Reservoir flood prevention water level early warning system and method

Abstract

The invention relates to the field of flood prevention water level early warning, and particularly discloses a reservoir flood prevention water level early warning system and method.The reservoir flood prevention water level early warning method comprises the steps that firstly, rainfall data along an upstream river are obtained in real time, and the water level change conditions of a reservoir inlet and a key river reach are monitored; performing data analysis on the collected rainfall data and water level data by adopting an artificial intelligence technology based on deep learning so as to capture space-time correlation characteristics among the rainfall data of each monitoring point along the river and space-time correlation characteristics among the water level data of each monitoring point in a water level monitoring area, and taking the rainfall data as a main variable so as to obtain a time-space correlation characteristic of the rainfall data of each monitoring point in the water level monitoring area; the water level data are used as covariables, and the influence mechanism of rainfall change on water level change is revealed by performing time sequence interaction analysis on the water level data and the covariables, so that the prediction of the reservoir water level change trend is realized. Therefore, the response speed and accuracy of the flood prevention water level early warning of the reservoir can be effectively improved, the dependence on human resources is reduced, and meanwhile, the maintenance cost of a traditional hydrological model is reduced.

Description

Technical Field

[0001] This application relates to the field of flood control water level warning, and more specifically, to a reservoir flood control water level warning system and method. Background Art

[0002] With the intensification of global climate change, extreme weather events occur frequently, and the risk of flood disasters faced by reservoirs is also increasing continuously. As an important infrastructure for flood control and disaster reduction and water resources management, the water level change of reservoirs is directly related to the life and property safety of downstream residents and the stability of the ecological environment. Traditional reservoir monitoring and warning systems often rely on manual observation and empirical judgment, and there are many limitations, such as slow response time, low warning accuracy, and high consumption of human resources. Therefore, how to accurately predict the change trend of reservoir water level and timely issue warning information has become a key issue in flood control work.

[0003] In this regard, the invention patent with the publication number of CN118430190A discloses a reservoir flood control water level warning system and method, which predicts and analyzes the change trend of the reservoir water level within a certain period of time in the future by real-time monitoring of rainfall data along the upstream river and the water level changes at the reservoir entrance and key river sections, and evaluates the risk level of the reservoir encountering flood disasters within a certain period of time according to the water level prediction result, and issues corresponding warning prompts.

[0004] However, the above method mainly predicts the change of reservoir water level by constructing a distributed hydrological model. Although certain results have been achieved in predicting the change of reservoir water level, due to the construction and parameter setting of the hydrological model highly depending on a large amount of historical data and empirical knowledge, including various geographical information and historical observation data of the upstream basin, the construction and adjustment process of the model is complex and time-consuming, and in the face of a complex and changeable natural environment, the model is difficult to quickly adapt and update, resulting in limited adaptability and generalization ability.

[0005] Therefore, an optimized reservoir flood control water level warning system and method are expected. Summary of the Invention

[0006] To solve the above technical problems, the present application is proposed. Embodiments of the present application provide a reservoir flood control water level warning system and method. First, rainfall data along the upstream river is obtained in real time, and the water level changes at the reservoir entrance and key river sections are monitored. Then, artificial intelligence technology based on deep learning is used to analyze the collected rainfall data and water level data to capture the spatio-temporal correlation characteristics among the rainfall data at each monitoring point along the river, and the spatio-temporal correlation characteristics among the water level data at each monitoring point within the water level monitoring area. Taking the rainfall data as the main variable and the water level data as the covariate, through time series interaction analysis of the two, the influence mechanism of rainfall changes on water level changes is revealed, so as to predict the change trend of the reservoir water level. In this way, the response speed and accuracy of reservoir flood control water level warning can be effectively improved, the dependence on human resources can be reduced, and the maintenance cost of traditional hydrological models can be reduced at the same time.

[0007] According to one aspect of the present application, a reservoir flood control water level warning method is provided, which includes: Obtain rainfall data along the upstream river in real time, and monitor the water level changes at the reservoir entrance and key river sections; Based on the obtained rainfall data and water level data, predict and analyze the change trend of the reservoir water level within a certain period of time in the future to obtain a water level prediction result; According to the water level prediction result, evaluate the risk level of the reservoir encountering flood disasters within a certain period of time in the future, and issue corresponding warnings; According to the water level prediction result and risk assessment, formulate and execute a flood discharge scheduling plan.

[0008] According to another aspect of the present application, a reservoir flood control water level warning system is provided, which includes: A data structured processing module for structuring the rainfall data and water level data to obtain a rainfall time series aggregation matrix along the upstream river and a full water level monitoring domain time series aggregation matrix; A main variable correlation encoding module for using the rainfall time series aggregation matrix along the upstream river as the main variable and performing correlation encoding on the rainfall time series aggregation matrix along the upstream river to obtain a rainfall time series correlation implicit feature vector along the river; A covariate correlation encoding module for using the full water level monitoring domain time series aggregation matrix as the covariate and performing correlation encoding on the full water level monitoring domain time series aggregation matrix to obtain a full water level monitoring domain water level time series implicit encoding aggregation matrix; A dynamic interaction analysis module for performing main-covariate dynamic interaction based on a fast scanning mechanism on the rainfall time series correlation implicit feature vector along the river and the full water level monitoring domain water level time series implicit encoding aggregation matrix to obtain a reservoir water level main-covariate time series interaction response encoding vector; A water level prediction generation module, configured to generate the water level prediction result based on the main-covariate time series interaction response coding vector of the reservoir water level.

[0009] Compared with the prior art, the reservoir flood control water level early warning system and method provided by the present application first obtains the rainfall data along the upstream river in real time, and monitors the water level changes at the reservoir inlet and key river sections. Then, it uses artificial intelligence technology based on deep learning to analyze the collected rainfall data and water level data, so as to capture the spatio-temporal correlation characteristics among the rainfall data at each monitoring point along the river, and the spatio-temporal correlation characteristics among the water level data at each monitoring point within the water level monitoring area. Taking the rainfall data as the main variable and the water level data as the covariate, through the time series interaction analysis of the two, the influence mechanism of rainfall changes on water level changes is revealed, so as to realize the prediction of the change trend of the reservoir water level. In this way, the reaction speed and accuracy of reservoir flood control water level early warning can be effectively improved, the dependence on human resources can be reduced, and the maintenance cost of traditional hydrological models can be reduced at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] By describing the embodiments of the present application in more detail with reference to the accompanying drawings, the above and other objects, features and advantages of the present application will become more obvious. The accompanying drawings are used to provide a further understanding of the embodiments of the present application, and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation to the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0011] Figure 1 It is a flowchart of the reservoir flood control water level early warning method according to an embodiment of the present application.

[0012] Figure 2 It is a schematic diagram of data flow of the reservoir flood control water level early warning method according to an embodiment of the present application.

[0013] Figure 3 It is a flowchart of sub-step S1 of the reservoir flood control water level early warning method according to an embodiment of the present application.

[0014] Figure 4 It is a flowchart of sub-step S4 of the reservoir flood control water level early warning method according to an embodiment of the present application.

[0015] Figure 5 It is a flowchart of sub-step S41 of the reservoir flood control water level early warning method according to an embodiment of the present application.

[0016] Figure 6 It is a block diagram of the reservoir flood control water level early warning system according to an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0017] As shown in this application and the claims, unless the context clearly indicates otherwise, words such as "a", "an", "one", and / or "the" are not specifically singular and may also include the plural. Generally speaking, the terms "comprising" and "including" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list. The method or device may also include other steps or elements.

[0018] Although this application makes various references to certain modules in the system according to the embodiments of this application, however, any number of different modules can be used and run on the user terminal and / or the server. The modules are only illustrative, and different aspects of the system and method can use different modules.

[0019] Flowcharts are used in this application to illustrate the operations performed by the system according to the embodiments of this application. It should be understood that the operations before or below do not necessarily need to be executed precisely in order. On the contrary, various steps can be processed in reverse order or simultaneously as needed. At the same time, other operations can also be added to these processes, or one or several steps can be removed from these processes.

[0020] Next, example embodiments according to this application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, rather than all the embodiments of this application. It should be understood that this application is not limited by the example embodiments described here.

[0021] It is worth noting that in this application, all actions of obtaining data are carried out on the premise of complying with the corresponding data protection regulations and policies of the country where it is located and obtaining the authorization given by the owner of the corresponding device.

[0022] As mentioned in the above background art, the patent CN118430190A proposes a reservoir flood control water level warning method, which includes: obtaining rainfall data along the upstream river in real time, and monitoring the water level changes at the reservoir entrance and key river sections; predicting and analyzing the future water level change trend of the reservoir within a certain period of time based on the obtained rainfall data and water level data to obtain a water level prediction result; evaluating the risk level of the reservoir encountering flood disasters within a certain period of time according to the water level prediction result, and issuing corresponding warnings; formulating and implementing a flood discharge scheduling plan based on the water level prediction result and risk assessment.

[0023] The above reservoir flood control water level warning method mainly predicts the change of the reservoir water level by constructing a distributed hydrological model. Although certain achievements have been made in predicting the change of the reservoir water level, due to the construction and parameter setting of the hydrological model highly depending on a large amount of historical data and empirical knowledge, including various geographical information and historical observation data of the upstream basin, the process of model construction and adjustment is complex and time-consuming. Moreover, when facing the complex and changeable natural environment, the model is difficult to quickly adapt and update, resulting in limited adaptability and generalization ability. To address this technical problem, this application proposes an optimized reservoir flood control water level warning method. Firstly, it obtains the rainfall data along the upstream river in real time and monitors the water level changes at the reservoir inlet and key river sections. Then, it uses artificial intelligence technology based on deep learning to analyze the collected rainfall data and water level data to capture the spatio-temporal correlation characteristics among the rainfall data at each monitoring point along the river, as well as the spatio-temporal correlation characteristics among the water level data at each monitoring point within the water level monitoring area. Taking the rainfall data as the main variable and the water level data as the covariate, through the time series interaction analysis of the two, it reveals the influence mechanism of rainfall change on water level change, so as to realize the prediction of the change trend of the reservoir water level. In this way, it can effectively improve the response speed and accuracy of the reservoir flood control water level warning, reduce the dependence on human resources, and at the same time reduce the maintenance cost of the traditional hydrological model.

[0024] With the support of modern information technology, the process of obtaining the rainfall data along the upstream river in real time first depends on a widely distributed network of rainfall monitoring stations. These stations are usually equipped with high-precision rain gauges, such as tipping bucket rain gauges, ultrasonic rain gauges or laser rain gauges, etc., which can accurately measure the precipitation and record the time series information of precipitation. The selection and layout location of the rain gauges are crucial because they directly affect the representativeness and accuracy of the data. The ideal layout principle is to cover the key areas of the entire basin, especially those sections that have a significant impact on the reservoir water level, such as the confluence of main tributaries, low-lying terrain areas, and areas prone to flood disasters historically. In addition, considering that different types of rain gauges may exhibit different performance characteristics under different climate conditions, in actual applications, the method of taking the average value by combining multiple types of rain gauges may be adopted to ensure the comprehensiveness and reliability of data collection.

[0025] In addition to the ground fixed monitoring points, mobile monitoring devices also play an important role in obtaining the rainfall data in real time. For example, unmanned aerial vehicles (UAVs) equipped with miniaturized and portable rain sensors can quickly inspect a large area in a short time; while vehicle-mounted weather stations can quickly reach the designated location for temporary monitoring in case of emergency. Such flexible and mobile monitoring means can not only supplement the blank areas between fixed stations, but also respond promptly to sudden heavy rainfall events, thus greatly enhancing the flexibility and emergency response ability of the overall monitoring system.

[0026] For the monitoring of water level changes at the reservoir entrance and key river sections, various types of automated water level monitoring devices can be deployed at the reservoir entrance and key river sections. These devices include, but are not limited to, float-type water level gauges, pressure-type water level gauges, and radar water level gauges, etc. The float-type water level gauge utilizes the principle that the float floating on the water surface moves up and down with the water level, and converts mechanical motion into an electrical signal output; the pressure-type water level gauge is based on the principle of hydrostatics and measures the pressure at the bottom of the water to determine the water level height; the radar water level gauge uses non-contact microwave radar technology and calculates the water level through the time difference between the transmitted and received reflected waves. These different types of water level gauges each have their own advantages. For example, the float-type is suitable for shallow waters, the pressure-type is applicable to deep water environments, and the radar-type features high precision and maintenance-free characteristics. Selecting a suitable instrument combination according to the specific application scenario and environmental conditions can ensure the accuracy and reliability of the monitoring data.

[0027] In addition to traditional water level gauges, high-definition cameras can also be installed near some important control points such as dams, spillways, and other important structures. These cameras are not only used for visual monitoring of the water surface state and water flow velocity, but also can assist in verifying whether the data provided by other sensors is reasonable. For example, when the water level gauge shows an abnormal increase in the water level, it is possible to confirm whether there is an actual flood threat or data error caused by sensor failure by viewing the video images. In addition, the cameras are equipped with infrared night vision functions, enabling effective monitoring even at night or under low light conditions. All camera devices are connected to the central processing center, and the captured images can be transmitted in real time and analyzed and processed to quickly respond to any emergencies.

[0028] In summary, through the above steps, rainfall data along the upstream river and water level changes at the reservoir entrance and key river sections can be systematically obtained, thus laying a solid foundation for subsequent data analysis and processing.

[0029] Figure 1 It is a flowchart of the reservoir flood control water level warning method according to an embodiment of the present application. Figure 2 It is a schematic diagram of data flow of the reservoir flood control water level warning method according to an embodiment of the present application. As Figure 1 and Figure 2As shown, the reservoir flood control water level warning method includes the steps of: S1, performing structured processing on the rainfall data and the water level data to obtain a time series aggregation matrix of rainfall along the upstream river and a time series aggregation matrix of the entire water level monitoring area; S2, using the time series aggregation matrix of rainfall along the upstream river as the main variable, performing correlation coding on the time series aggregation matrix of rainfall along the upstream river to obtain a time series correlation implicit feature vector of rainfall along the river; S3, using the time series aggregation matrix of the entire water level monitoring area as the covariate, performing correlation coding on the time series aggregation matrix of the entire water level monitoring area to obtain a time series implicit coding aggregation matrix of the water level in the entire water level monitoring area; S4, performing main-covariate dynamic interaction based on the fast scanning mechanism on the time series correlation implicit feature vector of rainfall along the river and the time series implicit coding aggregation matrix of the water level in the entire water level monitoring area to obtain a time series interaction response coding vector of the main-covariate of the reservoir water level; S5, generating the water level prediction result based on the time series interaction response coding vector of the main-covariate of the reservoir water level.

[0030] In the above reservoir flood control water level warning method, in step S1, the rainfall data and the water level data are subjected to structured processing to obtain a time series aggregation matrix of rainfall along the upstream river and a time series aggregation matrix of the entire water level monitoring area. Among them, Figure 3 is a flowchart of sub-step S1 of the reservoir flood control water level warning method according to an embodiment of the present application. As Figure 3 shown, step S1 includes the steps of: S11, arranging the rainfall data in the monitoring point sample dimension and the time dimension to form the time series aggregation matrix of rainfall along the upstream river; S12, arranging the water level data in the time dimension and the position sample dimension to form the time series aggregation matrix of the entire water level monitoring area.

[0031] Specifically, in step S11, the rainfall data is arranged in the monitoring point sample dimension and the time dimension to form the time series aggregation matrix of rainfall along the upstream river. It should be understood that considering that the rainfall data has significant spatio-temporal correlation, that is, there is a mutual relationship between the rainfall at different times and different geographical locations. Therefore, in order to be able to understand the rainfall situation along the entire upstream river in a broader context, the present application further arranges the collected rainfall data in the monitoring point sample dimension and the time dimension to construct a matrix structure that can reflect the spatio-temporal correlation of rainfall between each monitoring point along the river, and obtain the time series aggregation matrix of rainfall along the upstream river, which helps to capture the change patterns of rainfall along the river in time and space subsequently.

[0032] Specifically, in step S12, the water level data is arranged into the full water level monitoring domain time series aggregation matrix according to the time dimension and the location sample dimension. It should be understood that since the water level monitoring data also has similar spatio-temporal correlation, in this application, the water level data is also arranged into the full water level monitoring domain time series aggregation matrix according to the time dimension and the location sample dimension, so as to capture the change rules of the water level at different time points and different geographical locations. Among them, the location sample dimension represents the upstream and downstream position sequence relationship of each water level monitoring location, while the time dimension reflects the change trend of the water level data at each location over time. Through this matrix structure, the change trend of the water level at each location and the interaction between different locations can be analyzed more accurately.

[0033] Next, since the rainfall along the upstream river is directly related to the inflow of the reservoir, in this application, the rainfall time series aggregation matrix along the upstream river is used as the main variable, and the full water level monitoring domain time series aggregation matrix is used as the covariate. By performing time series interaction analysis on the two, the influence of rainfall on water level changes is revealed, so as to predict the change trend of the reservoir water level.

[0034] Specifically, in the above reservoir flood control water level warning method, in step S2, the rainfall time series aggregation matrix along the upstream river is used as the main variable, and the rainfall time series aggregation matrix along the upstream river is subjected to correlation coding to obtain the rainfall time series correlation implicit feature vector along the river. In a specific example of this application, step S2 includes: inputting the rainfall time series aggregation matrix along the upstream river into the rainfall time series correlation feature extractor based on the transformer structure to obtain the rainfall time series correlation implicit feature vector along the river. It should be understood that in order to capture the correlation change patterns of the rainfall along the upstream river in time and space, so as to better understand the influence of rainfall on the reservoir water level changes, in this application, a rainfall time series correlation feature extractor based on the transformer structure is used to perform global correlation coding processing on the rainfall time series aggregation matrix along the upstream river to obtain the rainfall time series correlation implicit feature vector along the river. Those of ordinary skill in the art should know that the transformer structure is based on the self-attention mechanism, which can effectively handle the long-distance dependence problem in sequence data, allowing the model to simultaneously focus on the feature information of all input positions, so as to capture the time series correlation features between the rainfall data of each monitoring point in the rainfall time series aggregation matrix along the upstream river globally, effectively model the spatio-temporal correlation characteristics in the rainfall data, and then obtain a feature representation that can reflect the spatio-temporal change patterns of the rainfall along the river, providing richer feature information for the subsequent reservoir water level prediction task.

[0035] In the above reservoir flood control water level warning method, in step S3, using the time series aggregation matrix of the full water level monitoring domain as a covariate, the time series aggregation matrix of the full water level monitoring domain is associated encoded to obtain a time series implicit encoding aggregation matrix of the full water level monitoring domain water level. In a specific example of the present application, step S3 includes: inputting the time series aggregation matrix of the full water level monitoring domain into a sequence encoder based on an LSTM model to obtain the time series implicit encoding aggregation matrix of the full water level monitoring domain water level. It should be understood that considering that the change of the reservoir water level is not only affected by direct rainfall but also by the hydrological conditions in the entire basin, therefore, in order to capture the change rules of water level monitoring data at different time points and different geographical locations, the present application uses a sequence encoder based on an LSTM model to perform sequence encoding processing on the time series aggregation matrix of the full water level monitoring domain. Specifically, the LSTM model, that is, the long short-term memory network, can effectively process and remember the long-term dependence information in the sequence data through its internal gating mechanism (such as input gate, output gate, and forget gate). In the technical solution of the present application, using the LSTM model to perform forward propagation sequence encoding on the water level data at each water level monitoring position in the time series aggregation matrix of the full water level monitoring domain can effectively capture the water level change rules between different geographical locations by using the long-term memory ability of the LSTM model, fully understand the potential impact of the upstream water level change on the reservoir water level, and thus provide a more accurate basis for predicting the change trend of the reservoir water level. In specific implementation, using each row vector in the time series aggregation matrix of the full water level monitoring domain (that is, the water level data at each water level monitoring position) as an input sequence, through the layer-by-layer processing of multiple hidden layers of the LSTM model, a hidden state sequence containing the cumulative effect of the upstream water level change is finally obtained, and it is combined into a time series implicit encoding aggregation matrix of the full water level monitoring domain water level, so as to realize in-depth analysis of the water level change rules.

[0036] In the above reservoir flood control water level warning method, in step S4, a main-covariate dynamic interaction based on a fast scanning mechanism is performed on the rainfall time-series correlation implicit feature vector along the river and the water level time-series implicit coding aggregation matrix of the entire water level monitoring area to obtain a reservoir water level main-covariate time-series interaction response coding vector. That is, in order to reveal the interaction between rainfall and water level changes, the present application further performs an interaction response analysis on the rainfall time-series correlation implicit feature vector along the river and the water level time-series implicit coding aggregation matrix of the entire water level monitoring area to predict the change trend of the reservoir water level. In particular, considering that there may be different correlations between water level changes and rainfall at different geographical locations. Therefore, in order to improve the correlation and efficiency of the data interaction response analysis, the present application proposes a main-covariate dynamic interaction method based on a fast scanning mechanism, which quickly identifies and strengthens the key connection between rainfall and water level changes by learning the semantic correlation between the rainfall time-series correlation implicit feature vector along the river and the water level change characteristics of each water level monitoring location, while filtering out the interference of irrelevant or noisy data, so as to achieve an accurate prediction of the change trend of the reservoir water level. Among them, Figure 4 is a flowchart of sub-step S4 of the reservoir flood control water level warning method according to an embodiment of the present application. As Figure 4 shown, step S4 includes steps: S41, based on the semantic correlation distribution between the rainfall time-series correlation implicit feature vector along the river and the water level time-series implicit coding aggregation matrix of the entire water level monitoring area, screen out a subsequence of the water level monitoring area water level time-series implicit coding vector from the water level time-series implicit coding aggregation matrix of the entire water level monitoring area; S42, perform cross-domain interaction fusion on the rainfall time-series correlation implicit feature vector along the river and the subsequence of the water level monitoring area water level time-series implicit coding vector to obtain the reservoir water level main-covariate time-series interaction response coding vector.

[0037] Specifically, Figure 5 is a flowchart of sub-step S41 of the reservoir flood control water level warning method according to an embodiment of the present application. As Figure 5As shown, step S41 includes steps: S411, splitting the full water level monitoring domain water level time series implicit coding aggregation matrix according to the position sample dimension to obtain a sequence of water level monitoring domain water level time series implicit coding vectors; S412, calculating the mutual information between the rainfall time series correlation implicit feature vector along the river and each water level monitoring domain water level time series implicit coding vector in the sequence of water level monitoring domain water level time series implicit coding vectors to obtain a sequence of reservoir water level main-covariate fast semantic matching factors; S413, identifying the first maximum value and the second maximum value from the sequence of reservoir water level main-covariate fast semantic matching factors; S414, based on the positions of the first maximum value and the second maximum value in the sequence of reservoir water level main-covariate fast semantic matching factors, determining a subsequence of the water level monitoring domain water level time series implicit coding vectors that are quickly matched from the sequence of water level monitoring domain water level time series implicit coding vectors.

[0038] More specifically, S411 is expressed by the formula:

[0039] Where represents the sequence of water level monitoring domain water level time series implicit coding vectors, , , and respectively represent the first, second, th, th water level monitoring domain water level time series implicit coding vectors in the sequence of water level monitoring domain water level time series implicit coding vectors, is the number of water level monitoring domain water level time series implicit coding vectors.

[0040] That is, splitting the full water level monitoring domain water level time series implicit coding aggregation matrix into a sequence of water level monitoring domain water level time series implicit coding vectors according to the position sample dimension, where each vector corresponds to a specific water level monitoring position.

[0041] More specifically, S412 is expressed by the formula:

[0042]

[0043]

[0044] Where represents the rainfall time series correlation implicit feature vector along the river, represents the sigmoid normalization function, Denote the normalized rainfall time - series correlation implicit feature vector along the river, Denote the th normalized water - level monitoring domain water - level time - series implicit encoding vector, Denote the information entropy of the normalized rainfall time - series correlation implicit feature vector along the river, Denote the th information entropy of the normalized water - level monitoring domain water - level time - series implicit encoding vector, Denote the joint entropy between the normalized rainfall time - series correlation implicit feature vector along the river and the th normalized water - level monitoring domain water - level time - series implicit encoding vector, Denote the th reservoir water - level main - covariate fast semantic matching factor.

[0045] That is, by calculating the mutual information between each water - level monitoring domain water - level time - series implicit encoding vector and the rainfall time - series correlation implicit feature vector along the river respectively, the feature correlation between the two is evaluated, and a sequence of reservoir water - level main - covariate fast semantic matching factors is generated. It should be understood that the larger the mutual information value, the higher the correlation between the two; conversely, the smaller the mutual information value, the weaker the correlation between the two. In this way, not only can the water - level change features most relevant to rainfall changes be effectively identified, but also an objective standard can be provided for subsequent fast matching through quantification.

[0046] More specifically, S413 is expressed by the formula:

[0047]

[0048] where, Denote the maximum - value function, and Denote the first maximum value and the second maximum value respectively.

[0049] More specifically, S414 is expressed by the formula:

[0050]

[0051]

[0052] where, Denote the index corresponding to the maximum value, and respectively represent the position indices of the first maximum value and the second maximum value in the sequence of the reservoir water level main-covariate fast semantic matching factor, and respectively represent the water level monitoring domain water level time series implicit coding vectors corresponding to the first maximum value and the second maximum value, represents a subsequence of the water level monitoring domain water level time series implicit coding vector.

[0053] That is, two maximum values are queried from the sequence of the reservoir water level main-covariate fast semantic matching factor, and based on the positions of the two maximum values in the sequence, a subsequence of the water level monitoring domain water level time series implicit coding vector significantly related to rainfall is intercepted from the corresponding sequence of the water level monitoring domain water level time series implicit coding vector, realizing preliminary feature selection and filtering, so as to concentrate resources to process the most relevant interaction information, reduce unnecessary calculations, and at the same time reduce the noise interference in the water level change information.

[0054] Specifically, in a specific example of the present application, the step S42 includes: performing a linear transformation on the rainfall time series correlation implicit feature vector along the river to obtain a query vector and a value vector, and using the subsequence of the water level monitoring domain water level time series implicit coding vector as the subsequence of the key vector, and inputting the query vector, the value vector and the subsequence of the key vector into a cross-domain query coding module based on a transformer structure to obtain the reservoir water level main-covariate time series interaction response coding vector, which is expressed by the formula:

[0055]

[0056]

[0057] Among them, represents the th water level monitoring domain water level time series implicit coding vector, and respectively represent the query embedding matrix and the value embedding matrix, and respectively represent the query bias vector and the value bias vector, and respectively represent the query vector and the value vector, S represents the length of the th water level monitoring domain water level time series implicit coding vector, represents the transpose of the vector, represents the normalized exponential function, represents the matrix multiplication operation, represents the reservoir water level main-covariate time series interaction response coding vector.

[0058] That is, a query vector and a value vector are constructed based on the rainfall time-series correlation implicit feature vector along the river, and a subsequence of the water level time-series implicit coding vector in the selected water level monitoring domain is used as a subsequence of the key vector. Cross-domain query interaction is performed through a transformer, and the self-attention mechanism is used to learn the correlation and dependence relationship between rainfall information and water level change information, generating a reservoir water level main-covariate time-series interaction response coding vector, so as to achieve an in-depth understanding of the correlation response relationship between upstream rainfall changes and reservoir water level changes.

[0059] In the above reservoir flood control water level warning method, in step S5, based on the reservoir water level main-covariate time-series interaction response coding vector, the water level prediction result is generated. In a specific example of the present application, step S5 includes: inputting the reservoir water level main-covariate time-series interaction response coding vector into a water level prediction engine based on an RNN model to obtain the water level prediction result. Those of ordinary skill in the art should know that the RNN model is a recurrent neural network model, which can effectively remember and utilize the information of the previous moment through its internal loop structure to affect the output of the current moment. Here, the RNN model receives the reservoir water level main-covariate time-series interaction response coding vector as input, and by using its internal loop structure to remember and utilize the rainfall information and the correlation influence information of rainfall on water level changes contained in the reservoir water level main-covariate time-series interaction response coding vector, it can accurately predict the future change trend of the reservoir water level. In addition, since the RNN model can share parameters between each time step and maintain a continuous state update process, it can maintain coherence and consistency when predicting future values, making the output water level prediction result smoother, reducing abrupt fluctuations, and being closer to the gradual changes in the actual physical process, thereby further improving the stability and reliability of the prediction.

[0060] Here, in the case where the rainfall time-series correlation implicit feature vector along the river and the water level time-series implicit coding aggregation matrix of the entire water level monitoring domain respectively represent the rainfall time-series correlation feature along the river and the water level time-series correlation feature of each water level monitoring position, when performing semantic query coding based on the fast scanning mechanism, considering the fast scanning semantic query differences caused by the source time-series distribution differences, it is expected to improve the semantic consistent aggregation expression effect of the reservoir water level main-covariate time-series interaction response coding vector obtained through response semantic coding.

[0061] In a preferred example of the present application, before inputting the reservoir water level main-covariate time series interaction response coding vector into the water level prediction engine based on the RNN model to obtain the water level prediction result, semantic consistency aggregation optimization is performed on the reservoir water level main-covariate time series interaction response coding vector to obtain an optimized reservoir water level main-covariate time series interaction response coding vector, and the process includes the following steps: Reveal the internal structure of the reservoir water level main-covariate time series interaction response coding vector through independent component analysis paradigm parsing to obtain the main-covariate time series interaction response principal component correlation activation matrix and the main-covariate time series interaction response feature fluctuation characterization matrix, expressed as:

[0062]

[0063]

[0064] Among them, represents the reservoir water level main-covariate time series interaction response coding vector, and are respectively the eigenvalues at the -th and -th positions of the reservoir water level main-covariate time series interaction response coding vector, represents the value at the -th position of the main-covariate time series interaction response principal component correlation activation matrix, represents the value at the -th position of the main-covariate time series interaction response feature fluctuation characterization matrix.

[0065] Using the main-covariate time series interaction response principal component correlation activation matrix and the main-covariate time series interaction response feature fluctuation characterization matrix as high-dimensional directional embedding coding matrices respectively, perform high-dimensional non-linear embedding coding on the reservoir water level main-covariate time series interaction response coding vector to obtain the main-covariate time series interaction response semantic association excitation coding vector and the main-covariate time series interaction response semantic fluctuation excitation coding vector, expressed as:

[0066]

[0067] Among them, represents the main-covariate time series interaction response principal component correlation activation matrix, represents the main-covariate time series interaction response feature fluctuation characterization matrix, represents matrix multiplication, represents the rectified linear unit function, Represents the main-covariate time-series interaction response semantic association excitation coding vector, Represents the main-covariate time-series interaction response semantic fluctuation excitation coding vector.

[0068] Perform self-association coding on the main-covariate time-series interaction response semantic association excitation coding vector and the main-covariate time-series interaction response semantic fluctuation excitation coding vector to obtain the main-covariate time-series interaction response semantic structure consistency potential coding matrix, denoted as:

[0069] Wherein, Represents the length of the vector, Represents the transpose of the vector, Represents the normalization exponential function, Represents the main-covariate time-series interaction response semantic structure consistency potential coding matrix.

[0070] After feature fusion of the main-covariate time-series interaction response semantic association excitation coding vector and the main-covariate time-series interaction response semantic fluctuation excitation coding vector, input the result of feature fusion into the main-covariate time-series interaction response semantic structure consistency potential coding matrix for semantic structure consistency explicit modeling to obtain the optimized reservoir water level main-covariate time-series interaction response coding vector, denoted as:

[0071] Wherein, Represents the activation function, such as the sigmoid function, Represents feature fusion, such as concatenation, Represents the weight matrix, obtained through training, Represents the optimized reservoir water level main-covariate time-series interaction response coding vector.

[0072] Correspondingly, on the basis of analyzing the local region activation foci of the respective eigenvalues of the reservoir water level main-covariate time-series interaction response coding vector through the independent component analysis paradigm, use a high-dimensional non-linear embedding network to perform correlation analysis on the potential spatial structure similarity between the local structural features and the feature vector. Thus, optimize the sparsity of the representation vector by means of self-association to establish a mapping relationship between the construction element relevance and the system-level complexity, so as to promote the convergence of the reservoir water level main-covariate time-series interaction response coding vector towards the semantic consistency axis direction to enhance the semantic consistent aggregation expression effect of the reservoir water level main-covariate time-series interaction response coding vector. In this way, improve the accuracy of the water level prediction result obtained by inputting the improved reservoir water level main-covariate time-series interaction response coding vector into the water level prediction engine based on the RNN model.

[0073] Next, in order to evaluate the risk level of flood disasters in reservoirs in the future based on water level forecast results and issue corresponding warnings, the entire process needs to be integrated from multiple perspectives, including the establishment of a risk assessment indicator system, multi-source data analysis and integration, risk level classification and warning level setting, and warning information release mechanism. The following is a comprehensive integration and detailed description of this process: First, in order to accurately assess the risk level of reservoirs encountering flood disasters, a scientific and reasonable risk assessment indicator system must be established. This system should include but is not limited to the following aspects: Water level threshold: Setting different levels of warning water levels (such as warning water levels, dangerous water levels, etc.), these thresholds are determined based on the experience summary of historical flood events and professional analysis of the reservoir's carrying capacity; Flow changes: Pay attention to the changing trend of inflow flow, especially the sudden increase in flow in a short period of time, which is often a precursor to the arrival of the flood peak; Rainfall intensity and duration: The rainfall in the upstream basin directly determines the amount of water flowing into the reservoir. Heavy rainfall or long-term continuous rainfall will significantly increase the risk of floods; Topographic and geomorphic features: Analyze the terrain undulation of the upstream basin , slope, soil type and other factors, which will affect the infiltration rate of rainwater and the runoff formation process; vegetation coverage: good vegetation can slow down the speed of surface runoff and reduce the possibility of sediment entering the river; changes in climate patterns: considering the impact of global climate change, long-term climate trend forecasts are also instructive for short-term flood forecasts; impact of human activities: urbanization level, agricultural irrigation demand, etc. may change the characteristics of local water cycle and increase or reduce flood risks; emergency response capabilities: assess the ability of local governments and relevant departments to respond to emergencies, such as the professional quality of rescue teams and the status of material reserves.

[0074] After obtaining the data of the above indicators, multi-source data analysis and fusion are carried out. Here, relatively simple and intuitive methods can be used, such as linear regression models, statistical methods, etc. According to the results obtained from multi-source data analysis, the risks faced by the reservoir are quantitatively scored according to pre-defined rules, and different risk levels are divided accordingly. Generally speaking, it can be divided into four levels: low risk, medium risk, high risk and extremely high risk. Each level corresponds to a different warning level (such as blue warning, yellow warning, orange warning and red warning), and is accompanied by specific countermeasures. The setting of the warning level should take into account the actual situation and allow dynamic adjustment to adapt to changing environmental conditions.

[0075] To ensure that early warning information can be quickly and effectively disseminated to the target audience, it is necessary to build a multi-level and all-round information release network. This can be achieved through the following methods: SMS platform, sending short and clear early warning messages to residents in the affected areas via SMS; social media platforms, making full use of social channels such as Weibo and WeChat official accounts to expand the information coverage and promote community interaction; traditional media channels, continuing to play the role of traditional media such as TV and radio to ensure that information reaches all people; community bulletin boards, setting up bulletin boards in community centers or public places to post the latest early warning information; emergency broadcast systems, installing wireless transmission devices in specific areas, which can continue to work in the event of a power outage to ensure that important notices are transmitted without dead ends.

[0076] Once the early warning level is determined, the corresponding follow-up measures and contingency plans become particularly important. This includes an evacuation plan that clarifies details such as evacuation routes, assembly point locations, and transportation arrangements; pre-stocking sufficient daily necessities and emergency supplies in advance; organizing medical teams to be on standby at any time, preparing ambulances, first aid equipment, and medicines; planning post-disaster reconstruction plans, paying attention to ecological protection and restoration to avoid the occurrence of secondary disasters.

[0077] In summary, through the above series of measures, a closed-loop management system from risk assessment to early warning release and then to emergency response can be constructed to effectively prevent and reduce the losses caused by flood disasters. This process not only relies on advanced technologies and scientific methods but also requires multi-party cooperation and efficient communication to ensure that every link can be smoothly executed and make full preparations for possible natural disasters.

[0078] Then, based on the risk assessment results, formulate and execute a flood discharge scheduling plan. This process not only depends on accurate water level predictions and risk assessment results but also needs to fully consider various factors such as reservoir operation rules, the safety of downstream residents, ecological environment protection, and the rational utilization of water resources. During this process, it is necessary to ensure the timeliness and accuracy of information and maintain effective communication with all stakeholders.

[0079] First, after receiving the water level prediction results based on the rainfall data along the upstream river and the time-series aggregation matrix of the entire water level monitoring area, quickly analyze the latest weather forecast, the real-time monitored water level change trend, and other natural or human factors that may affect the water level in the next few days. Based on this information, determine whether it is necessary to initiate flood discharge operations and the initially set target flood discharge volume.

[0080] Next, to optimize the flood discharge strategy, mathematical programming methods can be applied to find the optimal flood discharge path. For example, through linear programming or dynamic programming techniques, a solution that can maximize the flood control effect while ensuring downstream safety can be found under the premise of meeting the legal regulations for reservoir operation (such as the minimum ecological flow requirement). At the same time, considering the influence of seasonal factors, the water level should be kept relatively high during the dry season for subsequent use, and the water level should be appropriately lowered before the rainy season to reserve space for possible heavy rains. In addition, simulation software can be used to test different flood discharge scenarios, evaluate their effects, and select the most suitable plan.

[0081] Once the specific flood discharge plan is determined, immediate preparations need to be made. This requires notifying relevant departments and individuals to make preparations according to the established procedures to ensure that all preparatory work is in place. When the risk assessment shows a high possibility of flood disasters, the pre-set emergency plan should be immediately activated. According to the requirements of the plan, organize personnel to inspect and maintain the flood discharge facilities to ensure that they are in good working condition; at the same time, arrange the emergency rescue team to standby so that they can respond quickly in case of emergencies.

[0082] Before officially starting the flood discharge, it should be confirmed again that all conditions are met, including but not limited to: the spillway gates have been comprehensively overhauled and can be opened and closed normally; the downstream communities have been fully informed of the upcoming flood discharge operation and have taken necessary preventive measures; the communication system is unobstructed and can receive feedback information from the scene at any time. Then, gradually open the spillway gates or other flood discharge facilities according to the pre-calculated flood discharge plan. During this process, special attention should be paid to monitoring the water flow velocity, water quality, etc. to prevent secondary pollution or other problems caused by the flood discharge. If any abnormal situation is found during the flood discharge period, such as the actual water level rising faster than expected, the flood discharge strategy needs to be adjusted flexibly to ensure the safety and controllability of the whole process.

[0083] To ensure transparency and public trust during the flood discharge process, information disclosure channels must be established to promptly inform the residents in the affected areas of the latest water situation information and preventive measures. Official statements can be issued through various media platforms, such as social media accounts, local TV stations, radio stations, etc., so that the public can learn about the latest developments in a timely manner. In addition, community meetings should be actively organized to listen to opinions and enhance the public's understanding and cooperation with flood control work. Good social communication can not only improve the public's sense of security but also help reduce unnecessary panic and social instability factors.

[0084] During the entire flood discharge process, it is also necessary to closely monitor the environmental changes inside and outside the reservoir, such as the changes in key indicators such as rainfall intensity, river water level, and reservoir water level. If the actual situation deviates from the expectation, it is necessary to respond quickly and flexibly adjust the flood discharge strategy to ensure that the ultimate goals, namely protecting the lives and property of downstream residents and maintaining the stability of the natural environment, are achieved.

[0085] Finally, a comprehensive review after the flood discharge is an essential step. Based on the actual situation and the data collected, evaluate the effectiveness and existing problems of the flood discharge operation, and update and iterate the relevant models and algorithms to improve the prediction accuracy and response speed. This is also one of the important means to continuously improve and perfect the flood discharge scheduling system. To sum up, formulating and implementing a flood discharge scheduling plan is a comprehensive task involving multidisciplinary knowledge and technical applications. Through scientific and reasonable planning, effective technical support, and good social coordination, the damage caused by flood disasters can be minimized to the greatest extent, and the safety of people's lives and property and the sustainable development of the natural environment can be ensured.

[0086] To sum up, the reservoir flood control water level early warning method based on the embodiments of the present application is elucidated. It first obtains the rainfall data along the upstream river in real time and monitors the water level changes at the reservoir entrance and key river sections. Then, it uses artificial intelligence technology based on deep learning to analyze the collected rainfall data and water level data to capture the spatio-temporal correlation characteristics among the rainfall data at each monitoring point along the river, and the spatio-temporal correlation characteristics among the water level data at each monitoring point within the water level monitoring area. Taking the rainfall data as the main variable and the water level data as the covariate, through the time series interaction analysis of the two, the influence mechanism of rainfall changes on water level changes is revealed, so as to realize the prediction of the change trend of the reservoir water level. In this way, the response speed and accuracy of the reservoir flood control water level early warning can be effectively improved, the dependence on human resources can be reduced, and at the same time, the maintenance cost of traditional hydrological models can be lowered.

[0087] Furthermore, a reservoir flood control water level early warning system is also provided.

[0088] Figure 6 The block diagram of the reservoir flood control water level early warning system according to the embodiments of the present application is as follows. As Figure 6As shown, the reservoir flood control water level warning system 100 according to an embodiment of the present application includes: a data structuring and processing module 110, configured to perform structuring and processing on rainfall data and water level data to obtain a rainfall time series aggregation matrix along the upstream river and a full water level monitoring domain time series aggregation matrix; a main variable correlation encoding module 120, configured to use the rainfall time series aggregation matrix along the upstream river as the main variable, and perform correlation encoding on the rainfall time series aggregation matrix along the upstream river to obtain a rainfall time series correlation implicit feature vector along the river; a covariate correlation encoding module 130, configured to use the full water level monitoring domain time series aggregation matrix as the covariate, and perform correlation encoding on the full water level monitoring domain time series aggregation matrix to obtain a full water level monitoring domain water level time series implicit encoding aggregation matrix; a dynamic interaction analysis module 140, configured to perform main-covariate dynamic interaction based on a fast scanning mechanism on the rainfall time series correlation implicit feature vector along the river and the full water level monitoring domain water level time series implicit encoding aggregation matrix to obtain a reservoir water level main-covariate time series interaction response encoding vector; and a water level prediction generation module 150, configured to generate the water level prediction result based on the reservoir water level main-covariate time series interaction response encoding vector.

[0089] Here, those skilled in the art can understand that the specific operations of each module in the above reservoir flood control water level warning system have been described in detail above with reference to Figures 1 to 5 the description of the reservoir flood control water level warning method, and therefore, the repeated description thereof will be omitted.

[0090] The basic principles of the present invention have been described above in conjunction with specific embodiments. However, it should be noted that the advantages, advantages, effects, etc. mentioned in the present invention are only examples and not limitations, and it cannot be considered that these advantages, advantages, effects, etc. are essential for each embodiment of the present invention. In addition, the specific details of the above embodiments are only for the purpose of illustration and easy understanding, rather than limitations. The above details do not limit the present invention to necessarily adopt the above specific details to implement.

[0091] In the above embodiments, the descriptions of each embodiment have their own focuses. For parts not detailed or recorded in a certain embodiment, reference can be made to the relevant descriptions of other embodiments. In several embodiments provided by the present invention, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are only illustrative. For example, the unit division is only a logical function division, and there may be other division methods in actual implementation. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0092] It is obvious to those skilled in the art that the present invention is not limited to the details of the above-described exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention. Therefore, in any regard, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the claims are intended to be embraced within the present invention. Any reference signs in the claims should not be construed as limiting the claims involved.

[0093] In addition, it is obvious that the term "comprising" does not exclude other elements or steps, and the singular does not exclude the plural. The multiple elements recited in the system claims can also be implemented by one element through software or hardware.

[0094] Finally, it should be noted that the above description has been given for purposes of illustration and description. In addition, the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A reservoir flood control water level early warning method, comprising: Real-time acquisition of rainfall data along the upstream river, and monitoring of water level changes at the reservoir entrance and key river sections; forecasting and analyzing the reservoir water level change trend in a certain period of time in the future based on the acquired rainfall data and water level data to obtain water level forecast results; assessing the risk level of flood disasters encountered by the reservoir in the future based on the water level forecast results, and issuing corresponding warnings; formulating and executing flood discharge scheduling plans based on the water level forecast results and risk assessment, which is characterized in that forecasting and analyzing the reservoir water level change trend in a certain period of time in the future based on the acquired rainfall data and water level data to obtain water level forecast results, including: Performing structured processing on the rainfall data and the water level data to obtain a rainfall time series aggregation matrix along the upstream river and a time series aggregation matrix for the entire water level monitoring domain; Taking the rainfall time series aggregation matrix along the upstream river as the main variable, the rainfall time series aggregation matrix along the upstream river is associated encoded to obtain the rainfall time series associated implicit feature vector along the river; Taking the time series aggregation matrix of the entire water level monitoring domain as a covariate, the time series aggregation matrix of the entire water level monitoring domain is associated encoded to obtain an implicitly encoded aggregation matrix of the water level time series of the entire water level monitoring domain; Performing a main-covariate dynamic interaction based on a fast scanning mechanism on the rainfall time series associated implicit feature vector along the river and the water level time series implicit coding aggregation matrix of the entire water level monitoring domain to obtain a reservoir water level main-covariate time series interactive response coding vector; The water level prediction result is generated based on the reservoir water level main-covariate time series interaction response encoding vector.

2. The reservoir flood control water level early warning method according to claim 1, characterized in that: The rainfall data and the water level data are structured to obtain a rainfall time series aggregation matrix along the upstream river and a time series aggregation matrix of the entire water level monitoring domain, including: Arranging the rainfall data into a rainfall time series aggregation matrix along the upstream river according to the monitoring point sample dimension and the time dimension; The water level data are arranged according to the time dimension and the location sample dimension into the full water level monitoring domain time series aggregation matrix.

3. The reservoir flood control water level early warning method according to claim 2, characterized in that: The rainfall time series aggregation matrix along the upstream river is subjected to association coding to obtain an implicit feature vector of rainfall time series association along the river, including: The rainfall time series aggregation matrix along the upstream river is input into the rainfall time series correlation feature extractor along the river based on the converter structure to obtain the rainfall time series correlation implicit feature vector along the river.

4. The reservoir flood control water level early warning method according to claim 3, characterized in that: The whole water level monitoring domain time series aggregation matrix is ​​associated encoded to obtain the whole water level monitoring domain water level time series implicit encoding aggregation matrix, including: The full water level monitoring domain time series aggregation matrix is ​​input into a sequence encoder based on the LSTM model to obtain the full water level monitoring domain water level time series implicit coding aggregation matrix.

5. The reservoir flood control water level early warning method according to claim 4, characterized in that: The main-covariate dynamic interaction based on the fast scanning mechanism is performed on the rainfall time series associated implicit feature vector along the river and the water level time series implicit coding aggregation matrix of the whole water level monitoring domain to obtain the reservoir water level main-covariate time series interactive response coding vector, including: Based on the semantic correlation distribution between the rainfall time series associated implicit feature vector along the river and the water level time series implicit coding aggregation matrix of the entire water level monitoring domain, a subsequence of the water level time series implicit coding vector of the water level monitoring domain is screened out from the water level time series implicit coding aggregation matrix of the entire water level monitoring domain; The subsequences of the implicit feature vector associated with the rainfall time series along the river and the implicit coding vector of the water level time series in the water level monitoring domain are cross-domain interactively fused to obtain the main-covariate time series interactive response coding vector of the reservoir water level.

6. The reservoir flood control water level early warning method according to claim 5, characterized in that: Based on the semantic correlation distribution between the rainfall time series associated implicit feature vector along the river and the water level time series implicit coding aggregation matrix of the entire water level monitoring domain, a subsequence of the water level time series implicit coding vector of the water level monitoring domain is screened out from the water level time series implicit coding aggregation matrix of the entire water level monitoring domain, including: Splitting the water level time series implicit coding aggregation matrix of the entire water level monitoring domain according to the position sample dimension to obtain a sequence of water level time series implicit coding vectors of the water level monitoring domain; Calculate the mutual information between the implicit feature vector associated with the rainfall time series along the river and each water level monitoring domain water level time series implicit coding vector in the sequence of the water level monitoring domain water level time series implicit coding vector to obtain a sequence of reservoir water level main-covariate fast semantic matching factors; identifying a first maximum value and a second maximum value from a sequence of primary-covariate fast semantic matching factors of the reservoir water level; Based on the positions of the first maximum value and the second maximum value in the sequence of the reservoir water level principal-covariate fast semantic matching factors, a subsequence of the water level monitoring domain water level timing implicit coding vector for fast matching is determined from the sequence of the water level monitoring domain water level timing implicit coding vector.

7. The reservoir flood control water level early warning method according to claim 6, characterized in that: The subsequences of the rainfall time series associated implicit feature vector along the river and the water level time series implicit coding vector of the water level monitoring domain are cross-domain interactively fused to obtain the reservoir water level main-covariate time series interactive response coding vector, including: The implicit feature vector associated with the rainfall time series along the river is linearly transformed to obtain a query vector and a value vector, and a subsequence of the implicit coding vector of the water level time series in the water level monitoring domain is used as a subsequence of the key vector. The query vector, the value vector and the subsequence of the key vector are input into a cross-domain query coding module based on a converter structure to obtain the reservoir water level principal-covariate time series interactive response coding vector.

8. The reservoir flood control water level early warning method according to claim 7, characterized in that: Based on the reservoir water level main-covariate time series interactive response coding vector, the water level prediction result is generated, including: The reservoir water level principal-covariate time series interaction response encoding vector is input into a water level prediction engine based on an RNN model to obtain the water level prediction result.

9. A reservoir flood control water level early warning system, characterized in that: include: A data structured processing module is used to perform structured processing on rainfall data and water level data to obtain a rainfall time series aggregation matrix along the upstream river and a time series aggregation matrix for the entire water level monitoring domain; A main variable association coding module is used to use the rainfall time series aggregation matrix along the upstream river as the main variable, and to perform association coding on the rainfall time series aggregation matrix along the upstream river to obtain an implicit feature vector of rainfall time series association along the river; A covariate association coding module is used to use the full water level monitoring domain time series aggregation matrix as a covariate, and to perform association coding on the full water level monitoring domain time series aggregation matrix to obtain a full water level monitoring domain water level time series implicit coding aggregation matrix; A dynamic interaction analysis module is used to perform a main-covariate dynamic interaction based on a fast scanning mechanism on the implicit characteristic vector of the rainfall time series associated with the river and the implicit coding aggregation matrix of the water level time series of the entire water level monitoring domain to obtain a main-covariate time series interaction response coding vector of the reservoir water level; The water level prediction generation module is used to generate the water level prediction result based on the reservoir water level main-covariate time series interactive response coding vector.

Images