System for determining running state of energy storage system

Abstract

The invention relates to the technical field of data processing, and discloses an energy storage system operation state determination system, which comprises a data acquisition module used for acquiring temperature and remote sensing images of a reservoir according to a fixed time interval; the data screening module is used for analyzing the temperatures of the reservoir at multiple moments, determining a temperature fluctuation section of the reservoir, and selecting the remote sensing image of the reservoir at the moment corresponding to the temperature fluctuation section as an effective image; the image analysis module is used for dividing the reservoir into a plurality of equal sub-regions according to the horizontal direction, analyzing each sub-region of the reservoir in the effective image through a detection model, and determining the pollution risk of each sub-region; and the joint verification module is used for determining the feature distribution of the pollution risk of the sub-regions of the reservoir according to the position of the water outlet of the reservoir, and verifying the pollution risk of each sub-region according to the feature distribution to obtain the verification judgment of the pollution risk of each sub-region.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and more specifically, to a system for determining the operating state of an energy storage system. Background Art

[0002] During the operation of a water storage system based on a reservoir, the pollution risk of the reservoir is directly related to the operating status of the hydroelectric power generation equipment. The pollution risk includes pollutants such as silt and particles in the reservoir. The existing technology generally monitors and evaluates the pollution risk of the reservoir in the water storage system based on a reservoir manually, but the monitoring efficiency is low and the evaluation results are inaccurate. Summary of the invention

[0003] The present invention provides a system for determining the operating status of an energy storage system, which solves the technical problems raised in the background technology.

[0004] The present invention provides a system for determining the operating state of an energy storage system, comprising:

[0005] The data acquisition module is used to collect the temperature of the water reservoir at fixed time intervals, and to obtain remote sensing images of the water reservoir at fixed time intervals through monitoring equipment;

[0006] A data screening module is used to analyze the temperature of the water reservoir at multiple times, determine the temperature fluctuation section of the water reservoir, and select the remote sensing image of the water reservoir at the time corresponding to the temperature fluctuation section as a valid image;

[0007] An image analysis module is used to divide the reservoir into a number of equal sub-areas in the horizontal direction, and analyze each sub-area of ​​the reservoir in the effective image through a preset detection model to determine the pollution risk of each sub-area;

[0008] The joint verification module is used to determine the characteristic distribution of the pollution risk of the sub-areas of the reservoir according to the location of the water outlet of the reservoir, and verify the pollution risk of each sub-area according to the characteristic distribution to obtain the verification judgment of the pollution risk of each sub-area, as follows:

[0009] If the pollution risk of each sub-area is successfully determined, the operating state of the energy storage system constructed based on the reservoir is determined according to the pollution risk of each sub-area;

[0010] If the pollution risk determination of each sub-area fails, the detection model is updated in reverse according to the feature distribution.

[0011] Furthermore, the temperature of the water reservoir at multiple times is analyzed, including:

[0012] The critical Rayleigh number is calculated based on the temperature of the reservoir at multiple times. The formula is as follows:

[0013]

[0014] Among them, R a represents the critical Rayleigh number, g represents the gravitational acceleration, b represents the thermal expansion coefficient of the water in the reservoir, T b represents the temperature of the water bottom of the reservoir, T s represents the temperature of the water surface in the reservoir, h represents the height difference between the water surface and the bottom of the reservoir, v represents the kinematic viscosity of the water in the reservoir, k represents the heat diffusion efficiency of the water in the reservoir, V w represents the wind speed on the water surface of the reservoir, r w represents the wind speed weight coefficient, t represents the fixed time interval, r e The weight coefficient representing the rate of change of the water surface temperature of the reservoir;

[0015] R a >1700 is taken as the starting time of the temperature fluctuation segment, and R a The time corresponding to <1700 is taken as the end time of the temperature fluctuation section.

[0016] Furthermore, each sub-region of the reservoir in the effective image is analyzed through a preset detection model, which includes an input layer, a cutting layer, a detection layer, and an output layer:

[0017] Input layer, used to input valid images into the detection model;

[0018] The cutting layer is used to cut the effective image according to the sub-region division of the reservoir to obtain several cutting graphs;

[0019] The detection layer is used to identify each cutting graph, including:

[0020] Step 1, construct the energy functional E(M0),

[0021] Wherein, E(M0) represents the energy functional, which is used to make a smoothness assumption on the pollution-free matrix M0 to be solved, m represents the number of rows of the cutting graph, n represents the number of columns of the cutting graph, M0 represents the pollution-free matrix, M0(i,j) represents the pixel value of the i-th row and j-th column in the pollution-free matrix, M0(i+1,j) represents the pixel value of the i+1-th row and j-th column in the cutting graph, M0(i-1,j) represents the pixel value of the i-1-th row and j-th column in the cutting graph, M0(i,j+1) represents the pixel value of the i-th row and j+1-th column in the cutting graph, and M0(i,j-1) represents the pixel value of the i-th row and j-1-th column in the cutting graph;

[0022] Step 2, solving the pollution-free matrix M0 by an optimization algorithm, with the constraint that the pollution-free matrix M0 and the cutting graph have the same size;

[0023] Step 3, for the pollution-free matrix M0, when the number of update iterations of the pollution-free matrix M0 is 0, it is equal to the cutting graph;

[0024] Step 4, construct the Lagrangian function: Among them, λ represents the Lagrange multiplier, C represents the sum of the number of pixels in the cutting graph;

[0025] Step 5, update and solve the pollution-free matrix M0 to obtain the updated pollution-free matrix M0. The update formula is as follows:

[0026]

[0027] Among them, (M0) t+1 (i,j) represents the pixel value of the i-th row and j-th column updated by the pollution-free matrix M0, (M0) t (i,j) represents the pixel value of the i-th row and j-th column of the pollution-free matrix M0, represents the discrete Laplace operator, η represents the update iteration step size, and t represents the number of update iterations;

[0028] Step 6, subtract the z-th cutting graph and the updated pollution-free matrix M0 point by point to obtain the deviation matrix S, and compare the element value of any element of the deviation matrix S with the preset eigenvalue. If the element value is less than the preset eigenvalue, it is replaced with 0; if the element value is greater than the preset eigenvalue, it is replaced with 1 to obtain a replacement matrix;

[0029] Step 7, performing connectivity detection on each element in the replacement matrix to obtain the number L of connected clusters in the replacement matrix;

[0030] Step 8: Calculate the contamination risk R of the zth cutting graph totle , the calculation formula is as follows:

[0031] R totle =MLP(ω1×‖S‖ F +ω2×L); where ‖S‖ F represents the Frobenius norm of the bias matrix S, ω1 and ω2 represent the primary weight parameter and the secondary weight parameter respectively, and MLP represents the multi-layer perceptron;

[0032] Output layer, used to output the contamination risk R of the z-th cutting graph totle .

[0033] Furthermore, according to the location of the water outlet of the reservoir, the characteristic distribution of the pollution risk of the sub-area of ​​the reservoir is determined, including:

[0034] Get the center coordinates (X z ,Y z), and the outlet coordinates (X0, Y0);

[0035] For each sub-area, calculate the Euclidean distance to the outlet through the center coordinates;

[0036] Establish the characteristic distribution formula, the formula is as follows:

[0037]

[0038] Among them, XT represents the distribution intensity of pollution risk characteristics in the zth sub-region, d z represents the Euclidean distance from the center of the zth sub-region to the outlet, p represents the risk intensity coefficient, R0 represents the maximum pollution risk of the sub-region, φ1 represents the exponential decay parameter, e represents the natural base, φ2 represents the linear decay weight, and d max represents the maximum distance between any two points in the reservoir;

[0039] According to the characteristic distribution intensity of each sub-area, the characteristic distribution of the pollution risk of the sub-area of ​​the reservoir is obtained.

[0040] Furthermore, the pollution risk of each sub-area is verified based on the characteristic distribution, including:

[0041] Taking an average value of multiple contamination risks of each sub-region outputted from a plurality of valid images according to the detection model, an average contamination risk of each sub-region is determined;

[0042] According to the maximum and minimum values ​​of the average pollution risk of each sub-region, the average pollution risk of each sub-region is normalized to obtain the actual distribution intensity of the pollution risk of each sub-region;

[0043] Calculate the sum of square differences between the actual distribution intensity of the corresponding pollution risk and the characteristic distribution intensity of the pollution risk, and compare the sum of square differences with a preset deviation threshold;

[0044] If the sum of the square differences is less than the preset deviation threshold, the pollution risk determination of each sub-region is successful, otherwise the pollution risk determination of each sub-region fails.

[0045] Furthermore, if the pollution risk of each sub-area is successfully determined, the operation state of the water storage system is determined according to the pollution risk of each sub-area, including:

[0046] Take q consecutive pollution risks from the pollution risks of each sub-area at multiple moments as training samples, and take the q+1th pollution risk as the sample label to construct the training data to train and obtain the prediction model;

[0047] The prediction model outputs the pollution risk of each sub-area at the next moment. If the pollution risk of at least one sub-area at the next moment is greater than the preset warning threshold, the operating status of the sub-area is determined to be abnormal.

[0048] Furthermore, if the pollution risk determination of each sub-area fails, the detection model is updated inversely according to the feature distribution, including:

[0049] The customized parameters in the detection model are reversely updated by the gradient descent method, so that the difference between the normalized value of the average pollution risk output by the detection model for each sub-region at multiple moments and the intensity of the distribution of the pollution risk characteristics of the sub-region is reduced; the customized parameters refer to the parameters set manually.

[0050] Furthermore, the monitoring device is provided with a memory, which changes the state of a storage medium of the memory through an electrical signal and stores remote sensing images at multiple moments.

[0051] The beneficial effects of the present invention are as follows: by accurately analyzing the pollution risks of reservoir sub-areas, combining remote sensing images, dynamic model optimization and regional feature distribution, an automated and intelligent energy storage system operation status assessment mechanism is established, which not only improves the system operation efficiency, but also responds to environmental changes in real time, and dynamically adjusts the model to ensure accurate judgment of pollution risks, providing scientific decision-making support for the management of hydraulic energy storage systems, and significantly improving the safety and reliability of hydraulic energy storage systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 It is a module diagram of a system for determining the operating status of an energy storage system of the present invention. DETAILED DESCRIPTION

[0053] The subject matter described herein will now be discussed with reference to example embodiments. It should be understood that the discussion of these embodiments is only to enable those skilled in the art to better understand and implement the subject matter described herein, and the functions and arrangements of the elements discussed may be changed without departing from the scope of protection of the contents of this specification. Each example may omit, replace or add various processes or components as needed. In addition, the features described relative to some examples may also be combined in other examples.

[0054] like Figure 1 As shown, a system for determining the operating status of an energy storage system includes:

[0055] The data acquisition module is used to collect the temperature of the water reservoir at fixed time intervals, and to obtain remote sensing images of the water reservoir at fixed time intervals through monitoring equipment;

[0056] A data screening module is used to analyze the temperature of the water reservoir at multiple times, determine the temperature fluctuation section of the water reservoir, and select the remote sensing image of the water reservoir at the time corresponding to the temperature fluctuation section as a valid image;

[0057] An image analysis module is used to divide the reservoir into a number of equal sub-areas in the horizontal direction, and analyze each sub-area of ​​the reservoir in the effective image through a preset detection model to determine the pollution risk of each sub-area;

[0058] The joint verification module is used to determine the characteristic distribution of the pollution risk of the sub-areas of the reservoir according to the location of the water outlet of the reservoir, and verify the pollution risk of each sub-area according to the characteristic distribution to obtain the verification judgment of the pollution risk of each sub-area, as follows:

[0059] If the pollution risk of each sub-area is successfully determined, the operating state of the energy storage system constructed based on the reservoir is determined according to the pollution risk of each sub-area;

[0060] If the pollution risk determination of each sub-area fails, the detection model is updated in reverse according to the feature distribution.

[0061] In one embodiment of the present invention, the temperature of the water reservoir at multiple times is analyzed, including:

[0062] The critical Rayleigh number is calculated based on the temperature of the reservoir at multiple times. The formula is as follows:

[0063]

[0064] Among them, R a represents the critical Rayleigh number, g represents the gravitational acceleration, b represents the thermal expansion coefficient of the water in the reservoir, T b represents the temperature of the water bottom of the reservoir, T s represents the temperature of the water surface in the reservoir, h represents the height difference between the water surface and the bottom of the reservoir, v represents the kinematic viscosity of the water in the reservoir, k represents the heat diffusion efficiency of the water in the reservoir, V w represents the wind speed on the water surface of the reservoir, r w represents the wind speed weight coefficient, t represents the fixed time interval, r e The weight coefficient representing the rate of change of the water surface temperature of the reservoir;

[0065] R a >1700 is taken as the starting time of the temperature fluctuation segment, and R a The time corresponding to <1700 is taken as the end time of the temperature fluctuation section.

[0066] Specifically, during normal time periods, due to the lack of significant temperature differences within the water body, convection may not occur, and particles and soil at the bottom will not be rolled up. In this case, if the pollution risk is judged, the result may be inaccurate due to the lack of information on the movement of real pollutants, leading to an underestimation of the pollution situation. However, when convection occurs, the silt and particles at the bottom of the reservoir will be carried up by the movement of the water flow, causing pollution to the upper water body, and may even enter the hydroelectric generator, threatening the safe operation of the equipment. The occurrence of convection is usually accompanied by significant temperature changes (such as natural convection caused by temperature differences). Therefore, it is very important to identify the time periods (temperature fluctuation sections) when convection occurs by analyzing temperature data. These time periods are the key moments when silt and particles are rolled up and cause pollution to the reservoir. By screening the temperature fluctuation sections, we can focus on the time periods when convection is significant and avoid analysis interference from irrelevant time periods.

[0067] In one embodiment of the present invention, each sub-region of the water reservoir in the effective image is analyzed by a preset detection model, and the detection model includes an input layer, a cutting layer, a detection layer and an output layer:

[0068] Input layer, used to input valid images into the detection model;

[0069] The cutting layer is used to cut the effective image according to the sub-region division of the reservoir to obtain several cutting graphs;

[0070] The detection layer is used to identify each cutting graph, including:

[0071] Step 1, construct the energy functional E(M0),

[0072] Wherein, E(M0) represents the energy functional, which is used to make a smoothness assumption on the pollution-free matrix M0 to be solved, m represents the number of rows of the cutting graph, n represents the number of columns of the cutting graph, M0 represents the pollution-free matrix, M0(i,j) represents the pixel value of the i-th row and j-th column in the pollution-free matrix, M0(i+1,j) represents the pixel value of the i+1-th row and j-th column in the cutting graph, M0(i-1,j) represents the pixel value of the i-1-th row and j-th column in the cutting graph, M0(i,j+1) represents the pixel value of the i-th row and j+1-th column in the cutting graph, and M0(i,j-1) represents the pixel value of the i-th row and j-1-th column in the cutting graph;

[0073] Step 2, solving the pollution-free matrix M0 by an optimization algorithm, with the constraint that the pollution-free matrix M0 and the cutting graph have the same size;

[0074] Step 3, for the pollution-free matrix M0, when the number of update iterations of the pollution-free matrix M0 is 0, it is equal to the cutting graph;

[0075] Step 4, construct the Lagrangian function: Among them, λ represents the Lagrange multiplier, C represents the sum of the number of pixels in the cutting graph;

[0076] Step 5, update and solve the pollution-free matrix M0 to obtain the updated pollution-free matrix M0. The update formula is as follows:

[0077]

[0078] Among them, (M0) t+1 (i,j) represents the pixel value of the i-th row and j-th column updated by the pollution-free matrix M0, (M0) t (i,j) represents the pixel value of the i-th row and j-th column of the pollution-free matrix M0, represents the discrete Laplace operator, η represents the update iteration step size, and t represents the number of update iterations;

[0079] Step 6, subtract the z-th cutting graph and the updated pollution-free matrix M0 point by point to obtain the deviation matrix S, and compare the element value of any element of the deviation matrix S with the preset eigenvalue. If the element value is less than the preset eigenvalue, it is replaced with 0; if the element value is greater than the preset eigenvalue, it is replaced with 1 to obtain a replacement matrix;

[0080] Step 7, performing connectivity detection on each element in the replacement matrix to obtain the number L of connected clusters in the replacement matrix;

[0081] Step 8: Calculate the contamination risk R of the zth cutting graph totle , the calculation formula is as follows:

[0082] R totle =MLP(ω1×‖S‖ F +ω2×L); where ‖S‖ F represents the Frobenius norm of the bias matrix S, ω1 and ω2 represent the primary weight parameter and the secondary weight parameter respectively, and MLP represents the multi-layer perceptron;

[0083] Output layer, used to output the contamination risk R of the z-th cutting graph totle .

[0084] Specifically, by dividing the reservoir into multiple sub-areas, the pollution risk of each area is evaluated independently, which improves the precision and accuracy of the overall analysis. The pollution-free matrix is ​​constructed through energy functionals, and then the pollution risk is calculated based on the deviation matrix to ensure that the evaluation process is mathematically rigorous. The detection model consists of an input layer, a cutting layer, a detection layer, and an output layer. The functions of each layer are as follows:

[0085] Input layer: Receives filtered valid images as input data. These images have been filtered out through the analysis of temperature fluctuation segments to obtain remote sensing images corresponding to key moments.

[0086] The features of the input image may include pollution distribution characteristics of the water body (such as sediment diffusion).

[0087] Cutting layer: The effective image is cut according to the sub-areas of the reservoir to generate cutting diagrams of multiple sub-areas. Each cutting diagram corresponds to a sub-area in the reservoir, which facilitates the subsequent individual analysis of each area and quantification of pollution risks.

[0088] Detection layer: Calculate the contamination risk for each cutting graph and use multiple mathematical tools, such as energy functional, Lagrangian function and discrete Laplace operator, to deeply analyze the characteristics of sub-area contamination.

[0089] Output layer: Output the pollution risk value of each sub-area as the basis for subsequent pollution risk verification and system operation status assessment.

[0090] In one embodiment of the present invention, determining the characteristic distribution of the pollution risk of the sub-area of ​​the reservoir according to the location of the water outlet of the reservoir includes:

[0091] Get the center coordinates (X z ,Y z ), and the outlet coordinates (X0, Y0);

[0092] For each sub-area, calculate the Euclidean distance to the outlet through the center coordinates;

[0093] Establish the characteristic distribution formula, the formula is as follows:

[0094]

[0095] Among them, XT represents the distribution intensity of pollution risk characteristics in the zth sub-region, d z represents the Euclidean distance from the center of the zth sub-region to the outlet, p represents the risk intensity coefficient, R0 represents the maximum pollution risk of the sub-region, φ1 represents the exponential decay parameter, e represents the natural base, φ2 represents the linear decay weight, and d max represents the maximum distance between any two points in the reservoir;

[0096] According to the characteristic distribution intensity of each sub-area, the characteristic distribution of the pollution risk of the sub-area of ​​the reservoir is obtained.

[0097] Specifically, by calculating the distance from each sub-area to the outlet and combining it with the comprehensive attenuation model of pollution risk, the characteristic distribution intensity of the reservoir pollution risk is determined. This method not only takes into account the influence of spatial distance, but also combines the inverse, exponential and linear attenuation laws, providing a scientific basis for subsequent risk verification and system optimization. This distribution calculation method has the advantages of accuracy, efficiency and strong adaptability.

[0098] In one embodiment of the present invention, the pollution risk of each sub-area is verified according to the characteristic distribution, including:

[0099] Taking an average value of multiple contamination risks of each sub-region outputted from a plurality of valid images according to the detection model, an average contamination risk of each sub-region is determined;

[0100] According to the maximum and minimum values ​​of the average pollution risk of each sub-region, the average pollution risk of each sub-region is normalized to obtain the actual distribution intensity of the pollution risk of each sub-region;

[0101] Calculate the sum of square differences between the actual distribution intensity of the corresponding pollution risk and the characteristic distribution intensity of the pollution risk, and compare the sum of square differences with a preset deviation threshold;

[0102] If the sum of the square differences is less than the preset deviation threshold, the pollution risk determination of each sub-region is successful, otherwise the pollution risk determination of each sub-region fails.

[0103] Specifically, the evaluation results of the detection model are quantitatively verified by calculating the average value of the pollution risk, normalizing it, and summing the squared differences. By setting the deviation threshold, the reliability and accuracy of the detection model can be effectively judged, thereby ensuring that the evaluation results of the pollution risk of each sub-area are accurate.

[0104] In one embodiment of the present invention, if the pollution risk of each sub-area is successfully determined, the operation state of the water storage energy storage system is determined according to the pollution risk of each sub-area, including:

[0105] Take q consecutive pollution risks from the pollution risks of each sub-area at multiple moments as training samples, and take the q+1th pollution risk as the sample label to construct the training data to train and obtain the prediction model;

[0106] The prediction model outputs the pollution risk of each sub-area at the next moment. If the pollution risk of at least one sub-area at the next moment is greater than the preset warning threshold, the operating status of the sub-area is determined to be abnormal.

[0107] Specifically, the pollution risk of each sub-area is predicted by training the time series prediction model, and compared with the warning threshold, to detect potential operational anomalies in advance. Its advantages are: using historical data for trend prediction, enhancing the system's foresight and early warning capabilities. Independent analysis of each sub-area improves the precision of pollution risk assessment. Automated anomaly detection reduces human intervention and improves the intelligence level of the system.

[0108] In one embodiment of the present invention, if the pollution risk determination of each sub-region fails, the detection model is reversely updated according to the feature distribution, including:

[0109] The customized parameters in the detection model are reversely updated by the gradient descent method, so that the difference between the normalized value of the average pollution risk output by the detection model for each sub-region at multiple moments and the intensity of the distribution of the pollution risk characteristics of the sub-region is reduced; the customized parameters refer to the parameters set manually.

[0110] In one embodiment of the present invention, the monitoring device is equipped with a memory, and the memory stores remote sensing images at multiple moments by changing the state of a storage medium of the memory through electrical signals.

[0111] Specifically, by reversely updating the custom parameters in the detection model, the model can make the pollution risk assessment of each sub-area more accurate. The custom parameters are optimized through the gradient descent method to dynamically reduce the difference between the actual pollution risk distribution and the feature distribution, and realize the intelligent adjustment of the detection model.

[0112] The above describes an embodiment of the present embodiment, but the present embodiment is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present embodiment, ordinary technicians in this field can also make many forms, all of which are within the protection of the present embodiment.

Claims

1. A system for determining the operating status of an energy storage system, characterized in that: include: The data acquisition module is used to collect the temperature of the water reservoir at fixed time intervals, and to obtain remote sensing images of the water reservoir at fixed time intervals through monitoring equipment; A data screening module is used to analyze the temperature of the water reservoir at multiple times, determine the temperature fluctuation section of the water reservoir, and select the remote sensing image of the water reservoir at the time corresponding to the temperature fluctuation section as a valid image; An image analysis module is used to divide the reservoir into a number of equal sub-areas in the horizontal direction, and analyze each sub-area of ​​the reservoir in the effective image through a preset detection model to determine the pollution risk of each sub-area; The joint verification module is used to determine the characteristic distribution of the pollution risk of the sub-areas of the reservoir according to the location of the water outlet of the reservoir, and verify the pollution risk of each sub-area according to the characteristic distribution to obtain the verification judgment of the pollution risk of each sub-area, as follows: If the pollution risk of each sub-area is successfully determined, the operating state of the energy storage system constructed based on the reservoir is determined according to the pollution risk of each sub-area; If the pollution risk determination of each sub-area fails, the detection model is updated in reverse according to the feature distribution.

2. The energy storage system operating status determination system according to claim 1, characterized in that: Analyze the temperature of the water reservoir at multiple times, including: The critical Rayleigh number is calculated based on the temperature of the reservoir at multiple times. The formula is as follows: Among them, R a represents the critical Rayleigh number, g represents the gravitational acceleration, b represents the thermal expansion coefficient of the water in the reservoir, T b represents the temperature of the water bottom of the reservoir, T s represents the temperature of the water surface in the reservoir, h represents the height difference between the water surface and the bottom of the reservoir, v represents the kinematic viscosity of the water in the reservoir, k represents the heat diffusion efficiency of the water in the reservoir, V w represents the wind speed on the water surface of the reservoir, r w represents the wind speed weight coefficient, t represents the fixed time interval, r e The weight coefficient representing the rate of change of the water surface temperature of the reservoir; R a >1700 is taken as the starting time of the temperature fluctuation segment, and R a The time corresponding to <1700 is taken as the end time of the temperature fluctuation section.

3. The energy storage system operating status determination system according to claim 2, characterized in that: Each sub-region of the reservoir in the valid image is analyzed through a preset detection model, which includes an input layer, a cutting layer, a detection layer, and an output layer: Input layer, used to input valid images into the detection model; The cutting layer is used to cut the effective image according to the sub-region division of the reservoir to obtain several cutting graphs; The detection layer is used to identify each cutting graph, including: Step 1, construct the energy functional E(M0), Wherein, E(M0) represents the energy functional, which is used to make a smoothness assumption on the pollution-free matrix M0 to be solved, m represents the number of rows of the cutting graph, n represents the number of columns of the cutting graph, M0 represents the pollution-free matrix, M0(i,j) represents the pixel value of the i-th row and j-th column in the pollution-free matrix, M0(i+1,j) represents the pixel value of the i+1-th row and j-th column in the cutting graph, M0(i-1,j) represents the pixel value of the i-1-th row and j-th column in the cutting graph, M0(i,j+1) represents the pixel value of the i-th row and j+1-th column in the cutting graph, and M0(i,j-1) represents the pixel value of the i-th row and j-1-th column in the cutting graph; Step 2, solving the pollution-free matrix M0 by an optimization algorithm, with the constraint that the pollution-free matrix M0 and the cutting graph have the same size; Step 3, for the pollution-free matrix M0, when the number of update iterations of the pollution-free matrix M0 is 0, it is equal to the cutting graph; Step 4, construct the Lagrangian function: Among them, λ represents the Lagrange multiplier, C represents the sum of the number of pixels in the cutting graph; Step 5, update and solve the pollution-free matrix M0 to obtain the updated pollution-free matrix M0. The update formula is as follows: Among them, (M0) t+1 (i,j) represents the pixel value of the i-th row and j-th column updated by the pollution-free matrix M0, (M0) t (i,j) represents the pixel value of the i-th row and j-th column of the pollution-free matrix M0, represents the discrete Laplace operator, η represents the update iteration step size, and t represents the number of update iterations; Step 6, subtract the z-th cutting graph and the updated pollution-free matrix M0 point by point to obtain the deviation matrix S, and compare the element value of any element of the deviation matrix S with the preset eigenvalue. If the element value is less than the preset eigenvalue, it is replaced with 0; if the element value is greater than the preset eigenvalue, it is replaced with 1 to obtain a replacement matrix; Step 7, performing connectivity detection on each element in the replacement matrix to obtain the number L of connected clusters in the replacement matrix; Step 8: Calculate the contamination risk R of the zth cutting graph totle , the calculation formula is as follows: R totle =MLP(ω1×‖S‖ F +ω2×L); where ‖S‖ F represents the Frobenius norm of the bias matrix S, ω1 and ω2 represent the primary weight parameter and the secondary weight parameter respectively, and MLP represents the multi-layer perceptron; Output layer, used to output the contamination risk R of the z-th cutting graph totle .

4. The energy storage system operating status determination system according to claim 3, characterized in that: According to the location of the water outlet of the reservoir, the characteristic distribution of the pollution risk of the sub-area of ​​the reservoir is determined, including: Get the center coordinates (X z ,Y z ), and the outlet coordinates (X0, Y0); For each sub-area, calculate the Euclidean distance to the outlet through the center coordinates; Establish the characteristic distribution formula, the formula is as follows: Among them, XT represents the distribution intensity of pollution risk characteristics in the zth sub-region, d z represents the Euclidean distance from the center of the zth sub-region to the outlet, p represents the risk intensity coefficient, R0 represents the maximum pollution risk of the sub-region, φ1 represents the exponential decay parameter, e represents the natural base, φ2 represents the linear decay weight, and d max represents the maximum distance between any two points in the reservoir; According to the characteristic distribution intensity of each sub-area, the characteristic distribution of the pollution risk of the sub-area of ​​the reservoir is obtained.

5. The energy storage system operating status determination system according to claim 4, characterized in that: The pollution risk of each sub-area is verified based on the characteristic distribution, including: Taking an average value of multiple contamination risks of each sub-region outputted from a plurality of valid images according to the detection model, an average contamination risk of each sub-region is determined; According to the maximum and minimum values ​​of the average pollution risk of each sub-region, the average pollution risk of each sub-region is normalized to obtain the actual distribution intensity of the pollution risk of each sub-region; Calculate the sum of square differences between the actual distribution intensity of the corresponding pollution risk and the characteristic distribution intensity of the pollution risk, and compare the sum of square differences with a preset deviation threshold; If the sum of the square differences is less than the preset deviation threshold, the pollution risk determination of each sub-region is successful, otherwise the pollution risk determination of each sub-region fails.

6. The energy storage system operating status determination system according to claim 5, characterized in that: If the pollution risk of each sub-area is successfully determined, the operating status of the water storage system is determined according to the pollution risk of each sub-area, including: Take q consecutive pollution risks from the pollution risks of each sub-area at multiple moments as training samples, and take the q+1th pollution risk as the sample label to construct the training data to train and obtain the prediction model; The prediction model outputs the pollution risk of each sub-area at the next moment. If the pollution risk of at least one sub-area at the next moment is greater than the preset warning threshold, the operating status of the sub-area is determined to be abnormal.

7. The energy storage system operating status determination system according to claim 6, characterized in that: If the pollution risk determination of each sub-area fails, the detection model is updated in reverse according to the feature distribution, including: The customized parameters in the detection model are reversely updated by the gradient descent method, so that the difference between the normalized value of the average pollution risk output by the detection model for each sub-region at multiple moments and the intensity of the distribution of the pollution risk characteristics of the sub-region is reduced; the customized parameters refer to the parameters set manually.

8. The energy storage system operating status determination system according to claim 1, characterized in that: The monitoring device is equipped with a memory, which changes the state of the memory storage medium through electrical signals and stores remote sensing images at multiple moments.

Images