Ecological scheduling method for fish habitat habitat protection

Abstract

The invention discloses an ecological scheduling method for fish habitat habitat protection. The ecological scheduling method comprises the following steps of collecting basic data about water area ecological hydraulic characteristic parameters and fish ecological habits, constructing an evaluation system, formulating an ecological scheduling scheme, and performing adaptive regulation and control. According to the method, dynamic thresholds of key ecological parameters such as flow velocity and water temperature in different development periods are quantified by establishing a hydraulic-ecological response curved surface. And the requirements of fishes on water flow conditions in different development stages can be known more accurately. And XAJ distributed hydrological model and LSTM neural network hybrid forecasting is constructed, the operation water level of the reservoir is controlled in advance according to the result of the evaluation system, a scheduling strategy is adjusted in time, and sudden ecological requirements are met.

Description

Technical Field

[0001] This invention belongs to the field of ecological restoration technology, and in particular relates to an ecological scheduling method for the protection of fish habitats. Background Technology

[0002] Fish are a vital component of river ecosystems, and their survival directly impacts the stability and balance of the entire ecosystem. Protecting fish habitats helps maintain the biodiversity of river ecosystems and ensures their health and sustainable development. With increasing human activity, river ecosystems face unprecedented pressure. Factors such as water conservancy projects, water pollution, and overfishing have severely damaged fish habitats. The protection of fish habitats is closely related to water resource management. Through rational ecological regulation, water resource allocation can be optimized, water resource utilization efficiency can be improved, and damage to the river ecosystem can be reduced.

[0003] In existing technologies, ecological parameter assessments typically use fixed thresholds, neglecting the dynamic needs of species at different developmental stages. This leads to inaccurate assessment results that fail to reflect the true water flow requirements of fish at different developmental stages. Furthermore, in the unique habitats of plateau regions, river water volume is also affected by glacial meltwater, making the adjustment of reservoir operation strategies another problem that needs to be addressed. Therefore, this paper proposes an ecological scheduling method for fish habitat protection.

[0004] Patent application CN116149187A discloses an ecological scheduling method for the protection of fish habitats at the tail of a reservoir. Taking the concentrated spawning period of fish as the research period, it constructs a habitat suitability evaluation index system by utilizing the hydraulic parameter requirements of fish spawning on the habitat, and clarifies the main scheduling needs of fish spawning. By determining the water level-discharge relationship of the main control sections and typical representative sections, the scheduling threshold is determined. Through hydrological forecasting, the reservoir pre-stores or pre-releases water to keep the reservoir operating water level within the scheduling threshold, thereby controlling the water level fluctuation of the tail section of the reservoir and solving the ecological scheduling problem of the tail section of the reservoir.

[0005] Patent application CN120013068A discloses a method for determining the ecological water volume of seasonal rivers based on the life cycle of key species. This method can provide basic support and scientific basis for setting ecological water demand targets, allocating and managing ecological water demand in seasonal rivers in my country, and promote the health and ecological environment recovery of seasonal rivers.

[0006] Although the aforementioned patent application documents disclose the use of Bernoulli's equation for steady non-uniform flow and Manning's formula, the prior art documents do not address the calculation of the impact of glacial meltwater on rivers, and their methods are not applicable to rivers in regions containing glaciers.

[0007] Patent application CN119313285A discloses an environmental situational awareness data engineering system based on multi-source data fusion. The system's UEB model is applicable to various climatic and topographical conditions, exhibiting high accuracy and reliability in complex high-altitude and cold environments. The UEB model can help analyze and predict the impact of human activities on environmental factors, as well as the impact of environmental change on human behavior, deeply exploring the interaction between human activities and environmental change, and providing insights for decision-makers. Although this system incorporates the Manning formula and analyzes information including that of special environmental factors such as glaciers, it is not intended to address the technical problems of fish ecological environment restoration. Summary of the Invention

[0008] To address the aforementioned technical problems, this invention provides an ecological scheduling method for the protection of fish habitats.

[0009] The present invention is achieved through the following technical solutions.

[0010] This invention provides an ecological scheduling method for fish habitat protection, comprising the following steps:

[0011] Step 1: Basic Data Collection: Obtain basic data on the ecological and hydraulic characteristics of aquatic waters and the ecological habits of fish;

[0012] Step 2: Constructing the evaluation system: The evaluation system includes a habitat suitability evaluation index system and a hydrodynamic model; among them, the habitat suitability evaluation index system is constructed by utilizing the correlation between fish spawning and reproduction and river hydraulic elements; a hydrodynamic model of the reservoir tail section is constructed based on the Bernoulli equation for stable non-uniform flow; at the same time, a hydraulic-ecological response surface is established to determine the three-dimensional flow velocity threshold, water temperature window, and dissolved oxygen safety threshold for different fish species during key developmental stages;

[0013] Step 3: Develop an ecological scheduling plan: Construct a hybrid forecasting system combining the XAJ distributed hydrological model and LSTM neural network, control the reservoir's operating water level in advance based on the evaluation system results, and formulate different water level control strategies according to the different life stages and ecological needs of fish.

[0014] Step 4: Adaptive regulation: In response to the special habitat of the plateau region, establish a glacier mass balance model, predict snowmelt water volume, adjust reservoir operation strategy, and control the vertical mixing intensity of water body through deep hole-surface hole combined discharge to address dissolved oxygen stratification.

[0015] Preferably, the water ecological hydraulic characteristic parameters in step one include water level, flow rate, flow velocity, water temperature, and water quality. These water ecological hydraulic characteristic parameters are obtained by deploying a multi-parameter water quality sensor array and constructing an automatic cross-sectional flow monitoring station using an ADCP flow measurement system. At the same time, a water level-flow relationship curve library is established.

[0016] Preferably, the basic data on fish ecological habits in step one includes life history characteristics, geographical distribution characteristics, resource quantity, conservation necessity, dam construction impact, economic value, and spawning type. Environmental DNA technology is applied to monitor fish population dynamics and obtain fish ecological habit data. At the same time, a three-dimensional spawning ground characteristic model is established to understand the hydraulic conditions required for fish spawning.

[0017] Preferably, the specific steps in step two for constructing a habitat suitability evaluation index system based on the correlation between fish spawning and reproduction and river hydraulic elements are as follows:

[0018] Determining ecological factors and their weights: Four ecological factors, namely flow velocity, water level, water temperature and dissolved oxygen, were selected as evaluation indicators. The weights of these ecological factors were determined through statistical analysis.

[0019] Establish fuzzy sets: For each ecological factor, establish a corresponding fuzzy set. The fuzzy set is used to describe the different states or levels of the ecological factor.

[0020] Determine the membership function: Determine a triangular membership function for each fuzzy set. The triangular membership function is used to describe the degree to which the evaluation factor belongs to a certain fuzzy set.

[0021] Constructing a fuzzy comprehensive evaluation model: Based on fuzzy sets and triangular membership functions, a fuzzy comprehensive evaluation model is constructed. The fuzzy comprehensive evaluation model comprehensively considers the influence of multiple ecological factors to comprehensively evaluate the habitat suitability of fish habitats.

[0022] Calculate the fuzzy comprehensive evaluation results: Use the fuzzy comprehensive evaluation model to calculate the fuzzy comprehensive evaluation results for each evaluation object.

[0023] Preferably, the specific steps in step two for constructing the hydrodynamic model of the reservoir tail section based on the Bernoulli equation for steady non-uniform flow are as follows:

[0024] Determine the Bernoulli equation for steady nonuniform flow: The Bernoulli equation for steady nonuniform flow is expressed as:

[0025]

[0026] Application to the tail end of the reservoir: In the hydrodynamic model of the tail end of the reservoir, it is assumed that the fluid is incompressible and the density ρ is constant. For open channels or rivers, the fluid pressure term is ignored, and the Bernoulli equation is further simplified to:

[0027]

[0028] In equations (1) and (2), z1 and z2 are the heights at two locations upstream and downstream of the reservoir tail section, respectively; p1 and p2 are the fluid pressures at these two locations, respectively; ρ is the fluid density; g is the acceleration due to gravity; v1 and v2 are the flow velocities at these two locations, respectively; and h is the flow velocity at these two locations, respectively. f This is due to head loss caused by frictional resistance.

[0029] Preferably, the construction of the hydrodynamic model involves: collecting geometric feature data of the river section, measuring the flow velocity distribution, and using the Manning formula to estimate h. f The estimation formula is expressed as:

[0030]

[0031] Where n is the Manning roughness coefficient, Q is the flow rate, A is the cross-sectional area, and R is the hydraulic radius. Where P is the wetted perimeter. For a rectangular channel, the wetted perimeter P = 2W + 2D, where W is the channel width and D is the channel depth. The simplified Bernoulli equation is solved using the finite element method to obtain the water level changes of each section in the river segment under different inflow and operating water level conditions.

[0032] Preferably, step two involves establishing a hydraulic-ecological response surface to determine the three-dimensional flow velocity threshold, water temperature window, and dissolved oxygen safety threshold for different fish species during critical developmental stages, including the following steps:

[0033] Constructing a hydraulic-ecological response model: The collected data were used to construct a hydraulic-ecological response model using a support vector machine, with hydraulic characteristics as the independent variable and the key developmental period of fish as the dependent variable;

[0034] Determine the response surface: Through model calculation, the response of fish during key developmental periods under different combinations of hydraulic characteristics is obtained, and the calculation results are visualized as a three-dimensional response surface using the response surface method;

[0035] Determine the thresholds: Based on the response surface, determine the three-dimensional flow velocity thresholds, water temperature windows, and dissolved oxygen safety thresholds for different fish species during their critical developmental stages.

[0036] Preferably, in step three, a hybrid forecasting system combining the XAJ distributed hydrological model and LSTM neural network is constructed. Based on the evaluation system results, the reservoir's operating water level is controlled in advance. Different water level control strategies are formulated according to the different life stages and ecological needs of fish, including the following steps:

[0037] Constructing an XAJ distributed hydrological model: Collect rainfall, topography, vegetation, and soil data within the watershed and preprocess them. Based on the actual conditions of the watershed, set the model parameters, input the data into the model, and run the model to obtain the predicted flow at the watershed outlet.

[0038] Building an LSTM neural network: Design the structure of the LSTM neural network, including the input layer, hidden layer and output layer, and train the LSTM neural network using historical traffic data to obtain the optimal model parameters;

[0039] Hybrid forecast model construction: The XAJ distributed hydrological model and LSTM neural network are combined to construct a hybrid forecast model. The flow data output by the hydrological model is normalized and used as the input of the LSTM neural network to construct the hybrid forecast model. The hybrid forecast model is validated using a validation dataset to evaluate the predictive performance of the model.

[0040] Preemptive control of reservoir operation: A hybrid forecasting model is used to predict future flow changes. Based on the predicted flow changes, combined with the reservoir's storage and release capacity, future water level changes are calculated using the following formula:

[0041]

[0042] Where ΔH represents the change in water level, ΔV represents the change in reservoir storage, and A represents the surface area of ​​the reservoir.

[0043] Develop water level control strategies: During the non-spawning period of fish, maintain a steady-state mode to keep the water level relatively stable; during the spawning period of fish, when the water temperature reaches the standard, use a pulsed peak-forming mode to increase the water release from the reservoir in a short period of time to form a water flow peak, simulating the flood pulse phenomenon in natural rivers; during the fish egg hatching period, use a slow-release maintenance mode to slowly lower the water level of the reservoir.

[0044] Preferably, the glacier mass balance model described in step four is expressed as follows:

[0045] Q melt =k·A snow (T max -T threshold ) 1.8 (5)

[0046] Where k is the topography-climate coupling coefficient, calibrated by inversion from historical snowmelt data, and A snow It is the effective snow cover area, T max The highest temperature of the day, T threshold This is the temperature at which snow melting begins.

[0047] Preferably, in step four, for dissolved oxygen stratification, the vertical mixing intensity of the water is controlled by a combined deep-hole and surface-hole discharge strategy. This strategy includes surface-hole discharge, deep-hole discharge, and mixing intensity control. Surface-hole discharge releases oxygen-rich surface water to increase the oxygen content of the downstream water. Deep-hole discharge extracts low-temperature water from the bottom layer, which is then aerated through the spillway. Mixing intensity control is achieved by adjusting the opening ratio of the two holes to ensure the uniformity of vertical mixing of the water at the outlet. The aeration efficiency model is expressed as:

[0048]

[0049] Among them, H fall Q represents the discharge drop height, and Q represents the single-orifice flow rate.

[0050] The beneficial effects of this invention are as follows:

[0051] This invention quantifies the dynamic thresholds of key ecological parameters such as flow velocity and water temperature at different developmental stages by establishing a hydraulic-ecological response surface. This helps to more accurately understand the water flow requirements of fish at different developmental stages. A hybrid forecasting system combining an XAJ distributed hydrological model and an LSTM neural network is constructed. Based on the evaluation results, the reservoir's operating water level can be controlled in advance, and scheduling strategies can be adjusted in a timely manner to cope with sudden ecological demands. Attached Figure Description

[0052] Figure 1 This is a flowchart illustrating the steps of the present invention. Detailed Implementation

[0053] The technical solution of the present invention is further described below, but the scope of protection is not limited to what is described.

[0054] Example:

[0055] like Figure 1 As shown, an ecological scheduling method for fish habitat protection includes the following steps:

[0056] Step 1: Basic Data Collection: Obtain basic data on the ecological and hydraulic characteristics of aquatic waters and the ecological habits of fish;

[0057] Step Two: Constructing an Evaluation System: The evaluation system includes a habitat suitability evaluation index system and a hydrodynamic model. Specifically, the habitat suitability evaluation index system is constructed by utilizing the correlation between fish spawning and reproduction and river hydraulic elements. This correlation is based on the response relationship between eco-hydrological indicators and fish spawning behavior. A comprehensive evaluation of the habitat suitability of fish habitats is conducted by analyzing ecological factors in the river. Based on the Bernoulli equation for stable non-uniform flow, a hydrodynamic model for the reservoir tail section is constructed. This model is used to calculate the water level and its changes at the main representative sections under different inflow and operating water level conditions. Simultaneously, a hydraulic-ecological response surface is established to determine the three-dimensional velocity threshold, water temperature window, and dissolved oxygen safety threshold for different fish species during their critical developmental stages.

[0058] Step 3: Develop an ecological scheduling plan: Construct a hybrid forecasting system combining the XAJ distributed hydrological model and LSTM neural network, control the reservoir's operating water level in advance based on the evaluation system results, and formulate different water level control strategies according to the different life stages and ecological needs of fish.

[0059] Step 4: Adaptive regulation: In response to the special habitat of the plateau region, establish a glacier mass balance model, predict snowmelt water volume, adjust reservoir operation strategy, and control the vertical mixing intensity of water body through deep hole-surface hole combined discharge to address dissolved oxygen stratification.

[0060] Both the ecological scheduling plan in step three and the adaptive regulation in step four involve adjustments to the reservoir's operational strategy. Step four addresses adjustments to the reservoir's operational strategy, particularly in high-altitude regions. The ecological scheduling plan, on the other hand, formulates specific water level control strategies based on the needs of aquatic organisms such as fish, aiming to achieve a balance between reservoir operation and ecological environmental protection. The ecological scheduling plan forms the basis for the reservoir operational strategy in step four.

[0061] The water body's ecological hydraulic characteristics include water level, flow rate, flow velocity, water temperature, and water quality. These parameters are acquired by deploying a multi-parameter water quality sensor array and using an ADCP flow measurement system to construct an automatic cross-sectional flow monitoring station. Simultaneously, a water level-flow relationship curve library is established to better understand the water body's hydraulic characteristics.

[0062] The basic data on fish ecological habits include life history characteristics, geographical distribution characteristics, resource quantity, conservation necessity, dam construction impact, economic value, and spawning type. Environmental DNA (eDNA) technology is used to monitor fish population dynamics and obtain fish ecological habit data. At the same time, a three-dimensional spawning ground characteristic model is established to understand the hydraulic conditions required for fish spawning.

[0063] The specific steps for constructing a habitat suitability evaluation index system by utilizing the correlation between fish spawning and reproduction and river hydraulic elements are as follows:

[0064] Determining ecological factors and their weights: Four ecological factors, namely flow velocity, water level, water temperature and dissolved oxygen, were selected as evaluation indicators. The weights of these ecological factors were determined through statistical analysis.

[0065] Establish fuzzy sets: For each ecological factor, establish a corresponding fuzzy set. The fuzzy set is used to describe the different states or levels of the ecological factor, such as "high", "medium", "low", etc. These states or levels can be further subdivided and defined according to the actual situation;

[0066] Determine the membership function: Determine a triangular membership function for each fuzzy set. The triangular membership function is used to describe the degree to which the evaluation factor belongs to a certain fuzzy set.

[0067] Constructing a fuzzy comprehensive evaluation model: Based on fuzzy sets and triangular membership functions, a fuzzy comprehensive evaluation model is constructed. The fuzzy comprehensive evaluation model comprehensively considers the influence of multiple ecological factors to comprehensively evaluate the habitat suitability of fish habitats.

[0068] Calculate the fuzzy comprehensive evaluation results: Use the fuzzy comprehensive evaluation model to calculate the fuzzy comprehensive evaluation results for each evaluation object.

[0069] The specific steps for constructing the hydrodynamic model of the reservoir tail section based on Bernoulli's equation for steady non-uniform flow are as follows:

[0070] Determining the Bernoulli equation for steady nonuniform flow: The Bernoulli equation for steady nonuniform flow is the fundamental principle describing the conservation of energy in fluids during steady nonuniform flow. The Bernoulli equation for steady nonuniform flow is expressed as:

[0071]

[0072] Application to the tail section of the reservoir: In the hydrodynamic model of the tail section, it is assumed that the fluid is incompressible and the density ρ is constant. For open channels or rivers, the fluid pressure term is ignored because its influence is relatively small compared to gravitational potential energy and kinetic energy. The Bernoulli equation is further simplified as follows:

[0073]

[0074] In equations (1) and (2), z1 and z2 are the heights (water levels) at two locations upstream and downstream of the reservoir tail section, respectively; p1 and p2 are the fluid pressures at these two locations, respectively; ρ is the fluid density; g is the acceleration due to gravity; v1 and v2 are the flow velocities at these two locations, respectively; and h is the flow velocity at these two locations, respectively. f This is due to head loss caused by frictional resistance;

[0075] Constructing a hydrodynamic model: Collect geometric feature data of the river section (such as length, width, depth, etc.), measure the flow velocity distribution, and use the Manning formula to estimate h. fThe estimation formula is expressed as:

[0076]

[0077] Where n is the Manning roughness coefficient, Q is the flow rate, A is the cross-sectional area, and R is the hydraulic radius. Where P is the wetted perimeter. For a rectangular channel, the wetted perimeter P = 2W + 2D, where W is the channel width and D is the channel depth. The simplified Bernoulli equation is solved using the finite element method to obtain the water level changes of each section in the river segment under different inflow and operating water level conditions.

[0078] The process of establishing a hydraulic-ecological response surface to determine the three-dimensional flow velocity threshold, water temperature window, and dissolved oxygen safety threshold for different fish species during critical developmental stages includes the following steps:

[0079] Constructing a hydraulic-ecological response model: The collected data are used to construct a hydraulic-ecological response model using support vector machines. This model should be able to reflect the relationship between hydraulic characteristics and fish ecological habits. Hydraulic characteristics such as flow velocity, water temperature, and dissolved oxygen are used as independent variables, and key developmental periods of fish (such as spawning, hatching, and juvenile growth) are used as dependent variables.

[0080] Determine the response surface: Through model calculation, the response of fish during key developmental periods under different combinations of hydraulic characteristics is obtained, and the calculation results are visualized as a three-dimensional response surface using the Response Surface Methodology (RSM).

[0081] Determine the thresholds: Based on the response surface, determine the three-dimensional flow velocity thresholds, water temperature windows, and dissolved oxygen safety thresholds for different fish species during their critical developmental stages.

[0082] The construction of a hybrid forecasting system combining the XAJ distributed hydrological model and LSTM neural network, and the pre-control of reservoir operating water levels based on the evaluation system results, along with the formulation of different water level control strategies according to the different life stages and ecological needs of fish, includes the following steps:

[0083] Constructing the XAJ Distributed Hydrological Model: The XAJ distributed hydrological model is a physics-based hydrological model that can simulate rainfall-runoff processes within a watershed. This model predicts the flow at the watershed outlet by considering factors such as topography, vegetation, and soil within the watershed, as well as the spatial and temporal distribution of rainfall. It collects rainfall, topography, vegetation, and soil data within the watershed and preprocesses them. Based on the actual conditions of the watershed, it sets the model parameters, such as soil permeability coefficient and vegetation transpiration coefficient. The input data is then fed into the model, and the model is run to obtain the predicted flow at the watershed outlet.

[0084] Building an LSTM neural network: LSTM (Long Short-Term Memory) neural network is a special type of recurrent neural network that can handle long-term dependencies in time series data. The structure of the LSTM neural network is designed, including an input layer, a hidden layer, and an output layer. The number of LSTM units in the hidden layer needs to be adjusted according to the actual situation. The LSTM neural network is trained using historical traffic data to obtain the optimal model parameters.

[0085] Hybrid forecast model construction: The XAJ distributed hydrological model and LSTM neural network are combined to construct a hybrid forecast model. The flow data output by the hydrological model is normalized and used as the input of the LSTM neural network to construct the hybrid forecast model. The hybrid forecast model is validated using a validation dataset to evaluate the predictive performance of the model.

[0086] Preemptive control of reservoir operation: A hybrid forecasting model is used to predict future flow changes. Based on the predicted flow changes, combined with the reservoir's storage and release capacity, future water level changes are calculated using the following formula:

[0087]

[0088] Where ΔH represents the change in water level, ΔV represents the change in reservoir storage, and A represents the surface area of ​​the reservoir.

[0089] Develop water level control strategies: During the non-spawning period of fish, maintain a steady-state mode to keep the water level relatively stable and avoid drastic fluctuations that could adversely affect the fish's ecological environment. During the spawning period of fish, when the water temperature reaches the standard, use a pulsed peak-forming mode to increase the amount of water released from the reservoir in a short period of time to create a water flow peak, simulating the flood pulse phenomenon in natural rivers. This water flow change can stimulate fish spawning and improve the reproductive success rate. During the fish egg hatching period, use a slow-release maintenance mode to gradually lower the water level of the reservoir to reduce the scouring and damage to the fish eggs.

[0090] The glacier mass balance model is expressed as follows:

[0091] Q melt =k·A snow (T max -T threshold ) 1.8 (5)

[0092] Where k is the topographic-climate coupling coefficient (0.05–0.15 mm / ℃·d), calibrated by inversion from historical snowmelt data, A snow It is the effective snow cover area (km2), interpreted using MODIS / Terra satellite snow cover product (500m resolution), T maxThe highest temperature of the day (°C) is predicted based on meteorological station observations and the WRF downscaling model. threshold This is the snow melting start temperature (usually taken as 0-1℃, but needs to be corrected based on actual observations in high-altitude areas).

[0093] The proposed method for controlling dissolved oxygen stratification utilizes a combined deep-hole and surface-hole discharge strategy to manage vertical mixing intensity. This strategy includes surface-hole discharge, deep-hole discharge, and mixing intensity control. Surface-hole discharge releases oxygen-rich surface water (DO ≥ 6 mg / L) to increase the oxygen content of downstream water. Deep-hole discharge extracts low-temperature water (4–8°C) from the bottom layer and aerates it through the spillway (increasing DO by 2–3 mg / L). Mixing intensity control adjusts the opening ratio of the two holes to ensure uniform vertical mixing of the water at the outlet. The aeration efficiency model is expressed as follows:

[0094]

[0095] Among them, H fall Let Q be the discharge drop height (m), and Q be the single-hole flow rate (m3 / s).

[0096] The following conclusions are drawn based on the combined deep-hole and surface-hole drainage strategy.

[0097] Gate combination Traffic allocation ratio Water mixing degree Applicable Scenarios The aperture is fully open. 70% surface flow Low Egg spawning stimulation Central hole dominant 50% mid-laminar flow middle Maintaining feeding grounds for juvenile fish Bottom hole fine adjustment 30% of the underlying flow high Insulation during winter

Claims

1. An ecological scheduling method for fish habitat protection, characterized in that, Includes the following steps: Step 1: Basic Data Collection: Obtain basic data on the ecological and hydraulic characteristics of aquatic waters and the ecological habits of fish; Step 2: Constructing the evaluation system: The evaluation system includes a habitat suitability evaluation index system and a hydrodynamic model; among them, the habitat suitability evaluation index system is constructed by utilizing the correlation between fish spawning and reproduction and river hydraulic elements; a hydrodynamic model of the reservoir tail section is constructed based on the Bernoulli equation for stable non-uniform flow; at the same time, a hydraulic-ecological response surface is established to determine the three-dimensional flow velocity threshold, water temperature window, and dissolved oxygen safety threshold for different fish species during key developmental stages; Step 3: Develop an ecological scheduling plan: Construct a hybrid forecasting system combining the XAJ distributed hydrological model and LSTM neural network, control the reservoir's operating water level in advance based on the evaluation system results, and formulate different water level control strategies according to the different life stages and ecological needs of fish. Step 4: Adaptive regulation: In response to the special habitat of the plateau region, establish a glacier mass balance model, predict snowmelt water volume, adjust reservoir operation strategy, and control the vertical mixing intensity of water body through deep hole-surface hole combined discharge to address dissolved oxygen stratification.

2. The ecological scheduling method for fish habitat protection as described in claim 1, characterized in that: The water ecological hydraulic characteristic parameters in step one include water level, flow rate, flow velocity, water temperature, and water quality. These water ecological hydraulic characteristic parameters are obtained by deploying a multi-parameter water quality sensor array and constructing an automatic cross-sectional flow monitoring station using an ADCP flow measurement system. At the same time, a water level-flow relationship curve library is established.

3. The ecological scheduling method for fish habitat protection as described in claim 1, characterized in that: The basic data on fish ecological habits in step one include life history characteristics, geographical distribution characteristics, resource quantity, conservation necessity, dam construction impact, economic value, and spawning type. Environmental DNA technology is used to monitor fish population dynamics and obtain fish ecological habit data. At the same time, a three-dimensional spawning ground characteristic model is established to understand the hydraulic conditions required for fish spawning.

4. The ecological scheduling method for fish habitat protection as described in claim 1, characterized in that: The specific steps in step two of constructing a habitat suitability evaluation index system based on the correlation between fish spawning and reproduction and river hydraulic factors are as follows: Determining ecological factors and their weights: Four ecological factors, namely flow velocity, water level, water temperature and dissolved oxygen, were selected as evaluation indicators. The weights of these ecological factors were determined through statistical analysis. Establish fuzzy sets: For each ecological factor, establish a corresponding fuzzy set. The fuzzy set is used to describe the different states or levels of the ecological factor. Determine the membership function: Determine a triangular membership function for each fuzzy set. The triangular membership function is used to describe the degree to which the evaluation factor belongs to a certain fuzzy set. Constructing a fuzzy comprehensive evaluation model: Based on fuzzy sets and triangular membership functions, a fuzzy comprehensive evaluation model is constructed. The fuzzy comprehensive evaluation model comprehensively considers the influence of multiple ecological factors to comprehensively evaluate the habitat suitability of fish habitats. Calculate the fuzzy comprehensive evaluation results: Use the fuzzy comprehensive evaluation model to calculate the fuzzy comprehensive evaluation results for each evaluation object.

5. The ecological scheduling method for fish habitat protection as described in claim 1, characterized in that: The specific steps for constructing the hydrodynamic model of the reservoir tail section based on the Bernoulli equation for steady non-uniform flow in step two are as follows: Determine the Bernoulli equation for steady nonuniform flow: The Bernoulli equation for steady nonuniform flow is expressed as: Application to the tail end of the reservoir: In the hydrodynamic model of the tail end of the reservoir, it is assumed that the fluid is incompressible and the density ρ is constant. For open channels or rivers, the fluid pressure term is ignored, and the Bernoulli equation is further simplified to: In equations (1) and (2), z1 and z2 are the heights at two different locations in the reservoir tail section, p1 and p2 are the fluid pressures at these two locations, ρ is the fluid density, g is the acceleration due to gravity, v1 and v2 are the flow velocities at these two locations, and h is the flow velocity. f This is due to head loss caused by frictional resistance.

6. The ecological scheduling method for fish habitat protection as described in claim 5, characterized in that: The construction of the hydrodynamic model involves: collecting geometric feature data of the river section, measuring the velocity distribution, and using the Manning formula to estimate h. f The estimation formula is expressed as: Where n is the Manning roughness coefficient, Q is the flow rate, A is the cross-sectional area, and R is the hydraulic radius. Where P is the wetted perimeter. For a rectangular channel, the wetted perimeter P = 2W + 2D, where W is the channel width and D is the channel depth. The simplified Bernoulli equation is solved using the finite element method to obtain the water level changes of each section in the river segment under different inflow and operating water level conditions.

7. The ecological scheduling method for fish habitat protection as described in claim 1, characterized in that: Step two involves establishing a hydraulic-ecological response surface to determine the three-dimensional flow velocity threshold, water temperature window, and dissolved oxygen safety threshold for different fish species during critical developmental stages. This includes the following steps: Constructing a hydraulic-ecological response model: The collected data were used to construct a hydraulic-ecological response model using a support vector machine, with hydraulic characteristics as the independent variable and the key developmental period of fish as the dependent variable; Determine the response surface: Through model calculation, the response of fish during key developmental periods under different combinations of hydraulic characteristics is obtained, and the calculation results are visualized as a three-dimensional response surface using the response surface method; Determine the thresholds: Based on the response surface, determine the three-dimensional flow velocity thresholds, water temperature windows, and dissolved oxygen safety thresholds for different fish species during their critical developmental stages.

8. The ecological scheduling method for fish habitat protection as described in claim 1, characterized in that: Step three involves constructing a hybrid forecasting system combining the XAJ distributed hydrological model and LSTM neural network. Based on the evaluation results, the reservoir's operating water level is controlled in advance. Different water level control strategies are formulated according to the different life stages and ecological needs of fish species. This includes the following steps: Constructing an XAJ distributed hydrological model: Collect rainfall, topography, vegetation, and soil data within the watershed and preprocess them. Based on the actual conditions of the watershed, set the model parameters, input the data into the model, and run the model to obtain the predicted flow at the watershed outlet. Building an LSTM neural network: Design the structure of the LSTM neural network, including the input layer, hidden layer and output layer, and train the LSTM neural network using historical traffic data to obtain the optimal model parameters; Hybrid forecast model construction: The XAJ distributed hydrological model and LSTM neural network are combined to construct a hybrid forecast model. The flow data output by the hydrological model is normalized and used as the input of the LSTM neural network to construct the hybrid forecast model. The hybrid forecast model is validated using a validation dataset to evaluate the predictive performance of the model. Preemptive control of reservoir operation: A hybrid forecasting model is used to predict future flow changes. Based on the predicted flow changes, combined with the reservoir's storage and release capacity, future water level changes are calculated using the following formula: Where ΔH represents the change in water level, ΔV represents the change in reservoir storage, and A represents the surface area of ​​the reservoir. Develop water level control strategies: During the non-spawning period of fish, maintain a steady-state mode to keep the water level relatively stable; during the spawning period of fish, when the water temperature reaches the standard, use a pulsed peak-forming mode to increase the water release from the reservoir in a short period of time to form a water flow peak, simulating the flood pulse phenomenon in natural rivers; during the fish egg hatching period, use a slow-release maintenance mode to slowly lower the water level of the reservoir.

9. An ecological scheduling method for fish habitat protection as described in claim 1, characterized in that: The glacier mass balance model described in step four is expressed as follows: Q melt =k·A snow (T max -T threshold ) 1.8 (5) Where k is the topography-climate coupling coefficient, calibrated by inversion from historical snowmelt data, and A snow It is the effective snow cover area, T max The highest temperature of the day, T threshold This is the temperature at which snow melting begins.

10. An ecological scheduling method for fish habitat protection as described in claim 1, characterized in that: In step four, dissolved oxygen stratification is addressed by controlling the vertical mixing intensity of the water body through a combined deep-hole and surface-hole discharge strategy. This strategy includes surface-hole discharge, deep-hole discharge, and mixing intensity control. Surface-hole discharge releases oxygen-rich surface water to increase the oxygen content of the downstream water body. Deep-hole discharge extracts low-temperature water from the bottom layer and aerates it through the spillway. The mixing intensity control adjusts the opening ratio of the two holes to ensure the vertical mixing uniformity of the water at the outlet. The aeration efficiency model is expressed as: Among them, H fall Q represents the discharge drop height, and Q represents the single-orifice flow rate.

Images