A water conservancy project safety monitoring and early warning method and system based on artificial intelligence

Abstract

The present invention relates to a method and system for monitoring and warning safety of water conservancy projects based on artificial intelligence, the method comprising: gridding the space where sluices and pump stations are located; allocating sluices and pump stations to grids according to their positions and positional relationships, and making each grid have only one or less sluice or pump station; constructing a warning object prediction analysis model; using the warning object prediction analysis model to determine the warning object prediction parameters of the sluice or pump station; and constructing an intelligent analysis model to perform safety monitoring and warning of water conservancy projects. The present invention combines the prediction model of sluices and pump stations with the warning model based on artificial intelligence based on the positional relationship, thereby reducing the pressure on the construction and training of artificial intelligence models and greatly improving the efficiency and feasibility of monitoring and warning.

Description

Technical Field

[0001] The present invention belongs to the field of artificial intelligence technology, and in particular relates to a water conservancy project safety monitoring and early warning method and system based on artificial intelligence. Background Art

[0002] Before the 1980s, the early warning models of water conservancy projects at home and abroad were mainly concentrated in dams and reservoirs, and there were few studies on safety early warning models for sluices and pumping stations. In addition, the safety early warning models of dams and reservoirs generally used traditional models including statistical models, deterministic models and hybrid models. After the 1980s, the relevant models developed vertically, including gray models, neural network models, fuzzy clustering analysis models, etc. The previous hybrid model was a hybrid form of deterministic models and statistical models, and the statistical model was mainly based on multiple linear regression, stepwise regression or partial least squares regression. The data feature extraction of the statistical model is not enough: At present, the statistical model can use relevant statistical analysis methods, such as multiple linear regression, stepwise regression or partial least squares regression, to analyze and count features, but the corresponding complex water conservancy project data, especially the advanced features in the data, are still not enough to learn, and a new method is needed to learn and count advanced features.

[0003] With the development of artificial intelligence technology, neural network models, especially artificial neural network (ANN) models, are widely used in water quality data soft measurement, data anomaly monitoring and time series prediction in the early warning process of water gates and pumping stations in water conservancy projects. ANN models can predict and warn the operating status of water gates and pumping stations by learning the mapping relationship between input and output data. Some researchers have explored water level prediction models based on hydrodynamic-neural network coupling, combining the simulated data of hydrodynamic models with neural network models, such as BP neural networks and long short-term memory (LSTM) networks, to improve the accuracy of water level prediction. Data quality can be improved and the feasibility of prediction can be enhanced by data preprocessing technology and selecting appropriate filtering methods. In the field of water quality early warning, the application of ANN models depends on the quality of monitoring data. In order to improve the performance of the model, researchers usually perform soft measurement, anomaly monitoring and time series prediction of data. These models can handle nonlinear and non-stationary water quality data and provide more accurate prediction results than traditional statistical models. However, there are also some challenges in the application of ANN models, such as poor interpretability of the model and high requirements for hardware resources. Future research directions include establishing reasonable and explainable ANN models, combining data-driven and process-based mechanism models, and designing new ANN model structures to improve prediction accuracy. In general, neural network models play an important role in the early warning process of sluices and pumping stations. Through continuous optimization and improvement, these models are expected to play a greater role in future water environment monitoring and early warning systems. The problems existing in the prior art can be summarized as follows: First, the sluice or pumping station is considered separately without comprehensive consideration, and functionally, there is a close relationship between the two existing in the same control space or water system; Second, the traditional analysis method is not combined with artificial intelligence technology, which brings great pressure to the training of artificial intelligence models and the amount of data required, resulting in the problem of difficulty in achieving insufficient accuracy; Based on the above problems, the present invention combines the prediction model of the sluice and pumping station with the early warning model based on artificial intelligence based on the position relationship, which reduces the pressure on the construction and training of the artificial intelligence model and greatly improves the efficiency and feasibility of monitoring and early warning. Summary of the invention

[0004] In order to solve the above problems in the prior art, the present invention proposes a water conservancy project safety monitoring and early warning method and system based on artificial intelligence, the method comprising:

[0005] Step S1: Obtain the positional relationship between the sluice gate and the pump station, take the minimum rectangular space surrounding the space where the sluice gate and the pump station are located as the continuous space for grid division, grid the space where the sluice gate and the pump station are located, and evenly divide the continuous space into discrete grids; allocate the sluice gate and the pump station to the grid according to their position and positional relationship, and make each grid have less than or equal to one sluice gate or pump station; use row and column code (i, j) as the identification of the grid and the elements therein; that is, some grids do not have sluice gates or pump stations;

[0006] Step S11: Get the position relationship between the sluice gate and the pump station, and set the number of grids to or , set its initial number and cutoff number; the initial values ​​of N and M are based on the number of sluice gates and number of pumping stations The sum of Set, assuming that M is less than N, so that ;but , otherwise swap the initial values ​​of the two; if the two are equal, set ; Set the cutoff number equal to ;

[0007] Step S12: Randomly move the entire grid. When there are more than one sluice gate or pump station in each independent grid, continue to move randomly until the movement cutoff condition is met; when there is only one sluice gate or pump station in each independent grid, calculate the average distance between each sluice gate or pump station and all boundary nodes of the independent grid where it is located; when the distance averages are all less than the distance threshold, proceed to step S2; otherwise, continue to move randomly until the movement cutoff condition is met; after the movement cutoff condition is met, proceed to step S13;

[0008] Step S13: Determination or Whether the value reaches or exceeds the cutoff number, if not, increase the N and M values, set N=N+1, M=M+1; return to step S12; otherwise, the automatic gridding fails, and automatic or manual adjustment is performed to complete the gridding, or the shape of the grid is changed and step S1 is executed again;

[0009] Step S2: construct an early warning object prediction analysis model based on vertical displacement, temperature causes and historical time-dependent displacement change influence components; use the early warning object prediction analysis model to determine the early warning object prediction parameters of the sluice or pump station ;

[0010] Step S3: Construct an intelligent analysis model. The input of the intelligent analysis model is a joint input matrix, which is used to jointly identify the monitoring parameters of the sluice and pump station and their positional relationship. The output of the intelligent analysis model is a K-element output vector, each element of which is used to indicate the warning value of the corresponding warning object, and K is the total amount of the sluice and pump station. The element at the (i, j)th position in the joint input matrix corresponds to the (i, j)th grid and its warning object parameters. When there is a warning object in the grid, if the warning object is a pump station, the warning object parameters are set. ; If the warning object is a sluice, set the warning object parameters ; Obtain the current water pressure and water level difference data of the sluice and water pump, the time-dependent displacement change data and the temperature monitoring data, and use the early warning object prediction analysis model to calculate the current early warning object prediction parameters ; Use the current warning object prediction parameters to calculate the warning object parameters The parameter value and the warning object parameter , and set the parameter value of the warning object parameter of the empty grid to 0, construct the current joint input matrix as the input of the intelligent analysis model, obtain the K-element output vector as the output value, and perform water conservancy project safety monitoring and early warning based on the output value.

[0011] Furthermore, the minimum rectangular space surrounding the space where the sluice gate and the pump station are located is used as a continuous space for grid division.

[0012] Furthermore, the minimum rectangular space is rotated so that the upper left corner of the minimum rectangular space is located upstream of the water flow and the side opposite to the upper left corner is located downstream of the water flow; the grid with the smallest row and column code is located upstream relative to the grid with the largest row and column.

[0013] Furthermore, the grid is a square grid or a rectangular grid.

[0014] Furthermore, the movement cutoff condition is that the number of random moves reaches a preset value.

[0015] Furthermore, the distance threshold is a dynamic distance threshold, and the dynamic distance threshold is set equal to the grid area. ;in: is the grid area, is the adjustment coefficient.

[0016] A water conservancy project safety monitoring and early warning system based on artificial intelligence, wherein the water conservancy project safety monitoring and early warning system based on artificial intelligence is used to implement the above-mentioned water conservancy project safety monitoring and early warning method based on artificial intelligence.

[0017] A water conservancy project safety monitoring and early warning device based on artificial intelligence, wherein the water conservancy project safety monitoring and early warning device based on artificial intelligence is used to implement the above-mentioned water conservancy project safety monitoring and early warning method based on artificial intelligence.

[0018] A water conservancy project safety monitoring and early warning module based on artificial intelligence, wherein the water conservancy project safety monitoring and early warning module based on artificial intelligence is used to implement the above-mentioned water conservancy project safety monitoring and early warning method based on artificial intelligence.

[0019] A water conservancy project safety monitoring and early warning server based on artificial intelligence, characterized in that the water conservancy project safety monitoring and early warning server based on artificial intelligence is used to implement the above-mentioned water conservancy project safety monitoring and early warning method based on artificial intelligence.

[0020] The beneficial effects of the present invention include:

[0021] (1) Based on the location relationship, the data of sluice gates and pumping stations are combined. A dual-model structure is adopted, which combines the warning object prediction and analysis model of sluice gates and pumping stations with the warning model based on artificial intelligence. The traditional warning object prediction and analysis model is used to provide training and data for real-time warning, thereby reducing the pressure on the construction and training of artificial intelligence models and greatly improving the efficiency and feasibility of monitoring and early warning.

[0022] (2) The data of sluice gates and pumping stations are combined in a grid manner, so that the final matrix contains the topological position relationship between sluice gates and pumping stations. The two different types of data, sluice gates and pumping stations, are represented in the same matrix by numerical separation using positive and negative values, thereby integrating the two types of data into one matrix, reducing the number of neurons in the input layer of the neural network model, thereby reducing the complexity of training and the amount of data required. Through the K-element output vector, safety monitoring and early warning can be performed on all pumping stations and sluice gates at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of the present application, but do not constitute an improper limitation of the present invention. In the drawings:

[0024] Figure 1 Schematic diagram of the artificial intelligence-based water conservancy project safety monitoring and early warning method provided by the present invention. DETAILED DESCRIPTION

[0025] The present invention will be described in detail below in conjunction with the accompanying drawings and specific embodiments, wherein the illustrative embodiments and descriptions are only used to explain the present invention but are not intended to limit the present invention.

[0026] The present invention proposes a water conservancy project safety monitoring and early warning method and system based on artificial intelligence, as shown in the attached Figure 1 As shown, the method comprises the following steps:

[0027] Step S1: Obtain the positional relationship between the sluice gate and the pump station, grid the space where the sluice gate and the pump station are located, and evenly divide the continuous space into discrete grids; allocate the sluice gate and the pump station to the grids according to their positions and positional relationships, and make each grid have less than or equal to one sluice gate or pump station; use row and column codes (i, j) as identifiers of the grids and the elements therein; that is, some grids do not have sluice gates or pump stations;

[0028] Preferably: the minimum rectangular space surrounding the space where the sluice gate and the pump station are located is used as the continuous space for grid division; the minimum rectangular space is rotated so that the upper left corner of the minimum rectangular space is located upstream of the water flow and the side opposite to the upper left corner is located downstream of the water flow; at this time, the grid with the smallest row and column code is located upstream relative to the grid with the largest row and column code;

[0029] Preferably: the grid is a square grid, a rectangular grid, an irregular Voronoi diagram or a Delaunay triangle;

[0030] Preferably, the grid is continuous or discontinuous. When there is a large space between the sluice and the pump station, no grid may be provided in the space.

[0031] The step S1 specifically includes the following steps:

[0032] Step S11: Get the position relationship between the sluice gate and the pump station, and set the number of grids to or , set its initial number and cutoff number;

[0033] Preferred: The initial values ​​of N and M are based on the number of sluice gates. and number of pumping stations The sum of Set, assuming that M is less than N, so that ;but , otherwise swap the initial value settings of the two; set the cutoff number equal to ; If the two are equal, set ;

[0034] Preferred: Set the cutoff number equal to ;

[0035] Alternative: Set the cutoff number equal to ;

[0036] Step S12: Randomly move the entire grid. When there are more than one sluice gate or pump station in each independent grid, continue to move randomly until the moving cutoff condition is met; when there is only one sluice gate or pump station in each independent grid, calculate the average distance between each sluice gate or pump station and all boundary nodes of the independent grid where it is located (for example, a square grid has 4 boundary nodes); when the distance means are all less than the distance threshold (or more than a certain proportion of the distance means are less than the distance threshold), go to step S2; otherwise, continue to move randomly until the moving cutoff condition is met; after the moving cutoff condition is met, go to step S13;

[0037] After entering step S2, the number of grids and the pumping stations or sluices contained in each grid are determined and are in the best (local optimal) position;

[0038] The movement cut-off condition is that the number of random moves reaches a preset value;

[0039] The distance threshold is a dynamic distance threshold, and the dynamic distance threshold is set equal to the grid area. ;in: is the grid area ( or ), is the adjustment factor; for example: ;

[0040] Step S13: Determination (or ) value reaches or exceeds the cutoff number, if not, increase the N and M values, set N=N+1, M=M+1 (or increase the M and N values ​​in rounds, that is, only increase one of them each time); return to step S12; otherwise, the automatic gridding fails, and automatic or manual adjustment is performed to complete the gridding, or the shape of the grid is changed and step S1 is executed again;

[0041] Preferred: automatic adjustment by assigning sluice gates or pumping stations to the nearest grid cells;

[0042] Step S2: Determining the prediction parameters of the warning object of the sluice or pump station ; Specifically include the following steps:

[0043] Step S21: constructing a warning object prediction and analysis model based on vertical displacement, temperature causes and historical time-dependent displacement change influencing components; using the warning object prediction and analysis model to determine the warning object prediction parameters of the sluice or pump station; wherein: the warning object is the sluice and pump station; specifically comprising the following steps:

[0044] Step S22: Use the historical water pressure and water level difference data of the sluice and pump station to fit and construct the water pressure parameters of the warning object ; Specifically: Use the following formulas (1) and (2) to construct the water pressure parameters of the warning object: ;in: is the water pressure parameter of the pumping station; is the water pressure parameter of the sluice gate; , , is the fitting coefficient, , , are the difference between the upstream water depth, downstream water depth and reservoir water level (average water level of upstream and downstream) during monitoring and the average reservoir water level (average water level of upstream and downstream) j days before monitoring, is the monitoring number;

[0045] (1);

[0046] (2);

[0047] Alternative: It is the loop variable used in the fitting process; the initial value of the loop variable is 1;

[0048] The water pressure parameters need to consider reservoir water pressure, dam foundation uplift pressure, dam body uplift pressure, etc. Compared with the dam, the water level difference between the upstream and downstream of the sluice and pump station project is smaller. The present invention also considers the influence of the upstream and downstream water depths, and adopts the combination of reservoir water pressure and dam body uplift pressure to fit and express the water pressure parameters of the sluice and pump station project;

[0049] Step S23: Use historical temperature monitoring data to construct temperature parameters of the warning object ; Specifically: Use the following formula (3) to construct the temperature parameter of the warning object ;

[0050] (3);

[0051] in: , , is the fitting coefficient, is the time number (in days), i.e. the cumulative number of days between the monitoring time and the start time. is the average temperature on day t, Yes is the monitoring number;

[0052] Alternative: It is the loop variable or water layer number used in the fitting process; the initial value of the loop variable is set to 1; it increases by 1 after completing a calculation;

[0053] Step S24: Use historical time-dependent displacement change data to construct time-dependent displacement parameters for the warning object ; Specifically: Use the following formulas (4) and (5) to construct the temperature parameters of the warning object ; Where: , is the fitting coefficient, is the time number, is a preset constant value, is an intermediate variable;

[0054] (4);

[0055] (5);

[0056] Preferred: ;For example: ;

[0057] Considering that the reasons for the additional time-effect of the project are very complex, it comprehensively reflects the creep and plastic deformation of the structural concrete and the base, as well as the compression deformation of the geological structure; for projects that generally operate normally, the change law of the time-effect displacement is rapid in the early stage and gradually stabilizes in the later stage; in order to fully reflect the influence of the time-effect factors of initial water storage and multi-year operation, a combination of logarithmic function and linear function is used for fitting expression;

[0058] Step S25: Use the warning object prediction analysis model to construct warning object prediction parameters When the warning object is a pump station, the warning object prediction analysis model is set to: ; When the warning object is a sluice gate, the warning object prediction analysis model is set as: ;

[0059] Preferred: vertical displacement unit is mm, water level unit is m, temperature unit is Celsius, time unit is day;

[0060] Alternatively: when the warning object is a pump station, the warning object prediction analysis model is set to: ; When the warning object is a sluice gate, the warning object prediction analysis model is set as: ;in: is the fitting constant;

[0061] Further: the warning object prediction and analysis model uses the following formula (6) to construct the warning object prediction parameters of the vertical displacement of the sluice gate; the warning object prediction and analysis model uses the following formula (7) to construct the warning object prediction parameters of the vertical displacement parameters of the pump station;

[0062] (6);

[0063] (7);

[0064] Step S3: Construct an intelligent analysis model. The input of the intelligent analysis model is a joint input matrix, which is used to jointly identify the monitoring parameters of the sluice and pump station and their positional relationship. The output of the intelligent analysis model is a K-element output vector, each element of which is used to indicate the warning value of the corresponding warning object, and K is the total amount of the sluice and pump station. The element at the (i, j)th position in the joint input matrix corresponds to the (i, j)th grid and its warning object parameters. When there is a warning object in the grid, if the warning object is a pump station, the warning object parameters are set. ; If the warning object is a sluice, set the warning object parameters ; Obtain the current water pressure and water level difference data of the sluice and water pump, the time-dependent displacement change data and the temperature monitoring data, and use the early warning object prediction analysis model to calculate the current early warning object prediction parameters ; Use the current warning object prediction parameters to calculate the warning object parameters The parameter value and the warning object parameter , and set the parameter value of the warning object parameter of the empty grid to 0, construct the current joint input matrix as the input of the intelligent analysis model, obtain the K-element output vector as the output value, and perform water conservancy project safety monitoring and early warning based on the output value; when the parameter value is 0, it will not bring calculation pressure to the intelligent analysis model;

[0065] Preferably: the K-element output vector uniformly encodes the sluice gate and the pump station, and the element value of the k-th element in the K-element output vector indicates the warning value of the sluice gate or pump station encoded as the k value;

[0066] Preferably: the intelligent analysis model is an artificial intelligence model;

[0067] Alternatively: the intelligent analysis model is a neural network model;

[0068] The construction of the intelligent analysis model specifically includes the following steps:

[0069] Step S3A1: Perform data preprocessing and normalize the element values ​​in the joint input matrix to a uniform range, for example: -1 to 1; at this time, all grids with pump stations are positive values, grids with sluices are negative values, and empty grid element values ​​are 0; that is, when there is an early warning object in the grid, if the early warning object is a pump station, set the early warning object parameter ; If the warning object is a sluice, set the warning object parameters ; Set to 0 when there is no pump station and sluice gate;

[0070] Step S3A2: vectorize the joint input matrix in row and column order to form a 1-dimensional vector input to the input layer, and the input dimension of the input layer is 1; that is, set 1 neuron; set the first convolution layer after the input layer, and use: use 32 3x3 convolution kernels, the step size is 1, and use the ReLU activation function; set the first pooling layer thickly, use 2x2 maximum pooling, and the step size is 2; set the second convolution layer after the first pooling layer, use 64 3x3 convolution kernels, the step size is 1, and use the ReLU activation function; set the second pooling layer after the second convolution layer, use 2x2 maximum pooling, and the step size is 2; then set the fully connected layer 1: containing 128 nodes; finally set the output layer, output the K-element vector, set K neurons, each neuron corresponds to an element in the K-element vector; use the softmax activation function; the element values ​​in the K-element vector are continuous values;

[0071] Step S3A3: After completing the construction of the intelligent analysis model, compile the model and use the classification cross entropy loss function for training;

[0072] Preferred: Use Adam optimizer for parameter optimization;

[0073] Preferred: Evaluate model performance on a validation set;

[0074] Preferred: Perform final evaluation on the test set;

[0075] Preferred: Deploy the trained intelligent analysis model to the safety monitoring and early warning of water conservancy projects based on artificial intelligence, analyze the operation data of sluices and pumping stations in real time, and issue early warnings to each sluice and pumping station in real time based on the element value of the kth element in the K-element output vector;

[0076] Based on the same inventive concept, the present invention also provides a water conservancy project safety monitoring and early warning system based on artificial intelligence, the system includes a method for executing the above-mentioned water conservancy project safety monitoring and early warning method based on artificial intelligence;

[0077] Based on the same inventive concept, the present invention also provides a water conservancy project safety monitoring and early warning server based on artificial intelligence, the system includes a system for executing the above-mentioned water conservancy project safety monitoring and early warning method based on artificial intelligence;

[0078] Based on the same inventive concept, the present invention also provides a water conservancy project safety monitoring and early warning platform based on artificial intelligence, the system includes a system for executing the above-mentioned water conservancy project safety monitoring and early warning method based on artificial intelligence;

[0079] A computer program (also referred to as a program, software, software application, script, or code) can be written in any form of a programming language, including an assembled or interpreted language, a declarative or procedural language, and it can be deployed in any form, including as a stand-alone program or as a module, component, subroutine, object, or other unit suitable for use in a computing environment. A computer program may, but need not, correspond to a file in a file system. A program can be stored in a portion of a file that holds other programs or data (e.g., one or more scripts stored in a markup language document), in a single file dedicated to the program, or in multiple collaborative files (e.g., files storing one or more modules, subroutines, or code portions). A computer program can be deployed to execute on one computer or on multiple computers located at one site or distributed across multiple sites and interconnected by a communication network.

[0080] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0081] The present invention is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0082] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0083] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0084] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A water conservancy project safety monitoring and early warning method based on artificial intelligence, characterized in that: The method comprises: Step S1: Obtain the positional relationship between the sluice gate and the pump station, take the minimum rectangular space surrounding the space where the sluice gate and the pump station are located as the continuous space for grid division, grid the space where the sluice gate and the pump station are located, and evenly divide the continuous space into discrete grids; allocate the sluice gate and the pump station to the grid according to their position and positional relationship, and make each grid have less than or equal to one sluice gate or pump station; use row and column code (i, j) as the identification of the grid and the elements therein; that is, some grids do not have sluice gates or pump stations; Step S11: Get the position relationship between the sluice gate and the pump station, and set the number of grids to or , set its initial number and cutoff number; the initial values ​​of N and M are based on the number of sluice gates and number of pumping stations The sum of Set, assuming that M is less than N, so that ;but , otherwise swap the initial values ​​of the two; if the two are equal, set ; Set the cutoff number equal to ; Step S12: Randomly move the entire grid. When there are more than one sluice gate or pump station in each independent grid, continue to move randomly until the movement cutoff condition is met; when there is only one sluice gate or pump station in each independent grid, calculate the average distance between each sluice gate or pump station and all boundary nodes of the independent grid where it is located; when the distance averages are all less than the distance threshold, proceed to step S2; otherwise, continue to move randomly until the movement cutoff condition is met; after the movement cutoff condition is met, proceed to step S13; The moving cutoff condition is that the number of random moves reaches the preset value; Step S13: Determination or Whether the value reaches or exceeds the cutoff number, if not, increase the N and M values, set N=N+1, M=M+1; Return to step S12; Otherwise, the automatic gridding fails, and automatic or manual adjustment is performed to complete the gridding, or the shape of the grid is changed and step S1 is executed again; Step S2: construct an early warning object prediction analysis model based on vertical displacement, temperature causes and historical time-dependent displacement change influence components; use the early warning object prediction analysis model to determine the early warning object prediction parameters of the sluice or pump station ; Step S3: Construct an intelligent analysis model. The input of the intelligent analysis model is a joint input matrix, which is used to jointly identify the monitoring parameters of the sluice and pump station and their positional relationship. The output of the intelligent analysis model is a K-element output vector, each element of which is used to indicate the warning value of the corresponding warning object, and K is the total amount of the sluice and pump station. The element at the (i, j)th position in the joint input matrix corresponds to the (i, j)th grid and its warning object parameters. When there is a warning object in the grid, if the warning object is a pump station, the warning object parameters are set. ; If the warning object is a sluice gate, set the warning object parameters ; Obtain the current water pressure and water level difference data of the sluice and water pump, the time-dependent displacement change data and the temperature monitoring data, and use the early warning object prediction analysis model to calculate the current early warning object prediction parameters ; Calculate the warning object parameters using the current warning object prediction parameters The parameter value and the warning object parameter , and set the parameter value of the warning object parameter of the empty grid to 0, construct the current joint input matrix as the input of the intelligent analysis model, obtain the K-element output vector as the output value, and perform water conservancy project safety monitoring and early warning based on the output value.

2. The artificial intelligence-based water conservancy project safety monitoring and early warning method according to claim 1 is characterized in that: The minimum rectangular space surrounding the space where the sluice gate and pump station are located is used as the continuous space for grid division.

3. The artificial intelligence-based water conservancy project safety monitoring and early warning method according to claim 2 is characterized in that: The minimum rectangular space is rotated so that the upper left corner of the minimum rectangular space is located upstream of the water flow and the side opposite to the upper left corner is located downstream of the water flow; the grid with the smallest row and column code is located upstream relative to the grid with the largest row and column.

4. The artificial intelligence-based water conservancy project safety monitoring and early warning method according to claim 3 is characterized in that: The grid is a square grid or a rectangular grid.

5. The artificial intelligence-based water conservancy project safety monitoring and early warning method according to claim 4 is characterized in that: The distance threshold is a dynamic distance threshold, and the dynamic distance threshold is set equal to the grid area. ;in: is the grid area, is the adjustment coefficient.

6. A water conservancy project safety monitoring and early warning system based on artificial intelligence, characterized in that: The artificial intelligence-based water conservancy project safety monitoring and early warning system is used to implement the artificial intelligence-based water conservancy project safety monitoring and early warning method described in any one of claims 1 to 5.

7. A water conservancy project safety monitoring and early warning device based on artificial intelligence, characterized in that: The artificial intelligence-based water conservancy project safety monitoring and early warning device is used to implement the artificial intelligence-based water conservancy project safety monitoring and early warning method described in any one of claims 1 to 5.

8. A water conservancy project safety monitoring and early warning module based on artificial intelligence, characterized in that: The artificial intelligence-based water conservancy project safety monitoring and early warning module is used to implement the artificial intelligence-based water conservancy project safety monitoring and early warning method described in any one of claims 1 to 5.

9. A water conservancy project safety monitoring and early warning server based on artificial intelligence, characterized in that: The artificial intelligence-based water conservancy project safety monitoring and early warning server is used to implement the artificial intelligence-based water conservancy project safety monitoring and early warning method described in any one of claims 1 to 5.

Images