Flood peak evolution path and dam break risk early warning method and system

Abstract

The invention provides a flood peak evolution path and dam break risk early warning method and system. According to the method, a time-space coupling water conservancy data set is constructed by fusing multi-dimensional water conservancy monitoring data and regional rainfall prediction information, and a basin topology perception model is established. And further performing mode matching on the parameters and a historical dam break event characteristic spectrum, and generating a dynamic response strategy set including a flood storage and detention area capacity allocation scheme by correcting a space weight of a matching result. And finally, based on a watershed topology perception model, matching the strategy set with historical dam break features, realizing accurate identification of a flood peak evolution path and graded early warning of dam break risks, outputting a water conservancy analysis report containing risk grades, and realizing closed-loop management from data fusion and dynamic prediction to risk decision. According to the technical scheme provided by the invention, the efficiency and accuracy of intelligent analysis of the water conservancy data can be improved.

Description

Technical Field

[0001] This application relates to the technical field of intelligent analysis of water conservancy data, and particularly relates to a method and system for warning of flood peak evolution path and dam break risk. Background Art

[0002] In the fields of basin flood control and water conservancy project scheduling, with the frequent occurrence of extreme meteorological events and the exacerbation of the problem of aging water conservancy facilities, the need for dynamic and high-precision risk assessment and real-time decision support is becoming increasingly urgent. Currently, it is necessary to break through the bottleneck of spatio-temporal coupling modeling of multi-source heterogeneous data, accurately quantify the synergistic effects of hydrodynamic propagation, geological deformation, and meteorological driving, and achieve dynamic tracking of the flood peak evolution path and cross-space recursion of risk parameters. At the same time, it is necessary to deeply integrate historical disaster patterns with the real-time basin state to generate an adaptive response strategy that combines physical mechanisms and data-driven.

[0003] Currently, in response to the above needs, existing technical solutions mainly adopt a multi-modal data fusion framework based on deep learning. For example, long short-term memory networks are used to extract the temporal correlation features of meteorological precipitation and hydrological monitoring data, combined with convolutional neural networks to mine spatial distribution laws, and the change trend of regional runoff is predicted through end-to-end training, and a risk assessment report is generated by matching with the historical dam break event database. Such methods have realized the statistical modeling of the relationship between precipitation and runoff to a certain extent and can provide a coarse-grained prediction of the flood peak arrival time.

[0004] However, the existing solutions have significant limitations. First, the monitoring data of riverbed deformation is not integrated, resulting in the model being unable to perceive the dynamic impact of river channel scouring and silting on the flood propagation path, especially the prediction error is significantly amplified in areas with severe riverbed deformation. Second, spatio-temporal modeling relies on discrete feature splicing, lacking the embedding of hydrodynamic propagation rules based on the basin topology structure, making it difficult to characterize the cascading diffusion and energy attenuation characteristics of flood waves in complex river networks. Finally, risk assessment and strategy generation still rely on static historical templates, without introducing a weight correction mechanism for real-time hydrodynamic state and topological correlation, resulting in the disconnection between emergency strategies and dynamic parameters such as precipitation-driven deformation and water flow propagation delay in the current basin, and the practical operability is limited. Summary of the Invention

[0005] This application provides a method and system for warning of flood peak evolution path and dam break risk to solve the problems of low efficiency and poor accuracy in the intelligent analysis of water conservancy data in the prior art.

[0006] In a first aspect, this application provides a method for warning of flood peak evolution path and dam break risk, including:

[0007] Collect multi-dimensional water conservancy monitoring data and perform spatio-temporal dimension fusion with the regional precipitation prediction information provided by the meteorological service platform to generate a spatio-temporal coupled water conservancy data set;

[0008] Construct a basin topology perception model based on the spatio-temporal coupled water conservancy data set, take the correlation fluctuation between the riverbed deformation trend and the water flow dynamic characteristics of upstream and downstream nodes in the multi-dimensional water conservancy monitoring data as the identification benchmark for the flood peak evolution path, and generate runoff mutation probability parameters for different geographical units through the regional precipitation prediction information;

[0009] Perform pattern matching according to the runoff mutation probability parameters and the characteristic map of the historical dam break events in the preset basin, and generate a set of dynamic response strategies for water conservancy projects based on the spatial weights of the corrected pattern matching results of the identification benchmark;

[0010] Match the characteristic map of the historical dam break events in the basin and the flood storage and detention area capacity allocation plan in the set of dynamic response strategies for water conservancy projects, and generate an intelligent analysis report of water conservancy data including risk levels through the basin topology perception model.

[0011] Optionally, the constructing a basin topology perception model based on the spatio-temporal coupled water conservancy data set, taking the correlation fluctuation between the riverbed deformation trend and the water flow dynamic characteristics of upstream and downstream nodes in the multi-dimensional water conservancy monitoring data as the identification benchmark for the flood peak evolution path, and generating runoff mutation probability parameters for different geographical units through the regional precipitation prediction information includes:

[0012] Based on the spatial position relationship in the spatio-temporal coupled water conservancy data set, establish a basin topology perception model, and the connection weights of each upstream and downstream node of the basin topology perception model are determined by the covariance relationship between the water flow dynamic characteristics and the riverbed deformation trend between adjacent upstream and downstream nodes;

[0013] Extract the time series gradient of the riverbed deformation trend of each upstream and downstream node in the basin topology perception model to obtain the dynamic coupling coefficient between the riverbed deformation trend and the water flow dynamic characteristics of the upstream and downstream nodes;

[0014] Identify the correlation fluctuation between the water flow dynamic characteristics of the upstream and downstream nodes and the riverbed deformation trend according to the distribution characteristics of the dynamic coupling coefficient, and mark the synchronous mutation section of the dynamic coupling coefficient of continuous nodes in the correlation fluctuation as the candidate set of the flood peak evolution path;

[0015] Based on the spatial distribution density of the regional precipitation prediction information, superimpose the precipitation intensity gradient on the geographical units covered by the candidate set of the flood peak evolution path to generate runoff mutation probability parameters.

[0016] Optionally, the identifying the correlation fluctuation between the water flow dynamic characteristics of the upstream and downstream nodes and the riverbed deformation trend according to the distribution characteristics of the dynamic coupling coefficient includes:

[0017] Traverse the dynamic coupling coefficients and the riverbed deformation trends of the upstream and downstream nodes in the watershed topology awareness model. If the downstream mutation amplitude attenuation occurs between adjacent upstream and downstream nodes within the same time period, it is marked as a potential propagation segment;

[0018] Connect the potential propagation segments together. When there are multiple upstream segments for the upstream and downstream nodes, select the best upstream segment based on the collaborative constraints of the attenuation gradient of the dynamic coupling coefficient, the attenuation gradient of the riverbed deformation trend, and the distance between the upstream and downstream nodes to generate a candidate propagation chain;

[0019] Calculate the ratio of the comprehensive attenuation rate of the attenuation gradient of the dynamic coupling coefficient and the attenuation gradient of the riverbed deformation trend between adjacent upstream and downstream nodes in the candidate propagation chain. If the ratio exceeds the preset threshold, truncate the candidate propagation chain;

[0020] Based on the synchronization of the attenuation rates of the dynamic coupling coefficient and the riverbed deformation trend, identify the best truncated candidate propagation chain segment as the most strongly synchronized associated fluctuation.

[0021] Optionally, the connecting the potential propagation segments together. When there are multiple upstream segments for the upstream and downstream nodes, select the best upstream segment based on the collaborative constraints of the attenuation gradient of the dynamic coupling coefficient, the attenuation gradient of the riverbed deformation trend, and the distance between the upstream and downstream nodes to generate a candidate propagation chain includes:

[0022] Traverse the potential propagation segments of the upstream nodes, and extract the attenuation gradient of the dynamic coupling coefficient, the attenuation gradient of the riverbed deformation trend, and the distance between the upstream and downstream nodes of each upstream segment;

[0023] Based on the inverse relationship between the attenuation gradient of the dynamic coupling coefficient and the distance between the upstream and downstream nodes, and the linear relationship between the attenuation gradient of the riverbed deformation trend and the distance between the upstream and downstream nodes, calculate the collaborative constraint weight of each upstream segment;

[0024] If there are multiple upstream segments, select the maximum value of the collaborative constraint weight and the corresponding value that satisfies the attenuation gradient of the dynamic coupling coefficient being less than the attenuation gradient of the riverbed deformation trend as the best upstream segment;

[0025] Connect the best upstream segment with the upstream and downstream nodes, and transfer the attenuation gradient of the dynamic coupling coefficient, the attenuation gradient of the riverbed deformation trend, the distance between the upstream and downstream nodes, and the collaborative constraint weight to the downstream node, and perform iteration to generate a candidate propagation chain.

[0026] Optionally, the collecting multi-dimensional water conservancy monitoring data and performing spatio-temporal dimension fusion with the regional precipitation prediction information provided by the meteorological service platform to generate a spatio-temporal coupled water conservancy data set includes:

[0027] Obtain the time series of the riverbed deformation trend of multi-dimensional water conservancy monitoring data, the propagation delay matrix of the water flow dynamic characteristics, and the spatial topological coordinates, and extract the time stamp and the spatial distribution density of geographical units of the regional precipitation prediction information provided by the meteorological service platform;

[0028] Establish a spatial grid mapping relationship based on the spatial topological coordinates and the spatial distribution density of geographical units, and divide the time series and the time stamp into synchronous time windows;

[0029] According to the change characteristics of the propagation delay matrix within the synchronous time window, dynamically adjust the spatial weight allocation ratio of the time series and the regional precipitation prediction information;

[0030] Based on the spatial weight allocation ratio, perform parameter coupling on the riverbed deformation trend, the propagation delay matrix, and the regional precipitation prediction information within the spatial grid mapping relationship to generate a spatio-temporal coupled water conservancy data set.

[0031] Optionally, the establishing a spatial grid mapping relationship based on the spatial topological coordinates and the spatial distribution density of geographical units, and dividing the time series and the time stamp into synchronous time windows includes:

[0032] Define grid cells covering the basin based on the superposition relationship between the spatial topological coordinates and the geographical unit distribution density;

[0033] Within the grid cells, match the spatial attribution of the riverbed deformation data points of the riverbed deformation trend and the precipitation prediction data points of the regional precipitation prediction information according to the time series of the riverbed deformation trend and the precipitation distribution density of the regional precipitation prediction information;

[0034] According to the matching result of the spatial attribution, establish a time-axis interpolation rule for the riverbed deformation trend and the precipitation distribution density;

[0035] Based on the time difference between nodes of the propagation delay matrix of the water flow dynamic characteristics and the time-axis interpolation rule, divide the time stamps of the time series and the regional precipitation prediction information into synchronous time windows.

[0036] Optionally, the performing pattern matching on the runoff mutation probability parameter and the characteristic atlas of the historical dam-break events in the basin, and generating a dynamic response strategy set for water conservancy projects based on the spatial weight of the corrected pattern matching result of the recognition benchmark includes:

[0037] Perform pattern matching on the runoff mutation probability parameter and the characteristic atlas of the historical dam-break events in the basin, extract the characteristic vectors of the runoff mutation probability parameter and the historical dam-break events in the basin, and calculate the multi-dimensional similarity within the geographical unit;

[0038] Based on the spatial propagation rules of the recognition benchmark, correct the spatial weights of the multi-dimensional similarity, bind the corrected spatial weights to the response strategy template of the historical dam-break events in the basin, and select the best binding result as the initial response strategy;

[0039] According to the topological level of the flood peak evolution path, perform propagation delay compensation on the flood storage and detention area capacity parameters of the initial response strategy to generate a dynamic response strategy set for water conservancy projects.

[0040] In a second aspect, the present application provides a flood peak evolution path and dam-break risk warning system, including:

[0041] An acquisition module that acquires multi-dimensional water conservancy monitoring data, performs spatio-temporal dimension fusion with the regional precipitation prediction information provided by the meteorological service platform, and generates a spatio-temporal coupled water conservancy data set;

[0042] An identification module that constructs a basin topology perception model based on the spatio-temporal coupled water conservancy data set, uses the correlation fluctuation between the riverbed deformation trend and the water flow dynamic characteristics of upstream and downstream nodes in the multi-dimensional water conservancy monitoring data as the recognition benchmark for the flood peak evolution path, and generates runoff mutation probability parameters for different geographical units through the regional precipitation prediction information;

[0043] A generation module that performs pattern matching according to the runoff mutation probability parameters and the characteristic map of historical dam-break events in the preset basin, and corrects the spatial weights of the pattern matching results based on the recognition benchmark to generate a dynamic response strategy set for water conservancy projects;

[0044] A matching module that matches the characteristic map of historical dam-break events in the basin and the flood storage and detention area capacity allocation scheme in the dynamic response strategy set for water conservancy projects, and generates an intelligent analysis report of water conservancy data including risk levels through the basin topology perception model.

[0045] In a third aspect, an embodiment of the present application provides a computing device, including a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a method for warning of flood peak evolution path and dam-break risk as described in the first aspect above.

[0046] In a fourth aspect, an embodiment of the present application provides a computer storage medium storing a computer program, and when the computer program is executed by a computer, it implements a method for warning of flood peak evolution path and dam-break risk as described in the first aspect.

[0047] In the embodiments of the present application, multi-dimensional water conservancy monitoring data is collected and fused with the regional precipitation prediction information provided by the meteorological service platform in the spatio-temporal dimension to generate a spatio-temporal coupled water conservancy data set. A basin topology perception model is constructed based on the spatio-temporal coupled water conservancy data set. The correlation fluctuation between the riverbed deformation trend and the dynamic characteristics of water flow at upstream and downstream nodes in the multi-dimensional water conservancy monitoring data is used as the identification benchmark for the flood peak evolution path. And the runoff mutation probability parameters of different geographical units are generated through the regional precipitation prediction information. Pattern matching is performed according to the runoff mutation probability parameters and the characteristic atlas of historical dam-break events in the basin. And based on the spatial weight of the corrected pattern matching result of the identification benchmark, a dynamic response strategy set for water conservancy projects is generated. The characteristic atlas of historical dam-break events in the basin is matched with the flood detention and retarding area capacity allocation scheme in the dynamic response strategy set for water conservancy projects, and an intelligent water conservancy data analysis report including risk levels is generated through the basin topology perception model.

[0048] The technical solution of the present application has the following beneficial effects:

[0049] The present application collects multi-dimensional water conservancy monitoring data (riverbed deformation trend, water flow dynamic characteristics) and the regional precipitation prediction information provided by the meteorological service platform to obtain a spatio-temporal coupled water conservancy data set. A basin topology perception model is constructed based on this data set, and the correlation fluctuation between the riverbed deformation trend and the dynamic characteristics of water flow at upstream and downstream is used as the identification benchmark for the flood peak evolution path. According to the pattern matching result of the runoff mutation probability parameter and the characteristic atlas of historical dam-break events, combined with the correction of the spatial weight of the flood peak evolution path, a dynamic response strategy set is generated. Finally, the historical dam-break characteristics are matched with the flood detention and retarding area capacity allocation scheme in the strategy set, and an intelligent water conservancy analysis report including risk levels is output. Through the dynamic coupling modeling of hydrodynamic propagation and riverbed deformation, combined with the quantification of runoff mutation probability driven by precipitation, the present application realizes the accurate identification of the flood peak evolution path and the cross-space recursion of risk parameters, improves the spatio-temporal resolution of basin risk assessment, enhances the physical consistency and decision interpretability of the dynamic response strategy of water conservancy projects, provides a multi-source collaborative intelligent analysis framework for flood control scheduling and disaster warning, and improves the decision support ability of flood control resource scheduling and disaster warning.

[0050] Furthermore, based on the spatial location relationship of the spatio-temporal coupled water conservancy dataset, a basin topology perception model is constructed. The node connection weights are dynamically determined by the covariation relationship between the water flow dynamic characteristics of adjacent nodes and the riverbed deformation trend. The dynamic coupling coefficient is calculated in combination with the water flow dynamic characteristics. The associated wave propagation chain is identified according to the distribution characteristics of the dynamic coupling coefficient, and the synchronous mutation section of continuous nodes is selected as the candidate set of flood peak evolution paths. The spatial distribution density of precipitation prediction is superimposed on the candidate set coverage area to generate runoff mutation probability parameters that integrate the coupling effect of hydrodynamic and deformation and precipitation gradient. Through the dynamic coupling modeling of hydrodynamic propagation and riverbed deformation, the candidate sections of flood peak evolution paths are accurately identified; combined with the spatial heterogeneity driven by precipitation, the runoff mutation risk parameters are quantified, the spatio-temporal resolution of basin risk assessment and the physical interpretability of emergency strategies are improved, and a decision-making basis with both mechanism and data-driven characteristics is provided for flood control scheduling.

[0051] These aspects or other aspects of the present application will be more clearly understood in the following description of the embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0053] Figure 1 The flowchart of a method for warning of flood peak evolution path and dam break risk provided by the present application is shown;

[0054] Figure 2 The structural schematic diagram of a system for warning of flood peak evolution path and dam break risk provided by the present application is shown;

[0055] Figure 3 The structural schematic diagram of a computing device provided by the present application is shown. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0056] In order to enable those skilled in the art to better understand the solutions of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.

[0057] In some of the processes described in the specification, claims, and the above-mentioned drawings of this application, a plurality of operations appear in a specific order. However, it should be clearly understood that these operations may not be executed in the order in which they appear herein or may be executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish different operations, and the serial numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions such as "first" and "second" in this article are used to distinguish different messages, devices, modules, etc., do not represent a sequence, and do not limit that "first" and "second" are of different types.

[0058] The technical solution of this application is applicable to the scenario of tracing the source of cross-border river pollution, solves the problem of data correlation break caused by the migration of pollutants among water, sediment and biota, focuses on the spatio-temporal fusion of multi-source data and the dynamic response of the basin, and collects hydraulic data such as water level, flow velocity, and riverbed deformation in real time through the Internet of Things, and performs spatio-temporal interpolation fusion with meteorological precipitation prediction to construct a basin topology model to analyze the coupling relationship between the riverbed and the water flow. A distributed hydrological model is used to calculate the probability of runoff mutation in geographical units, and dynamic strategies such as flood discharge scheduling and reservoir capacity regulation are generated through spatial weight correction. Integrate multi-objective optimization and fuzzy evaluation, iterate and simulate the capacity allocation scheme of flood detention areas, and output an intelligent analysis report with risk grading.

[0059] Next, the technical solutions in the embodiments of this application will be clearly and completely described in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are only a part of the embodiments of this application, rather than all the embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative efforts belong to the scope of protection of this application.

[0060] Figure 1 The flowchart of a method for warning of flood peak evolution path and dam break risk is provided for the embodiments of this application, as Figure 1 shown, the method includes:

[0061] 101. Collect multi-dimensional hydraulic monitoring data, and perform spatio-temporal dimension fusion with the regional precipitation prediction information provided by the meteorological service platform to generate a spatio-temporal coupled hydraulic data set;

[0062] In this step, the multi-dimensional hydraulic monitoring data includes the trend of riverbed deformation (time series data of riverbed elevation changes monitored by sensors), the dynamic characteristics of water flow (time series of parameters such as flow velocity, flow rate, and water level), and the spatial topological coordinates (longitude and latitude of monitoring points and the connection relationship between upstream and downstream nodes).

[0063] Regional precipitation prediction information refers to the precipitation intensity, spatial distribution density, and timestamp of different geographical units within a future time period provided by a meteorological service platform.

[0064] Spatio-temporal dimension fusion refers to the process of integrating and analyzing data from different sources in two dimensions: time and space.

[0065] The spatio-temporal coupled water conservancy dataset is a parameterized dataset formed by fusing water conservancy monitoring data and precipitation prediction information within spatio-temporal grid cells, containing parameters such as riverbed deformation, water flow propagation delay, and precipitation intensity that are synchronized in time.

[0066] In the embodiments of this application, first, real-time hydrological data is collected using a sensor network deployed within the basin. At the same time, precipitation prediction information accurate to small regions is obtained from the meteorological service platform. Then, a spatio-temporal fusion algorithm is used to process the two types of data to ensure that data from different sources are aligned in time series and geographical coordinates, forming a comprehensive spatio-temporal coupled water conservancy dataset. In this process, the key lies in ensuring data consistency and accuracy, and usually, technical means such as data cleaning and calibration are adopted to improve data quality.

[0067] Taking the middle and lower reaches of the Yangtze River as an example, the real-time flow velocity data of Yichang, Hankou, and Datong hydrological stations are integrated, combined with the riverbed deformation data obtained from lidar scanning of Dongting Lake area, and combined with the precipitation prediction grid data for the next three days from Wuhan to Nanjing provided by the Central Meteorological Observatory. First, using these multi-source data, a coupled dataset covering the Jingjiang and Wanjiang sections is generated through spatio-temporal interpolation technology. The time resolution of this dataset reaches 1 hour, enabling fine capture of short-term changes, and the spatial accuracy reaches 1 square kilometer, ensuring the accuracy of geographical information. This comprehensive dataset not only contains detailed water flow velocity and riverbed terrain information but also integrates high-resolution precipitation forecasts, providing solid data support for flood risk assessment and early warning within the basin. In this way, a comprehensive monitoring and analysis of flood dynamics in the middle and lower reaches of the Yangtze River region are achieved, which helps to formulate effective flood control strategies in a timely manner to protect the lives and property of residents along the coast.

[0068] 102. Construct a basin topology perception model based on the spatio-temporal coupled water conservancy dataset, take the correlation fluctuations between the riverbed deformation trend and the dynamic characteristics of water flow at upstream and downstream nodes in the multi-dimensional water conservancy monitoring data as the identification benchmark for the flood peak evolution path, and generate runoff mutation probability parameters for different geographical units through the regional precipitation prediction information;

[0069] In this step, the basin topology perception model is used to identify the correlation fluctuations between the riverbed deformation trend and the dynamic characteristics of water flow at upstream and downstream nodes, and use this as the identification benchmark for the flood peak evolution path.

[0070] The trend of riverbed deformation refers to the pattern of changes in the river bottom topography over time. The riverbed can change its shape due to natural factors or human activities.

[0071] The dynamic characteristics of water flow refer to the changing patterns of water flow in time and space, including but not limited to flow velocity, flow direction, flow rate, and the changes of these parameters over time and position.

[0072] The flood peak evolution path describes the process of how the water flow propagates from the upstream to the downstream during the flood peak period.

[0073] The runoff mutation probability parameter is an indicator of the likelihood of sudden water flow changes occurring in different geographical units, calculated based on the above model combined with precipitation prediction information.

[0074] In the embodiments of this application, based on the constructed spatio-temporal coupled water conservancy data set, a watershed topology awareness model is developed using mathematical modeling and simulation techniques. The model learns the relationship between riverbed deformation and water flow characteristics by analyzing patterns in historical data, and then predicts the possible evolution path of the flood peak. For each geographical unit, its runoff mutation probability parameter is updated according to the latest precipitation prediction, which usually involves complex hydraulic calculations and statistical analysis methods.

[0075] Continuing with the above case, for the Jingjiang section in the middle reaches of the Yangtze River, by constructing and applying a watershed risk management model, the impact of changes in the discharge of the Three Gorges Reservoir on the downstream water level can be accurately identified. Specifically, when the model detects that the discharge of the Three Gorges Reservoir increases by every 5000 cubic meters per second (m 3 / s), it is predicted that the water level at the downstream Jianli Station will rise by 0.8 to 1.2 meters after about 8 hours. At the same time, combining the regional precipitation prediction information provided by the meteorological service platform, especially for a specific grid area in the eastern part of Jingzhou, the probability that the cumulative precipitation will reach 80 millimeters within the next 3 hours is evaluated to be 65%. Based on this forecast data and considering factors such as the soil saturation and topographic characteristics of the region, the runoff mutation probability in this area is calculated to be as high as 72%. This means that in a short period of time, due to the sharp increase in rainfall and the superimposed impact of upstream discharge, the risk of sudden flood events has increased significantly. In view of this situation, the system automatically triggers the yellow warning threshold.

[0076] 103. Perform pattern matching according to the runoff mutation probability parameter and the preset characteristic atlas of historical dam break events in the watershed, and generate a set of dynamic response strategies for water conservancy projects based on the spatial weights of the corrected pattern matching results of the recognition benchmark;

[0077] In this step, the characteristic atlas of historical dam break events is a database containing dam break locations, breach widths, flood propagation speeds, etc.

[0078] Pattern matching uses a convolutional neural network to extract the similarity between current hydrological parameters and historical cases.

[0079] Spatial weight correction uses a geographically weighted regression model to adjust the risk impact factors in different regions.

[0080] The dynamic response strategy set for water conservancy projects is a set of specific countermeasures formulated for potential flood events, including but not limited to evacuation route planning, levee reinforcement suggestions, etc.

[0081] In the embodiment of the present application, the runoff mutation probability parameter obtained in the previous step is used to compare and analyze it with the historical atlas recording the details of past dam break events, find similarities and evaluate the risk level in the current situation. Specifically, first, collect and organize the data of historical dam break events to construct a detailed atlas containing the characteristics of various dam break events. These characteristics include not only hydrological parameters such as water level changes and water flow velocities, but also factors such as geographical environment and meteorological conditions. Then, through machine learning algorithms or statistical methods, the currently monitored runoff mutation probability parameter is pattern-matched with the historical atlas to identify the closest historical event and its countermeasures. Based on the spatial weight of the matching result, a set of highly targeted and flexibly adjustable dynamic response strategy sets for water conservancy projects are formulated.

[0082] When a water level rise pattern similar to the levee breach at Hankou Station in 1954 (0.28 meters per hour for 6 hours) is detected in the Wuhan section, combined with the radar detection data of the flood control wall at Hankou River Beach, the response strategy is further refined and adjusted. Specifically, through radar detection technology, obtain the real-time status information of the flood control wall, such as structural integrity, potential weak points, etc. Based on these data, re-evaluate the effectiveness and feasibility of the "reinforce the secondary levee" strategy. After comprehensive analysis, it is found that the confidence level of the "reinforce the secondary levee" strategy has increased, indicating that under the current conditions, strengthening the existing flood control facilities is more prioritized and effective than immediately evacuating the residential areas in the north of the Yangtze River.

[0083] 104. Match the characteristic atlas of historical dam break events in the basin with the flood storage and detention area capacity allocation plan in the dynamic response strategy set for water conservancy projects, and generate an intelligent analysis report of water conservancy data containing the risk level through the basin topology perception model.

[0084] In this step, the flood storage and detention area capacity allocation plan uses a multi-objective optimization algorithm to dynamically select the optimal combination in the flood diversion area.

[0085] The risk level calculates the inundation probability of key cities under different scenarios through Monte Carlo simulation.

[0086] The intelligent analysis report of water conservancy data summarizes all relevant information, including risk assessment results, emergency response suggestions, etc., to provide support for decision-making.

[0087] In the embodiments of the present application, by combining the basin topology perception model with the dynamic response strategy set of water conservancy projects, the capacity of flood detention areas is scientifically planned. First, detailed geographical information and hydrological data are obtained through the basin topology perception model, including topographical features, the distribution of existing water conservancy facilities and their operating states, etc. These data provide a solid foundation for evaluating the actual conditions of each region. When formulating a detailed water volume regulation plan, factors such as topographical features and existing facilities are fully considered. For example, the topographical characteristics of different flood detention areas, such as slope, soil type, and vegetation cover, are analyzed to determine the areas most suitable for temporarily storing floodwater, and the natural terrain is used to mitigate the impact of floods. At the same time, the states and capacities of existing flood control facilities in the region, including dikes, reservoirs, pumping stations, etc., are evaluated. Understanding the current conditions of these facilities is crucial for formulating an effective water volume regulation plan to ensure that the reservoir capacity can be emptied in advance when necessary to cope with the upcoming flood. Based on the above analysis, the usage plan of the flood detention area is further optimized to ensure that it can meet the requirements of rapid response in emergency situations and also take into account the requirements of long-term flood control planning. Finally, all the analysis results are sorted out and compiled into a detailed intelligent analysis report of water conservancy data.

[0088] When it is predicted that the water level in the Jiujiang section may exceed the guaranteed water level by 1.5 meters, the model quickly activates the emergency plan and preferentially calls on the 320 million cubic meters of reservoir capacity in the Poyang Lake flood detention area. This decision is based on a comprehensive assessment. Considering that although the Honghu flood diversion area has a larger capacity, its activation will involve more cultivated land, which may lead to serious agricultural losses and long-term recovery costs. In contrast, the Poyang Lake flood detention area has significant advantages in terms of geographical location, environmental impact, and operational convenience, and can effectively relieve the flood pressure without affecting the main agricultural land. At the same time, the risk report generated by the model marks the risk levels of the affected areas in detail. Especially for Hukou County, the risk report marks it as the red risk level, which means that this area faces extremely high flood threats. It is estimated that this flood event will affect approximately 230,000 people, and the situation is extremely severe. According to this high risk level, the report recommends that the local government immediately activate the emergency evacuation plan and complete the safety transfer of all residents in the threatened areas within 8 hours.

[0089] In summary, steps 101 to 104 achieve the intelligent prevention and control of flood disasters in the middle and lower reaches of the Yangtze River through a four-step closed-loop process. First, by constructing a spatio-temporal coupling data set, the accuracy of flood prediction is significantly improved. Second, using the basin topology perception model, not only can the flood peak evolution path be quickly identified, but also the time required for identification is greatly shortened, enabling the early warning system to provide more timely and effective early warning information before the flood arrives. This method not only considers the current water conservancy situation but also draws on historical experience to ensure the effectiveness and adaptability of response measures. The finally generated risk report supports one-hourly updated emergency decision-making, provides detailed data support and action suggestions, and realizes the full-chain intelligence from data collection to action instructions. This entire set of processes not only improves the ability to respond to flood disasters but also lays a solid foundation for long-term water resource management and disaster prevention and mitigation planning, effectively protecting the lives and property safety of residents along the coast. Through this comprehensive and refined management method, the flood control system in the middle and lower reaches of the Yangtze River becomes more scientific, efficient, and reliable.

[0090] To solve the problems of insufficient accuracy in identifying the flood peak evolution path and lag in runoff mutation early warning in the Yangtze River Basin, this solution uses a spatio-temporal coupling water conservancy data set to construct a basin topology perception model. By analyzing the covariation relationship between the water flow dynamic characteristics and the riverbed deformation trend between upstream and downstream nodes, the connection weights are determined, and the temporal gradient of the riverbed deformation trend is extracted to calculate the dynamic coupling coefficient. Based on these coefficients, the associated fluctuations are identified, and the synchronous mutation sections are marked as the candidate set of the flood peak evolution path. In some embodiments, in step 102, constructing the basin topology perception model based on the spatio-temporal coupling water conservancy data set, using the associated fluctuations between the riverbed deformation trend and the water flow dynamic characteristics of upstream and downstream nodes in the multi-dimensional water conservancy monitoring data as the identification benchmark for the flood peak evolution path, and generating runoff mutation probability parameters for different geographical units through the regional precipitation prediction information, further includes:

[0091] 201. Based on the spatial position relationship in the spatio-temporal coupling water conservancy data set, establish a basin topology perception model, and the connection weights of each upstream and downstream node of the basin topology perception model are determined by the covariation relationship between the water flow dynamic characteristics of the adjacent upstream and downstream nodes and the riverbed deformation trend;

[0092] In step 201, the spatial position relationship refers to the relative position and mutual connection between different entities in the geographical space. The basin topology perception model is a dynamic network model constructed based on the spatial connection relationship of water system nodes. The water flow dynamic characteristics of upstream and downstream nodes include real-time monitoring parameters such as flow velocity, flow rate, and water level fluctuations. The covariation relationship of the riverbed deformation trend refers to the spatio-temporal correlation between the riverbed elevation change rate and the water flow parameters between upstream and downstream nodes.

[0093] In the embodiments of the present application, a topological network is constructed based on the confluence points of the main stream hydrological stations and tributaries of the Yangtze River. The nodes include twelve control sections such as Yichang, Jingzhou, and Hankou. The graph convolutional network algorithm is used to calculate the Pearson correlation coefficient between the riverbed elevation change rate (extracted from the monthly lidar scan data) and the water level fluctuation (hourly monitoring data) between adjacent nodes (such as Yichang and Jingzhou), and the correlation coefficient is normalized to the connection weight from zero to one. The network weight is dynamically updated every six hours to reflect the impact of river channel scouring and silting changes on water flow transmission.

[0094] 202. Extract the temporal gradient of the riverbed deformation trend of each upstream and downstream node in the watershed topological perception model to obtain the dynamic coupling coefficient between the riverbed deformation trend and the dynamic characteristics of water flow in the upstream and downstream nodes;

[0095] In step 202, the watershed topological perception model is a mathematical model used to describe and analyze the hydrological processes and their spatial relationships within a watershed. It provides a framework for comprehensively understanding the flood propagation mechanism by simulating the terrain, water flow paths, river networks, and the relationships between various monitoring points within the watershed. The temporal gradient of the riverbed deformation trend is the acceleration of the riverbed elevation change calculated by the sliding window method. The dynamic coupling coefficient is a quantitative index characterizing the matching degree between the riverbed deformation rate and the water level fluctuation intensity.

[0096] In the embodiments of the present application, the lidar scan data of the Dongting Lake area in the past thirty days is extracted, and the second derivative of the riverbed elevation change of each node is calculated with a three-day window to accurately capture the dynamic change characteristics of the riverbed topography. The gradient boosting tree algorithm is used, which is excellent in dealing with complex non-linear relationships due to its high efficiency and accuracy. The input parameters include the current period riverbed deformation gradient, the change rate of upstream incoming water flow, and the sediment deposition amount, which comprehensively reflect the interaction between the riverbed morphology and the water flow conditions. By training the gradient boosting tree model, a dynamic coupling coefficient matrix is generated, and its numerical range is from zero to one. The larger the value, the stronger the correlation between the riverbed change and the water flow dynamics.

[0097] 203. Identify the correlation fluctuations between the dynamic characteristics of water flow in the upstream and downstream nodes and the riverbed deformation trend according to the distribution characteristics of the dynamic coupling coefficient, and mark the synchronous mutation section of the dynamic coupling coefficient of the continuous nodes in the correlation fluctuations as the candidate set of the flood peak evolution path;

[0098] In step 203, the correlation fluctuation refers to the synchronous abnormal change of the dynamic coupling coefficient between continuous nodes. The synchronous mutation section is a set of river reaches where the dynamic coupling coefficient exceeds the threshold and mutations occur in more than three spatially continuous nodes. The candidate set of the flood peak evolution path is a set of areas identified according to the change pattern of the dynamic coupling coefficient that may experience rapid flood peak movement.

[0099] In the embodiments of the present application, a sliding scan is performed on the topological network of the Yichang-Jiujiang section to detect key nodes with a dynamic coupling coefficient exceeding 0.75. A detailed river basin network diagram is constructed through the river basin topology perception model, and the entire river basin is gradually scanned using the sliding window technique to identify nodes with a dynamic coupling coefficient significantly higher than the threshold. These nodes indicate strong water flow interactions between adjacent regions. The density clustering algorithm is used to identify spatially continuous groups of mutation nodes and label them as the candidate set of flood peak evolution paths. For example, in the Jianli-Chenglingji section, the dynamic coupling coefficients of five consecutive nodes suddenly increase to 0.82, indicating a high connectivity and fluidity during flood propagation in this river section, and it is determined as a potential flood peak channel. This method not only improves the accuracy of flood peak evolution path recognition but also provides a scientific basis for flood prevention early warning, helps formulate effective countermeasures in advance, and reduces losses caused by floods. The finally generated risk report supports rapid response and precise decision-making to ensure the safety of residents along the coast.

[0100] 204. Based on the spatial distribution density of the regional precipitation prediction information, superimpose the precipitation intensity gradient on the geographical units covered by the candidate set of flood peak evolution paths to generate a runoff mutation probability parameter.

[0101] In step 204, the precipitation intensity gradient is the change rate of rainfall per unit area in meteorological prediction data. The runoff mutation probability parameter is a regional runoff surge risk value that combines the spatial distribution of precipitation and the characteristics of the flood peak path. The spatial distribution density generally refers to the quantity or intensity of a certain phenomenon per unit area.

[0102] In the embodiments of the present application, to accurately evaluate the flood risk in the area covered by the flood peak candidate path, an overlay analysis is performed on the kilometer-level grid covered by the flood peak candidate path and the meteorological prediction rainfall intensity distribution map. A convolutional neural network is used to process each grid and calculate its runoff increment probability. The input parameters include the average value of the dynamic coupling coefficient of the candidate path, the cumulative rainfall of the grid, and the surface infiltration coefficient. These parameters comprehensively reflect the influence of topography, hydrological conditions, and precipitation on runoff. Through the trained convolutional neural network model, the probability value that the runoff increment of each grid exceeds the safety threshold within the next six hours is output, generating a five-level risk parameter map. This map can not only clearly show the risk levels of different regions but also provide a scientific basis for formulating precise flood prevention strategies.

[0103] The following is a specific example:

[0104] Taking the section from Jingjiang to Wanjiang in the middle and lower reaches of the Yangtze River as an example, the flow velocities monitored by Yichang, Hankou, and Datong hydrological stations have been continuously increasing, and lidar scans show that the riverbed in the Dongting Lake area is showing an accelerating scouring trend. The calculated results of the constructed basin topology perception model show that the covariance relationship weight of the section from Yichang to Jingzhou has increased significantly, reflecting that the incision of the riverbed has enhanced the water level conduction efficiency. Combining the extracted time-series gradient data of riverbed deformation with the upstream water inflow changes, it is identified that the dynamic coupling coefficient of the section from Jianli to Hankou reaches the peak value. Through density clustering, synchronous mutations of five consecutive nodes in this section are detected and marked as the main path of flood peak evolution. After superimposing the predicted heavy rainfall distribution by the meteorological department, it is calculated that the probability of runoff mutation in the area covered by this path in the next six hours exceeds the threshold, triggering a hierarchical warning and generating an emergency dispatching plan.

[0105] In summary, steps 201 to 204 achieve a collaborative mechanism through dynamic topology modeling, gradient coupling analysis, path clustering identification, and meteorological fusion evaluation, significantly improving the prediction accuracy of the flood peak evolution path in the middle and lower reaches of the Yangtze River. The calculation of covariance weight quantifies the interaction between the riverbed and water flow, the dynamic coupling coefficient reveals the enhanced effect of river channel erosion and deposition on flood peak transmission, synchronous mutation detection accurately locates high-risk river sections, and the precipitation superposition model realizes spatial hierarchical warning of runoff mutation probability. The overall solution shortens the flood peak path identification time by 40%, improves the accuracy of runoff mutation warning, secures a key decision-making window for flood diversion scheduling, and effectively reduces the risk of levee breach and emergency costs.

[0106] To solve the problem of insufficient quantification of the correlation after the propagation path of abnormal fluctuations in the river channel is truncated, this solution further details the process of identifying the associated fluctuations between upstream and downstream nodes, including traversing the dynamic coupling coefficient and the riverbed deformation trend to mark potential propagation segments, selecting the best upstream segment according to specific constraints to generate candidate propagation chains, and truncating the propagation chains that do not meet the preset threshold by calculating the comprehensive attenuation rate ratio, so as to find the associated fluctuations with the strongest synchrony. In some embodiments, the identifying the associated fluctuations between the dynamic characteristics of water flow and the riverbed deformation trend of the upstream and downstream nodes according to the distribution characteristics of the dynamic coupling coefficient in step 203 includes:

[0107] 301. Traverse the dynamic coupling coefficient and the riverbed deformation trend of each upstream and downstream node in the basin topology perception model. If the downstream mutation amplitude attenuates within the same time period for adjacent upstream and downstream nodes, mark it as a potential propagation segment;

[0108] In step 301, the dynamic coupling coefficient is used to quantify the degree of correlation of the water flow velocity change between adjacent hydrological stations. The riverbed deformation trend refers to the change of the riverbed topography over time due to natural or human factors. The potential propagation segment refers to the continuous river section where abnormal fluctuation attenuation occurs between upstream and downstream nodes. The mutation amplitude attenuation refers to the phenomenon that the change amplitude of the water flow velocity or water level of the downstream node within the same time period is reduced compared with that of the upstream node between adjacent upstream and downstream nodes.

[0109] In the embodiment of the present application, first, the real-time flow velocity data of each hydrological monitoring station in the basin and the riverbed topographic data obtained by regular lidar scanning are collected. Then, the dynamic coupling coefficient between adjacent stations is calculated, and the trend of riverbed deformation is analyzed. If it is found that the mutation amplitude of the downstream node attenuates compared with that of the upstream node, this section is marked as a potential propagation section. Through this marking mechanism, the areas that may have an important impact on the flood propagation path are preliminarily screened out.

[0110] 302. Connect the potential propagation sections together. When there are multiple upstream sections between the upstream and downstream nodes, the best upstream section is selected based on the collaborative constraints of the attenuation gradient of the dynamic coupling coefficient, the attenuation gradient of the riverbed deformation trend, and the distance between the upstream and downstream nodes to generate a candidate propagation chain.

[0111] In step 302, the attenuation gradient includes the change slope of the dynamic coupling coefficient and the spatial derivative of the riverbed deformation rate. The collaborative constraint refers to a multi-objective optimization model constructed by integrating the gradient change rate and the geographical distance. The candidate propagation chain refers to a chain-like structure composed of multiple potential propagation sections that may represent the actual flood propagation path.

[0112] In the embodiment of the present application, by comparing the attenuation gradients between different upstream sections and considering the influence of geographical distance, the most suitable path is selected to connect the upstream and downstream nodes. Specifically, a weighted evaluation method is adopted, comprehensively considering the attenuation rate of the dynamic coupling coefficient, the attenuation rate of the riverbed deformation trend, and the physical distance between the upstream and downstream nodes, to ensure that the generated candidate propagation chain can reflect the actual path of flood propagation to the greatest extent. Finally, based on these evaluation results, a candidate propagation chain with a high credibility is generated.

[0113] 303. Calculate the ratio of the comprehensive attenuation rate of the attenuation gradient of the dynamic coupling coefficient and the attenuation gradient of the riverbed deformation trend between adjacent upstream and downstream nodes in the candidate propagation chain. If the ratio exceeds a preset threshold, truncate the candidate propagation chain.

[0114] In step 303, the ratio of the comprehensive attenuation rate is the ratio of the vector modulus of the dynamic coupling attenuation gradient to the riverbed deformation attenuation gradient. The preset threshold is obtained by calibrating through the inversion of historical flood events. The associated fluctuation is used to describe the high synchronization of the attenuation rates of the dynamic coupling coefficient and the riverbed deformation trend between the candidate propagation chain segments after the best truncation during the flood propagation process.

[0115] In the embodiments of the present application, by analyzing the data of each pair of adjacent upstream and downstream nodes, the comprehensive attenuation rate ratio of the attenuation gradient of the dynamic coupling coefficient and the attenuation gradient of the riverbed deformation trend is calculated. If this ratio exceeds a preset threshold, it is considered that this part of the propagation chain needs to be truncated. Statistical analysis methods are used to evaluate the synchrony and correlation between segments to ensure the selection of the optimal propagation path. Through multiple iterations and verifications, it is finally determined which propagation chain segments need to be truncated.

[0116] 304. Based on the synchrony of the attenuation rates of the dynamic coupling coefficient and the riverbed deformation trend, identify the candidate propagation chain segments for optimal truncation as the most strongly correlated fluctuations.

[0117] In step 304, the synchrony is characterized by the Pearson correlation coefficient between the attenuation rate of the dynamic coupling coefficient and the attenuation rate of the riverbed deformation. Optimal truncation means splitting the candidate propagation chain according to certain criteria to remove those parts whose fluctuation characteristics are not sufficient to continue maintaining the original propagation mode, ensuring that the remaining parts can more accurately reflect the actual path of flood propagation.

[0118] In the embodiments of the present application, a detailed synchrony analysis is performed on the remaining propagation chain segments, and correlation analysis techniques are used to evaluate the synchrony index of each segment. Specifically, by calculating the correlation coefficient between the dynamic coupling coefficient and the attenuation rate of the riverbed deformation trend, it is determined whether the two are highly synchronous. According to the synchrony index, determine which parts can be used as the most effective flood propagation paths and further study and apply them as key propagation chain segments.

[0119] The following is a specific example:

[0120] Taking the middle and lower reaches of the Yangtze River as an example, using the real-time flow velocity information of three hydrological stations in Yichang, Hankou, and Datong, combined with the riverbed terrain change data obtained by lidar scanning in the Dongting Lake area, a basin topology perception model is constructed. First, by analyzing the real-time flow velocity data of each station, potential propagation segments with decaying mutation amplitudes are marked. Then, based on the attenuation rates of the dynamic coupling coefficient and the riverbed deformation trend, candidate propagation chains are generated, and the best upstream segment is selected to connect the upstream and downstream nodes. Subsequently, the comprehensive attenuation rate ratio of each segment in the candidate propagation chain is calculated, and the propagation chains that do not meet the requirements are truncated. Finally, through synchrony analysis, the propagation chain segments with the strongest correlated fluctuations are identified to optimize the flood warning system. Based on this information, the flood control strategy is adjusted, and more scientific and reasonable flood control measures are formulated to protect the lives and property of the residents along the coast.

[0121] In summary, steps 301 to 304 achieve the precise identification of the propagation path of abnormal fluctuations through the collaborative analysis of dynamic coupling and riverbed deformation. Compared with traditional single-factor methods, its comprehensive identification accuracy is improved, and it can effectively distinguish the fluctuation variations caused by natural attenuation and human interference. Especially when dealing with complex river systems with multiple tributaries, it can accurately capture the spatio-temporal evolution law of the main propagation path. Through the quantitative evaluation of the synchronization intensity, it provides a reliable basis for tracing the source of the path for flood evolution prediction, and at the same time supports the precise positioning of river regulation projects, significantly enhancing the response ability to abnormal fluctuation events in the basin.

[0122] To solve the problem of insufficient continuity of parameter transfer in the construction of the propagation chain for multi-tributary rivers, this solution details how to select the best upstream section based on the collaborative constraints of the dynamic coupling coefficient, the trend of riverbed deformation, and the distance between upstream and downstream nodes in the case of multiple upstream sections. By calculating the collaborative constraint weights of each upstream section, the upstream section with the maximum weight and meeting specific conditions is selected as the best section, and the candidate propagation chain is generated iteratively. In some embodiments, when connecting the potential propagation sections together in step 302, when there are multiple upstream sections for the upstream and downstream nodes, the best upstream section is selected based on the collaborative constraints of the attenuation gradient of the dynamic coupling coefficient, the attenuation gradient of the riverbed deformation trend, and the distance between upstream and downstream nodes to generate a candidate propagation chain, including:

[0123] 401. Traverse the potential propagation sections of the upstream nodes, and extract the attenuation gradient of the dynamic coupling coefficient, the attenuation gradient of the riverbed deformation trend, and the distance between upstream and downstream nodes of each upstream section;

[0124] In step 401, the attenuation gradient of the dynamic coupling coefficient is used to describe the change of the correlation of water flow velocity between adjacent hydrological stations with distance; the attenuation gradient of the riverbed deformation trend reflects the degree of change of the riverbed topography over time and its change with distance; the distance between upstream and downstream nodes refers to the actual distance between two nodes.

[0125] In the embodiments of the present application, first, the spatial topological connection relationship of each upstream section is extracted through a geographic information system, the attenuation gradient of the dynamic coupling coefficient of each section (the change slope of the coupling coefficient within the window) is calculated using the sliding window method, and combined with the differential result of the digital elevation model generated from lidar point cloud data, the attenuation gradient of the riverbed deformation trend (the difference in deformation rate between adjacent nodes divided by the distance) is extracted. Finally, a structured data table containing three types of parameters is formed.

[0126] 402. Calculate the collaborative constraint weight of each upstream section based on the inverse relationship between the attenuation gradient of the dynamic coupling coefficient and the distance between upstream and downstream nodes, and the linear relationship between the attenuation gradient of the riverbed deformation trend and the distance between upstream and downstream nodes;

[0127] In step 402, the collaborative constraint weight is a comprehensive index used to evaluate the importance of the upstream section in flood propagation. The inverse relationship is that the dynamic coupling attenuation gradient is positively correlated with the reciprocal of the spacing, reflecting that the interaction intensity decays faster during long-distance propagation. The linear relationship is that the riverbed deformation attenuation gradient changes proportionally with the spacing, which is determined by the sediment transport continuity.

[0128] In the application embodiment, a multi-objective normalization method is adopted to construct a weight calculation model. The inverse weighting is performed on the attenuation gradient of the dynamic coupling coefficient to highlight the advantages of the strong coupling section at close range. The linear weighting is performed on the attenuation gradient of the riverbed deformation trend to strengthen the influence of the river reach with a high deformation gradient. By normalizing the range difference of the two types of weights, all weight values fall within a unified range (such as between 0 and 1) to eliminate the influence of different dimensions. Finally, the normalized two types of weights are summed to generate the collaborative constraint weight value. This method not only considers the correlation and intensity of the water flow velocity change but also combines the influence of the riverbed topography change, comprehensively reflecting the complexity and potential risks of each section in flood propagation, providing solid data support and technical guarantee for formulating scientific and reasonable flood control strategies, and improving the accuracy and reliability of the flood warning system.

[0129] 403. If there are multiple upstream sections, select the corresponding value with the maximum collaborative constraint weight and satisfying that the attenuation gradient of the dynamic coupling coefficient is less than the attenuation gradient of the riverbed deformation trend as the optimal upstream section;

[0130] In step 403, the collaborative constraint weight is a normalized evaluation index that comprehensively combines the two attenuation gradient and spacing relationships, used to quantify the contribution degree of the upstream section to the downstream fluctuation propagation. The optimal upstream section is the upstream path with the highest collaborative constraint weight on the premise of satisfying that the dynamic coupling attenuation is slower than the riverbed deformation attenuation.

[0131] In the application embodiment of the present application, for the bifurcated node with multiple upstream sections, first screen the candidate sections where the dynamic coupling attenuation gradient is less than the riverbed deformation attenuation gradient to exclude the coupling-dominated abnormal fluctuations (such as gate regulation interference). Subsequently, select the upstream section corresponding to the maximum collaborative constraint weight among the qualified sections and optimize the threshold boundary conditions by the golden section method.

[0132] 404. Connect the optimal upstream section with the upstream and downstream nodes, and transfer the attenuation gradient of the dynamic coupling coefficient, the attenuation gradient of the riverbed deformation trend, the upstream and downstream node spacing, and the collaborative constraint weight to the downstream node, and perform iterative execution to generate a candidate propagation chain.

[0133] In step 404, parameter passing takes the decay gradient, spacing, and weight calculated by the current node as the input parameters for the downstream node, realizing the recursive expansion of the propagation chain. The candidate propagation chain is generated by analyzing and screening potential propagation segments level by level, and can reflect the possibility and characteristics of flood propagation from upstream to downstream.

[0134] In the embodiment of the present application, the iterative backtracking algorithm is adopted. After connecting the best upstream segment to the current node, the parameters are mapped to the neighborhood grid of the downstream node through spatial interpolation. The iteration terminates when the propagation chain extends to the end node or the parameters do not meet the decay condition. It realizes the gradual transmission and accumulation of information, which helps to construct a complete flood propagation path.

[0135] The following is a specific example:

[0136] In the analysis of the Yichang to Datong section in the middle and lower reaches of the Yangtze River, after the sudden increase in flow velocity is detected at the Yichang Station: The dynamic coupling decay gradient of 0.008 / km and the riverbed deformation decay gradient of 0.12 m / km / day in the Yichang to Jingzhou section (spacing 120 km) are extracted, and the corresponding gradients in the Jingzhou to Hankou section (spacing 240 km) are 0.005 / km and 0.09 m / km / day. The collaborative constraint weight of the Yichang to Jingzhou section is calculated as 0.76 (inverse weight 0.58 + linear weight 0.18), and the weight of the Jingzhou to Hankou section is 0.61. The Yichang to Jingzhou section with a qualified dynamic coupling decay gradient (0.008 < 0.12) is selected, and its weight of 0.76 is higher than that of the Jingzhou to Hankou section. The parameters of the Yichang to Jingzhou section are passed to the Hankou Station, and after iterative calculation, the Hankou to Jiujiang section is connected, and finally a candidate propagation chain from Yichang to Jingzhou to Hankou to Jiujiang is generated.

[0137] To sum up, steps 401 to 404 solve the key technical problems of optimizing the propagation path of multi-branch rivers through the collaborative constraint modeling of dynamic coupling and riverbed deformation decay gradient. Its innovation lies in: constructing a weight evaluation system based on the inverse ratio and linear relationship, quantifying the coupling effect of hydrology and topography; realizing the continuous tracking of long-distance propagation chains through the iterative parameter passing mechanism. In the application in the middle and lower reaches of the Yangtze River, the main propagation path from Yichang to Jingzhou to Hankou is accurately identified, excluding the interference of the Dongting Lake tributary, and the path matching degree is improved compared with the traditional method, providing a scientific basis for the precise layout of flood control projects.

[0138] To solve the problem of insufficient accuracy in the fusion of multi-source information in the spatio-temporal coupling analysis of water conservancy data, this solution focuses on collecting multi-dimensional water conservancy monitoring data and regional precipitation prediction information for spatio-temporal fusion, establishing a spatial grid mapping relationship, dividing synchronous time windows, adjusting the spatio-temporal weight distribution ratio according to the change characteristics of the propagation delay matrix, realizing parameter coupling, and generating a spatio-temporal coupling water conservancy data set. In some embodiments, in step 101, collecting multi-dimensional water conservancy monitoring data and performing spatio-temporal dimension fusion with the regional precipitation prediction information provided by the meteorological service platform to generate a spatio-temporal coupling water conservancy data set includes:

[0139] 501. Obtain the time series of the riverbed deformation trend, the propagation delay matrix of the water flow dynamic characteristics, and the spatial topological coordinates of the multi-dimensional water conservancy monitoring data, and extract the time stamp and the spatial distribution density of the geographical unit of the regional precipitation prediction information provided by the meteorological service platform;

[0140] In step 501, the riverbed deformation trend refers to the change of the riverbed topography over time. The propagation delay matrix describes the time required for the propagation of water flow dynamic characteristics between different monitoring points. The spatial topological coordinates refer to the specific positions of each monitoring point in the geographical space. The spatial distribution density of the geographical unit represents the spatial density of precipitation prediction information in a certain area.

[0141] In the embodiments of the present application, the water flow velocity and water level data of the hydrological station are collected through the water conservancy Internet of Things, and the daily change sequence of the riverbed deformation is generated by combining lidar scanning. The cross-correlation analysis method is used to calculate the propagation time difference of the sudden increase events of the water flow velocity between adjacent stations, and the propagation delay matrix is constructed. The meteorological service platform is called to obtain the precipitation prediction data for the next 72 hours, and its time label and the rainfall distribution density of the geographical grid are extracted.

[0142] 502. Establish a spatial grid mapping relationship based on the spatial topological coordinates and the spatial distribution density of the geographical unit, and divide the time series and the time stamp into synchronous time windows;

[0143] In step 502, the spatial grid mapping relationship is to perform a spatial overlay analysis on the river channel spatial topological coordinates and the meteorological geographical unit grid, and establish a position correspondence table between the hydrological stations and the meteorological grids. The synchronous time window refers to grouping the data in different time periods according to certain rules for easy comparison and analysis.

[0144] In the embodiments of the present application, the spatial overlay tool of the geographic information system is used to perform buffer analysis (radius 5 km) on the coordinates of the main stream hydrological stations of the Yangtze River and the meteorological grids to determine the meteorological grid numbers associated with each station; align the precipitation prediction time stamp (hourly level) with the hydrological data time stamp (minute level), and divide the synchronous window based on the minimum time granularity (15 minutes).

[0145] 503. Dynamically adjust the spatial weight allocation ratio between the timing sequence and the regional precipitation prediction information according to the variation characteristics of the propagation delay matrix within the synchronization time window.

[0146] In step 503, the variation characteristics of the propagation delay matrix refer to the fluctuation amplitude and spatio-temporal correlation of the delay values within the time window. The variation characteristics refer to the patterns, trends, or properties exhibited by a variable or a set of variables over a specific time period. In the field of water conservancy and flood management, it is usually used to describe the variation of hydrological parameters (such as water flow velocity, water level, precipitation, etc.) over time or other factors. The spatial weight allocation ratio characterizes the influence weights of precipitation prediction data and hydrological monitoring data on the coupling result.

[0147] In the embodiment of the present application, a machine learning algorithm is used to analyze the variation pattern of the propagation delay matrix within the synchronization time window, and dynamically adjust the spatial weight allocation ratio between the timing sequence and the precipitation prediction information. This process requires training and verification on a large amount of historical data to ensure the scientificity and rationality of the weight adjustment.

[0148] 504. Based on the spatial weight allocation ratio, perform parameter coupling on the riverbed deformation trend, propagation delay matrix, and regional precipitation prediction information within the spatial grid mapping relationship to generate a spatio-temporal coupled water conservancy data set.

[0149] In step 504, parameter coupling is to fuse the riverbed deformation, water flow delay, and precipitation prediction data according to the weights into a spatio-temporal correlated data set. The spatio-temporal coupled water conservancy data set refers to a data set obtained by integrating and correlating water conservancy-related data from different sources and of different types (such as the timing sequence of the riverbed deformation trend, the propagation delay matrix of the water flow dynamic characteristics, the regional precipitation prediction information, etc.) in the time and space dimensions.

[0150] In the embodiment of the present application, by comprehensively considering the spatial weight allocation ratio, all relevant parameters are integrated together to form a water conservancy data set containing spatio-temporal information. This step utilizes big data processing technology and advanced mathematical models to ensure the integrity and reliability of the data set.

[0151] The following is a specific example:

[0152] For the Yichang-Hankou-Datong section in the middle and lower reaches of the Yangtze River, obtain the daily rate sequence of riverbed deformation at Yichang Station (0.1 to +0.3 m / day), the propagation delay matrix (6 hours from Yichang to Hankou, 9 hours from Hankou to Datong), and the precipitation prediction grid data for the next 24 hours in the Dongting Lake area (maximum rainfall 50 mm / hour). Establish the association between Yichang Station and 3 meteorological grids through spatial grid mapping, and divide 15-minute synchronous windows. Analyzing the delay matrix reveals that the delay volatility in the Hankou-Datong section exceeds the threshold, and the precipitation weight is dynamically adjusted to 0.4. Couple the deformation data (weight 0.6), the delay matrix (weight 0.3), and the precipitation data (weight 0.4) for each grid cell to generate a spatio-temporal dataset containing the coupled values of deformation, delay, and precipitation, revealing the correlation law between the accelerated riverbed erosion downstream of the strong rainfall area (such as the east bank of Dongting Lake) and the shortened delay at Hankou Station.

[0153] In summary, steps 501 to 504 construct a high-precision spatio-temporal coupling analysis model by integrating multi-source water conservancy data and meteorological prediction information. Based on the weight allocation mechanism of dynamic time warping and fuzzy logic, it adaptively responds to the spatio-temporal variations of water flow propagation characteristics. By combining spatial grid mapping and Kriging interpolation, the problem of spatial scale mismatch between hydrological and meteorological data is solved. The parameter coupling model quantifies the interaction effect between riverbed deformation and water flow delay driven by precipitation. In the application in the middle and lower reaches of the Yangtze River, it successfully predicts the local riverbed erosion downstream of Hankou Station and the acceleration of flood peak propagation caused by heavy rainfall. The spatio-temporal coupling accuracy is improved compared with traditional methods, providing reliable data support for real-time early warning and engineering intervention in basin flood control and dispatch.

[0154] To solve the problem of insufficient spatio-temporal alignment accuracy of multi-source data in the basin, this solution realizes the spatial attribution matching of the riverbed deformation trend and regional precipitation prediction information by establishing grid cells covering the basin. First, define the grid cells based on the superposition relationship between the spatial topological coordinates and the distribution density of geographical units. Then, within each grid cell, match the corresponding riverbed deformation data points and precipitation prediction data points according to the time series of the riverbed deformation trend and the precipitation distribution density. Next, establish the time axis interpolation rule based on these matching results to fill in the data gaps and ensure continuity. Finally, combine the time difference of the propagation delay matrix of the water flow dynamic characteristics and the time axis interpolation rule to divide the time stamps of the time series and precipitation prediction information into synchronous time windows for subsequent analysis. In some embodiments, step 502, which establishes the spatial grid mapping relationship based on the spatial topological coordinates and the spatial distribution density of geographical units and divides the time series and time stamps into synchronous time windows, includes:

[0155] 601. Define grid cells covering the basin based on the superposition relationship between the spatial topological coordinates and the distribution density of geographical units;

[0156] In step 601, the spatial topological coordinates refer to the specific positions of each monitoring point in the geographical space. The geographical unit distribution density represents the spatial density of data points within a certain area. The grid unit is obtained by dividing the research area into multiple small areas according to certain rules to facilitate data analysis and processing.

[0157] In the embodiment of the present application, first, using geographic information system technology, the entire basin is meshed according to the spatial topological coordinates of each monitoring point and the geographical unit distribution density. The purpose of doing this is to more accurately capture the changes within different geographical regions, thereby laying a foundation for subsequent data analysis.

[0158] 602. Within the grid unit, according to the time series of the riverbed deformation trend and the precipitation distribution density of the regional precipitation prediction information, match the spatial attribution of the riverbed deformation data points of the riverbed deformation trend and the precipitation prediction data points of the regional precipitation prediction information;

[0159] In step 602, the riverbed deformation trend refers to the change of the riverbed topography over time. The precipitation distribution density describes the spatial distribution characteristics of the precipitation amount within a certain period. The spatial attribution matching is to map the riverbed deformation data points and the precipitation prediction data points into the same grid unit to establish a spatial position correspondence relationship.

[0160] In the embodiment of the present application, the spatial interpolation method and the spatio-temporal data mining algorithm are used to match the spatial distribution of the riverbed deformation data with the time series and the precipitation prediction information. By analyzing the data points within each grid unit, it is determined which specific riverbed deformation or precipitation prediction data set they belong to, ensuring the consistency and accuracy of the data.

[0161] 603. According to the matching result of the spatial attribution, establish the time-axis interpolation rule between the riverbed deformation trend and the precipitation distribution density;

[0162] In step 603, the time-axis interpolation rule is to define the interpolation method for time series alignment in view of the time resolution difference between the riverbed deformation and the precipitation data (such as minute level and hour level). The matching result is the process and final output of spatially attributing the riverbed deformation data points of the riverbed deformation trend and the precipitation prediction data points of the regional precipitation prediction information within the grid unit according to specific rules or algorithms.

[0163] In the embodiment of the present application, time series analysis methods are used to construct the time-axis interpolation rule based on the known data points and their spatial attribution. These rules are used to fill in the data gaps to ensure that all data points can be compared and analyzed within the same time window. In this way, the dynamic relationship between the riverbed deformation trend and the precipitation distribution can be more accurately reflected.

[0164] 604. Based on the time difference between nodes of the propagation delay matrix with respect to the dynamic characteristics of water flow and the time axis interpolation rule, divide the time series and the time stamps of regional precipitation prediction information into synchronous time windows.

[0165] In step 604, the synchronous time window is a time segment dynamically adjusted according to the water flow propagation delay and is used to uniformly analyze time series data. The time difference refers to the time interval of the change in water flow characteristics between different monitoring points in the propagation delay matrix based on the dynamic characteristics of water flow.

[0166] In the embodiments of the present application, a machine learning algorithm is used to analyze the propagation delay matrix and, in combination with the time axis interpolation rule, divide synchronous time windows. This process requires training and verification of a large amount of historical data to ensure the scientificity and rationality of the time window division. The finally generated data set can comprehensively reflect the water conservancy conditions and their change laws within the basin.

[0167] The following is a specific example:

[0168] In the application of the section from Yichang to Datong in the middle and lower reaches of the Yangtze River, 2185 1-kilometer grid cells are generated based on the superposition of hydrological station coordinates and meteorological grids. The bed deformation rate (0.15 m / day) and precipitation density (50 mm / hour) of the grid where Yichang Station is located are matched, and the bed deformation rate (0.08 m / day) of the Hankou Station grid corresponds to the precipitation density (30 mm / hour). The hourly deformation data of Hankou Station is downsampled to the minute level to synchronize the time stamps of the precipitation data of Datong Station. According to the maximum propagation delay (15 hours) of the section from Yichang to Datong, 5-hour windows are divided, and it is identified that the scouring rate of the river bed downstream of Hankou increases within the strong precipitation window (the 3rd - 8th hours), revealing the delayed correlation that the deformation reaches the maximum value 6 hours after the precipitation peak.

[0169] To sum up, steps 601 to 604 achieve the precise spatio-temporal fusion of hydrometeorological data through spatial grid attribution matching and dynamic time window division, use the Thiessen polygon grid superposition technology to unify the spatial scale, establish an adaptive interpolation rule to eliminate the time resolution difference, and capture the hydraulic dynamic process based on the sliding window of the propagation delay. In the application in the middle and lower reaches of the Yangtze River, the delayed enhancement effect of precipitation events on river bed scouring is successfully quantified, the spatio-temporal coupling accuracy is improved, providing a reliable data basis for the chain warning of flood disasters.

[0170] To solve the problem of flood risk management in the basin, this solution generates a dynamic response strategy set for water conservancy projects by pattern-matching the runoff mutation probability parameters with the characteristic atlas of historical dam-break events in the basin. First, the runoff mutation probability parameters are pattern-matched with the characteristic atlas of historical dam-break events, feature vectors are extracted, and the multi-dimensional similarity within the geographical unit is calculated. Then, based on the spatial propagation rules of the recognition benchmark, the spatial weights of the multi-dimensional similarity are corrected, and the corrected spatial weights are bound to the response strategy template of historical dam-break events, and the best matching result is selected as the initial response strategy. Finally, according to the topological level of the flood peak evolution path, the propagation delay compensation is performed on the flood storage and detention area capacity parameters in the initial response strategy, so as to generate a comprehensive set of dynamic response strategies for water conservancy projects to ensure the effectiveness and timeliness of the response measures. In some embodiments, in step 103, pattern-matching the runoff mutation probability parameters with the preset characteristic atlas of historical dam-break events in the basin, and based on the spatial weights of the corrected pattern-matching results of the recognition benchmark, generating a dynamic response strategy set for water conservancy projects, includes:

[0171] 701. Pattern-match the runoff mutation probability parameters with the preset characteristic atlas of historical dam-break events in the basin, extract the feature vectors of the runoff mutation probability parameters and historical dam-break events in the basin, and calculate the multi-dimensional similarity within the geographical unit;

[0172] In step 701, the runoff mutation probability parameters are used to describe the possibility of runoff change within a certain period. The characteristic atlas is a data model constructed based on historical data and reflecting the characteristics of dam-break events. The feature vectors contain the main feature information of specific events. The multi-dimensional similarity refers to the degree of similarity between different geographical units in multiple dimensions.

[0173] In the embodiments of the present application, first, the real-time flow velocity data and riverbed deformation data of each monitoring point are collected and analyzed, and the runoff mutation probability parameters are extracted. Then, machine learning algorithms (such as support vector machines or neural networks) are used to pattern-match these parameters with the characteristic atlas of historical dam-break events, extract the corresponding feature vectors, and calculate the multi-dimensional similarity within the geographical unit. Through this method, it is possible to identify which areas have a high risk of dam break.

[0174] 702. Based on the spatial propagation rules of the recognition benchmark, correct the spatial weights of the multi-dimensional similarity, bind the corrected spatial weights to the response strategy template of the historical dam-break events in the basin, and select the best binding result as the initial response strategy;

[0175] In step 702, the spatial propagation rule for identifying the benchmark refers to the weight adjustment criterion determined according to the geographical distribution and the law of water flow propagation. The spatial propagation rule is a set of criteria or models used to describe how disasters spread between different geographical locations, based on the characteristics of geographical spatial distribution, when analyzing and predicting natural disasters such as floods. The response strategy template is a set of pre-determined response measures for different types of dam break events. The initial response strategy refers to a series of preliminary response measures immediately taken after identifying potential flood risks, aiming to minimize the losses caused by disasters and protect the safety of people in the affected areas.

[0176] In the embodiments of the present application, the spatio-temporal data analysis technology is adopted to correct the spatial weight of the multi-dimensional similarity according to the spatial propagation rule of the identification benchmark. Then, in combination with the response strategy template of historical dam break events, the best binding result is selected as the initial response strategy through an optimization algorithm (such as a genetic algorithm). This step ensures that the response strategy not only considers the current situation but also draws on historical experience, improving the effectiveness of the response measures.

[0177] 703. According to the topological level of the flood peak evolution path, perform propagation delay compensation on the flood detention area capacity parameter of the initial response strategy to generate a set of dynamic response strategies for water conservancy projects.

[0178] In step 703, the topological level of the flood peak evolution path refers to the different regions passed through and their mutual relationships during the propagation of the flood from the upstream to the downstream. The propagation delay compensation is to ensure that each flood detention area can be activated at the correct time to maximize its role.

[0179] In the embodiments of the present application, a flood evolution simulation model is used to analyze the topological level of the flood peak evolution path and calculate the propagation delay. Then, according to this delay information, the flood detention area capacity parameter in the initial response strategy is adjusted to generate the final set of dynamic response strategies for water conservancy projects. This process ensures that all flood control facilities can operate in coordination to minimize the losses caused by floods.

[0180] The following is a specific example:

[0181] Taking the middle and lower reaches of the Yangtze River as an example, integrate the real-time flow velocity data of three hydrological stations in Yichang, Hankou, and Datong, and combine the riverbed topographic change data obtained by lidar scanning in the Dongting Lake area to establish a basin risk management model. First, by analyzing the real-time data of each monitoring point, extract the runoff mutation probability parameters, and perform pattern matching with the characteristic atlas of historical dam-break events to identify high-risk areas. Then, use spatio-temporal data analysis technology to correct the spatial weights of multi-dimensional similarity, and combine with the historical response strategy template to select the best initial response strategy. Finally, according to the topological hierarchy of the flood propagation path, perform propagation delay compensation on the storage and detention flood area capacity parameters to generate a set of dynamic response strategies for water conservancy projects. Adjust flood control strategies based on this information, formulate more scientific and reasonable flood control measures, and protect the lives and property of residents along the coast.

[0182] In summary, steps 701 to 703 significantly improve the accuracy and timeliness of flood risk management. By integrating multi-source data and performing spatio-temporal coupling analysis, the flood warning system can more accurately capture the development trend of floods, timely issue warning information, greatly reduce the losses caused by flood disasters, and at the same time provide strong data support and technical guarantee for basin comprehensive management. This method not only improves the ability to respond to sudden flood events, but also lays a solid foundation for long-term water resources management and disaster prevention and mitigation planning.

[0183] Figure 2 This application embodiment provides a structural schematic diagram of a flood peak propagation path and dam-break risk warning system, as Figure 2 shown, the system includes:

[0184] A collection module 21, which collects multi-dimensional water conservancy monitoring data, and performs spatio-temporal dimension fusion with the regional precipitation prediction information provided by the meteorological service platform to generate a spatio-temporal coupled water conservancy data set;

[0185] An identification module 22, which constructs a basin topology perception model based on the spatio-temporal coupled water conservancy data set, takes the correlation fluctuation between the riverbed deformation trend and the water flow dynamic characteristics of upstream and downstream nodes in the multi-dimensional water conservancy monitoring data as the identification benchmark for the flood peak propagation path, and generates runoff mutation probability parameters for different geographical units through the regional precipitation prediction information;

[0186] A generation module 23, which performs pattern matching according to the runoff mutation probability parameters and the preset characteristic atlas of historical dam-break events in the basin, and corrects the spatial weights of the pattern matching results based on the identification benchmark to generate a set of dynamic response strategies for water conservancy projects;

[0187] A matching module 24, which matches the characteristic atlas of historical dam-break events in the basin and the storage and detention flood area capacity allocation scheme in the dynamic response strategy set for water conservancy projects, and generates an intelligent water conservancy data analysis report including risk levels through the basin topology perception model.

[0188] Figure 2 The described peak flood evolution path and dam-break risk warning system can execute Figure 1 For the peak flood evolution path and dam-break risk warning method described in the above embodiments, its implementation principle and technical effects will not be elaborated further. For the peak flood evolution path and dam-break risk warning system in the above embodiments, the specific manners in which each module and unit perform operations have been described in detail in the embodiments related to the method, and will not be elaborated here.

[0189] In a possible design, Figure 2 The peak flood evolution path and dam-break risk warning system of the above embodiments can be implemented as a computing device, such as Figure 3 As shown, the computing device may include a storage component 31 and a processing component 32;

[0190] The storage component 31 stores one or more computer instructions, where the one or more computer instructions are called and executed by the processing component 32.

[0191] The processing component 32 is used for the above Figure 1 The peak flood evolution path and dam-break risk warning method of the above embodiments.

[0192] Among them, the processing component 32 may include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component may also be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic components for executing the above method.

[0193] The storage component 31 is configured to store various types of data to support operations on the terminal. The storage component can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0194] Of course, the computing device will certainly also include other components, such as input / output interfaces, display components, communication components, etc.

[0195] The input / output interface provides an interface between the processing component and the peripheral interface module, and the above peripheral interface module may be an output device, an input device, etc.

[0196] The communication component is configured to facilitate communication between a computing device and other devices in a wired or wireless manner, etc.

[0197] Among them, the computing device can be a physical device or an elastic computing host provided by a cloud computing platform, etc. At this time, the computing device can refer to a cloud server. The above processing component, storage component, etc. can be basic server resources rented or purchased from a cloud computing platform.

[0198] The embodiment of the present application also provides a computer storage medium storing a computer program, and when the computer program is executed by a computer, it can implement the above Figure 1 A method for warning of flood peak evolution path and dam break risk in the shown embodiment.

[0199] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described systems, devices, and units can refer to the corresponding processes in the foregoing method embodiments, and will not be described herein again.

[0200] The device embodiments described above are merely illustrative. 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 may be distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative labor.

[0201] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the essence of the above technical solution, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0202] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, and are not intended to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of each embodiment of the present application.

Claims

1. A method and system for flood peak evolution path and dam-break risk warning, characterized in that Including: Collecting multi-dimensional water conservancy monitoring data, and performing spatio-temporal dimension fusion with the regional precipitation prediction information provided by the meteorological service platform to generate a spatio-temporal coupled water conservancy data set; Constructing a basin topology perception model based on the spatio-temporal coupled water conservancy data set, using the correlation fluctuation between the riverbed deformation trend and the water flow dynamic characteristics of upstream and downstream nodes in the multi-dimensional water conservancy monitoring data as the identification benchmark for the flood peak evolution path, and generating runoff mutation probability parameters for different geographical units through the regional precipitation prediction information; Performing pattern matching according to the runoff mutation probability parameters and the characteristic map of the historical dam-break events in the preset basin, and generating a dynamic response strategy set for water conservancy projects based on the spatial weight of the corrected pattern matching results of the identification benchmark; Matching the characteristic map of the historical dam-break events in the basin and the flood storage and detention area capacity allocation plan in the dynamic response strategy set for water conservancy projects, and generating an intelligent analysis report of water conservancy data including risk levels through the basin topology perception model.

2. The method according to claim 1, characterized in that, The constructing a basin topology perception model based on the spatio-temporal coupled water conservancy data set, using the correlation fluctuation between the riverbed deformation trend and the water flow dynamic characteristics of upstream and downstream nodes in the multi-dimensional water conservancy monitoring data as the identification benchmark for the flood peak evolution path, and generating runoff mutation probability parameters for different geographical units through the regional precipitation prediction information, includes: Establishing a basin topology perception model based on the spatial position relationship in the spatio-temporal coupled water conservancy data set, and determining the connection weights of each upstream and downstream node in the basin topology perception model by the covariance relationship between the water flow dynamic characteristics and the riverbed deformation trend of adjacent upstream and downstream nodes; Extracting the time series gradient of the riverbed deformation trend of each upstream and downstream node in the basin topology perception model to obtain the dynamic coupling coefficient between the riverbed deformation trend and the water flow dynamic characteristics of upstream and downstream nodes; Identifying the correlation fluctuation between the water flow dynamic characteristics and the riverbed deformation trend of upstream and downstream nodes according to the distribution characteristics of the dynamic coupling coefficient, and marking the synchronous mutation section of the dynamic coupling coefficient of continuous nodes in the correlation fluctuation as the candidate set of the flood peak evolution path; Based on the spatial distribution density of the regional precipitation prediction information, superimposing the precipitation intensity gradient on the geographical units covered by the candidate set of the flood peak evolution path to generate runoff mutation probability parameters.

3. The method according to claim 2, wherein The identifying the correlation fluctuation between the water flow dynamic characteristics and the riverbed deformation trend of upstream and downstream nodes according to the distribution characteristics of the dynamic coupling coefficient, includes: Traversing the dynamic coupling coefficient and the riverbed deformation trend of each upstream and downstream node in the basin topology perception model, and if the downstream mutation amplitude attenuation occurs in adjacent upstream and downstream nodes within the same time period, marking it as a potential propagation section; Connecting the potential propagation sections together, and when there are multiple upstream sections for the upstream and downstream nodes, selecting the best upstream section based on the collaborative constraints of the attenuation gradient of the dynamic coupling coefficient, the attenuation gradient of the riverbed deformation trend and the distance between upstream and downstream nodes to generate a candidate propagation chain; Calculating the ratio of the comprehensive attenuation rate of the attenuation gradient of the dynamic coupling coefficient and the attenuation gradient of the riverbed deformation trend of adjacent upstream and downstream nodes in the candidate propagation chain, and if the ratio exceeds the preset threshold, truncating the candidate propagation chain; Based on the synchronization between the dynamic coupling coefficient and the attenuation rate of the riverbed deformation trend, identify the candidate propagation chain segment for the best truncation as the associated fluctuation with the strongest synchronization.

4. The method according to claim 3, wherein Connecting the potential propagation segments together, when there are multiple upstream segments for the upstream and downstream nodes, based on the collaborative constraints of the attenuation gradient of the dynamic coupling coefficient, the attenuation gradient of the riverbed deformation trend, and the distance between the upstream and downstream nodes, select the best upstream segment to generate a candidate propagation chain, including: Traverse the potential propagation segments of the upstream node, and extract the attenuation gradient of the dynamic coupling coefficient, the attenuation gradient of the riverbed deformation trend, and the distance between the upstream and downstream nodes for each upstream segment; Based on the inverse relationship between the attenuation gradient of the dynamic coupling coefficient and the distance between the upstream and downstream nodes, and the linear relationship between the attenuation gradient of the riverbed deformation trend and the distance between the upstream and downstream nodes, calculate the collaborative constraint weight for each upstream segment; If there are multiple upstream segments, select the maximum value of the collaborative constraint weight and the corresponding value that satisfies the attenuation gradient of the dynamic coupling coefficient being less than the attenuation gradient of the riverbed deformation trend as the best upstream segment; Connect the best upstream segment with the upstream and downstream nodes, and transfer the attenuation gradient of the dynamic coupling coefficient, the attenuation gradient of the riverbed deformation trend, the distance between the upstream and downstream nodes, and the collaborative constraint weight to the downstream node, and iterate to generate a candidate propagation chain.

5. The method according to claim 1, characterized in that Collect multi-dimensional water conservancy monitoring data, and perform spatio-temporal dimension fusion with the regional precipitation prediction information provided by the meteorological service platform to generate a spatio-temporal coupled water conservancy data set, including: Obtain the time series of the riverbed deformation trend, the propagation delay matrix of the water flow dynamic characteristics, and the spatial topological coordinates of the multi-dimensional water conservancy monitoring data, and extract the timestamp and the spatial distribution density of the geographical unit of the regional precipitation prediction information provided by the meteorological service platform; Based on the spatial topological coordinates and the spatial distribution density of the geographical unit, establish a spatial grid mapping relationship, and divide the time series and the timestamp into synchronous time windows; According to the change characteristics of the propagation delay matrix within the synchronous time window, dynamically adjust the spatial weight allocation ratio between the time series and the regional precipitation prediction information; Based on the spatial weight allocation ratio, perform parameter coupling on the riverbed deformation trend, the propagation delay matrix, and the regional precipitation prediction information within the spatial grid mapping relationship to generate a spatio-temporal coupled water conservancy data set.

6. The method according to claim 5, wherein Based on the spatial topological coordinates and the spatial distribution density of the geographical unit, establish a spatial grid mapping relationship, and divide the time series and the timestamp into synchronous time windows, including: Based on the superposition relationship between the spatial topological coordinates and the geographical unit distribution density, define the grid cells covering the basin; Within the grid cells, according to the time series of the riverbed deformation trend and the precipitation distribution density of the regional precipitation prediction information, match the spatial attribution of the riverbed deformation data points of the riverbed deformation trend and the precipitation prediction data points of the regional precipitation prediction information; According to the matching result of the spatial attribution, establish the time axis interpolation rule between the riverbed deformation trend and the precipitation distribution density. Based on the time difference between nodes of the propagation delay matrix with respect to the dynamic characteristics of water flow and the time axis interpolation rule, the time series and the timestamps of regional precipitation prediction information are divided into synchronous time windows.

7. The method according to claim 1, characterized in that, Performing pattern matching on the runoff mutation probability parameter and a preset characteristic map of historical dam break events in the basin, and generating a set of dynamic response strategies for water conservancy projects based on the spatial weights of the corrected pattern matching results of the recognition benchmark, including: Performing pattern matching on the runoff mutation probability parameter and a preset characteristic map of historical dam break events in the basin, extracting the characteristic vectors of the runoff mutation probability parameter and the historical dam break events in the basin, and calculating the multi-dimensional similarity within the geographical unit; Based on the spatial propagation rule of the recognition benchmark, correcting the spatial weights of the multi-dimensional similarity, binding the corrected spatial weights with the response strategy template of the historical dam break events in the basin, and selecting the best binding result as the initial response strategy; According to the topological level of the flood peak evolution path, compensating the storage and detention flood area capacity parameters of the initial response strategy for propagation delay to generate a set of dynamic response strategies for water conservancy projects.

8. A method for warning of flood peak evolution path and dam-break risk, characterized in that, Including: A collection module that collects multi-dimensional water conservancy monitoring data and performs spatio-temporal dimension fusion with the regional precipitation prediction information provided by the meteorological service platform to generate a spatio-temporal coupled water conservancy data set; An identification module that constructs a basin topology perception model based on the spatio-temporal coupled water conservancy data set, takes the correlation fluctuation between the riverbed deformation trend and the dynamic characteristics of water flow at upstream and downstream nodes in the multi-dimensional water conservancy monitoring data as the recognition benchmark for the flood peak evolution path, and generates runoff mutation probability parameters for different geographical units through the regional precipitation prediction information; A generation module that performs pattern matching on the runoff mutation probability parameter and a preset characteristic map of historical dam break events in the basin, and generates a set of dynamic response strategies for water conservancy projects based on the spatial weights of the corrected pattern matching results of the recognition benchmark; A matching module that matches the characteristic map of historical dam break events in the basin and the storage and detention flood area capacity allocation plan in the set of dynamic response strategies for water conservancy projects, and generates an intelligent analysis report of water conservancy data including risk levels through the basin topology perception model.

9. A computing device, characterized in that, Including a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a method for flood peak evolution path and dam break risk warning as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that, Stored with a computer program, when the computer program is executed by a computer, it implements a method for flood peak evolution path and dam break risk warning as described in any one of claims 1 to 7.

Images