An adaptive river ecological flow calculation method and system

Abstract

The application provides a kind of adaptive river ecological flow calculation method and system, it is related to image processing technical field, method includes: obtaining remote sensing image;Extract the ground feature elements in remote sensing image;Through Geohash coding algorithm, the geographic coordinates of ground feature elements are converted into binary Geohash code;Through Geohash coding and deep neural network nesting, construct GeohashNet network model;Through GeohashNet network model, adaptively extract the river water surface width in remote sensing image;According to the river water surface width, calculate the catchment area of different sections;According to the catchment area of different sections, calculate the ecological flow of different sections.In the present application, through Geohash coding and deep neural network nesting, considering that numerous ground features have the same regional characteristics on a particular spatial scale, so that the deep neural network can fully learn the complex spatial features, improve the accuracy of image ground feature extraction, and further improve the accuracy of river ecological flow calculation.

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, and in particular to a method and system for calculating adaptive river ecological flow. Background Art

[0002] Calculating river ecological flow is of great significance for ecology, the environment, and water resource management. Currently, this data is primarily obtained through manual measurement, which is time-consuming, labor-intensive, and inefficient. Compared to manual measurement, the application of drone technology to determine river ecological flow reduces human, financial, and material costs. Drone imagery can be used to obtain hydraulic parameters such as river width, gradient, and roughness, which can then be converted into river ecological flow.

[0003] Currently, when extracting features from remote sensing images, the images are typically divided into regular, small-sized blocks, which are then fed individually into deep neural networks (DNNs) for prediction. From a geographic location perspective, the underlying assumption is that these images are independent of each other and their location.

[0004] However, it is well known that many surface features share common regional characteristics at a certain spatial scale. In particular, Waldo Tobler stated in his First Law of Geography: "Everything is related to everything else, but near things are more related than distant things." Remote sensing imagery is an observational representation of surface features, and therefore, the semantic labeling of images also follows this First Law. This means that similar remote sensing images share certain underlying patterns, while images farther apart differ more significantly. Therefore, the current method of dividing images into regular, small-sized image blocks and then feeding each into a deep neural network for prediction is overly idealistic, ignoring the common regional characteristics of many surface features at a specific spatial scale. This results in low accuracy in image feature extraction, and consequently, low accuracy in river ecological flow calculations. Summary of the Invention

[0005] In order to solve the technical problem that the current method of dividing images into regular small-sized image blocks and then sending them individually into deep neural networks for prediction is too idealistic, ignoring that many surface objects have the same regional characteristics at a specific spatial scale, resulting in low accuracy in image object element extraction and thus low accuracy in river ecological flow calculation, the present invention provides an adaptive river ecological flow calculation method and system.

[0006] The technical solutions provided by the present invention are as follows:

[0007] First aspect

[0008] The present invention provides an adaptive river ecological flow calculation method, comprising:

[0009] S1: Acquire remote sensing images;

[0010] S2: extracting ground features from the remote sensing image;

[0011] S3: converting the geographic coordinates of the land feature into binary Geohash codes through a Geohash coding algorithm;

[0012] S4: Constructing a GeohashNet network model by nesting the Geohash code with a deep neural network;

[0013] S5: Adaptively extracting the river surface width in the remote sensing image through the GeohashNet network model;

[0014] S6: Calculating the catchment area of ​​different sections according to the width of the river surface;

[0015] S7: Calculate the ecological flow of different sections based on the catchment area of ​​different sections.

[0016] Second aspect

[0017] The present invention provides an adaptive river ecological flow calculation system, comprising: a processor and a memory for storing processor executable instructions; the processor is configured to call the instructions stored in the memory to execute the adaptive river ecological flow calculation method described in the first aspect.

[0018] Compared with the prior art, the above technical solution has at least the following beneficial effects:

[0019] In the present invention, the geographic coordinates of the land features extracted from the remote sensing image are converted into binary Geohash codes through the Geohash coding algorithm, and then the binary Geohash codes are embedded into the deep neural network. Considering that many surface features have the same regional characteristics at a specific spatial scale, the deep neural network can fully learn complex spatial features. There is no need to divide the remote sensing image into regular small-size image blocks and send them into the deep neural network separately, which improves the accuracy of image land feature extraction and thus improves the accuracy of river ecological flow calculation. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0021] Figure 1 A schematic flow chart of an adaptive river ecological flow calculation method provided by the present invention;

[0022] Figure 2 This is a structural schematic diagram of an adaptive river ecological flow calculation system provided by the present invention. DETAILED DESCRIPTION

[0023] To make the purpose, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions of the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the described embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0024] Unless otherwise defined, the technical or scientific terms used in the present invention shall have the usual meaning understood by persons of ordinary skill in the field to which the present invention belongs. The words "first", "second" and similar terms used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. Similarly, words such as "one", "an" or "the" do not indicate a quantity limitation, but rather indicate the presence of at least one. Words such as "include" or "comprise" mean that the elements or objects preceding the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. Words such as "connect" or "connected" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect.

[0025] It should be noted that the terms "up", "down", "left", "right", "front" and "back" used in the present invention are only used to indicate relative position relationships. When the absolute position of the object being described changes, the relative position relationship may also change accordingly.

[0026] Reference Manual Figure 1 , which shows a flow chart of an adaptive river ecological flow calculation method provided by the present invention.

[0027] An embodiment of the present invention provides an adaptive river ecological flow calculation method, comprising:

[0028] S1: Acquire remote sensing images.

[0029] Specifically, remote sensing images can be obtained by using drones equipped with cameras, and remote sensing images can also be obtained from satellite data.

[0030] S2: Extract ground features from remote sensing images.

[0031] Specifically, appropriate bands can be selected for feature extraction. Using these bands, image processing techniques are used to extract the features of the features. These features may include texture, shape, color, and more. Common feature extraction methods include gray-level co-occurrence matrix, principal component analysis, and histogram equalization. Subsequently, machine learning or deep learning methods are used to classify the image, assigning pixels in the image to different feature categories. Common classification methods include support vector machines (SVMs), random forests, and convolutional neural networks (CNNs).

[0032] S3: The geographic coordinates of the features are converted into binary Geohash codes through the Geohash coding algorithm.

[0033] Geohash encoding is a special geospatial indexing method that converts longitude and latitude into a series of character strings. It consists of two stages: first, converting longitude and latitude into binary encoding, and then converting them into character string encoding. The first encoding stage is achieved by performing a spatial binary division along the longitude and latitude axes. The algorithm recursively subdivides the longitude and latitude space into smaller and smaller grids until the required spatial accuracy is met. Therefore, the spatial division operation can produce encodings of arbitrary spatial accuracy. To clearly convey the correct meaning of the method, the codes generated in the first stage of the Geohash method are called binary Geohash codes. It is worth noting that different encoding accuracy will lead to different results in feature extraction. Binary Geohash codes provide the ability to flexibly control spatial accuracy. Through multi-scale hierarchical encoding, features within the same scale range have the same geocode, an advantage that decimal encoding does not have.

[0034] Furthermore, binary Geohash encoding allows for infinite partitioning of longitude and latitude space, allowing for arbitrary accuracy. As shown in the figure above, the first binary Geohash code is 1, and the red dot falls within the latitude interval [0°, 90°]. If the true latitude of the location is guessed to be 45°, the error range for a 1-bit latitude encoding is [-45°, 45°]. Note that the ±45° error represents the maximum error range, not the standard deviation. With more bits in the encoding, the error can be significantly reduced. As shown in the table below, each additional bit reduces the positioning error by approximately half.

[0035] In the present invention, in the context of extracting geographical features and calculating river ecological flow, Geohash coding helps to convert geographic location information into a processable form and embed it in the deep learning model to better learn and understand complex spatial features.

[0036] S4: Construct the GeohashNet network model by nesting Geohash encoding with deep neural network.

[0037] Specifically, Geohash coding can be nested with the feature space, parameter space and / or residual correction of deep neural networks to enhance the ability of deep neural networks to learn spatial features and improve the accuracy of the GeohashNet network model in extracting river surface width from remote sensing images.

[0038] Furthermore, within the feature space embedding, Geohash codes can be connected or concatenated with the feature space of a deep neural network. This can be achieved by concatenating the Geohash codes with the original features of the remote sensing image at a certain level, or by inputting the Geohash codes into neurons in a certain layer, blending them with the original features to form a new feature representation. This helps the model better understand geographic information and improves the accuracy of remote sensing image processing.

[0039] Furthermore, within the parameter space embedding, Geohash encoding can be connected to the weights and biases of a deep neural network. This means that the information in the Geohash encoding is used to adjust the weights and biases in the neural network to adapt to the spatial characteristics of the location. This approach allows the model to better consider location information during training.

[0040] Furthermore, in residual correction embedding, Geohash coding can be connected to the residual term of a deep neural network. This allows the information of the Geohash coding to be used to correct the residual term of each layer in the neural network to better capture subtle changes in spatial features.

[0041] In a possible implementation, S4 specifically includes: constructing a GeohashNet network model by nesting Geohash coding with the feature space of a deep neural network.

[0042] The specific way to embed Geohash coding into the feature space of deep neural network is as follows:

[0043] y=f(w g [x,g]+b)

[0044] Among them, y represents the output vector of the current layer, f() represents the activation function, and w g represents the feature embedding weight matrix, x represents the output vector of the previous layer, g represents the Geohash encoding vector, and b represents the bias term.

[0045] It should be noted that the feature embedding weight matrix w g It plays a key role in the model, combining the Geohash code with the output vector x and the bias term b of the previous layer. This combination allows the network to more flexibly learn the spatial characteristics of different geographic locations, including the distribution and shape of surface features, thereby improving its ability to model the geographic environment.

[0046] In this paper, Geohash encoding is a method for mapping geographic coordinates into binary codes. By embedding this code within the feature space of a deep neural network, it can fully utilize geographic location information, helping the model better understand the surface features of different locations and improving its perception of geographic spatial structure. Furthermore, the GeohashNet network model achieves adaptability to different locations by embedding geographic location information within the feature space of the neural network. This adaptability enables the model to better adapt to the topography and features of different regions, enhancing the model's generalization capabilities.

[0047] In a possible implementation, S4 specifically includes: constructing a GeohashNet network model by nesting Geohash coding with the parameter space of a deep neural network.

[0048] The specific way to embed Geohash coding into the parameter space of deep neural networks is as follows:

[0049] w=f(w p g T +b p )

[0050] y=f(w[x,g]+b)

[0051] Among them, w represents the weight matrix, f() represents the activation function, w prepresents the parameter nested weight matrix, g represents the Geohash encoding vector, (·) T represents the matrix bias, b p Represents the parameter nested bias term, y represents the output vector of the current layer, x represents the output vector of the previous layer, and b represents the bias term.

[0052] It should be noted that the parameter nested weight matrix w is used p , through the Geohash encoding vector g and matrix bias b p To dynamically adjust the weight matrix w, learn the weight vector w from the Geohash encoding vector g, and replace the weight vector in the linear transformation of the equational neural network, the model can make different degrees of weight adjustments to the input data of different geographical locations, so as to better adapt to the characteristics of land features in different regions.

[0053] In this invention, the adaptability of the parameter space can reduce the need for manual adjustments to neural network parameters. Through Geohash encoding and dynamic adjustment of the parameter space, the model can more flexibly learn geographic location information, reducing the burden of manual parameter adjustment in different geographical environments. Furthermore, the adaptability of geographic location information helps improve the model's generalization performance, making it more robust when applied to remote sensing imagery data from different regions. This is crucial for processing spatially varying data, such as different regions in river ecological flow calculations.

[0054] In a possible implementation, S4 specifically includes: nesting Geohash coding and residual correction of a deep neural network to construct a GeohashNet network model.

[0055] Among them, the specific way of nesting Geohash coding and residual correction of deep neural network is:

[0056] Δ=w r [x,g] T +b r

[0057] y residual =f(wx+b+Δ)

[0058] Where Δ represents the residual correction term, w r represents the residual correction weight matrix, x represents the output vector of the previous layer, g represents the Geohash encoding vector, [·] T represents the matrix bias, b r Represents the residual correction bias term, y residual Represents the residual output vector, f() represents the activation function, w represents the weight matrix, and b represents the bias term.

[0059] It is important to note that the residual correction mechanism allows the network to learn the residual (i.e., the difference between the output of the previous layer and the target output of the current layer), thereby facilitating the construction of deeper network structures. This helps improve the model's ability to model complex ground features and spatial characteristics. The residual correction structure helps alleviate the vanishing gradient problem, especially in deep networks. By introducing the residual term, gradients can be more easily propagated and the network can learn more complex feature representations.

[0060] In the present invention, geographic location information is introduced through Geohash coding and combined with the residual correction structure, so that the network can more effectively handle geographic space changes and more accurately extract geographic features such as river surface width.

[0061] Furthermore, the Adam optimization method is used during model training. Cyclic learning rate and cosine annealing schedule are used to accelerate the convergence of the model. Based on the backup of network weights at the end of each cycle, the snapshot ensembling method can further improve the accuracy of the model. The maximum learning rate of the cosine annealing schedule is 1×10 -3 , the minimum learning rate is 1×10 -6 .

[0062] Data augmentations were used to avoid overfitting, including random flipping in horizontal and vertical directions, random rotation, and random brightness jittering. Each image was preprocessed by subtracting 128 from the original pixel value and then dividing by 128. The weight decay regularization value was 1×10 -5 Batch Normalization (BN) is applied before each ReLU activation layer. This network model combines the U-Net and NASNet-Mobile models, where the NASNet model is obtained through neural architecture searching. This model is called the U-NASNetMobile model, and the U-NASNetMobile model with Geohash encoding is called GeohashNet. Compared with other DenseASPP and DeepLabv3 models, the U-NASNetMobile model has higher computational efficiency and takes up less GPU memory.

[0063] S5: Adaptively extract the river surface width from remote sensing images through the GeohashNet network model.

[0064] Specifically, after the GeohashNet network model training is completed, the new remote sensing image is input into the GeohashNet network model. The model will adaptively extract the width of the river surface and use its learning of geographic location information to identify the position and width of the river in the image.

[0065] S6: Calculate the catchment area of ​​different sections based on the width of the river surface.

[0066] Catchment area refers to the total area upstream of a given river or waterway into which all rainfall flows. This concept is a key parameter in hydrology, with important implications for hydrological models, flood forecasting, and water resources management.

[0067] Specifically, using the measured river channel width data combined with the channel geometry, the water surface area of ​​each cross-section can be calculated. This usually involves integrating the water surface width or other geometric calculation methods.

[0068] S7: Calculate the ecological flow of different sections based on the catchment area of ​​different sections.

[0069] In a possible implementation, S7 specifically includes calculating the ecological flow of different sections according to the following formula:

[0070]

[0071] Among them, R i_eco represents the ecological flow of section i, S i represents the catchment area of ​​section i, S eco represents the catchment area of ​​the reference section, r eco Represents the ecological flow of the reference section.

[0072] In the present invention, by using the ecological flow of the reference section as a benchmark, a relatively consistent indicator is provided for the evaluation of the ecological flow, which helps to ensure that a certain ecological flow level is maintained in the entire river or water area, which is beneficial to the stability of the ecosystem.

[0073] Compared with the prior art, the above technical solution has at least the following beneficial effects:

[0074] In the present invention, the geographic coordinates of the land features extracted from the remote sensing image are converted into binary Geohash codes through the Geohash coding algorithm, and then the binary Geohash codes are embedded into the deep neural network. Considering that many surface features have the same regional characteristics at a specific spatial scale, the deep neural network can fully learn complex spatial features. There is no need to divide the remote sensing image into regular small-size image blocks and send them into the deep neural network separately, which improves the accuracy of image land feature extraction and thus improves the accuracy of river ecological flow calculation.

[0075] In a possible implementation, after S7, the adaptive river ecological flow calculation method further includes:

[0076] S8: Obtain the biomass, density and diversity index of benthic animals in the river.

[0077] It's important to note that benthic animals are a crucial component of aquatic ecosystems, critically impacting water quality and ecological balance. Obtaining information on their biomass, density, and diversity provides a more comprehensive understanding of the health of river ecosystems.

[0078] S9: Calculate ecological indicator parameters based on the biomass, density and diversity index of benthic animals.

[0079] In a possible implementation, S9 specifically includes calculating an ecological indicator parameter according to the following formula:

[0080] S=λ1c+λ2ρ+λ3M

[0081] Among them, S represents the ecological indicator parameter, c represents the biomass of benthic animals, λ1 represents the weight coefficient of biomass, ρ represents the density of benthic animals, λ2 represents the weight coefficient of density, M represents the diversity index of benthic animals, and λ3 represents the weight coefficient of diversity index.

[0082] Among them, those skilled in the art can set the sizes of the biomass weight coefficient λ1, the density weight coefficient λ2, and the diversity index weight coefficient λ3 according to actual conditions, and the present invention does not limit them.

[0083] It should be noted that by calculating ecological indicators related to benthic fauna, more accurate indicators of ecosystem status can be obtained. Ecological indicators calculated based on benthic fauna data can be used to develop more effective river management strategies.

[0084] In a possible implementation, the weight coefficients of the biomass, density, and diversity index of benthic animals are determined with the goal of maximizing the correlation coefficient between the ecological indicator parameters and the ecological flow.

[0085] In this paper, by optimizing the weight coefficients, the correlation coefficient between ecological indicator parameters and ecological flows is maximized, thereby ensuring the strongest correlation between ecological indicator parameters and actual ecological flows. This helps to establish a more reliable model and improve the accuracy of ecosystem health assessment. Furthermore, by targeting maximum correlation, the determination of weight coefficients is more data-driven. This method can find the optimal weight distribution based on actual observational data through statistical methods, improving the fit between ecological indicator parameters and measured data.

[0086] In one possible implementation, the correlation coefficient between the ecological indicator parameter and the ecological flow is calculated as follows:

[0087]

[0088] Among them, r R,S represents the correlation coefficient between the ecological indicator parameter and the ecological flow, R represents the ecological flow, S represents the ecological indicator parameter, and n represents the total number of data points.

[0089] S10: Carry out river management based on ecological indicator parameters.

[0090] Specifically, we can analyze the changing trends of ecological indicator parameters and identify any unusual fluctuations or trends. This can lead to the identification of issues, such as water pollution, changes in benthic animal populations, and ecosystem disturbances. Understanding the root causes of these issues allows us to develop appropriate management strategies based on this understanding.

[0091] Alternatively, treatment rules can be established in advance. By setting reasonable target values ​​for ecological indicators and regularly tracking changes in these indicators, managers can assess the improvement or deterioration of the ecosystem and adjust management strategies to achieve the expected ecological benefits.

[0092] In this paper, ecological indicators are comprehensive measures of ecosystem health. By monitoring and analyzing these parameters, we can gain a more comprehensive understanding of the health of river ecosystems. These indicators can be used to identify potential problems and trends, allowing for early implementation of appropriate management measures.

[0093] Reference Manual Figure 2 , which shows a structural schematic diagram of an adaptive river ecological flow calculation system provided by the present invention.

[0094] The present invention also provides an adaptive river ecological flow calculation system 20, comprising: a processor 201 and a memory 202 for storing instructions executable by the processor 201. The processor 201 is configured to call the instructions stored in the memory 202 to execute the above-mentioned adaptive river ecological flow calculation method.

[0095] The adaptive river ecological flow calculation system 20 provided by the present invention can execute the above-mentioned adaptive river ecological flow calculation method and achieve the same or similar technical effects. To avoid repetition, the present invention will not elaborate on it.

[0096] Compared with the prior art, the above technical solution has at least the following beneficial effects:

[0097] In the present invention, the geographic coordinates of the land features extracted from the remote sensing image are converted into binary Geohash codes through the Geohash coding algorithm, and then the binary Geohash codes are embedded into the deep neural network. Considering that many surface features have the same regional characteristics at a specific spatial scale, the deep neural network can fully learn complex spatial features. There is no need to divide the remote sensing image into regular small-size image blocks and send them into the deep neural network separately, which improves the accuracy of image land feature extraction and thus improves the accuracy of river ecological flow calculation.

[0098] There are a few points to note:

[0099] (1) The drawings of the embodiments of the present invention only relate to the structures related to the embodiments of the present invention. Other structures may refer to conventional designs.

[0100] (2) For the sake of clarity, the thickness of layers or regions in the drawings used to describe the embodiments of the present invention are exaggerated or reduced, that is, these drawings are not drawn to scale. It is understood that when an element such as a layer, film, region, or substrate is referred to as being "on" or "under" another element, the element may be "directly" "on" or "under" the other element or intervening elements may be present.

[0101] (3) In the absence of conflict, the embodiments of the present invention and the features therein may be combined with each other to form new embodiments.

[0102] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. The protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. An adaptive river ecological flow calculation method, characterized in that: include: S1: Acquire remote sensing images; S2: extracting ground features from the remote sensing image; S3: converting the geographic coordinates of the land feature into binary Geohash codes through a Geohash coding algorithm; S4: Constructing a GeohashNet network model by nesting the Geohash code with a deep neural network; S5: Adaptively extracting the river surface width in the remote sensing image through the GeohashNet network model; S6: Calculating the catchment area of ​​different sections according to the width of the river surface; S7: Calculate the ecological flow of different sections based on the catchment area of ​​different sections; The S4 specifically comprises: constructing a GeohashNet network model by nesting the Geohash code with the feature space of the deep neural network; The specific way of embedding the Geohash code with the feature space of the deep neural network is as follows: y=f(w g [x,g]+b) Among them, y represents the output vector of the current layer, f() represents the activation function, and w g represents the feature embedding weight matrix, x represents the output vector of the previous layer, g represents the Geohash encoding vector, and b represents the bias term.

2. The adaptive river ecological flow calculation method according to claim 1 is characterized in that: The S4 specifically comprises: nesting the Geohash code with the parameter space of the deep neural network to construct a GeohashNet network model; The specific way of embedding the Geohash code and the parameter space of the deep neural network is as follows: w=f(w p g T +b p ) y=f(w[x,g]+b) Among them, w represents the weight matrix, f() represents the activation function, w g represents the parameter nested weight matrix, g represents the Geohash encoding vector, (·) T represents the matrix bias, b p Represents the parameter nested bias term, y represents the output vector of the current layer, x represents the output vector of the previous layer, and b represents the bias term.

3. The adaptive river ecological flow calculation method according to claim 1, characterized in that: The S4 specifically comprises: nesting the Geohash coding and the residual correction of the deep neural network to construct a GeohashNet network model; The specific way of nesting the Geohash coding and the residual correction of the deep neural network is as follows: Δ=w r [x,g] T +b r y residual =f(wx+b+Δ) Where Δ represents the residual correction term, w r represents the residual correction weight matrix, x represents the output vector of the previous layer, g represents the Geohash encoding vector, [·] T represents the matrix bias, b r Represents the residual correction bias term, y residual Represents the residual output vector, f() represents the activation function, w represents the weight matrix, and b represents the bias term.

4. The adaptive river ecological flow calculation method according to claim 1, characterized in that: The S7 is specifically: The ecological flow of different sections is calculated according to the following formula: Among them, R i_eco represents the ecological flow of section i, S i represents the catchment area of ​​section i, S eco represents the catchment area of ​​the reference section, r eco Represents the ecological flow of the reference section.

5. The adaptive river ecological flow calculation method according to claim 1, characterized in that: After S7, it also includes: S8: Obtain the biomass, density and diversity index of benthic animals in the river; S9: Calculate ecological indicator parameters based on the biomass, density and diversity index of benthic animals; S10: Perform river management according to the ecological indicator parameters.

6. The adaptive river ecological flow calculation method according to claim 5, characterized in that: The S9 is specifically: Calculate the ecological indicator parameters according to the following formula: S=λ1c+λ2ρ+λ3M Among them, S represents the ecological indicator parameter, c represents the biomass of benthic animals, λ1 represents the weight coefficient of biomass, ρ represents the density of benthic animals, λ2 represents the weight coefficient of density, M represents the diversity index of benthic animals, and λ3 represents the weight coefficient of diversity index.

7. The adaptive river ecological flow calculation method according to claim 6, characterized in that: With the goal of maximizing the correlation coefficient between the ecological indicator parameters and the ecological flow, the weight coefficients of the biomass, density and diversity index of benthic animals are determined.

8. The adaptive river ecological flow calculation method according to claim 7, characterized in that: The correlation coefficient between the ecological indicator parameter and the ecological flow is calculated as follows: Among them, r R,S represents the correlation coefficient between the ecological indicator parameter and the ecological flow, R represents the ecological flow, S represents the ecological indicator parameter, and n represents the total number of data points.

9. An adaptive river ecological flow calculation system, characterized in that: It comprises a processor and a memory for storing processor executable instructions; the processor is configured to call the instructions stored in the memory to execute the adaptive river ecological flow calculation method according to any one of claims 1 to 8.

Images