An artificial intelligence-based water quality detection method and system for water conservancy projects

Abstract

The application relates to the technical field of hydraulic engineering and discloses a water quality detection method and system for hydraulic engineering based on artificial intelligence. First multi-source data are collected for the hydraulic engineering, water quality parameter data, remote sensing image data and meteorological and hydrological data in the first multi-source data are aligned based on timestamps, and second multi-source data are obtained. Data preprocessing and feature extraction processing are performed on the second multi-source data, and target feature data are obtained. A multi-task space-time graph convolution network is designed, a genetic algorithm is used for parameter optimization, and a water quality detection model is constructed. The target feature data are input into the water quality detection model, and a water quality detection result is output through the water quality detection model. When the water quality detection result determines that the water quality is abnormal, a pollution diffusion simulation atlas is generated based on a space-time graph structure and a water quality diffusion law, and an emergency decision suggestion is provided according to a hydraulic engineering dispatching scheme. The application improves the accuracy and functional integrity of water quality detection.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of water conservancy engineering, and particularly relates to a water quality detection method and system for water conservancy engineering based on artificial intelligence. BACKGROUND

[0002] In the field of water conservancy engineering, water quality detection is a key link to ensure the safety and rational use of water resources. Traditional water quality detection methods usually rely on manual sampling and then laboratory analysis and detection. This approach has many drawbacks. On the one hand, the frequency of manual sampling is low, which cannot realize real-time monitoring of water quality and is difficult to discover sudden changes in water quality in time. On the other hand, the laboratory detection process is complex and time-consuming, and the detection results have a lag, which cannot provide timely and effective data support for real-time management and decision-making of water conservancy engineering. SUMMARY

[0003] The present application relates to the technical field of water conservancy engineering, and particularly relates to a water quality detection method and system for water conservancy engineering based on artificial intelligence.

[0004] The present application relates to the technical field of water conservancy engineering, and particularly relates to a water quality detection method and system for water conservancy engineering based on artificial intelligence.

[0005] The first multi-source data is collected for water conservancy engineering, and the water quality parameter data, remote sensing image data and meteorological and hydrological data in the first multi-source data are aligned based on timestamps to obtain second multi-source data;

[0006] The second multi-source data is subjected to data preprocessing and feature extraction processing to obtain target feature data;

[0007] A multi-task spatio-temporal graph convolution network is designed, and a genetic algorithm is used for parameter optimization to construct a water quality detection model;

[0008] The target feature data is input into the water quality detection model, and the water quality detection model outputs a water quality detection result;

[0009] When the water quality detection result determines that the water quality is abnormal, a pollution diffusion simulation atlas is generated based on the spatio-temporal graph structure and the water quality diffusion law, and an emergency decision-making suggestion is provided according to the water conservancy engineering dispatching scheme.

[0010] Optionally, in the first implementation manner of the first aspect, the data preprocessing and feature extraction processing of the second multi-source data to obtain the target feature data comprises:

[0011] The water quality parameter data in the second multi-source data is subjected to smoothing processing by using a sliding window method, and the Kalman filtering algorithm is combined to dynamically estimate and eliminate noise in the data, and the water quality parameter data after preprocessing is subjected to feature extraction by using principal component analysis to obtain water quality parameter features;

[0012] For the remote sensing image data in the second multi-source data, radiation correction and geometric correction are first performed, then a wavelet transform algorithm is used to extract high-frequency water quality abnormal signals, and a convolutional neural network is used to extract features from the preprocessed remote sensing image data to obtain remote sensing image features;

[0013] For the meteorological and hydrological data of the remote sensing image data in the second multi-source data, the physical relationship and statistical correlation between the meteorological and hydrological data and the water quality parameters are analyzed, meteorological and hydrological factors related to water quality changes are screened out, and statistical features of time series are extracted according to the screening results to obtain meteorological and hydrological features;

[0014] The water quality parameter features, the remote sensing image features, and the meteorological and hydrological features are integrated to finally obtain target feature data.

[0015] Optionally, in the second implementation manner of the first aspect of the present application, the data preprocessing and feature extraction processing of the second multi-source data to obtain target feature data comprises:

[0016] For the water quality parameter data in the second multi-source data, a sliding window method is used for smoothing processing, a Kalman filter algorithm is combined to dynamically estimate and eliminate noise in the data, and principal component analysis is used for feature extraction on the preprocessed water quality parameter data to obtain water quality parameter features;

[0017] For the remote sensing image data in the second multi-source data, radiation correction and geometric correction are first performed, then a wavelet transform algorithm is used to extract high-frequency water quality abnormal signals, and a convolutional neural network is used to extract features from the preprocessed remote sensing image data to obtain remote sensing image features;

[0018] For the meteorological and hydrological data of the remote sensing image data in the second multi-source data, the physical relationship and statistical correlation between the meteorological and hydrological data and the water quality parameters are analyzed, meteorological and hydrological factors related to water quality changes are screened out, and statistical features of time series are extracted according to the screening results to obtain meteorological and hydrological features;

[0019] The water quality parameter features, the remote sensing image features, and the meteorological and hydrological features are integrated to finally obtain target feature data.

[0020] Optionally, in the third implementation manner of the first aspect of the present application, the radiation correction and geometric correction are first performed on the remote sensing image data in the second multi-source data, then a wavelet transform algorithm is used to extract high-frequency water quality abnormal signals, and a convolutional neural network is used to extract features from the preprocessed remote sensing image data to obtain remote sensing image features, comprising:

[0021] The geometrically corrected remote sensing image data is converted into a digital matrix, the wavelet decomposition is performed on the remote sensing image data of each band, the approximate coefficients and the detail coefficients of different scales are obtained, and the high-frequency water quality abnormal signal is identified by analyzing the detail coefficients at different scales;

[0022] The threshold value processing is performed on the detail coefficients of each layer obtained by the wavelet decomposition to distinguish the signal and the noise, and the wavelet reconstruction is performed on the processed detail coefficients to generate an enhanced image;

[0023] The enhanced image is preprocessed and input into a convolutional neural network, different sizes of convolution kernels are used in the convolution layer to perform convolution operation on the input image to extract local features, the pooling layer performs down-sampling on the convolution output, and through the stacking of multiple layers of convolution and pooling operations, the abstract features from low level to high level are gradually extracted;

[0024] The feature maps of each layer are calculated through forward propagation, global average pooling is performed on the feature maps to generate a fixed-length feature vector, and the final remote sensing image feature is formed.

[0025] Optionally, in the fourth implementation manner of the first aspect of the present application, the multi-task spatio-temporal graph convolutional network at least includes a spatial branch, a temporal branch and a task branch.

[0026] Optionally, in the fifth implementation manner of the first aspect of the present application, the parameter optimization is performed by using the genetic algorithm to construct the water quality detection model, which comprises:

[0027] Each parameter in the parameter space is mapped to a chromosome code, and an initial population is randomly generated, each individual representing a combination of model parameters;

[0028] The tournament selection method is used to randomly select multiple individuals from the current population, compare the fitness values, select the optimal individual into the mating pool, and repeat the selection until there are a preset number of individuals in the mating pool;

[0029] The crossover operation is performed on the individuals in the mating pool, two parent individuals are randomly selected, and for each parameter gene site, it is determined whether to perform crossover with a crossover probability of 0.8, if the crossover is performed, a heuristic crossover method is used to generate the parameter value of the offspring, and the offspring individual is obtained;

[0030] For each gene site of each offspring individual, it is determined whether to mutate with a mutation probability of 0.01, if the mutation is performed, a new value is randomly generated within the value range of the parameter;

[0031] The individual with the highest fitness in the current population is directly entered into the next generation to replace the individual with the lowest fitness in the offspring;

[0032] The repeated selection, crossover, mutation and elite reservation operations are performed until a maximum iteration number is reached, and after the iteration ends, a parameter combination corresponding to a historical optimal individual is selected as a final optimization result, and a water quality detection model is constructed.

[0033] Optionally, in the sixth implementation form of the first aspect of the present application, the inputting the target feature data into the water quality detection model and outputting the water quality detection result through the water quality detection model comprises:

[0034] The target feature data is sorted according to timestamps to construct an input sequence, and a space-time graph is constructed based on a water conservancy engineering network topology structure;

[0035] The space-time graph is input into a spatial branch of the water quality detection model, a spatial dependency relationship between nodes is captured through a multi-layer graph convolution operation, a current node representation is updated through aggregation of neighborhood node information at each layer of graph convolution, a propagation mode of water quality in space is learned, a graph attention mechanism is applied to adaptively adjust weights of neighborhood nodes, and a node representation fused with spatial information is generated;

[0036] The node representation sequence output by the spatial branch is input into a temporal branch, time sequence information is processed through a bidirectional LSTM, a change trend of water quality over time is captured, a time attention mechanism is applied to automatically focus on important time points, and a feature representation fused with space-time information is generated;

[0037] The feature representation output by the spatial branch is input into a task branch, three sub-tasks of water quality classification, anomaly detection and pollution tracing are executed in parallel, and prediction results of the three sub-tasks are integrated to output the water quality detection result.

[0038] The second aspect of the present application provides a water quality detection system for water conservancy projects based on artificial intelligence, which comprises:

[0039] A data acquisition module is configured to acquire first multi-source data for a water conservancy project, align water quality parameter data, remote sensing image data and meteorological and hydrological data in the first multi-source data based on timestamps, and obtain second multi-source data.

[0040] A feature extraction module is configured to perform data preprocessing and feature extraction processing on the second multi-source data to obtain target feature data.

[0041] A model construction module is configured to design a multi-task space-time graph convolution network and perform parameter optimization through a genetic algorithm to construct a water quality detection model.

[0042] A water quality detection module is configured to input the target feature data into the water quality detection model and output a water quality detection result through the water quality detection model.

[0043] The decision suggestion module is configured to generate a pollution diffusion simulation atlas based on a spatio-temporal graph structure and a water quality diffusion rule when the water quality detection result determines that the water quality is abnormal, and provide an emergency decision suggestion according to a water conservancy project scheduling scheme.

[0044] A third aspect of the present application provides an artificial intelligence-based water quality detection device for water conservancy projects, which comprises a memory and at least one processor, and the memory stores instructions; the at least one processor invokes the instructions in the memory to enable the artificial intelligence-based water quality detection device for water conservancy projects to perform the steps of the artificial intelligence-based water quality detection method for water conservancy projects according to any one of the above.

[0045] A fourth aspect of the present application provides a computer-readable storage medium, which stores instructions, and the instructions are executed by a processor to implement the steps of the artificial intelligence-based water quality detection method for water conservancy projects according to any one of the above.

[0046] In the technical solution provided by the present application, first multi-source data is collected for water conservancy projects, water quality parameter data, remote sensing image data and meteorological and hydrological data in the first multi-source data are aligned based on timestamps to obtain second multi-source data; the second multi-source data is subjected to data preprocessing and feature extraction processing to obtain target feature data; a multi-task spatio-temporal graph convolution network is designed, and a genetic algorithm is used for parameter optimization to construct a water quality detection model; the target feature data is input into the water quality detection model, and a water quality detection result is output by the water quality detection model; when the water quality detection result determines that the water quality is abnormal, a pollution diffusion simulation atlas is generated based on a spatio-temporal graph structure and a water quality diffusion rule, and an emergency decision suggestion is provided according to a water conservancy project scheduling scheme; the present application provides rich and accurate data basis for water quality detection through comprehensive collection and deep fusion of multi-source data, and data processing and feature enhancement are performed to effectively improve data quality and feature expression capability; the multi-task spatio-temporal graph convolution network is used to fully learn the spatio-temporal features of water quality data, and water quality classification, anomaly detection and pollution tracing are simultaneously realized to improve the accuracy and functional integrity of water quality detection, realize accurate early warning and scientific decision making, and provide strong support for water quality management of water conservancy projects. BRIEF DESCRIPTION OF DRAWINGS

[0047] Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments with reference made to the accompanying drawings. The drawings are for purposes of illustration only and are not considered as limiting the application.

[0048] Figure 1 A flowchart of the artificial intelligence-based water quality detection method for water conservancy projects provided by the present application is shown in the figure.

[0049] Figure 2 A structure schematic diagram of a water conservancy project water quality detection system based on artificial intelligence provided by an embodiment of the present application is shown in the figure.

[0050] Figure 3 A structure schematic diagram of a water conservancy project water quality detection device based on artificial intelligence provided by an embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0051] The terms "first", "second", "third", "fourth" and the like in the description, claims, and drawings of the application, and those above and below (if any) are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the use of the terms so construed can be interchanged, such that the embodiments described herein can be carried out in another order than the one described here without departing from the principle of the application. Furthermore, the terms "comprising" or "including" and any of their derivatives, are intended to cover non-exclusive inclusions, such that a process, method, article, or apparatus that comprises, includes, has, or consists of any feature recited herein is not necessarily limited to those features that are clearly explicitly described herein but can include further steps or articles that are not expressly listed or suggested herein.

[0052] For the sake of understanding, the specific flow of the embodiments of the present application is described below, please refer to Figure 1 The flowchart of the water conservancy project water quality detection method based on artificial intelligence provided by the embodiment of the present application, which specifically includes the following steps:

[0053] Step 101, collecting first multi-source data for water conservancy project, aligning water quality parameter data, remote sensing image data and meteorological and hydrological data in the first multi-source data based on timestamp to obtain second multi-source data;

[0054] In this embodiment, the Internet of Things sensor network is deployed at the water inlet, water outlet, key nodes and pollution-prone areas of the water conservancy project to collect water quality parameter data such as water temperature, turbidity, conductivity and dissolved oxygen in real time. The sensor network uses low-power wide-area network communication technology to ensure stable data transmission to the data center; satellite / unmanned aerial vehicle remote sensing data is accessed to obtain water spectral feature data, including chlorophyll a concentration, suspended matter distribution and other information, and remote sensing image processing technology is used to perform radiation correction, geometric correction and other preprocessing on the obtained remote sensing image data to improve data quality, integrate meteorological data and hydrological data, align water quality parameter data, remote sensing data, meteorological and hydrological data based on timestamp, construct a dynamic environment correlation model, and comprehensively reflect the relationship between water quality of water conservancy project and environmental factors.

[0055] Step 102, data preprocessing and feature extraction processing are performed on the second multi-source data to obtain target feature data;

[0056] In this embodiment, for the water quality parameter data in the second multi-source data, a sliding window method is used for smoothing processing, combined with a Kalman filtering algorithm, to dynamically estimate and eliminate noise in the data, and principal component analysis is used for feature extraction on the preprocessed water quality parameter data to obtain water quality parameter features; for the remote sensing image data in the second multi-source data, radiation correction and geometric correction are first performed, and then a wavelet transform algorithm is used to extract high-frequency water quality abnormal signals, and a convolutional neural network is used for feature extraction on the preprocessed remote sensing image data to obtain remote sensing image features; for the meteorological and hydrological data of the remote sensing image data in the second multi-source data, the physical relationship and statistical correlation between the meteorological and hydrological data and the water quality parameters are analyzed, and meteorological and hydrological factors related to water quality changes are screened out, and according to the screening results, time series statistical features are extracted to obtain meteorological and hydrological features; the water quality parameter features, remote sensing image features, and meteorological and hydrological features are integrated to finally obtain target feature data.

[0057] In this embodiment, a sliding window method is used to smooth the water quality parameter data to generate a preliminary smoothed water quality parameter sequence; a Kalman filter model is constructed, and the smoothed water quality parameter sequence contains observation values of true signals and noise; a state vector and an error covariance matrix are initialized, and process noise covariance and observation noise covariance parameters are set; for each time point of water quality parameter, based on the state estimation value of the previous time and the preset state transition model, the state of the current time is predicted; according to the error covariance matrix and the process noise covariance, the prediction error covariance is updated to generate the prior state estimation and error covariance of the current time; the actual observation value and the predicted value of the current time are compared to calculate the residual, and based on the residual and the observation noise covariance, the Kalman gain is calculated; the Kalman gain is used to adjust the prior state estimation to generate the posterior state estimation of the current time, and the error covariance matrix is updated, and the data of all time points are repeatedly processed to obtain the denoised water quality parameter sequence; the denoised water quality parameter sequence is standardized, a covariance matrix of the standardized water quality parameter is constructed, and eigenvalue decomposition is performed to calculate the eigenvalues and the corresponding eigenvectors, the eigenvalues are arranged in descending order, and the eigenvectors corresponding to the first k largest eigenvalues are selected to obtain the water quality parameter features.

[0058] In this embodiment, the geometric corrected remote sensing image data is converted into a digital matrix, the wavelet decomposition is performed on the remote sensing image data of each band to obtain approximate coefficients and detail coefficients of different scales, the high-frequency water quality abnormal signal is identified by analyzing the detail coefficients under different scales, the threshold processing is performed on the detail coefficients obtained by the wavelet decomposition to distinguish the signal and the noise, the wavelet reconstruction is performed on the processed detail coefficients to generate an enhanced image, the enhanced image is preprocessed and input into a convolutional neural network, different sizes of convolution kernels are used in the convolution layer to perform convolution operation on the input image to extract local features, the pooling layer performs down-sampling on the convolution output, and the low-level to high-level abstract features are gradually extracted through stacking multiple layers of convolution and pooling operations; the feature maps of each layer are calculated through forward propagation, the global average pooling is performed on the feature maps to generate a fixed-length feature vector, and the final remote sensing image feature is formed.

[0059] Step 103, designing a multi-task space-time graph convolutional network, and performing parameter optimization by using a genetic algorithm to construct a water quality detection model;

[0060] In this embodiment, based on the network structure of water conservancy projects such as the connection relationship of rivers and reservoirs, a space-time graph is constructed. The monitoring points are taken as nodes of the graph, and the edge weights between the nodes are determined according to the spatial distance, water flow direction and flow rate. For each time step, the water quality parameters, meteorological and hydrological data and remote sensing features are taken as node features to construct a space-time graph sequence. In the time dimension, the nodes of adjacent time steps are connected through a time window to form a space-time dependency relationship.

[0061] In this embodiment, the multi-task space-time graph convolutional network at least includes a spatial branch, a temporal branch and a task branch.

[0062] In this embodiment, each parameter in the parameter space is mapped to a chromosome code, and an initial population is randomly generated, each individual representing a combination of model parameters. The tournament selection method is used to randomly select multiple individuals from the current population, compare the fitness values, and select the optimal individual into the mating pool. The selection is repeated until there are a predetermined number of individuals in the mating pool. The individuals in the mating pool are subjected to crossover operation, and two parent individuals are randomly selected. For each parameter gene site, a crossover probability of 0.8 is used to determine whether to perform crossover. If crossover occurs, a heuristic crossover method is used to generate the parameter value of the offspring, and the offspring individual is obtained. For each gene site of each offspring individual, a mutation probability of 0.01 is used to determine whether to mutate. If mutation occurs, a new value is randomly generated within the value range of the parameter. The individual with the highest fitness in the current population is directly entered into the next generation and replaces the individual with the lowest fitness in the offspring. The selection, crossover, mutation and elite preservation operations are repeated until the maximum number of iterations is reached. After the iteration is completed, the parameter combination corresponding to the historical optimal individual is selected as the final optimization result, and the water quality detection model is constructed.

[0063] Step 104, input the target feature data into the water quality detection model, and output the water quality detection result through the water quality detection model;

[0064] In this embodiment, the target feature data is sorted by timestamp to construct an input sequence, and a spatio-temporal graph is constructed based on the water conservancy engineering network topology; the spatio-temporal graph is input into the spatial branch of the water quality detection model, and the spatial dependence between nodes is captured through multi-layer graph convolution operation; each layer of graph convolution updates the current node representation by aggregating the neighborhood node information, learns the propagation mode of water quality in space, applies the graph attention mechanism to adaptively adjust the weight of the neighborhood nodes, and generates the node representation fused with spatial information; the node representation sequence output by the spatial branch is input into the time branch, the time sequence information is processed by bidirectional LSTM, the trend of water quality change over time is captured, the time attention mechanism is applied to automatically focus on important time points, and the feature representation fused with spatio-temporal information is generated; the feature representation output by the spatial branch is input into the task branch, and the water quality classification, anomaly detection and pollution source tracing three sub-tasks are executed in parallel, and the prediction results of the three sub-tasks are integrated to output the water quality detection result.

[0065] In this embodiment, water quality classification: the probability distribution of water quality level is output through the full connection layer and the softmax activation function, and the class with the highest probability is selected as the prediction result; anomaly detection: the anomaly score is output through the binary classifier, and if the anomaly score is greater than the threshold, the water quality is determined to be abnormal, and the anomaly confidence is calculated; pollution source tracing: based on the spatio-temporal propagation mode, the probability distribution of possible pollution source position is output through the full connection layer, and the most possible pollution source direction is identified.

[0066] Step 105, when the water quality detection result determines that the water quality is abnormal, a pollution diffusion simulation atlas is generated based on the spatio-temporal graph structure and the water quality diffusion law, and an emergency decision suggestion is provided according to the water conservancy engineering dispatching scheme.

[0067] In this embodiment, when the water quality detection result determines that the water quality is abnormal, the pollution propagation path is traced back based on the spatio-temporal graph structure, the node feature change at the time of abnormality is analyzed, the most possible pollution source position is determined by combining the pollution source tracing probability distribution, the key nodes and paths of pollution propagation are identified by comparing the differences of graph structures at different time steps, and the initial pollution diffusion model is established; the parameters related to water flow are extracted from the spatio-temporal graph, including the flow rate, flow direction, water level, etc. of each monitoring point, the historical hydrological data and real-time weather conditions are combined to construct a dynamic hydrodynamic model, the distribution rule of water flow parameters in space is learned through the graph convolution network, and the physical basis for pollution diffusion simulation is provided;

[0068] Based on historical pollution event data, a diffusion coefficient estimation model is trained to analyze the diffusion characteristics of different types of pollutants under different flow conditions, and a mapping relationship between pollutant characteristics and diffusion coefficient is established. The node features and edge weights in the space-time graph are used to dynamically adjust the diffusion coefficient, reflecting the spatial heterogeneity of the water environment.

[0069] An improved convection-diffusion equation is used in combination with a hydrodynamic model and a diffusion coefficient to simulate pollution diffusion on a space-time graph. The position of the pollution source is used as the initial condition, and the concentration distribution of the pollutant in the future time period is predicted through iterative calculation. The regulating effect of water conservancy facilities on water flow is considered, and the boundary conditions are dynamically adjusted during the simulation process.

[0070] According to the simulated pollutant concentration distribution, combined with the water quality standard threshold, the area is divided into high-risk, medium-risk and low-risk zones, and a risk heat map that evolves over time is generated to visually display the spatial range and temporal dynamics of pollution diffusion.

[0071] The historical scheduling scheme and real-time state data of the water conservancy project are retrieved, and an evaluation model of the relationship between the scheduling scheme and pollution diffusion is established. For different scheduling measures such as gate opening and closing, reservoir discharge, and pumping station pumping, the influence on pollution diffusion path and concentration is simulated, and the evaluation indicators include pollution diffusion speed reduction rate, high-risk area area reduction rate, and pollutant retention time.

[0072] Generate optimal emergency decision suggestions, such as recommending the best gate control scheme, reservoir discharge amount and timing, and pumping station pumping direction, and recommend corresponding physical / chemical / biological treatment methods based on the type and concentration of pollutants, and specify the best implementation time window for each measure.

[0073] Integrate pollution diffusion simulation results and emergency decision suggestions into a visual atlas, dynamic pollution diffusion heat map, mark different risk level areas, scheduling status and recommended adjustment scheme of key water conservancy facilities, emergency response timeline, display the best implementation order and time window of each measure, decision effect prediction, and compare pollution control effects under different schemes.

[0074] In this embodiment, the newly collected multi-source data is periodically added to the historical data set, and the data preprocessing, feature enhancement, retraining and optimization of the water quality detection model are performed again. At the same time, according to the feedback results in actual application and new water quality problems, the reinforcement learning early warning strategy model is adjusted to ensure that the entire water quality detection system can adapt to changing water quality and environmental conditions, and continuously maintain good detection, early warning and decision support performance.

[0075] Please refer to Figure 2 The structure diagram of the water quality detection system of the water conservancy project based on artificial intelligence provided by the embodiment of the present application, the system comprises:

[0076] a data collection module, configured to collect first multi-source data for the water conservancy project, align water quality parameter data, remote sensing image data and meteorological and hydrological data in the first multi-source data based on timestamps, and obtain second multi-source data;

[0077] a feature extraction module, configured to perform data preprocessing and feature extraction processing on the second multi-source data, and obtain target feature data;

[0078] a model construction module, configured to design a multi-task spatio-temporal graph convolution network, and perform parameter optimization by using a genetic algorithm, and construct a water quality detection model;

[0079] a water quality detection module, configured to input the target feature data into the water quality detection model, and output a water quality detection result by using the water quality detection model;

[0080] a decision suggestion module, configured to, when the water quality detection result determines that the water quality is abnormal, generate a pollution diffusion simulation atlas based on a spatio-temporal graph structure and a water quality diffusion law, and provide an emergency decision suggestion according to a water conservancy project scheduling scheme.

[0081] Figure 3 is a structure schematic diagram of a water quality detection equipment for water conservancy projects based on artificial intelligence provided by an embodiment of the present application. The water quality detection equipment for water conservancy projects based on artificial intelligence 600 can have relatively large differences due to different configurations or performances, and can include one or more processors (central processing units, CPUs) 610 (for example, one or more processors) and a memory 620, one or more storage media 630 (for example, one or more mass storage devices) storing application programs 633 or data 632. The memory 620 and the storage medium 630 can be temporary storage or persistent storage. The programs stored in the storage medium 630 can include one or more modules (not shown in the figure), and each module can include a series of instruction operations in the water quality detection equipment for water conservancy projects based on artificial intelligence 600. Further, the processor 610 can be configured to communicate with the storage medium 630, execute a series of instruction operations in the storage medium 630 on the water quality detection equipment for water conservancy projects based on artificial intelligence 600, so as to realize the method provided by the above embodiment.

[0082] The water quality detection equipment for water conservancy projects based on artificial intelligence 600 can also include one or more power supplies 640, one or more wired or wireless network interfaces 650, one or more input and output interfaces 660, and / or one or more operating systems 631, such as Windows Serve, Mac OS X, Unix, Linux, FreeBSD, etc. Those skilled in the art can understand that,Figure 3 The structure of the artificial intelligence-based water conservancy water quality detection device shown does not constitute a limitation on the computer device provided by the present application, and can include more or fewer components than the illustration, or combine certain components, or different component arrangements.

[0083] The present application also provides a computer readable storage medium, which can be a non-volatile computer readable storage medium, and can also be a volatile computer readable storage medium, and the computer readable storage medium has instructions stored therein, and when the instructions run on a computer, the computer executes the steps of the artificial intelligence-based water conservancy water quality detection method provided by the above embodiments.

[0084] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the above-described device or apparatus, unit can refer to the corresponding process in the foregoing method embodiments, which will not be repeated here.

[0085] The integrated unit, if realized in the form of a software function unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application essentially or the part that contributes to the prior art or the whole or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, and includes a number of instructions to make a computer device (which can be a personal computer, a server, or a network device, etc.) execute all or part of the steps of the method described in the various embodiments of the present application. The foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various program code storage media.

[0086] The basic principles, main features and advantages of the present application are shown and described above. Those skilled in the art should understand that the present application is not limited by the above embodiments, and the above embodiments and descriptions in the specification are only preferred examples of the present application and do not limit the present application. Without departing from the spirit and scope of the present application, various changes and improvements can be made to the present application, and these changes and improvements all fall within the scope of the claimed present application. The scope of protection of the present application is defined by the appended claims and their equivalents.

Claims

1. An artificial intelligence-based water quality detection method for hydraulic engineering, characterized in that, The method comprises the following steps: Collecting first multi-source data for water conservancy projects, aligning water quality parameter data, remote sensing image data, and meteorological and hydrological data in the first multi-source data based on timestamps to obtain second multi-source data; Performing data preprocessing and feature extraction processing on the second multi-source data to obtain target feature data; Designing a multi-task spatio-temporal graph convolution network and performing parameter optimization using a genetic algorithm to construct a water quality detection model; Inputting the target feature data into the water quality detection model and outputting a water quality detection result through the water quality detection model; When the water quality detection result determines that the water quality is abnormal, generating a pollution diffusion simulation atlas based on the spatio-temporal graph structure and the water quality diffusion law, and providing emergency decision suggestions according to the water conservancy project scheduling scheme; When the water quality detection result determines that the water quality is abnormal, backtracking the pollution propagation path based on the spatio-temporal graph structure, analyzing the node feature changes at the time of the abnormality, combining the pollution source tracing probability distribution to determine the most likely pollution source location, identifying the key nodes and paths of pollution propagation by comparing the differences in the graph structure at different time steps, establishing an initial pollution diffusion model; extracting water flow-related parameters from the spatio-temporal graph, including the flow rate, flow direction, and water level of each monitoring point, combining historical hydrological data and real-time weather conditions to construct a dynamic hydrodynamic model, and learning the spatial distribution of the water flow parameters through the graph convolution network to provide a physical basis for pollution diffusion simulation; Based on historical pollution event data, training a diffusion coefficient estimation model to analyze the diffusion characteristics of different types of pollutants under different water flow conditions, establishing a mapping relationship between pollutant characteristics and diffusion coefficients, and dynamically adjusting the diffusion coefficients using the node features and edge weights in the spatio-temporal graph to reflect the spatial heterogeneity of the water environment; Using an improved convection-diffusion equation, combining the hydrodynamic model and the diffusion coefficient, and simulating pollution diffusion on the spatio-temporal graph, taking the pollution source location as the initial condition, and predicting the concentration distribution of pollutants in the future time period through iterative calculation, considering the regulation of water conservancy facilities on water flow, and dynamically adjusting the boundary conditions during the simulation process; According to the simulated pollutant concentration distribution, combining the water quality standard threshold, dividing the area into high-risk, medium-risk, and low-risk zones, and generating a risk heat map that evolves over time to visually display the spatial range and temporal dynamics of pollution diffusion; Retrieving historical scheduling schemes and real-time state data of water conservancy projects, establishing an evaluation model of the relationship between scheduling schemes and pollution diffusion, simulating the influence of different scheduling measures on pollution diffusion path and concentration, and evaluating indicators including pollution diffusion speed reduction rate, high-risk area area reduction rate, and pollutant residence time; Generating optimal emergency decision suggestions, combining the type and concentration of pollutants, and recommending appropriate physical / chemical / biological treatment methods, and specifying the best implementation time window for each measure; The results of pollution diffusion simulation and emergency decision-making recommendations are integrated into a visual map, a dynamic pollution diffusion heat map, marking areas with different risk levels, the scheduling status of key water conservancy facilities and recommended adjustment plans, an emergency response timeline, showing the optimal implementation sequence and time window for each measure, prediction of decision-making effects, and comparison of pollution control effects under different plans.

2. The water quality detection method for water conservancy projects based on artificial intelligence according to claim 1, characterized in that, The process of preprocessing and feature extraction of the second multi-source data to obtain target feature data includes: For the water quality parameter data in the second multi-source dataset, a sliding window method is used for smoothing, combined with the Kalman filter algorithm to dynamically estimate and eliminate noise in the data, and principal component analysis is used to extract features from the preprocessed water quality parameter data to obtain water quality parameter features. For the remote sensing image data in the second multi-source data, radiometric and geometric corrections are first performed, then wavelet transform algorithm is used to extract high-frequency water quality anomaly signals, and convolutional neural network is used to extract features from the preprocessed remote sensing image data to obtain remote sensing image features. For the meteorological and hydrological data of remote sensing imagery in the second multi-source data, the physical relationship and statistical correlation between them and water quality parameters are analyzed, meteorological and hydrological factors related to water quality changes are screened out, and statistical features of time series are extracted based on the screening results to obtain meteorological and hydrological features. By integrating water quality parameter characteristics, remote sensing image characteristics, and meteorological and hydrological characteristics, the target characteristic data is finally obtained.

3. The water quality detection method for water conservancy projects based on artificial intelligence according to claim 2, characterized in that, The water quality parameter data in the second multi-source data is smoothed using a sliding window method, combined with a Kalman filter algorithm to dynamically estimate and eliminate noise in the data. Principal component analysis is then used to extract features from the preprocessed water quality parameter data, yielding water quality parameter features, including: The sliding window method was used to smooth the water quality parameter data, generating a preliminary smoothed water quality parameter sequence. A Kalman filter model was constructed, and the smoothed water quality parameter sequence contained observations of both real signals and noise. Initialize the state vector and error covariance matrix, and set the process noise covariance and observation noise covariance parameters; For each water quality parameter at each time point, the state at the current time is predicted based on the state estimate of the previous time point and the preset state transition model; Based on the error covariance matrix and the process noise covariance, update the prediction error covariance to generate the prior state estimate and error covariance at the current time. The actual observed value at the current moment is compared with the predicted value, the residual is calculated, and the Kalman gain is calculated based on the residual and the observation noise covariance. The prior state estimate is adjusted using Kalman gain to generate the posterior state estimate at the current time, and the error covariance matrix is ​​updated. This process is repeated for all time points to obtain the denoised water quality parameter sequence. The denoised water quality parameter sequence is standardized to construct the covariance matrix of the standardized water quality parameters. Eigenvalue decomposition is then performed to calculate the eigenvalues ​​and corresponding eigenvectors. The eigenvalues ​​are sorted in descending order, and the eigenvectors corresponding to the k largest eigenvalues ​​are selected to obtain the water quality parameter features.

4. The water quality detection method for water conservancy projects based on artificial intelligence according to claim 2, characterized in that, The remote sensing image data in the second multi-source data is first subjected to radiation correction and geometric correction, and then a wavelet transform algorithm is used to extract high-frequency water quality abnormal signal, and a convolutional neural network is used to extract features from the preprocessed remote sensing image data to obtain remote sensing image features, including: The geometrically corrected remote sensing image data is converted into a digital matrix, the wavelet of each band of remote sensing image data is decomposed to obtain approximate coefficients and detail coefficients of different scales, and by analyzing the detail coefficients at different scales, the high-frequency water quality abnormal signal of water quality is identified; The threshold value of each layer of detail coefficients obtained by wavelet decomposition is processed to distinguish signal and noise, and the processed detail coefficients are subjected to wavelet reconstruction to generate enhanced images; The preprocessed enhanced images are input into the convolutional neural network, different size convolution kernels are used in the convolution layer to perform convolution operation on the input image to extract local features, the pooling layer performs down-sampling on the convolution output, and by stacking multiple layers of convolution and pooling operations, abstract features from low to high levels are gradually extracted; The feature map of each layer is calculated by forward propagation, and the feature map is globally averaged to generate a fixed-length feature vector to form the final remote sensing image feature.

5. The water quality detection method for water conservancy projects based on artificial intelligence according to claim 1, characterized in that, The multi-task spatio-temporal graph convolution network at least includes a spatial branch, a temporal branch and a task branch.

6. The water quality detection method for water conservancy projects based on artificial intelligence according to claim 1, characterized in that, The genetic algorithm is used for parameter optimization to construct a water quality detection model, including: Each parameter in the parameter space is mapped to a chromosome code, an initial population is randomly generated, and each individual represents a combination of model parameters; A tournament selection method is used to randomly select multiple individuals from the current population, compare the fitness values, and select the optimal individual into the mating pool, and repeat the selection until there are a predetermined number of individuals in the mating pool; The individuals in the mating pool are subjected to crossover operation, two parent individuals are randomly selected, and for each parameter gene site, a crossover probability of 0.8 is used to determine whether to cross, and if crossed, a heuristic crossover method is used to generate the parameter value of the offspring to obtain the offspring individual; For each gene site of each offspring individual, a mutation probability of 0.01 is used to determine whether to mutate, and if mutated, a new value is randomly generated within the parameter value range; The individual with the highest fitness in the current population is directly entered into the next generation to replace the individual with the lowest fitness in the offspring; Repeat the selection, crossover, mutation and elite preservation operations until the maximum number of iterations is reached, and after the iteration is completed, the parameter combination corresponding to the historical optimal individual is selected as the final optimization result to construct the water quality detection model.

7. The water quality detection method for water conservancy projects based on artificial intelligence according to claim 1, characterized in that, The target feature data is input into the water quality detection model, and the water quality detection result is output by the water quality detection model, including: The target feature data is sorted by timestamp to construct an input sequence, and a spatio-temporal graph is constructed based on the network topology of the water conservancy project; The spatio-temporal graph is input into the spatial branch of the water quality detection model, and the spatial dependence between nodes is captured through multi-layer graph convolution operation, each layer of graph convolution updates the current node representation by aggregating the neighborhood node information, learns the propagation pattern of water quality in space, applies a graph attention mechanism to adaptively adjust the weights of the neighborhood nodes, and generates a node representation that integrates spatial information; The node representation sequence output by the space branch is input into the time branch, time sequence information is processed by using a bidirectional LSTM, a change trend of water quality over time is captured, a time attention mechanism is applied to automatically focus on important time points, and a feature representation integrating space-time information is generated; The feature representation output by the space branch is input into the task branch, three sub-tasks of water quality classification, anomaly detection and pollution tracing are executed in parallel, and prediction results of the three sub-tasks are integrated to output the water quality detection result.

8. An artificial intelligence-based water quality detection system for hydraulic engineering, which is used to realize an artificial intelligence-based water quality detection method for hydraulic engineering according to claim 1. The system comprises: The data acquisition module is configured to acquire first multi-source data for the water conservancy project, align water quality parameter data, remote sensing image data, and meteorological and hydrological data in the first multi-source data based on timestamps, and obtain second multi-source data. The feature extraction module is configured to perform data preprocessing and feature extraction processing on the second multi-source data to obtain target feature data. The model construction module is configured to design a multi-task space-time graph convolution network, perform parameter optimization by using a genetic algorithm, and construct a water quality detection model. The water quality detection module is configured to input the target feature data into the water quality detection model and output a water quality detection result by using the water quality detection model. The decision suggestion module is configured to generate a pollution diffusion simulation atlas based on a space-time graph structure and a water quality diffusion law when the water quality detection result determines that the water quality is abnormal, and provide an emergency decision suggestion according to a water conservancy project dispatching scheme.

9. An artificial intelligence-based water quality detection device for hydraulic engineering, characterized by, The water quality detection device based on artificial intelligence comprises a memory and at least one processor, and the memory stores instructions.

10. A computer-readable storage medium having stored thereon instructions, the computer-readable storage medium comprising: The instructions are executed by the processor to implement the steps of the water quality detection method based on artificial intelligence of any one of claims 1-7. The instructions are executed by the processor to implement the steps of the water quality detection method based on artificial intelligence of any one of claims 1-7.

Images