Fishway efficiency evaluation and diagnosis method and system based on AI ecological monitoring and medium

Abstract

The invention provides a fishway efficiency evaluation and diagnosis method and system based on AI ecological monitoring and a medium, and the method comprises the steps: firstly, collecting images, three-dimensional point clouds, water quality parameters and water flow data inside and outside a fishway in real time through the deployment of an underwater camera, a sonar, a water quality sensor and hydrological monitoring equipment, and forming multi-source heterogeneous data; secondly, through data cleaning and preprocessing and a preset AI identification technology, automatically counting the passing quantity and types of fishes; then, multi-dimensional evaluation is conducted on the fishway based on the structure, water quality, hydrology and ecology, and then the fishway health index is analyzed based on expert weight distribution; and finally, performing evaluation and problem diagnosis on the fishway energy efficiency according to the fishway health index, and generating a maintenance suggestion. According to the method, a monitoring-evaluation-diagnosis-optimization closed-loop system is constructed, dynamic management and quick response of fishway efficiency are achieved, and scientific support is provided for hydraulic engineering and ecological sustainable development.

Description

Technical Field

[0001] The present invention relates to the field of fishways, and more specifically, to a fishway performance evaluation and diagnosis method, system, and medium based on AI ecological monitoring. Background Art

[0002] Fishways, as crucial ecological compensation facilities in water conservancy projects, have the core function of helping migratory fish cross artificial barriers like dams and weirs, thereby maintaining river ecological continuity. However, traditional fishway performance assessment methods often rely on manual inspections or single-sensor testing, making it difficult to comprehensively capture multi-dimensional data on fish passage, fishway structure, water quality, and hydrology, leading to one-sided assessment results. Furthermore, the causes of fishway failure are complex and diverse, potentially involving complex factors such as structural design flaws, deteriorating water quality, unsuitable flow rates, or changes in ecological populations. Traditional methods lack the ability to collaboratively analyze these multiple factors, often leading to misjudgments or delayed responses.

[0003] In recent years, with the development of AI and Internet of Things technologies, some studies have attempted to introduce cameras or sonar equipment to monitor fish migration. However, there are problems with large data volumes, i.e., prominent data noise. In addition, due to the complexity of the data, it is difficult to unify a quantitative evaluation system and generate actionable diagnostic conclusions.

[0004] Therefore, there is an urgent need for a dynamic evaluation and diagnosis technology for fishway efficiency that can improve the energy efficiency of fishways and provide efficient ecological protection. Summary of the Invention

[0005] In view of the above problems, the purpose of the present invention is to provide a fishway efficiency assessment and diagnosis method, system, and medium based on AI ecological monitoring. The system collects images, three-dimensional point clouds, water quality parameters, and water flow data from inside and outside the fishway in real time. Through data cleaning and preprocessing, as well as preset AI recognition technology, it automatically counts the number and species of fish passing through. The system also conducts a multi-dimensional assessment of the fishway and analyzes the fishway health index based on expert weight distribution. The system also evaluates the energy efficiency of the fishway and diagnoses problems based on the fishway health index, and generates maintenance recommendations. The present invention establishes a closed-loop system of monitoring-assessment-diagnosis-optimization to achieve dynamic management and rapid response of fishway efficiency, providing scientific support for water conservancy projects and ecological sustainable development.

[0006] A first aspect of the present invention provides a fishway effectiveness evaluation and diagnosis method based on AI ecological monitoring, the method comprising:

[0007] Acquiring first multi-source heterogeneous information, including image information, three-dimensional point cloud information, water quality information, and water flow information;

[0008] Based on a preset data cleaning and preprocessing mechanism, obtaining second multi-source heterogeneous information according to the first multi-source heterogeneous information;

[0009] Generating corresponding traffic rate information according to fish types based on the second multi-source heterogeneous information;

[0010] generating multidimensional assessment information based on the second multi-source heterogeneous information and the traffic rate information, including structural assessment, water quality assessment, hydrological assessment, and ecological assessment;

[0011] Obtaining a fishway health index based on the multi-dimensional evaluation information based on a preset weighting algorithm;

[0012] If the fishway health index is lower than a preset health threshold, outputting diagnostic information based on a preset diagnostic model and the multidimensional evaluation information, including key factor analysis and problem probability distribution;

[0013] Based on the preset knowledge base and the diagnostic information, a diagnostic conclusion and maintenance suggestion are generated.

[0014] In this solution, the preset data cleaning and preprocessing mechanism specifically includes:

[0015] According to the image information, invalid image frames are eliminated based on the inter-frame difference method, and then valid image frames are extracted through the frame extraction strategy;

[0016] generating a three-dimensional model of the fishway based on the three-dimensional point cloud information, and then identifying structural abnormal areas using a clustering algorithm, including areas where the slope deviation is greater than a preset slope reference value and / or the fishway width is less than a preset width reference value;

[0017] According to the water quality information and the water flow information, according to the preset time series alignment and outlier removal, the noise data is smoothed using the preset sliding window method;

[0018] If there is missing data in the first multi-source heterogeneous information, interpolation is performed based on a preset linear interpolation or historical mean.

[0019] In this solution, the corresponding traffic rate information is generated according to the fish type based on the second multi-source heterogeneous information, specifically:

[0020] Based on a preset AI fish recognition model, target detection is performed on the fish in the image information;

[0021] Count the number and species of fish that pass through within a preset unit time and generate a fish passage log;

[0022] Analyzing the distribution density and movement trajectory of fish in the fishway based on the three-dimensional point cloud information and the fish passage log;

[0023] determining a fish passage bottleneck area based on the distribution density and the movement trajectory;

[0024] generating a water flow-traffic volume correlation model based on the water flow information and the fish passage log;

[0025] Based on the fish passage log, the passage rate is counted by fish category and a passage time series report is generated, which at least includes the daily passage curve and the fish species percentage.

[0026] In this solution, the multi-dimensional evaluation information is generated according to the second multi-source heterogeneous information and the traffic rate information, specifically:

[0027] The three-dimensional fishway model generated from the three-dimensional point cloud information is compared with a preset standard fishway model to obtain a structural assessment score based on preset structural dimensions, wherein the structural dimensions include at least slope, turning radius, bottleneck width, and siltation volume;

[0028] Real-time monitoring of water quality information, calculation of fish species survival indicators based on fish types, and obtaining a water quality assessment score, wherein the water quality information includes at least dissolved oxygen, water temperature, pH value, and turbidity data;

[0029] According to the water flow-volume correlation model, critical flow velocity values and abnormal water level periods are identified to obtain a hydrological assessment score;

[0030] Target detection is performed using the fish species, and the fish population diversity index and the rate of change in the number of target fish species are calculated to obtain an ecological assessment score.

[0031] In this solution, the fishway health index is obtained based on the multi-dimensional evaluation information based on the preset weighted algorithm, specifically:

[0032] Based on the preset assessment model, multi-dimensional assessment weights are obtained, including structural weight, water quality weight, hydrological weight and ecological weight;

[0033] The fishway health index is obtained by weighted summation based on the multi-dimensional evaluation information and the multi-dimensional evaluation weights.

[0034] In this solution, the preset diagnostic model outputs diagnostic information based on the multi-dimensional evaluation information, specifically:

[0035] Constructing a diagnostic model training dataset based on the fishway case database and its historical multi-dimensional evaluation data, and corresponding root cause labels, for training the diagnostic model;

[0036] Input the multi-dimensional evaluation information of the current fishway into the diagnostic model to obtain the contribution probability of each feature to the failure of the fishway;

[0037] Generate a problem probability distribution based on the contribution probabilities, and sort the key factors from high to low in terms of probability.

[0038] A second aspect of the present invention provides a fishway effectiveness evaluation and diagnosis system based on AI ecological monitoring, including a fishway effectiveness evaluation and diagnosis method program based on AI ecological monitoring. When the fishway effectiveness evaluation and diagnosis method program based on AI ecological monitoring is executed by the processor, the following steps are implemented:

[0039] Acquiring first multi-source heterogeneous information, including image information, three-dimensional point cloud information, water quality information, and water flow information;

[0040] Based on a preset data cleaning and preprocessing mechanism, obtaining second multi-source heterogeneous information according to the first multi-source heterogeneous information;

[0041] Generating corresponding traffic rate information according to fish types based on the second multi-source heterogeneous information;

[0042] generating multidimensional assessment information based on the second multi-source heterogeneous information and the traffic rate information, including structural assessment, water quality assessment, hydrological assessment, and ecological assessment;

[0043] Obtaining a fishway health index based on the multi-dimensional evaluation information based on a preset weighting algorithm;

[0044] If the fishway health index is lower than a preset health threshold, outputting diagnostic information based on a preset diagnostic model and the multidimensional evaluation information, including key factor analysis and problem probability distribution;

[0045] Based on the preset knowledge base and the diagnostic information, a diagnostic conclusion and maintenance suggestion are generated.

[0046] In this solution, the preset data cleaning and preprocessing mechanism specifically includes:

[0047] According to the image information, invalid image frames are eliminated based on the inter-frame difference method, and then valid image frames are extracted through the frame extraction strategy;

[0048] generating a three-dimensional model of the fishway based on the three-dimensional point cloud information, and then identifying structural abnormal areas using a clustering algorithm, including areas where the slope deviation is greater than a preset slope reference value and / or the fishway width is less than a preset width reference value;

[0049] According to the water quality information and the water flow information, according to the preset time series alignment and outlier removal, the noise data is smoothed using the preset sliding window method;

[0050] If there is missing data in the first multi-source heterogeneous information, interpolation is performed based on a preset linear interpolation or historical mean.

[0051] In this solution, the corresponding traffic rate information is generated according to the fish type based on the second multi-source heterogeneous information, specifically:

[0052] Based on a preset AI fish recognition model, target detection is performed on the fish in the image information;

[0053] Count the number and species of fish that pass through within a preset unit time and generate a fish passage log;

[0054] Analyzing the distribution density and movement trajectory of fish in the fishway based on the three-dimensional point cloud information and the fish passage log;

[0055] determining a fish passage bottleneck area based on the distribution density and the movement trajectory;

[0056] generating a water flow-traffic volume correlation model based on the water flow information and the fish passage log;

[0057] Based on the fish passage log, the passage rate is counted by fish category and a passage time series report is generated, which at least includes the daily passage curve and the fish species percentage.

[0058] The third aspect of the present invention provides a computer-readable storage medium, which includes a fishway efficiency evaluation and diagnosis method program based on AI ecological monitoring. When the fishway efficiency evaluation and diagnosis method program based on AI ecological monitoring is executed by a processor, the steps of the fishway efficiency evaluation and diagnosis method based on AI ecological monitoring as described in any one of the above items are implemented.

[0059] The present invention provides a fishway efficiency assessment and diagnosis method, system and medium based on AI ecological monitoring. First, by deploying underwater cameras, sonar, water quality sensors and hydrological monitoring equipment, images, three-dimensional point clouds, water quality parameters and water flow data inside and outside the fishway are collected in real time to form multi-source heterogeneous data; secondly, through data cleaning and preprocessing, as well as preset AI recognition technology, the number and species of fish passing through are automatically counted; then, the fishway is evaluated in multiple dimensions based on structure, water quality, hydrology and ecology, and the fishway health index is analyzed based on expert weight distribution; finally, the fishway energy efficiency is evaluated and problems are diagnosed based on the fishway health index, and maintenance recommendations are generated; the present invention constructs a closed-loop system of monitoring-assessment-diagnosis-optimization to achieve dynamic management and rapid response of fishway efficiency, providing scientific support for water conservancy projects and ecological sustainable development. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. It should be understood that the following drawings only illustrate certain embodiments of the present invention and therefore should not be regarded as limiting the scope.

[0061] Figure 1A flow chart showing a fishway effectiveness evaluation and diagnosis method based on AI ecological monitoring according to the present invention is shown;

[0062] Figure 2 The following is a flowchart showing an operation of a data cleaning and preprocessing mechanism provided by an embodiment of the present invention;

[0063] Figure 3 A flow chart showing a calculation of traffic rate information provided by an embodiment of the present invention is shown;

[0064] Figure 4 A block diagram of a fishway performance evaluation and diagnosis system based on AI ecological monitoring according to the present invention is shown. DETAILED DESCRIPTION

[0065] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 making creative efforts are within the scope of protection of the present invention.

[0066] Unless otherwise defined, all terms (including technical and scientific terms) used in the embodiments of the present invention have the same meaning as commonly understood by those skilled in the art to which the present invention belongs. It should also be understood that terms such as those defined in common dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology, and should not be interpreted in an idealized or extremely formal sense, unless explicitly defined in this manner in the embodiments of the present invention.

[0067] The terms "first," "second," and similar words used in the embodiments of the present invention do not indicate any order, quantity, or importance, but are only used to distinguish different components. Terms such as "a," "an," or "the" do not indicate a limit on quantity, but rather indicate the presence of at least one. Similarly, terms such as "include," "comprise," and "comprising" mean that the elements or objects preceding the term include the elements or objects listed after the term and their equivalents, without excluding other elements or objects.

[0068] "Connected" or "connected" and similar terms are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The preceding or subsequent steps of the methods of the embodiments of the present invention do not necessarily need to be performed in exact order. Instead, various steps may be performed in reverse order or simultaneously. Furthermore, other operations may be added to these processes, or one or more steps may be removed from these processes.

[0069] In addition, the functional modules in the various embodiments of the present invention may be integrated together to form an independent part, or each module may exist independently, or two or more modules may be integrated to form an independent part.

[0070] Figure 1 A flow chart of a fishway performance evaluation and diagnosis method based on AI ecological monitoring according to the present invention is shown.

[0071] like Figure 1 As shown, the first aspect of the present invention discloses a fishway effectiveness evaluation and diagnosis method based on AI ecological monitoring, the method comprising:

[0072] S102, acquiring first multi-source heterogeneous information, including image information, three-dimensional point cloud information, water quality information, and water flow information;

[0073] S104, obtaining second multi-source heterogeneous information based on the first multi-source heterogeneous information based on a preset data cleaning and preprocessing mechanism;

[0074] S106, generating corresponding traffic rate information according to fish types based on the second multi-source heterogeneous information;

[0075] S108, generating multi-dimensional evaluation information based on the second multi-source heterogeneous information and the traffic rate information, including structural evaluation, water quality evaluation, hydrological evaluation, and ecological evaluation;

[0076] S110, obtaining a fishway health index based on the multi-dimensional evaluation information based on a preset weighting algorithm;

[0077] S112, if the fishway health index is lower than a preset health threshold, outputting diagnostic information based on the multi-dimensional evaluation information based on a preset diagnostic model, including key factor analysis and problem probability distribution;

[0078] S114, based on a preset knowledge base and according to the diagnostic information, generating a diagnostic conclusion and maintenance suggestion.

[0079] It should be noted that underwater cameras, sonar equipment, water quality sensors, and hydrological monitoring equipment collect images, 3D structural point clouds, water quality parameters, and flow data within the fishway, respectively. These data are recorded as image information, 3D point cloud information, water quality information, and flow information. As an implementation method, the monitoring equipment transmits this data in real time to a terminal device or server backend, forming the first multi-source heterogeneous information to ensure synchronization and integrity. The data cleaning and preprocessing mechanism includes cleaning invalid image data, fusion identification of 3D point cloud data, noise processing of water quality and flow data, and completion of missing data, thereby generating the second multi-source heterogeneous information. Image data is processed using inter-frame differencing to remove invalid frames such as blurred or occluded images, retaining only valid video clips. The 3D point cloud data is then registered and fused with a standard fishway model to generate a 3D fishway model, marking abnormal areas such as slope deviations or siltation. Water quality and flow information are timestamp-aligned, and noise is smoothed using a sliding window method. The first multi-source heterogeneous information is detected for missing data. If so, interpolation is performed based on a preset linear interpolation or historical mean. The fish species and number in the image data are identified based on a preset AI fish recognition model. As an implementation, images are recognized based on a pre-trained YOLOv8 model. The number of fish passing through the image data per unit time is counted according to fish type, and the passage rate is calculated. Based on preset scoring rules, the structure, water quality, hydrology, and ecology of the fishway are scored separately to obtain a structural assessment, a water quality assessment, a hydrological assessment, and an ecological assessment, forming multidimensional assessment information. The structural assessment is used to assess the rationality of the fishway model, the water quality assessment is used to assess water quality, the hydrological assessment is used to assess water flow velocity and flow, and the ecological assessment is used to assess the ecological diversity of the fishway. Based on a preset weighting algorithm, the scores of the structural assessment, water quality assessment, hydrological assessment, and ecological assessment are weighted and summed according to dynamically assigned weights to obtain a fishway health index. If the fishway health index falls below a preset threshold, a diagnostic model is activated to evaluate and output diagnostic information. This model outputs the contribution probability of each factor, such as structural issues or water quality issues, and then generates a probability distribution and key factors. Finally, maintenance recommendations are generated based on a knowledge base including design specifications and ecological thresholds. This embodiment integrates multi-dimensional data to avoid misjudgment of a single factor and improve diagnostic accuracy. Real-time monitoring and early warning mechanisms enable rapid response to ecological risks.

[0080] Figure 2 The figure shows an operation flow chart of a data cleaning and preprocessing mechanism provided by an embodiment of the present invention.

[0081] According to an embodiment of the present invention, Figure 2 As shown, the preset data cleaning and preprocessing mechanism specifically includes:

[0082] S202, based on the image information, eliminating invalid image frames based on an inter-frame difference method, and then extracting valid image frames through a frame extraction strategy;

[0083] S204, generating a three-dimensional model of the fishway based on the three-dimensional point cloud information, and then identifying structural abnormal areas using a clustering algorithm, including areas where the slope deviation is greater than a preset slope reference value and / or the fishway width is less than a preset width reference value;

[0084] S206, based on the water quality information and the water flow information, perform preset time series alignment and outlier removal, and then use a preset sliding window method to smooth the noise data;

[0085] S208: If there is missing data in the first multi-source heterogeneous information, interpolation is performed based on a preset linear interpolation or historical mean.

[0086] It should be noted that, as an implementation method, this embodiment uses the OpenCV library to perform inter-frame difference analysis on video image data, removing consecutive similar frames, such as blurred segments. A high-definition image is extracted every 10 seconds for subsequent AI analysis and recognition, reducing the computational load. For the cleaning and preprocessing of 3D point cloud data, the ICP algorithm is used to register multi-view sonar point cloud data to generate a complete fishway model. DBSCAN clustering is then used to identify abnormal areas, such as areas with siltation point density exceeding a set threshold or areas with slope deviation exceeding a set threshold. For the cleaning and preprocessing of water quality and flow data, the data is first timestamped and then smoothed using a sliding window to smooth flow velocity fluctuations. Finally, the first multi-source heterogeneous information is checked for missing data. If missing data is present, it is interpolated based on a preset linear interpolation or historical mean. This embodiment reduces noise by cleaning the data, thereby improving the accuracy of subsequent analysis. Frame extraction of the image data is also performed to balance computing resources and detection accuracy, avoiding redundant calculations.

[0087] Figure 3 A flow chart for calculating traffic rate information provided by an embodiment of the present invention is shown.

[0088] According to an embodiment of the present invention, Figure 3 As shown, the corresponding traffic rate information is generated according to the fish category based on the second multi-source heterogeneous information, specifically:

[0089] S302, performing target detection on the fish in the image information based on a preset AI fish recognition model;

[0090] S304, counting the number and species of fish that pass through within a preset unit time and generating a fish passage log;

[0091] S306, analyzing the distribution density and movement trajectory of fish in the fishway based on the three-dimensional point cloud information and the fish passage log;

[0092] S308, determining a bottleneck area for fish passage based on the distribution density and the movement trajectory;

[0093] S310, generating a water flow-traffic volume correlation model based on the water flow information and the fish passage log;

[0094] S312, based on the fish passage log, the passage rate is counted according to the fish category, and a passage time series report is generated, which at least includes the daily passage curve and the fish species ratio.

[0095] It should be noted that, as an implementation method, a pre-trained YOLOv8 model is fine-tuned based on a local fish dataset to identify and count fish in image information. The number of each type of fish passing through in a preset unit time is recorded to generate a traffic log. The three-dimensional model is divided into grids of preset sizes based on the three-dimensional point cloud information, the frequency of fish staying in each grid is counted, and high-density areas and low-traffic areas are identified. Regression analysis is used to establish the relationship between flow velocity and traffic volume, determine the critical flow velocity threshold, and establish a quantitative relationship between water flow and traffic volume, providing a basis for flow velocity regulation. Based on the relationship between fish distribution density and flow velocity-traffic volume, the bottleneck area in the fishway is determined. Finally, the daily traffic volume is counted by fish species to generate a curve graph.

[0096] According to an embodiment of the present invention, generating multidimensional evaluation information based on the second multi-source heterogeneous information and the traffic rate information is specifically:

[0097] The three-dimensional fishway model generated from the three-dimensional point cloud information is compared with a preset standard fishway model to obtain a structural assessment score based on preset structural dimensions, wherein the structural dimensions include at least slope, turning radius, bottleneck width, and siltation volume;

[0098] Real-time monitoring of water quality information, calculation of fish species survival indicators based on fish types, and obtaining a water quality assessment score, wherein the water quality information includes at least dissolved oxygen, water temperature, pH value, and turbidity data;

[0099] According to the water flow-volume correlation model, critical flow velocity values and abnormal water level periods are identified to obtain a hydrological assessment score;

[0100] Target detection is performed using the fish species, and the fish population diversity index and the rate of change in the number of target fish species are calculated to obtain an ecological assessment score.

[0101] It should be noted that the health of a fishway is scored based on its structure, water quality, hydrology, and ecology, according to preset scoring rules. As an implementation, a score is assigned based on the deviation of the actual slope from the standard design value, the percentage of silt volume, and other factors, thereby generating a structural score. A score is assigned based on the deviation of dissolved oxygen content from the fish survival threshold, and based on water temperature data exceeding the tolerance range of the target fish species. A score is assigned based on the deviation of water flow and water level from the standard design values. An ecological score is assigned based on the number of fish species and fish populations, with changes in fish populations counted. This embodiment transforms complex issues into quantifiable indicators through a scoring system that comprehensively reflects the health of the fishway from four dimensions: structure, water quality, hydrology, and ecology.

[0102] According to an embodiment of the present invention, the fishway health index is obtained based on the multi-dimensional evaluation information based on the preset weighted algorithm, specifically:

[0103] Based on the preset assessment model, multi-dimensional assessment weights are obtained, including structural weight, water quality weight, hydrological weight and ecological weight;

[0104] The fishway health index is obtained by weighted summation based on the multi-dimensional evaluation information and the multi-dimensional evaluation weights.

[0105] It should be noted that a pre-set expert evaluation model dynamically assigns multi-dimensional assessment weights, including structural, water quality, hydrological, and ecological weights. As one example, when seasonal structural issues are identified as having the greatest impact on fishway performance, the structural weight is maximized. Finally, based on these dynamically assigned weights, the scores for structural, water quality, hydrological, and ecological assessments are weighted and summed to produce the Fishway Health Index.

[0106] According to an embodiment of the present invention, the output of diagnostic information based on the preset diagnostic model and the multi-dimensional evaluation information is specifically:

[0107] Constructing a diagnostic model training dataset based on the fishway case database and its historical multi-dimensional evaluation data, and corresponding root cause labels, for training the diagnostic model;

[0108] Input the multi-dimensional evaluation information of the current fishway into the diagnostic model to obtain the contribution probability of each feature to the failure of the fishway;

[0109] Generate a problem probability distribution based on the contribution probabilities, and sort the key factors from high to low in terms of probability.

[0110] It should be noted that historical cases and their root cause structure labels were collected as a dataset for training the diagnostic model. Features were extracted from these historical cases, including slope deviation, number of dissolved oxygen exceedances, peak flow velocity, and population decline rate. The model outputs the contribution probability of each feature, generating a probability distribution map. The extracted key factors are then ranked based on probability.

[0111] It is worth mentioning that the knowledge base construction and update process is also included, specifically:

[0112] Collect and filter fish ecology literature to obtain data on fish survival thresholds, including at least the minimum dissolved oxygen limit, water temperature tolerance range, and migratory habits;

[0113] Collect water conservancy project specifications to obtain fishway design standards, which are used to update the standard fishway model, including at least the upper limit of slope, minimum turning radius and recommended flow rate;

[0114] Import historical diagnosis cases and optimization plans according to the preset time period to generate a structured rule base;

[0115] According to the survival threshold data, the standard fishway model and the structured rule base, based on preset mapping rules, the data are stored in the knowledge base.

[0116] It is important to note that fish survival thresholds, such as dissolved oxygen requirements, water temperature tolerance, and migratory habits, were extracted from fish ecology literature. Structured data was constructed based on fish species and survival thresholds. Data such as upper slope limits and minimum turning radius were converted from water conservancy project specifications into database fields, and fishway design standards were regularly updated based on new regulations. Historical diagnostic cases and their corresponding optimization solutions were linked to generate a structured rule base. Finally, based on pre-set mapping rules and after expert review, the results were integrated into a knowledge base.

[0117] It is worth mentioning that it also includes real-time monitoring and early warning mechanisms, specifically:

[0118] Dynamically monitor the water quality information, and if the dissolved oxygen or water temperature exceeds the corresponding threshold value within the first preset time, a water quality warning is triggered;

[0119] Tracking and analyzing the passage rate information, if the target fish species are not detected passing through for more than a preset second time, a high-risk warning is issued;

[0120] If the fishway health index is lower than the health threshold, the diagnostic model is automatically activated and maintenance recommendations are generated.

[0121] It should be noted that, as an implementation, this example provides a real-time early warning mechanism. For example, if dissolved oxygen levels are less than 5 mg / L for two consecutive hours or the water temperature exceeds a preset threshold, a water quality warning is triggered and sent to the fishway administrator. If target fish species are not detected passing through for 30 consecutive days, the warning is marked as "high risk" and an emergency report is generated. If the fishway health index is less than 60 points, the system automatically accesses the knowledge base, matches historical cases, and generates maintenance recommendations.

[0122] It is worth mentioning that it also includes:

[0123] Analytically matching the optimization solutions in the knowledge base based on key factors to obtain adjustment strategies, which at least include structural adjustment, water quality improvement or flow rate regulation;

[0124] According to the probability distribution of the problem and based on the corresponding adjustment strategy, determine the priority of the specific adjustment method;

[0125] generating a detailed construction plan according to the adjustment strategy and the priority;

[0126] Output maintenance cost estimate and construction schedule based on the detailed construction plan.

[0127] It should be noted that, as an implementation, if the primary diagnosis is excessively steep slope, a slope adjustment plan is retrieved from the knowledge base; if the primary diagnosis is insufficient dissolved oxygen, a solution for adding aeration equipment is matched from the knowledge base. A comprehensive ranking is performed based on the contribution probability and the corresponding implementation cost. A detailed construction plan is then generated based on the adjustment strategy and priority, including process steps, materials, work quantity, and duration. Finally, the total maintenance cost is calculated based on the detailed construction plan.

[0128] Figure 4 A block diagram of a fishway performance evaluation and diagnosis system based on AI ecological monitoring according to the present invention is shown.

[0129] like Figure 4 As shown, the second aspect of the present invention discloses a fishway effectiveness evaluation and diagnosis system 4 based on AI ecological monitoring, including a memory 41 and a processor 42. The memory includes a fishway effectiveness evaluation and diagnosis method program based on AI ecological monitoring. When the fishway effectiveness evaluation and diagnosis method program based on AI ecological monitoring is executed by the processor, the following steps are implemented:

[0130] Acquiring first multi-source heterogeneous information, including image information, three-dimensional point cloud information, water quality information, and water flow information;

[0131] Based on a preset data cleaning and preprocessing mechanism, obtaining second multi-source heterogeneous information according to the first multi-source heterogeneous information;

[0132] Generating corresponding traffic rate information according to fish types based on the second multi-source heterogeneous information;

[0133] generating multidimensional assessment information based on the second multi-source heterogeneous information and the traffic rate information, including structural assessment, water quality assessment, hydrological assessment, and ecological assessment;

[0134] Obtaining a fishway health index based on the multi-dimensional evaluation information based on a preset weighting algorithm;

[0135] If the fishway health index is lower than a preset health threshold, outputting diagnostic information based on a preset diagnostic model and the multidimensional evaluation information, including key factor analysis and problem probability distribution;

[0136] Based on the preset knowledge base and the diagnostic information, a diagnostic conclusion and maintenance suggestion are generated.

[0137] It should be noted that underwater cameras, sonar equipment, water quality sensors, and hydrological monitoring equipment collect images, 3D structural point clouds, water quality parameters, and flow data within the fishway, respectively. These data are recorded as image information, 3D point cloud information, water quality information, and flow information. As an implementation method, the monitoring equipment transmits this data in real time to a terminal device or server backend, forming the first multi-source heterogeneous information to ensure synchronization and integrity. The data cleaning and preprocessing mechanism includes cleaning invalid image data, fusion identification of 3D point cloud data, noise processing of water quality and flow data, and completion of missing data, thereby generating the second multi-source heterogeneous information. Image data is processed using inter-frame differencing to remove invalid frames such as blurred or occluded images, retaining only valid video clips. The 3D point cloud data is then registered and fused with a standard fishway model to generate a 3D fishway model, marking abnormal areas such as slope deviations or siltation. Water quality and flow information are timestamp-aligned, and noise is smoothed using a sliding window method. The first multi-source heterogeneous information is detected for missing data. If so, interpolation is performed based on a preset linear interpolation or historical mean. The fish species and number in the image data are identified based on a preset AI fish recognition model. As an implementation, images are recognized based on a pre-trained YOLOv8 model. The number of fish passing through the image data per unit time is counted according to fish type, and the passage rate is calculated. Based on preset scoring rules, the structure, water quality, hydrology, and ecology of the fishway are scored separately to obtain a structural assessment, a water quality assessment, a hydrological assessment, and an ecological assessment, forming multidimensional assessment information. The structural assessment is used to assess the rationality of the fishway model, the water quality assessment is used to assess water quality, the hydrological assessment is used to assess water flow velocity and flow, and the ecological assessment is used to assess the ecological diversity of the fishway. Based on a preset weighting algorithm, the scores of the structural assessment, water quality assessment, hydrological assessment, and ecological assessment are weighted and summed according to dynamically assigned weights to obtain a fishway health index. If the fishway health index falls below a preset threshold, a diagnostic model is activated to evaluate and output diagnostic information. This model outputs the contribution probability of each factor, such as structural issues or water quality issues, and then generates a probability distribution and key factors. Finally, maintenance recommendations are generated based on a knowledge base including design specifications and ecological thresholds. This embodiment integrates multi-dimensional data to avoid misjudgment of a single factor and improve diagnostic accuracy. Real-time monitoring and early warning mechanisms enable rapid response to ecological risks.

[0138] According to an embodiment of the present invention, the preset data cleaning and preprocessing mechanism specifically includes:

[0139] According to the image information, invalid image frames are eliminated based on the inter-frame difference method, and then valid image frames are extracted through the frame extraction strategy;

[0140] generating a three-dimensional model of the fishway based on the three-dimensional point cloud information, and then identifying structural abnormal areas using a clustering algorithm, including areas where the slope deviation is greater than a preset slope reference value and / or the fishway width is less than a preset width reference value;

[0141] According to the water quality information and the water flow information, according to the preset time series alignment and outlier removal, the noise data is smoothed using the preset sliding window method;

[0142] If there is missing data in the first multi-source heterogeneous information, interpolation is performed based on a preset linear interpolation or historical mean.

[0143] It should be noted that, as an implementation method, this embodiment uses the OpenCV library to perform inter-frame difference analysis on video image data, removing consecutive similar frames, such as blurred segments. A high-definition image is extracted every 10 seconds for subsequent AI analysis and recognition, reducing the computational load. For the cleaning and preprocessing of 3D point cloud data, the ICP algorithm is used to register multi-view sonar point cloud data to generate a complete fishway model. DBSCAN clustering is then used to identify abnormal areas, such as areas with siltation point density exceeding a set threshold or areas with slope deviation exceeding a set threshold. For the cleaning and preprocessing of water quality and flow data, the data is first timestamped and then smoothed using a sliding window to smooth flow velocity fluctuations. Finally, the first multi-source heterogeneous information is checked for missing data. If missing data is present, it is interpolated based on a preset linear interpolation or historical mean. This embodiment reduces noise by cleaning the data, thereby improving the accuracy of subsequent analysis. Frame extraction of the image data is also performed to balance computing resources and detection accuracy, avoiding redundant calculations.

[0144] According to an embodiment of the present invention, the generation of corresponding traffic rate information according to fish types based on the second multi-source heterogeneous information is specifically as follows:

[0145] Based on a preset AI fish recognition model, target detection is performed on the fish in the image information;

[0146] Count the number and species of fish that pass through within a preset unit time and generate a fish passage log;

[0147] Analyzing the distribution density and movement trajectory of fish in the fishway based on the three-dimensional point cloud information and the fish passage log;

[0148] determining a fish passage bottleneck area based on the distribution density and the movement trajectory;

[0149] generating a water flow-traffic volume correlation model based on the water flow information and the fish passage log;

[0150] Based on the fish passage log, the passage rate is counted by fish category and a passage time series report is generated, which at least includes the daily passage curve and the fish species percentage.

[0151] It should be noted that, as an implementation method, a pre-trained YOLOv8 model is fine-tuned based on a local fish dataset to identify and count fish in image information. The number of each type of fish passing through in a preset unit time is recorded to generate a traffic log. The three-dimensional model is divided into grids of preset sizes based on the three-dimensional point cloud information, the frequency of fish staying in each grid is counted, and high-density areas and low-traffic areas are identified. Regression analysis is used to establish the relationship between flow velocity and traffic volume, determine the critical flow velocity threshold, and establish a quantitative relationship between water flow and traffic volume, providing a basis for flow velocity regulation. Based on the relationship between fish distribution density and flow velocity-traffic volume, the bottleneck area in the fishway is determined. Finally, the daily traffic volume is counted by fish species to generate a curve graph.

[0152] According to an embodiment of the present invention, generating multidimensional evaluation information based on the second multi-source heterogeneous information and the traffic rate information is specifically:

[0153] The three-dimensional fishway model generated from the three-dimensional point cloud information is compared with a preset standard fishway model to obtain a structural assessment score based on preset structural dimensions, wherein the structural dimensions include at least slope, turning radius, bottleneck width, and siltation volume;

[0154] Real-time monitoring of water quality information, calculation of fish species survival indicators based on fish types, and obtaining a water quality assessment score, wherein the water quality information includes at least dissolved oxygen, water temperature, pH value, and turbidity data;

[0155] According to the water flow-volume correlation model, critical flow velocity values and abnormal water level periods are identified to obtain a hydrological assessment score;

[0156] Target detection is performed using the fish species, and the fish population diversity index and the rate of change in the number of target fish species are calculated to obtain an ecological assessment score.

[0157] It should be noted that the health of a fishway is scored based on its structure, water quality, hydrology, and ecology, according to preset scoring rules. As an implementation, a score is assigned based on the deviation of the actual slope from the standard design value, the percentage of silt volume, and other factors, thereby generating a structural score. A score is assigned based on the deviation of dissolved oxygen content from the fish survival threshold, and based on water temperature data exceeding the tolerance range of the target fish species. A score is assigned based on the deviation of water flow and water level from the standard design values. An ecological score is assigned based on the number of fish species and fish populations, with changes in fish populations counted. This embodiment transforms complex issues into quantifiable indicators through a scoring system that comprehensively reflects the health of the fishway from four dimensions: structure, water quality, hydrology, and ecology.

[0158] According to an embodiment of the present invention, the fishway health index is obtained based on the multi-dimensional evaluation information based on the preset weighted algorithm, specifically:

[0159] Based on the preset assessment model, multi-dimensional assessment weights are obtained, including structural weight, water quality weight, hydrological weight and ecological weight;

[0160] The fishway health index is obtained by weighted summation based on the multi-dimensional evaluation information and the multi-dimensional evaluation weights.

[0161] It should be noted that a pre-set expert evaluation model dynamically assigns multi-dimensional assessment weights, including structural, water quality, hydrological, and ecological weights. As one example, when seasonal structural issues are identified as having the greatest impact on fishway performance, the structural weight is maximized. Finally, based on these dynamically assigned weights, the scores for structural, water quality, hydrological, and ecological assessments are weighted and summed to produce the Fishway Health Index.

[0162] According to an embodiment of the present invention, the output of diagnostic information based on the preset diagnostic model and the multi-dimensional evaluation information is specifically:

[0163] Constructing a diagnostic model training dataset based on the fishway case database and its historical multi-dimensional evaluation data, and corresponding root cause labels, for training the diagnostic model;

[0164] Input the multi-dimensional evaluation information of the current fishway into the diagnostic model to obtain the contribution probability of each feature to the failure of the fishway;

[0165] Generate a problem probability distribution based on the contribution probabilities, and sort the key factors from high to low in terms of probability.

[0166] It should be noted that historical cases and their root cause structure labels were collected as a dataset for training the diagnostic model. Features were extracted from these historical cases, including slope deviation, number of dissolved oxygen exceedances, peak flow velocity, and population decline rate. The model outputs the contribution probability of each feature, generating a probability distribution map. The extracted key factors are then ranked based on probability.

[0167] It is worth mentioning that the knowledge base construction and update process is also included, specifically:

[0168] Collect and filter fish ecology literature to obtain data on fish survival thresholds, including at least the minimum dissolved oxygen limit, water temperature tolerance range, and migratory habits;

[0169] Collect water conservancy project specifications to obtain fishway design standards, which are used to update the standard fishway model, including at least the upper limit of slope, minimum turning radius and recommended flow rate;

[0170] Import historical diagnosis cases and optimization plans according to the preset time period to generate a structured rule base;

[0171] According to the survival threshold data, the standard fishway model and the structured rule base, based on preset mapping rules, the data are stored in the knowledge base.

[0172] It is important to note that fish survival thresholds, such as dissolved oxygen requirements, water temperature tolerance, and migratory habits, were extracted from fish ecology literature. Structured data was constructed based on fish species and survival thresholds. Data such as upper slope limits and minimum turning radius were converted from water conservancy project specifications into database fields, and fishway design standards were regularly updated based on new regulations. Historical diagnostic cases and their corresponding optimization solutions were linked to generate a structured rule base. Finally, based on pre-set mapping rules and after expert review, the results were integrated into a knowledge base.

[0173] It is worth mentioning that it also includes real-time monitoring and early warning mechanisms, specifically:

[0174] Dynamically monitor the water quality information, and if the dissolved oxygen or water temperature exceeds the corresponding threshold value within the first preset time, a water quality warning is triggered;

[0175] Tracking and analyzing the passage rate information, if the target fish species are not detected passing through for more than a preset second time, a high-risk warning is issued;

[0176] If the fishway health index is lower than the health threshold, the diagnostic model is automatically activated and maintenance recommendations are generated.

[0177] It should be noted that, as an implementation, this example provides a real-time early warning mechanism. For example, if dissolved oxygen levels are less than 5 mg / L for two consecutive hours or the water temperature exceeds a preset threshold, a water quality warning is triggered and sent to the fishway administrator. If target fish species are not detected passing through for 30 consecutive days, the warning is marked as "high risk" and an emergency report is generated. If the fishway health index is less than 60 points, the system automatically accesses the knowledge base, matches historical cases, and generates maintenance recommendations.

[0178] It is worth mentioning that it also includes:

[0179] Analytically matching the optimization solutions in the knowledge base based on key factors to obtain adjustment strategies, which at least include structural adjustment, water quality improvement or flow rate regulation;

[0180] According to the probability distribution of the problem and based on the corresponding adjustment strategy, determine the priority of the specific adjustment method;

[0181] generating a detailed construction plan according to the adjustment strategy and the priority;

[0182] Output maintenance cost estimate and construction schedule based on the detailed construction plan.

[0183] It should be noted that, as an implementation, if the primary diagnosis is excessively steep slope, a slope adjustment plan is retrieved from the knowledge base; if the primary diagnosis is insufficient dissolved oxygen, a solution for adding aeration equipment is matched from the knowledge base. A comprehensive ranking is performed based on the contribution probability and the corresponding implementation cost. A detailed construction plan is then generated based on the adjustment strategy and priority, including process steps, materials, work quantity, and duration. Finally, the total maintenance cost is calculated based on the detailed construction plan.

[0184] The third aspect of the present invention provides a computer-readable storage medium, which includes a fishway efficiency evaluation and diagnosis method program based on AI ecological monitoring. When the fishway efficiency evaluation and diagnosis method program based on AI ecological monitoring is executed by a processor, the steps of the fishway efficiency evaluation and diagnosis method based on AI ecological monitoring as described in any one of the above items are implemented.

[0185] In summary, the present invention provides a fishway efficiency evaluation and diagnosis method, system and medium based on AI ecological monitoring. First, by deploying underwater cameras, sonar, water quality sensors and hydrological monitoring equipment, images, three-dimensional point clouds, water quality parameters and water flow data inside and outside the fishway are collected in real time to form multi-source heterogeneous data; secondly, through data cleaning and preprocessing, as well as preset AI recognition technology, the number and species of fish passing through are automatically counted; then, the fishway is evaluated in multiple dimensions based on structure, water quality, hydrology and ecology, and the fishway health index is analyzed based on expert weight distribution; finally, the fishway energy efficiency is evaluated and problems are diagnosed based on the fishway health index, and maintenance recommendations are generated; the present invention constructs a closed-loop system of monitoring-evaluation-diagnosis-optimization to achieve dynamic management and rapid response of fishway efficiency, providing scientific support for water conservancy projects and ecological sustainable development.

[0186] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0187] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention are intended to be within the scope of protection of the present invention.

Claims

1. A fishway performance evaluation and diagnosis method based on AI ecological monitoring, characterized in that: The method comprises: Acquiring first multi-source heterogeneous information, including image information, three-dimensional point cloud information, water quality information, and water flow information; Based on a preset data cleaning and preprocessing mechanism, obtaining second multi-source heterogeneous information according to the first multi-source heterogeneous information; Generating corresponding traffic rate information according to fish types based on the second multi-source heterogeneous information; generating multidimensional assessment information based on the second multi-source heterogeneous information and the traffic rate information, including structural assessment, water quality assessment, hydrological assessment, and ecological assessment; Obtaining a fishway health index based on the multi-dimensional evaluation information based on a preset weighting algorithm; If the fishway health index is lower than a preset health threshold, outputting diagnostic information based on a preset diagnostic model and the multidimensional evaluation information, including key factor analysis and problem probability distribution; Based on the preset knowledge base and the diagnostic information, a diagnostic conclusion and maintenance suggestion are generated.

2. The fishway performance evaluation and diagnosis method based on AI ecological monitoring according to claim 1 is characterized in that: The preset data cleaning and preprocessing mechanism specifically includes: According to the image information, invalid image frames are eliminated based on the inter-frame difference method, and then valid image frames are extracted through the frame extraction strategy; generating a three-dimensional model of the fishway based on the three-dimensional point cloud information, and then identifying structural abnormal areas using a clustering algorithm, including areas where the slope deviation is greater than a preset slope reference value and / or the fishway width is less than a preset width reference value; According to the water quality information and the water flow information, according to the preset time series alignment and outlier removal, the noise data is smoothed using the preset sliding window method; If there is missing data in the first multi-source heterogeneous information, interpolation is performed based on a preset linear interpolation or historical mean.

3. The fishway performance evaluation and diagnosis method based on AI ecological monitoring according to claim 1 is characterized in that: The method of generating corresponding traffic rate information according to fish types based on the second multi-source heterogeneous information is specifically as follows: Based on a preset AI fish recognition model, target detection is performed on the fish in the image information; Count the number and species of fish that pass through within a preset unit time and generate a fish passage log; Analyzing the distribution density and movement trajectory of fish in the fishway based on the three-dimensional point cloud information and the fish passage log; determining a fish passage bottleneck area based on the distribution density and the movement trajectory; generating a water flow-traffic volume correlation model based on the water flow information and the fish passage log; Based on the fish passage log, the passage rate is counted by fish category and a passage time series report is generated, which at least includes the daily passage curve and the fish species percentage.

4. The fishway performance evaluation and diagnosis method based on AI ecological monitoring according to claim 3 is characterized in that: The generating of multi-dimensional evaluation information according to the second multi-source heterogeneous information and the traffic rate information is specifically as follows: The three-dimensional fishway model generated from the three-dimensional point cloud information is compared with a preset standard fishway model to obtain a structural assessment score based on preset structural dimensions, wherein the structural dimensions include at least slope, turning radius, bottleneck width, and siltation volume; Real-time monitoring of water quality information, calculation of fish species survival indicators based on fish types, and obtaining a water quality assessment score, wherein the water quality information includes at least dissolved oxygen, water temperature, pH value, and turbidity data; According to the water flow-volume correlation model, critical flow velocity values and abnormal water level periods are identified to obtain a hydrological assessment score; Target detection is performed using the fish species, and the fish population diversity index and the rate of change in the number of target fish species are calculated to obtain an ecological assessment score.

5. The fishway performance evaluation and diagnosis method based on AI ecological monitoring according to claim 1 is characterized in that: The fishway health index is obtained based on the multi-dimensional evaluation information based on the preset weighted algorithm, specifically: Based on the preset assessment model, multi-dimensional assessment weights are obtained, including structural weight, water quality weight, hydrological weight and ecological weight; The fishway health index is obtained by weighted summation based on the multi-dimensional evaluation information and the multi-dimensional evaluation weights.

6. The fishway performance evaluation and diagnosis method based on AI ecological monitoring according to claim 1 is characterized in that: The preset diagnostic model outputs diagnostic information according to the multi-dimensional evaluation information, specifically: Constructing a diagnostic model training dataset based on the fishway case database and its historical multi-dimensional evaluation data, and corresponding root cause labels, for training the diagnostic model; Input the multi-dimensional evaluation information of the current fishway into the diagnostic model to obtain the contribution probability of each feature to the failure of the fishway; Generate a problem probability distribution based on the contribution probabilities, and sort the key factors from high to low in terms of probability.

7. A fishway performance evaluation and diagnosis system based on AI ecological monitoring, characterized by: The system includes a memory and a processor. The memory includes a fishway effectiveness evaluation and diagnosis method program based on AI ecological monitoring. When the fishway effectiveness evaluation and diagnosis method program based on AI ecological monitoring is executed by the processor, the following steps are implemented: Acquiring first multi-source heterogeneous information, including image information, three-dimensional point cloud information, water quality information, and water flow information; Based on a preset data cleaning and preprocessing mechanism, obtaining second multi-source heterogeneous information according to the first multi-source heterogeneous information; Generating corresponding traffic rate information according to fish types based on the second multi-source heterogeneous information; generating multidimensional assessment information based on the second multi-source heterogeneous information and the traffic rate information, including structural assessment, water quality assessment, hydrological assessment, and ecological assessment; Obtaining a fishway health index based on the multi-dimensional evaluation information based on a preset weighting algorithm; If the fishway health index is lower than a preset health threshold, outputting diagnostic information based on a preset diagnostic model and the multidimensional evaluation information, including key factor analysis and problem probability distribution; Based on the preset knowledge base and the diagnostic information, a diagnostic conclusion and maintenance suggestion are generated.

8. The fishway performance evaluation and diagnosis system based on AI ecological monitoring according to claim 7 is characterized in that: The preset data cleaning and preprocessing mechanism specifically includes: According to the image information, invalid image frames are eliminated based on the inter-frame difference method, and then valid image frames are extracted through the frame extraction strategy; generating a three-dimensional model of the fishway based on the three-dimensional point cloud information, and then identifying structural abnormal areas using a clustering algorithm, including areas where the slope deviation is greater than a preset slope reference value and / or the fishway width is less than a preset width reference value; According to the water quality information and the water flow information, according to the preset time series alignment and outlier removal, the noise data is smoothed using the preset sliding window method; If there is missing data in the first multi-source heterogeneous information, interpolation is performed based on a preset linear interpolation or historical mean.

9. The fishway performance evaluation and diagnosis system based on AI ecological monitoring according to claim 7 is characterized in that: The method of generating corresponding traffic rate information according to fish types based on the second multi-source heterogeneous information is specifically as follows: Based on a preset AI fish recognition model, target detection is performed on the fish in the image information; Count the number and species of fish that pass through within a preset unit time and generate a fish passage log; Analyzing the distribution density and movement trajectory of fish in the fishway based on the three-dimensional point cloud information and the fish passage log; determining a fish passage bottleneck area based on the distribution density and the movement trajectory; generating a water flow-traffic volume correlation model based on the water flow information and the fish passage log; Based on the fish passage log, the passage rate is counted by fish category and a passage time series report is generated, which at least includes the daily passage curve and the fish species percentage.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: The computer-readable storage medium includes a fishway efficiency evaluation and diagnosis method program based on AI ecological monitoring. When the fishway efficiency evaluation and diagnosis method program based on AI ecological monitoring is executed by a processor, the steps of the fishway efficiency evaluation and diagnosis method based on AI ecological monitoring as described in any one of claims 1 to 6 are implemented.

Images