognition system framework processes the collected biological images, automatically identifies and classifies marine organisms through machine learning algorithms, and inputs the recognition results and other types of data into the intelligent data fusion and predictive analysis framework IDFPAF. According to the analysis results, the data collection and processing strategies are adjusted to optimize the image recognition and data fusion algorithms.

[0073] Among them, analysis, storage and feedback are: using the IDFPAF framework to conduct in-depth analysis of the fused data, including statistical analysis and trend prediction, and storing the analysis results and original data in the basic database.

[0074] The marine biological image intelligent recognition system specifically refers to: a framework for biological image recognition design, which is mainly used to improve the efficiency and accuracy of automatic recognition of biological images. Its core purpose is to flexibly solve the in-situ image recognition problem of organisms through image similarity comparison. It is suitable for long-term biological monitoring and research in complex marine environments.

[0075] The formula involved in the marine biological image intelligent recognition system is: ;in, Represents an image, Represents the feature representation of the positive sample of this image, Represents the similarity between two features, usually calculated using dot product or cosine similarity. is a temperature parameter that controls the scaling of the similarity scores, is the total number of images in the batch.

[0076] The specific steps of the marine organism image intelligent recognition system are as follows: S1.1. Select a powerful feature extractor, SEResNext, as the basic model to extract useful feature representations from the images of plankton.

[0077] S1.2. Train the feature extractor using a supervised contrastive learning method. This process involves selecting a set of positive and negative samples, comparing their feature representations, and optimizing the model parameters so that the model can better distinguish different plankton images.

[0078] S1.3, by comparing the similarity of the features of the image to be identified with those of the images in the retrieval library, determine the specific category of the image to be identified. This step may involve an efficient retrieval algorithm to quickly find the most similar image or image set.

[0079] S1.4. Provide an interface that allows users to check and verify recognition results and expand the search library as needed to continuously improve and enhance recognition performance.

[0080] Among them, the formula involved in the data fusion prediction and analysis system mentioned is: ;in is the predicted value, is the number of decision trees, It is The prediction results of the tree for the input variables.

[0081] The marine biological species diversity monitoring system also includes a basic database 3, which stores the data collected by the data collection and processing module 2 in the basic database 3 and is used to store species data, environmental parameters, socio-economic data and literature information.

[0082] Among them, the species data mentioned in the basic database include the classification, distribution, quantity and ecological habits of various organisms. These data are the basis for identifying ecosystem changes and assessing the health of biodiversity; environmental parameters include water temperature, salinity, dissolved oxygen content, pH value, which are crucial to understanding the physical and chemical conditions of marine ecosystems; socio-economic data include fishery catch, sea area utilization, and pollution emission data; literature information is the existing knowledge and latest progress in marine biodiversity research.

[0083] Among them, species data include the types, numbers, distribution, and ecological habits of marine organisms, assess biodiversity and ecosystem health, identify key species, and protect priority areas.

[0084] Among them, environmental chemical elements include water temperature, salinity, dissolved oxygen, pH value, nutrient concentration and the content of harmful substances. These parameters are monitored to assess environmental quality and identify pollution sources.

[0085] Among them, the correlation analysis data between biology and chemistry analyzes the relationship between biological data and environmental chemical elements, reveals the impact of specific environmental factors on biodiversity, and understands the functions of ecosystems and the effects of organisms on environmental changes.

[0086] Among them, social and economic development data include fishery catches, marine tourism income, pollution emissions, and information on sea area utilization, reflecting the impact of human activities on the marine ecosystem.

[0087] Among them, literature information includes scientific research articles, reports, and policy documents, which provide the latest research results on marine biodiversity, ecosystem management and protection. Designers and implementers of monitoring systems can understand current scientific trends and policy backgrounds to guide monitoring activities and protection decisions.

[0088] The marine biological species diversity monitoring system further comprises a data analysis module 4 , which performs statistical analysis and model simulation on the data in the basic database 3 .

[0089] Among them, the formula involved in data statistical analysis technology is: Shannon-Wiener diversity index: ;in, It is The relative abundance of species, is the total number of species, Express The result of natural logarithm transformation of the proportions.

[0090] Among them, the formula involved in the model simulation statistical technical analysis is: Logistic growth model: ;in, is the population size, is the intrinsic growth rate, is the environmental carrying capacity, It's time. Indicates population size Over time The rate of change of is the growth rate of the population.

[0091] The marine biological species diversity monitoring system also includes an indicator system construction module 5, which is used to construct an evaluation indicator system reflecting the functional status of the estuary ecosystem based on the characteristics of functional and structural indicator organisms, and selects the SEI index ecological evaluation method for standardization.

[0092] Among them, the characteristics of functional indicator organisms: functional indicator organisms mainly reflect the ecological processes, functions and services, material circulation, energy flow, and ecosystem productivity in the ecosystem; the characteristics of structural indicator organisms: structural indicator organisms reflect the structural characteristics of the ecosystem, species diversity, community structure, and the relationships between species.

[0093] The marine biological species diversity monitoring system also includes a model library module 6, which is used to analyze and process basic data. By analyzing the basic data, the model library reveals the changing trends of marine biological populations and their ecological environment. By analyzing the correlation between environmental parameters and biological data, it helps to identify the impact of different environmental factors on marine biodiversity. By considering species richness, uniformity, and species relationship indicators, it evaluates and monitors the health status and biodiversity level of marine ecosystems, grasps the changes of specific species in time and space scales, reflects the trend of environmental changes, and provides scientific support for biodiversity resource management and protection. These models include: target layer, criterion layer, factor layer, indicator layer, and assignment of related indicators.

[0094] Among them, the target layer: the degree of environmental change, aims to quantify the impact of environmental change on marine biodiversity, and its expression is: ;in Represents the degree of environmental change. Represents structural indicators, Represents functional indicators.

[0095] Criteria layer: Construction of structural indicators and functional indicators, structural indicators Including species richness, evenness, functional indicators Covering primary productivity and nutrient cycle, the expression of the criterion layer is: ; ;in is the index value of the structural index and are the values ​​of various functional indicators, for Weight and for Weight.

[0096] Factor layer: The purpose is to identify the species that are most sensitive or representative to environmental changes. The factor layer function is: ;in, Representative species dataset, is the screening threshold or criterion.

[0097] Indicator layer: Specific monitoring indicators for selected species, expressed as: ;in, To monitor the index value, For the dataset of selected species, is the environmental parameter.

[0098] Among them, the mathematical expression of the data analysis process is: ;in, and They are the weights of structural indicators and functional indicators in environmental change assessment.

[0099] The marine biological species diversity monitoring system also includes a monitoring area module 7, which is used to optimize the monitoring area data and display the evaluation results and spatial distribution.

[0100] The optimization steps are:

[0101] S2.1. Evaluate the effectiveness of existing monitoring indicators, identify key indicators that can accurately reflect the health status of estuarine ecosystems, introduce new monitoring indicators, and improve the comprehensiveness and accuracy of monitoring based on the latest scientific research results and technological developments.

[0102] Among them, the effectiveness of existing monitoring indicators is usually evaluated by analyzing the accuracy, sensitivity and representativeness of these indicators in reflecting ecological changes, predicting environmental trends and guiding conservation decisions.

[0103] Among them, the new monitoring indicators include biodiversity index based on the latest scientific research findings, ecosystem service function assessment, ecological pressure response model, and changes in environmental parameters monitored using advanced technology.

[0104] S2.2. Use modern technology such as remote sensing, automatic monitoring stations and drones to improve monitoring efficiency and data accuracy, and apply big data and machine learning technology to improve the efficiency and accuracy of data analysis.

[0105] S2.3. Scientifically plan the layout of monitoring sites to ensure that monitoring points cover key and representative ecological areas, and flexibly adjust the monitoring frequency according to ecological events and environmental changes, especially during key ecological periods and in areas sensitive to changes.

[0106] S2.4. Summarize and standardize monitoring experience, form a set of standardized and systematic monitoring technical specifications, carry out standardized training and promotion, improve the professional level of monitoring personnel, and ensure the standardization and commercialization of monitoring activities.

[0107] The above describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only preferred examples of the present invention and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. Marine biological species diversity monitoring system, characterized in that: include: A species selection and classification module (1), wherein the species selection and classification module (1) is used for automatic selection and classification of species in the ocean, and uses an adaptive deep learning framework ADLF technology to automatically select and classify representative species, thereby improving the efficiency and accuracy of species identification; A data collection and processing module (2), wherein the data collection and processing module (2) is used to collect historical survey and latest monitoring data, and uses a marine biological image intelligent recognition system and a data fusion prediction analysis system to realize automatic and efficient data processing and species identification, and stores these data in a basic database (3). The basic database (3) is used to store species data, environmental parameters, socio-economic data and literature information, and then performs data statistical analysis and model simulation on the data through a data analysis module (4); The formula involved in the marine life image intelligent recognition system is: ； in, Represents an image, Represents the feature representation of the positive sample of this image, Represents the similarity between two features. is a temperature parameter, is the total number of images in the batch; The specific steps of the marine life image intelligent recognition system are: S1.

1. Select feature extractor, SEResNext as the base model to extract useful feature representations from biological images; S1.

2. Train the feature extractor using contrastive learning. This process involves selecting a set of positive and negative samples, comparing their feature representations, and optimizing the model parameters so that the model can better distinguish different biological images. S1.3, by comparing the similarity of the features of the image to be identified with those of the images in the retrieval library, determine the specific category of the image to be identified; S1.

4. Provide an interface that allows users to check and verify recognition results and expand the search library as needed to continuously improve and enhance recognition performance; An indicator system construction module (5), wherein the indicator system construction module (5) is used to construct an evaluation indicator system reflecting the functional status of the estuarine ecosystem based on the characteristics of functional and structural indicator organisms, and selects the SEI index ecological evaluation method for standardization; A model library module (6), wherein the model library module (6) includes a model for building environmental changes, structural indicators and functional indicators; A monitoring area module (7), wherein the monitoring area module (7) is used to optimize monitoring area data and display evaluation results and spatial distribution.

2. The marine biological species diversity monitoring system according to claim 1, characterized in that: The species selection and classification module (1) is based on data preprocessing and enhancement, an adaptive deep learning framework ADLF, and species identification and classification methods.

3. The marine biological species diversity monitoring system according to claim 1, characterized in that: The data collection and processing module (2) includes data collection and preprocessing, biological image recognition and data fusion, analysis storage and feedback; Among them, the formula involved in the marine life image intelligent recognition system is: ； in, Represents an image, Represents the feature representation of the positive sample of this image, Represents the similarity between two features. is a temperature parameter, is the total number of images in the batch.

4. The marine biological species diversity monitoring system according to claim 1, characterized in that: The data in the basic database (3) include species, environmental chemical elements, biological and chemical correlation analysis, social and economic development, and literature information.

5. The marine biological species diversity monitoring system according to claim 1, characterized in that: The data analysis module (4) is implemented through data statistical analysis and model simulation technology. Among them, the formula involved in data statistical analysis technology is: Shannon-Wiener Diversity Index: ; in, It is The relative abundance of species, is the total number of species, Express The result of natural logarithm transformation of proportions; Among them, the formula involved in the model simulation statistical technical analysis is: Logistic Growth Model: ; in, is the population size, is the intrinsic growth rate, is the environmental carrying capacity, It's time. Indicates population size Over time The rate of change.

6. The marine biological species diversity monitoring system according to claim 1, characterized in that: The indicator system construction module (5) is used to construct an evaluation indicator system that reflects the functional status of the estuary ecosystem, and selects the SEI index, an ecological evaluation method, for standardization.

7. The marine species diversity monitoring system according to claim 1, characterized in that: The model library module (6) is used to analyze and process basic data to help identify the impact of different environmental factors on marine biodiversity.

8. The marine biological species diversity monitoring system according to claim 1, characterized in that: The monitoring area module (7) is used to verify, supplement and improve the estuary ecological monitoring and evaluation system established in the above work, and further optimize the monitoring indicators, monitoring technology, station layout, monitoring frequency, evaluation indicators and evaluation standards; The optimization steps are: S8.

1. Evaluate the effectiveness of existing monitoring indicators, identify key indicators that can accurately reflect the health of estuarine ecosystems, introduce new monitoring indicators, and improve the comprehensiveness and accuracy of monitoring based on the latest scientific research results and technological developments; Among them, the effectiveness of existing monitoring indicators is evaluated by analyzing the accuracy, sensitivity and representativeness of these indicators in reflecting ecological changes, predicting environmental trends and guiding conservation decisions; The new monitoring indicators include biodiversity indices based on the latest scientific research findings, ecosystem service function assessments, ecological pressure response models, and changes in environmental parameters monitored using advanced technologies; S8.

2. Use modern technology such as remote sensing, automatic monitoring stations and drones to improve monitoring efficiency and data accuracy, and apply big data and machine learning techniques to improve the efficiency and accuracy of data analysis; S8.

3. Scientifically plan the layout of stations to ensure that monitoring points cover key and representative ecological areas, and flexibly adjust the monitoring frequency according to ecological events and environmental changes, in key ecological periods and areas sensitive to changes; S8.

4. Summarize and standardize monitoring experience, form a set of standardized and systematic monitoring technical specifications, conduct standardized training and promotion, improve the professional level of monitoring personnel, and ensure the standardization and commercialization of monitoring activities.

9. The marine species diversity monitoring system according to claim 7, characterized in that: The model library module (6) includes: a target layer, a criterion layer, a factor layer, an indicator layer, and the assignment of related indicators; in, Target layer: the degree of environmental change, the purpose is to quantify the impact of environmental change on marine biodiversity. The target layer expression is: ； in Represents the degree of environmental change; represents structural indicators; Represents functional indicators; Criteria layer: Construction of structural indicators and functional indicators, structural indicators Including species richness, evenness, functional indicators Covering primary productivity and nutrient cycle, the expression of the criterion layer is: ； ； in are the values ​​of various indicators of structural indicators; are the values ​​of various indicators of functional indicators; for Weight, for Weight; represents the total number of structural indicators; Indicates the total number of functional indicators; Factor layer: The purpose is to identify the species that are most sensitive or representative to environmental changes. The factor layer function is: ； in, represents the factor layer function; Representative species dataset; is the screening threshold or criteria; Indicator layer: Specific monitoring indicators for selected species, expressed as: ； in, is the monitoring indicator value; Datasets for selected species; is the environmental parameter; express The functional relationship between Among them, the mathematical expression of the data analysis process is: ； in, The weight of the representative structural indicators in the environmental change assessment, The weight of the representative functional indicators in the environmental change assessment, is the weighted sum of structural indicators, is the weighted sum of the functional indicators, The total number of structural indicators, Representative The weight of each structural indicator, Representative The value of the structural indicator, The total number of functional indicators, Representative The weight of each functional indicator, Representative The value of a functional indicator.

10. The marine biological species diversity monitoring system according to claim 1, characterized in that: in, The formula involved in the data fusion prediction analysis system mentioned is: ； in is the predicted value, is the number of decision trees, It is The prediction results of the tree for the input variables; the formula involved in the data statistical analysis is: Shannon-Wiener Diversity Index: ; in, It is The relative abundance of species, is the total number of species, Express The result of natural logarithm transformation of proportions; Among them, the formula involved in the model simulation statistical analysis is: Logistic Growth Model: ; in, is the population size, is the intrinsic growth rate, is the environmental carrying capacity, It's time. Indicates population size Over time The rate of change.

Images