Levee breach plugging method, system, electronic device, and storage medium

Abstract

Embodiments of the present application provide a dike breach plugging method and system, electronic equipment and storage medium, belonging to the technical field of automatic rescue. The method reconstructs a three-dimensional flow field model according to the collected water surface flow data of the breach area; extracts target features from the three-dimensional flow field model to obtain a flow field feature matrix; optimizes the first target function as the optimization target, and performs decision optimization according to the flow field feature matrix, the regional throwing state data and the available material data to obtain an optimal throwing scheme, wherein the first target function represents at least one of the minimum flow rate energy or the minimum material loss; the second target function is used as the optimization target, and the optimal throwing scheme and the transportation resource data are used for transportation decision optimization to obtain an optimal transportation scheduling scheme, and the optimal transportation scheduling scheme is used for transportation resource scheduling, and the second target function represents the maximum transportation efficiency. The application can improve the efficiency of dike breach plugging.

Description

Technical Field

[0001] The present application relates to the field of automated emergency rescue technology, and in particular to a dike breach sealing method, system, electronic equipment and storage medium. Background Art

[0002] Levees are crucial infrastructure for river flood control and safety. A breach not only causes enormous economic losses but can also pose a serious threat to human life. Therefore, the fastest possible rescue and sealing of breaches is crucial. In current disaster relief efforts, levee breach sealing relies primarily on the on-site experience of rescue personnel, who visually estimate breach size, flow velocity distribution, riprap specifications and quantities, and the scheduling of plugging resources such as sand and gravel. This traditional approach has significant shortcomings: First, relying on manual experience can lead to inaccurate judgments of complex on-site flow velocities and inaccurate estimates of riprap size and material usage, making decision-making highly subjective and potentially delaying rescue efforts. Second, manual decision-making makes it difficult to quickly and efficiently coordinate rescue resources, requiring ad hoc assessments of sand and gravel reserves and scheduling, impacting rescue efficiency and the ultimate effectiveness of the blockade.

[0003] In recent years, drones and intelligent monitoring technologies have been gradually applied to flood prevention and emergency response. For example, embankment breach monitoring devices based on drones and video image analysis use drones equipped with high-definition cameras to capture real-time video footage of the breach area and transmit the video signal back to a monitoring center, where technicians visually analyze the breach to guide on-site rescue operations. However, these technologies are limited to video capture and manual identification, requiring manual personnel to make qualitative rescue decisions based on on-site video footage. Embankment breach plugging and spraying methods are unreliable and cannot be implemented in an orderly manner, resulting in inefficient rescue operations. Summary of the Invention

[0004] The main purpose of the embodiments of the present application is to propose a dike breach sealing method, system, electronic device and storage medium, aiming to improve the efficiency of dike breach sealing.

[0005] To achieve the above-mentioned purpose, one aspect of an embodiment of the present application provides a dike breach sealing method, comprising the following steps:

[0006] Collect water surface flow data in the breach area;

[0007] Performing virtual three-dimensional space reconstruction on the breach area according to the water surface flow data to obtain a three-dimensional flow field model;

[0008] Performing target feature extraction on the three-dimensional flow field model to obtain a flow field feature matrix;

[0009] Taking a first objective function as an optimization goal, optimizing the casting decision according to the flow field characteristic matrix, the regional casting state data, and the available material data to obtain an optimal casting solution, wherein the first objective function represents at least one of minimizing flow velocity energy or minimizing material loss;

[0010] Taking the second objective function as the optimization target, the transportation decision is optimized according to the optimal throwing plan and transportation resource data to obtain the optimal transportation scheduling plan, and the transportation resources are scheduled according to the optimal transportation scheduling plan, wherein the second objective function represents the maximization of transportation efficiency.

[0011] In some embodiments, the water surface flow data includes a water surface flow video, and performing virtual three-dimensional spatial reconstruction of the breach area based on the water surface flow data to obtain a three-dimensional flow field model includes the following steps:

[0012] Preprocessing each frame image in the water surface flow video to obtain preprocessed continuous frame images;

[0013] Extracting water surface ripple feature points from the continuous frame images using a scale-invariant feature transformation algorithm;

[0014] Determine the water surface flow velocity distribution data according to the position change of the water surface ripple feature points at each position in the previous and next frames;

[0015] A three-dimensional flow field model is constructed in a virtual three-dimensional space according to the water surface flow velocity distribution data.

[0016] In some embodiments, the water surface flow data further includes water flow monitoring data from a water surface sensor, and the virtual three-dimensional spatial reconstruction of the breach area based on the water surface flow data to obtain a three-dimensional flow field model further includes the following steps:

[0017] Determining the flow velocity distribution of a corresponding area in the water surface flow velocity distribution data according to the collection location of the water flow monitoring data;

[0018] Correcting the flow velocity distribution in the corresponding area based on the water flow monitoring data;

[0019] The corrected water surface flow velocity distribution data is obtained according to the corrected flow velocity distribution in each area.

[0020] In some embodiments, extracting target features from the three-dimensional flow field model to obtain a flow field feature matrix comprises the following steps:

[0021] Partitioning the three-dimensional flow field model to obtain regional flow field models of different flow velocity areas;

[0022] Respectively, the target feature extraction of each area flow field model is performed to obtain the flow field feature vectors of different flow velocity areas;

[0023] The flow field feature matrix is determined according to the flow field feature vectors of different flow velocity areas.

[0024] In some embodiments, the first objective function is taken as an optimization target, and a throwing decision optimization is performed according to the flow field feature matrix, the area throwing state data and the available material data to obtain an optimal throwing scheme, including the following steps:

[0025] The throwing material constraint of each flow velocity area is determined according to the flow velocity of different flow velocity areas in the flow field feature matrix;

[0026] The throwing decision scheme of each flow velocity area is determined according to the area throwing state data, the available material data, the flow field feature vector and the throwing material constraint;

[0027] The throwing decision scheme is optimized to obtain the optimal throwing scheme, taking the first objective function as an optimization target.

[0028] In some embodiments, the throwing decision scheme includes throwing area coordinates, stone particle size and throwing quantity, and the stone particle size is calculated by the following formula:

[0029] ;

[0030] Wherein, represents the stone particle size, represents the maximum flow velocity of the area, is a stability coefficient, is the relative density of stone, is the acceleration of gravity.

[0031] In some embodiments, the second objective function is taken as an optimization target, and a transportation decision optimization is performed according to the optimal throwing scheme and transportation resource data to obtain an optimal transportation scheduling scheme, including the following steps:

[0032] The material transportation task of each carrying resource is determined according to the optimal throwing scheme and the transportation resource data, and the material transportation task includes transportation target material, transportation quantity, transportation starting point and transportation endpoint;

[0033] The transportation path is obtained by performing path planning on the carrying resource according to the material transportation task, taking the minimum transportation time as the target;

[0034] The transportation scheduling decision scheme is determined according to the material transportation task and the transportation path of each carrying resource;

[0035] Taking the second objective function as the optimization target, the transportation scheduling decision plan is optimized to obtain the optimal transportation scheduling plan.

[0036] To achieve the above objectives, another aspect of the present application provides a dike breach sealing system, comprising:

[0037] The first module is used to collect water surface flow data in the breach area;

[0038] The second module is used to reconstruct the breach area in a virtual three-dimensional space according to the water surface flow data to obtain a three-dimensional flow field model;

[0039] The third module is used to extract target features from the three-dimensional flow field model to obtain a flow field feature matrix;

[0040] A fourth module is configured to optimize the casting decision based on the flow field characteristic matrix, the regional casting state data, and the available material data, using the first objective function as an optimization target to obtain an optimal casting solution, wherein the first objective function represents at least one of minimizing flow velocity energy or minimizing material loss;

[0041] The fifth module is used to optimize the transportation decision based on the optimal throwing plan and transportation resource data with the second objective function as the optimization target, to obtain the optimal transportation scheduling plan, and to schedule the transportation resources according to the optimal transportation scheduling plan, wherein the second objective function represents the maximization of transportation efficiency.

[0042] To achieve the above-mentioned purpose, another aspect of an embodiment of the present application proposes an electronic device, which includes a memory, a processor, a program stored on the memory and executable on the processor, and a data bus for realizing connection and communication between the processor and the memory. When the program is executed by the processor, the method described in the above embodiment is implemented.

[0043] To achieve the above-mentioned purpose, another aspect of an embodiment of the present application proposes a storage medium, which is a computer-readable storage medium used for computer-readable storage. The storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the method described in the above embodiment.

[0044] The present application proposes a dike breach sealing method, system, electronic device, and storage medium. The method first collects water surface flow data in the breach area, reconstructs the breach area in a virtual three-dimensional space based on the water surface flow data to obtain a three-dimensional flow field model, then extracts target features from the three-dimensional flow field model to obtain a flow field feature matrix, and then uses a first objective function as an optimization target to optimize the casting decision based on the flow field feature matrix, regional casting state data, and available material data to obtain an optimal casting plan. The first objective function represents at least one of minimizing flow velocity energy or minimizing material loss. The second objective function is used as an optimization target to optimize the transportation decision based on the optimal casting plan and transportation resource data to obtain an optimal transportation scheduling plan, and to schedule transportation resources based on the optimal transportation scheduling plan. The second objective function represents maximizing transportation efficiency. The present application can simultaneously construct a three-dimensional flow field model, determine the optimal casting plan that can reduce flow velocity and logistics loss based on the feature data in the three-dimensional flow field model, and then determine the optimal material transportation scheduling plan that can improve rescue efficiency based on the optimal casting plan, thereby improving the efficiency of dike breach sealing. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 This is a flow chart of a dike breach sealing method provided by an embodiment of the present application;

[0046] Figure 2 This is a schematic diagram of the hardware structure of the electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0047] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0048] It should be noted that although the system is divided into functional modules and the flowcharts illustrate a logical sequence, in some cases, the steps shown or described may be performed in a different order than the module division in the system or the order in the flowcharts. The terms "first," "second," and so on in the specification, claims, and drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or precedence.

[0049] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this application pertains. The terms used herein are for the purpose of describing the embodiments of this application only and are not intended to limit this application.

[0050] The embodiments of the present application provide a dike breach sealing method, system, electronic device and storage medium, aiming to improve the efficiency of dike breach sealing.

[0051] The dike breach sealing method, system, electronic device and storage medium provided in the embodiments of the present application are specifically described through the following embodiments. First, the dike breach sealing method in the embodiments of the present application is described.

[0052] The dike breach sealing method provided in the embodiment of the present application relates to the field of automated emergency rescue technology. The dike breach sealing method provided in the embodiment of the present application can be applied to a terminal, can be applied to a server side, or can be software running in a terminal or a server side. In some embodiments, the terminal can be a smart phone, a tablet computer, a laptop computer, a desktop computer, etc.; the server side can be configured as an independent physical server, or as a server cluster or distributed system composed of multiple physical servers, or as a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms; the software can be an application that implements the dike breach sealing method, etc., but is not limited to the above forms.

[0053] The present application can be used in many general or special computer system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, and the like. The present application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, and the like that perform specific tasks or implement specific abstract data types. The present application can also be practiced in distributed computing environments in which tasks are performed by remote processing devices connected via a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media, including storage devices.

[0054] According to some embodiments of the present application, the dike breach sealing method of the embodiment of the present application can be implemented on a system architecture consisting of a perception layer, a decision layer, and an execution layer, as follows:

[0055] The perception layer is fundamental to achieving rapid emergency decision-making. Specifically, it involves using a swarm of drones to quickly take off and hover above a dike breach. Using a high-precision Doppler radar sensor flowmeter and a real-time kinematic positioning system (RTK), the drones can obtain real-time information about the water velocity and direction in the breach area, as well as the spatial coordinates of the corresponding measuring points. Each drone is also equipped with a high-definition camera to capture real-time video of the water surface flow in the breach area. Advanced image recognition and computer vision technologies are used to assist in analyzing the velocity and direction distribution characteristics of the flow field, further improving the accuracy of water flow parameter measurements and the comprehensiveness of spatial coverage. This multi-sensor data fusion approach enables fast, efficient, and accurate data acquisition, completely eliminating the large measurement errors and low efficiency inherent in traditional manual measurement methods.

[0056] The decision-making layer is responsible for the important task of generating emergency rescue plans for sealing dike breaches. The traditional method of formulating emergency rescue plans relies on on-site analysis by experts, which is usually time-consuming and inefficient, and misses the golden period for emergency rescue. The embodiment of the present application proposes an intelligent and rapid emergency rescue decision-making generation mechanism that can automatically generate reliable emergency throwing plans. Specifically, the intelligent decision-making engine of the decision-making layer adopts a technical architecture that combines advanced deep reinforcement learning (DRL) and long short-term memory network (LSTM). Based on historical rescue data and real-time flow field data, it automatically and quickly predicts and generates emergency throwing plans. The specific method for generating emergency throwing plans will be introduced in the subsequent embodiments.

[0057] The execution layer mainly realizes the rapid dispatch of emergency materials and the intelligent management and deployment of on-site construction resources. The embodiment of this application introduces advanced V2X vehicle-road collaborative technology in the execution layer to realize real-time dynamic collaborative management of transport vehicles, material storage depots and emergency sites. The execution layer technology platform quickly and automatically accesses the local emergency material reserve database based on the emergency plan data output by the decision-making layer, and grasps the quantity and location distribution of local available emergency resources in real time. It also automatically generates the best material transportation scheduling plan based on factors such as real-time road traffic conditions, vehicle transportation capacity, and material loading and unloading time. The specific material transportation scheduling plan generation method is introduced in the subsequent embodiments.

[0058] Figure 1 This is an optional flow chart of the dike breach sealing method provided in the embodiment of the present application. Figure 1 The method may include but is not limited to steps S101 to S105.

[0059] Step S101, collecting water surface flow data in the breach area;

[0060] Step S102, reconstructing the breach area in a virtual three-dimensional space based on the water surface flow data to obtain a three-dimensional flow field model;

[0061] Step S103, target feature extraction is performed on the three-dimensional flow field model to obtain a flow field feature matrix;

[0062] Step S104, a first objective function is taken as an optimization target, and a throwing decision optimization is performed according to the flow field feature matrix, the regional throwing state data and the available material data to obtain an optimal throwing scheme, the first objective function representing at least one of a minimum flow velocity energy or a minimum material loss amount;

[0063] Step S105, a second objective function is taken as an optimization target, and a transportation decision optimization is performed according to the optimal throwing scheme and the transportation resource data to obtain an optimal transportation scheduling scheme, and the transportation resource scheduling is performed according to the optimal transportation scheduling scheme, wherein the second objective function represents a maximum transportation efficiency.

[0064] In step S101 of some embodiments, water surface flow data of the breach area can be collected through a perception layer, the water surface flow data including water surface flow video and water flow monitoring data, the water surface flow video being collected through a camera on a drone, and the water flow monitoring data being collected through a flow velocity sensor arranged on a water surface monitoring point. The water surface flow video includes continuous frame images, image collection time and image collection coordinates. The water flow monitoring data includes flow velocity data collected at each monitoring point, monitoring point position and collection time.

[0065] In an embodiment, multiple drones fly in an autonomously designed high-efficiency scanning path to achieve rapid and comprehensive coverage of the breach area. The scanning path of the drone adopts a grid scanning path optimization algorithm, and the path optimization algorithm is as shown in the following formula:

[0066] ;

[0067] wherein, represents the length of the optimized path, represents the path scanning coordinates, and the path optimization algorithm realizes automatic planning of the shortest path of the drone.

[0068] The drone simultaneously collects water flow data of each measurement point in real time through a flow velocity sensor on the water surface during flight, and each measurement point data is defined as a four-dimensional vector:

[0069] ;

[0070] wherein, is a spatial coordinate, is a water flow velocity vector (including flow velocity size and direction).

[0071] At the same time, the onboard high-definition camera collects high-definition water surface flow velocity videos in the breach area in real time. The water surface flow velocity videos can be transmitted to the data processing platform in real time through the 5G network. The data processing platform uses image recognition and computer vision technology to analyze the water surface texture characteristics in real time to quickly obtain the flow velocity field distribution on the breach surface.

[0072] In step S102 of some embodiments, by analyzing the water surface flow data, the water flow distribution data of the breach area is accurately calculated, and the water flow distribution data is represented in a virtual three-dimensional space to obtain a three-dimensional flow field model. The three-dimensional flow field model is a mathematical model used to describe the motion state of water flow in three-dimensional space. It uses mathematical equations and physical laws to depict the distribution and evolution laws of physical quantities such as velocity, pressure, turbulent kinetic energy of the fluid in three spatial dimensions and time (t). The three-dimensional flow field model can be used to extract the flow field features required for subsequent emergency decision-making, thereby improving the accuracy and comprehensiveness of the flow field feature extraction. Specifically, the embodiment of the present application can collect a terrain image of the breach area, initialize a terrain structure model of the breach area in a virtual three-dimensional space based on the terrain image, and then map the water flow distribution data to the terrain structure model to obtain a three-dimensional flow model.

[0073] In step S103 of some embodiments, target features refer to defined flow field-related features required for casting decisions, including but not limited to velocity-time features and energy features. The flow field feature matrix includes multiple flow field feature vectors, each of which represents velocity-time features and energy features at a specific location.

[0074] In step S104 of some embodiments, a throwing decision is made based on the flow field characteristic matrix, regional throwing status data, and available material data to obtain a throwing plan, and then the throwing plan is optimized with the first objective function as the optimization target to obtain the optimal throwing plan. The throwing plan is used to characterize the specifications of regional rescue materials (such as stone particle size), throwing amount, throwing position and sequence, and expected construction completion time. The first objective function represents at least one of minimizing flow velocity energy or minimizing material loss. The throwing plan is evaluated based on the first objective function, and the throwing plan is adjusted based on the evaluation result. Finally, the throwing plan that can minimize the sum of any number of flow velocity energy, material loss, or rescue time is selected as the optimal throwing plan.

[0075] In step S105 of some embodiments, a casting decision is made based on the optimal casting plan and transportation resource data to obtain a transportation scheduling plan. The transportation scheduling plan is then optimized using the second objective function as the optimization objective to obtain the optimal transportation scheduling plan. Transportation resource data includes the location distribution of materials, the location distribution of transportation resources (such as vehicles), the number of available emergency resources in each area, the carrying capacity of transportation resources, the efficiency of material loading and unloading, and real-time road traffic conditions. The transportation scheduling plan is used to represent the material carrying capacity, type of material carried, and transportation path of each transportation resource. The second objective function represents maximizing transportation efficiency. The transportation scheduling plan is evaluated based on the second objective function and adjusted based on the evaluation results. Finally, the transportation scheduling plan that maximizes transportation efficiency is selected as the optimal transportation scheduling plan.

[0076] According to some embodiments of the present application, the water surface flow data includes a water surface flow video, and step S102 may include but is not limited to the following steps:

[0077] Step S201, preprocessing each frame image in the water surface flow video to obtain preprocessed continuous frame images;

[0078] Step S202, extracting water surface ripple feature points from continuous frame images using a scale-invariant feature transformation algorithm;

[0079] Step S203, determining the water surface flow velocity distribution data based on the position changes of the water surface ripple feature points at each position in the previous and next frames;

[0080] Step S204: constructing a three-dimensional flow field model in a virtual three-dimensional space according to the water surface flow velocity distribution data.

[0081] In this embodiment, image recognition and computer vision technology are used to analyze the water surface texture features in real time to quickly obtain the flow velocity field distribution on the breach surface. The specific image flow field calculation algorithm includes the following steps:

[0082] Image preprocessing: Perform noise reduction, defogging, contrast enhancement and other preprocessing on each frame of the real-time water surface flow video to improve the image clarity and feature significance.

[0083] Feature point extraction and matching: The scale-invariant feature transform (SIFT) algorithm is used to extract and match water surface ripple feature points from consecutive frame images, and the water flow displacement is determined by the position changes of the water surface ripple feature points in the previous and next frames.

[0084] Calculate surface velocity using the optical flow method: The Lucas-Kanade optical flow algorithm is used to analyze the displacement of water surface ripple feature points in a continuous image sequence to achieve rapid calculation of the velocity vector field. The optical flow velocity calculation formula is expressed as:

[0085] ;

[0086] in, 、 is the grayscale gradient component of the image in space, is the grayscale gradient of the image over time, Represents the displacement velocity components of the water surface ripple feature points in the image in the horizontal and vertical directions.

[0087] Water surface velocity calibration and conversion: To accurately obtain the actual water surface velocity, the system uses the UAV RTK positioning system to obtain the correspondence between the real coordinates of the water surface and the image pixel coordinates, and determines the conversion coefficient between the image pixels and the real space scale. , convert the image optical flow velocity into the real water surface velocity. The conversion formula is as follows:

[0088] ;

[0089] in, is the actual water surface velocity (m / s), is the image scale calibration coefficient (m / pixel), 、 is the image pixel displacement speed calculated by the optical flow method.

[0090] The water surface velocity distribution data obtained through the above processing process is used to construct a three-dimensional flow field model in a virtual three-dimensional space.

[0091] The reconstruction of the three-dimensional flow field model uses advanced spatial data interpolation technology (such as Kriging interpolation), which can fully utilize the dense spatial measurement point data obtained by the drone (i.e., water surface velocity distribution data) to finely restore the actual on-site water flow conditions through spatial interpolation. The algorithm formula is as follows:

[0092] ;

[0093] in, Indicates the flow velocity at the i-th position, weight coefficient Solved by semivariogram:

[0094] ;

[0095] in, c(h) is the semivariogram function value, which indicates the spatial variation of the spatial variable (here, flow velocity) with distance (interval is h), and is used to calculate the weight coefficient in spatial interpolation methods (such as Kriging interpolation); h It is the distance between two spatial measurement points or the spatial lag distance (Lag distance); N(h) Indicates the distanceh The number of point pairs, that is, in spatial measurement data, the distance h The number of distances between a pair of measurement points; V ( x i ) indicates position x i The measured value at the measurement point, which specifically represents the flow velocity value at the measurement point in this embodiment; V ( x i +h ) indicates position x i +h The measured value at the distance measuring point x i For distance h The flow velocity value of another measuring point at .

[0096] At the same time, the system further applies advanced turbulence analysis technology (such as the k-ε turbulence model) to accurately identify and locate key hydrodynamic features such as the mainstream channel, high-speed area, and vortex area in the breach area, providing an accurate basis for the next casting decision. The turbulence analysis calculation formula is as follows:

[0097] Turbulent kinetic energy (k) equation:

[0098] ;

[0099] in, r is the fluid density (unit: kg / m³); k is the turbulent kinetic energy (unit: m² / s²), which represents the average kinetic energy of turbulence per unit mass of fluid; t is time (unit: s); u i Indicates the speed i Directional component (unit: m / s); x i 、x j Respectively represent i、j Direction coordinates (unit: m); P k represents the turbulent kinetic energy generation term (unit: kg / (m·s³)), which describes the rate of turbulent kinetic energy generation caused by the mean velocity gradient; e is the turbulent kinetic energy dissipation rate (unit: m² / s³), which describes the dissipation rate of turbulent kinetic energy in the process of converting it into internal energy (heat); m is the dynamic viscosity coefficient of the fluid (unit: Pa·s); m tThe turbulent viscosity coefficient (unit: Pa·s), also known as eddy viscosity or eddy viscosity coefficient, represents the momentum diffusion effect caused by turbulent fluctuation effect. s k The empirical constant of the turbulent model, which is a dimensionless quantity, is generally determined by experiment or numerical simulation.

[0100] The turbulent kinetic energy dissipation rate (ε) equation:

[0101] ;

[0102] wherein, C 1ε 、C 2ε Both are model empirical constants (dimensionless), and in k-e The turbulent model usually takes an empirical recommended value (for example, the value C 1ε =1.44, C 2ε =1.92); P k The generation term of turbulent kinetic energy (unit: kg / (m·s³)), which represents the rate of turbulent kinetic energy due to fluid shear effect; m t The turbulent viscosity coefficient (unit: Pa·s), which reflects the additional momentum transport effect caused by turbulent flow; s ε The turbulent model empirical constant (dimensionless), which can be 1.3.

[0103] The three-dimensional flow model includes water surface flow velocity distribution, turbulent kinetic energy and turbulent kinetic energy dissipation rate of key areas, and improves the accuracy of three-dimensional flow field analysis through the three-dimensional flow model, and provides accurate data basis for the next step decision.

[0104] According to some embodiments of the present application, the water surface flow data further includes water flow monitoring data of the water surface sensor, and step S102 can further include but is not limited to the following steps:

[0105] Step S301, determining the flow velocity distribution of the corresponding area in the water surface flow velocity distribution data according to the collection position of the water flow monitoring data;

[0106] Step S302, correcting the flow velocity distribution of the corresponding area based on the water flow monitoring data;

[0107] Step S303, obtaining the corrected water surface flow velocity distribution data according to the corrected flow velocity distribution of each area.

[0108] In this embodiment, after obtaining the water surface velocity distribution data based on water surface flow video analysis, it can be verified and fused with the water flow monitoring data measured by water surface sensors (such as radar current meters), thereby improving the reliability and spatial resolution of flow field parameter measurements and providing more accurate and comprehensive on-site data support for emergency plan decisions.

[0109] Specifically, each drone collects water surface flow videos in real time, and water flow monitoring data is measured by a radar current meter. The above data is transmitted to the data processing platform in real time through the wireless 5G network. After analyzing the water surface flow video to obtain water surface flow velocity distribution data, the water flow monitoring data is integrated into the water surface flow velocity distribution data to correct the video measurement error to obtain more accurate water surface flow velocity distribution data. Then, a three-dimensional flow field model reflecting the actual flow state on site is constructed based on the corrected water surface flow velocity distribution data.

[0110] The fusion and correction process for water surface velocity distribution data is as follows: Flow meters are placed at regular intervals along the river channel. The river channel can be divided into several zones based on the distribution of the flow meters, with each zone corresponding to a flow meter. Based on these zone divisions, the flow velocity distributions for the corresponding zones can be extracted from the water surface velocity distribution data from the video. The flow velocity distribution for the corresponding zone in the water surface velocity distribution data is determined based on the acquisition location of the water flow monitoring data. The flow velocity distribution for the corresponding zone is then corrected based on the water flow monitoring data. The correction method involves extracting the video-measured flow velocity at the corresponding acquisition location from the flow velocity distribution of the zone. The video-measured flow velocity at that acquisition location is then weightedly fused with the water flow monitoring data (from the flow meter) to obtain a fused flow velocity value for that acquisition location. A correction ratio is then determined based on the fused flow velocity value and the video-measured flow velocity. The flow velocity distribution for the zone is then corrected based on the correction ratio. The corrected flow velocity distributions for each zone are combined to produce the corrected water surface velocity distribution data.

[0111] According to some embodiments of the present application, step S103 may include but is not limited to the following steps:

[0112] Step S401, partitioning the three-dimensional flow field model to obtain regional flow field models of different flow velocity regions;

[0113] Step S402, extracting target features from the flow field models of each region to obtain flow field feature vectors of regions with different flow velocities;

[0114] Step S403: determining a flow field characteristic matrix according to the flow field characteristic vectors of different flow velocity areas.

[0115] In this embodiment, in order to provide reliability of sealing, different sealing strategies will be adopted for different breach situations. The breach situations are distinguished by the flow velocity of the region. For example, large stones need to be thrown for the mainstream area (high-speed area), and small and medium-sized stones need to be thrown for the slow-flow area (low-speed area). Therefore, in order to make targeted decisions on appropriate throwing schemes for different flow velocity areas in the future, the embodiment of the present application partitions the three-dimensional flow field model to obtain regional flow field models of different flow velocity areas, and then performs target feature extraction on each regional flow field model to obtain flow field feature vectors of different flow velocity areas, and combines the flow field feature vectors of different flow velocity areas to determine the flow field feature matrix.

[0116] According to some embodiments of the present application, step S104 may include but is not limited to the following steps:

[0117] Step S501, determining the material throwing constraint of each flow velocity area according to the flow velocity of different flow velocity areas in the flow field characteristic matrix;

[0118] Step S502, determining a casting decision plan for each flow rate area based on the regional casting state data, available material data, flow field characteristic vectors, and casting material constraints;

[0119] Step S503: Optimize the casting decision plan with the first objective function as the optimization target to obtain the optimal casting plan.

[0120] In this embodiment, after obtaining the three-dimensional flow field model, the entire breach area is divided into the mainstream area (high speed area), the transition area (medium speed area) and the slow flow area (low speed area) based on the flow velocity. Then, according to the throwing strategy based on different flow velocity areas, the throwing material constraints of each flow velocity area can be determined. The throwing material constraints include the throwing stone particle size and throwing density constraints.

[0121] Furthermore, in addition to using flow velocity as a basis for dividing flow zones, a long-short-term memory neural network can be used to predict flow evolution and scour pit locations in real time to identify future scour pit areas. The flow field feature vectors of these scour pit areas can then be segmented and extracted, allowing for optimization of the casting strategy for this area during subsequent casting decisions, thereby improving rescue effectiveness. Furthermore, based on the flow field feature matrix (including topographic features), areas with casting slopes greater than a preset value are identified. A constructed collapse classification model is used to predict the collapse probability of these areas in real time. When the collapse probability exceeds a threshold, the area is identified as a hazardous area. The flow field feature vectors of these hazardous areas are then segmented and extracted, allowing for optimization of the casting strategy for this area during subsequent casting decisions, thereby improving rescue effectiveness. The collapse classification model's input data can include topographic data and flow velocity time series.

[0122] Furthermore, in order to ensure the safety and convenience of construction, the embodiment of the present application can also plan warning areas and vehicle passing areas in combination with the three-dimensional flow field model, and display the planned warning areas and vehicle passing areas on the terminal of the on-site staff. By planning warning areas for users, risks to personnel and equipment caused by sudden changes in water flow can be prevented.

[0123] The particle size and weight of the thrown material can be used to estimate the critical impact particle size based on empirical formulas (such as the Izbash formula), thereby determining the thrown material constraints. In some examples, the throwing strategies for different flow rate areas are described as follows:

[0124] Mainstream area (high-speed area): The size and amount of large rocks need to be thrown, and intensive throwing is used to form a core layer of anti-impact;

[0125] Transition zone (medium speed zone): requires medium-sized and heavy stones to be thrown, as well as the size and amount of auxiliary wire cages;

[0126] Slow flow area (low speed area): mainly throwing small and medium-sized stones, supplemented by manual cooperation to trim the slope surface.

[0127] The decision-making layer input data for the embodiment of the present application includes flow field feature vectors (characterizing the flow velocity, flow direction, water depth, and other characteristics of each area), material throwing constraints (characterizing the constraint range of various parameters of the thrown material in the flow velocity area), regional throwing status data (characterizing the status of the area where the throwing has been completed), and available material data (characterizing the storage status of emergency materials in the area). The decision-making layer outputs the optimal throwing plan, which characterizes the specifications (such as stone particle size), throwing amount, throwing location, throwing sequence, or expected construction completion time of the emergency material in each area (including the different flow velocity areas divided as described above, predicted scour pit areas, or dangerous areas, etc.). For example, the decision-making layer identifies the most dangerous area in the flow field and calculates and determines the particle size, quantity, throwing location, and throwing sequence of the emergency sealing material based on real-time flow velocity parameters, breach cross-section conditions, and emergency material characteristics.

[0128] The decision layer includes a decision model and an optimization model. The decision model is used to calculate the particle size of the material suitable for emergency rescue and plugging in the corresponding area and the quantity of the required material based on the flow rate parameters, breach cross-section shape and material properties in the input data. For example, the state space of the decision model is defined as St={flow field characteristic vector, state of the thrown area, remaining available materials}, and the action space is defined as At={throwing area coordinates (x, y), stone particle size (D), throwing amount (Q)}. In the decision-making process, the embodiment of the present application uses the improved Xie Cai formula to accurately determine the particle size of the rescue stone, and the formula is expressed as follows:

[0129] ;

[0130] in, Indicates the stone particle size, represents the maximum flow velocity in the region, is the stability coefficient, is the relative density of stone, is the acceleration due to gravity.

[0131] In the decision-making process, the total amount of emergency stone is calculated as follows:

[0132] ;

[0133] in, Q total Indicates the total amount of thrown stones required for emergency sealing construction (volume, unit: m³);

[0134] A i Indicates the i The horizontal area of ​​the area to be blocked (unit: m²); H i Indicates the i The thickness of stone throwing required for each blocking area (unit: m); β Indicates the material loss coefficient during the construction process (dimensionless), which is used to take into account the actual loss of stone during the construction process (generally based on experience, the value is 5%~15%);

[0135] n Indicates the total number of areas to be sealed, that is, the total number of areas that require zoning and throwing construction.

[0136] The optimization model is implemented based on a deep learning algorithm. The optimization model is used to evaluate the casting scheme given by the decision model, and continuously optimize the decision model based on the evaluation results, so that the decision model can give the optimal casting scheme. The optimization model can also be used to predict and dynamically correct the progress and effect of future plugging construction in real time, thereby improving the real-time accuracy and adaptability of the rescue plan. It has been verified that the entire decision optimization process of the embodiment of the present application takes only a dozen minutes from the completion of data collection to the generation of the plan, which improves the timeliness of the rescue decision and meets the needs of the "golden period" of rescue. In order to achieve the optimal decision-making effect, the reward function (i.e., the first objective function) is defined as:

[0137] ;

[0138] in, 、 、 are all optimized weight coefficients, represents the flow energy, For the expected time of emergency construction, The amount of material loss.

[0139] According to some embodiments of the present application, step S105 may include but is not limited to the following steps:

[0140] Step S601: Determine the material transportation task for each transport resource based on the optimal casting plan and transport resource data. The material transportation task includes the transport target material, transport volume, transport starting point, and transport destination.

[0141] Step S602: Taking the minimum transportation time as the goal, the transportation resources are route-planned according to the material transportation task to obtain a transportation route;

[0142] Step S603: Determine a transportation scheduling decision plan based on the material transportation tasks and transportation routes of each transportation resource;

[0143] Step S604: Taking the second objective function as the optimization target, the transportation scheduling decision plan is optimized to obtain the optimal transportation scheduling plan.

[0144] In this embodiment, the execution layer can optimize the optimal transportation scheduling plan based on the optimal throwing plan output by the decision layer, so as to schedule the transportation resources according to the optimal transportation scheduling plan. Specifically,

[0145] The execution layer utilizes V2X vehicle-to-infrastructure collaboration technology to achieve real-time, dynamic collaborative management of transport vehicles, material storage depots, and emergency response sites. Based on the optimal drop plan data output by the decision layer and automatic access to the local emergency material reserve database (which stores transport resource data), the execution layer maintains a real-time understanding of the quantity and location of available emergency resources. Furthermore, it automatically generates an optimal transport scheduling plan based on factors such as real-time road traffic conditions, vehicle transport capacity, and material loading and unloading times.

[0146] In terms of transport route planning, the embodiments of this application utilize an intelligent route optimization algorithm based on minimum time entropy. This algorithm rapidly and intelligently optimizes transport routes and scheduling plans based on comprehensive factors such as the urgency of the material transport task, the road's real-time traffic capacity, and the remaining carrying capacity of transport vehicles. This ensures that emergency supplies are delivered safely and efficiently to the rescue site in the shortest possible time. On-site construction personnel can view the progress and location of material transport in real time through the interactive interface provided by the system, and provide feedback on on-site construction progress and needs, forming a dynamic closed loop between on-site needs and material transport scheduling.

[0147] The transport scheduling scheme in this embodiment includes material transport tasks and transport routes for each transport resource. Material transport tasks include target materials (characterized by material type, specifications, etc.), transport volume, and transport origin and destination. Material transport tasks are implemented using a task decision model based on a deep learning network. After the material transport tasks are determined by the task decision model, a material transport path optimization model can be used to plan the routes for the transport origin and destination within the material transport tasks. Specifically, the material transport path optimization model, which targets minimum transport time entropy, is expressed as follows:

[0148] ;

[0149] Among them, the path selection probability Calculated based on real-time road conditions, distance, vehicle load and other factors.

[0150] The optimization model is then used to evaluate the transport scheduling plan. Based on the evaluation results, the task decision model and the material transport path optimization model are adjusted so that the task decision model and the material transport path optimization model jointly output the optimal transport scheduling plan that maximizes transport efficiency. The transport efficiency evaluation function (i.e., the second objective function) is defined in the optimization model as follows:

[0151] ;

[0152] in, Indicates the Time taken by vehicle for transportation; Indicates the vehicle loading time, Indicates vehicle unloading time. The system evaluates and adjusts vehicle transportation strategies in real time to ensure timely arrival of materials at the construction site, enabling efficient and coordinated rescue efforts.

[0153] According to some embodiments of the present application, the three technical layers mentioned above—the perception layer, the decision layer, and the execution layer—are closely integrated through the 5G network and Mobile Edge Computing (MEC) technology. All perception data, decision instructions, execution status, and construction progress are transmitted in real time to the edge computing node, where data is rapidly processed and analyzed, and then fed back to each layer to form a coordinated response mechanism. For example, during a rescue operation, the data sensed by the drone can be updated in real time to the decision layer, which dynamically updates the rescue plan based on the latest data. At the same time, the transportation scheduling system also updates the transportation route and resource allocation in real time based on the decision plan, achieving a closed data loop and real-time linkage and collaboration among the three layers, significantly improving the efficiency of information sharing and the speed of on-site response during the rescue process. Through the technical architecture and implementation methods proposed in the embodiments of the present application, intelligent, precise, and automated levee breach rescue is achieved, providing a new technical solution for levee rescue and a replicable and scalable technical reference for the field of emergency rescue technology.

[0154] According to some embodiments of the present application, the breach rescue and plugging implementation process of the embodiments of the present application is specifically as follows: S11, system startup and drone perception task scheduling.

[0155] When a dike breach occurs, the system activates emergency procedures. The emergency command center sends a start command, and the drone swarm automatically plans a flight path and quickly flies over the designated rescue site to collect data. During this data collection process, the drones transmit sensory data in real time to the data center via the 5G communication network.

[0156] S12, data fusion and real-time 3D flow field modeling.

[0157] After receiving the real-time data from the drone swarm, the data center (edge ​​computing node) immediately initiates a 3D flow field model reconstruction program and performs data fusion. Using a spatial interpolation algorithm, the 3D flow field is rapidly reconstructed. A turbulence model is then used to analyze the flow pattern and identify critical hazardous areas, creating a real-time flow field situation map.

[0158] S13, intelligent decision-making layer solution generation.

[0159] After receiving real-time flow field data, the decision-making layer quickly loads historical rescue experience and model parameters, and uses deep reinforcement learning and LSTM fusion models to automatically generate rescue throwing plans, including clear rescue material specifications, throwing quantity, throwing area, throwing sequence, and expected construction completion time.

[0160] S14, dynamic material scheduling at the intelligent execution layer.

[0161] Upon receiving the decision, the execution layer automatically accesses the local emergency material reserve database, providing real-time information on material distribution and transport vehicle locations. It then optimizes and dispatches vehicle transport routes based on real-time traffic conditions. Intelligent dispatch algorithms swiftly deliver emergency materials to the construction site, ensuring the coordinated and precise delivery of all resources.

[0162] S15, dynamic feedback and optimization adjustment of the rescue process.

[0163] During the emergency rescue construction process, on-site personnel use the terminal to provide real-time feedback on the rescue progress and changes in on-site needs. The system updates and adjusts the rescue plan and resource allocation strategy in real time to achieve dynamic closed-loop management and control, ensuring that the rescue work is completed smoothly, efficiently and accurately.

[0164] According to some embodiments of the present application, the embodiments of the present application have the following beneficial effects:

[0165] First, the rescue methods of related technologies rely on manual visual inspection, water gauge readings or single-point equipment measurements to determine the flow velocity, water depth and water flow morphology at the breach. This traditional measurement method has a slow response and is difficult to accurately restore the overall hydrodynamic structure in a complex breach flow field, which can easily lead to decision-making errors. The embodiment of the present application uses a cluster of drones equipped with high-precision Doppler radar sensor flowmeters, RTK real-time positioning systems and high-definition cameras to automatically plan the optimal scanning path over the breach area of ​​the embankment, and quickly and efficiently obtain three-dimensional flow field data in the breach area. The drone measures the flow velocity, flow direction and spatial coordinates of the breach area in real time, and simultaneously collects high-definition images of the water surface. The system uses image recognition and optical flow calculation technology (SIFT feature matching and Lucas-Kanade optical flow method) to analyze the motion information of the water surface image texture in real time, calculate the water surface flow velocity distribution data, and fuse and cross-check with the radar flow meter measurement data, effectively improving the flow velocity measurement accuracy and data reliability. In addition, spatial interpolation algorithms (such as Kriging interpolation) and the improved k-ε turbulence model are used to achieve rapid reconstruction of the three-dimensional water flow field data of the breach and high-precision flow state analysis, providing a reliable data basis for accurate decision-making on emergency rescue plans.

[0166] Second, the riprap rescue decision-making process in the related art usually relies on the experience evaluation of on-site experts and temporary meeting discussions. It often takes too long from information collection, evaluation and analysis to scheme formulation, and it is difficult to meet the demand for making scientific decisions within the "golden rescue period". Especially under the conditions of night or poor communication, the risk of decision delay is greater. The embodiment of the present application adopts a transfer learning decision engine constructed by fusing deep reinforcement learning (DRL) and long short-term memory network (LSTM), which quickly and automatically identifies the dangerous area of the breach, and intelligently determines the key parameters (throwing area, throwing sequence, stone particle size, throwing quantity) of the plugging rescue scheme. By combining historical rescue experience with real-time data, a reward function (objective function) is designed for the optimization of the throwing scheme, which improves the speed and accuracy of the rescue scheme generation, and ensures that the rescue plugging operation can be quickly executed within the "golden rescue period".

[0167] Third, the embodiment of the present application proposes a rescue throwing stone particle size calculation method based on the improved Xiecai formula, which accurately determines the safe particle size of the thrown stone combined with real-time flow rate data, realizes the safety and stability of the dike plugging. Based on the comprehensive formula optimized by flow rate energy loss, construction time and material loss, the total stone quantity required for rescue plugging construction is accurately determined, and the plugging construction effect and resource utilization efficiency are improved.

[0168] Fourth, in the related art, the dispatching of rescue materials from the storage point to the scene usually adopts a fixed path or manual instruction, without considering real-time road conditions, task urgency, vehicle remaining capacity and other multi-dimensional factors, resulting in the problem that rescue resources cannot dynamically match the scene demand, and there are problems such as waste of materials and delay of dispatching. The embodiment of the present application proposes a rescue material transportation path optimization model with minimum time entropy as the optimization objective, and realizes real-time dynamic and accurate scheduling of rescue resources by combining V2X vehicle-road cooperative communication technology, which significantly improves the transportation efficiency and scheduling accuracy of rescue resources. The embodiment of the present application proposes a dynamic feedback and scheduling adjustment method based on a transportation efficiency evaluation model, which realizes real-time dynamic cooperation between material scheduling and on-site construction, so that rescue materials can arrive at the rescue scene in a timely and efficient manner.

[0169] Fifth, the embodiment of the present application constructs a rescue intelligent decision-making closed-loop system covering the three-level linkage of "perception - decision - execution", shares the data of the perception, decision and execution layers in real time, dynamically adjusts the rescue scheme, and realizes intelligent dynamic closed-loop control of the whole process of the rescue scene. The construction site of the rescue construction can feed back the construction progress and resource demand in real time, and the intelligent decision-making and execution system dynamically updates the rescue scheme and resource scheduling strategy accordingly, which improves the overall timeliness and accuracy of the dike breach rescue disposal.

[0170] The embodiment of the present application also proposes a dike breach plugging system, which comprises:

[0171] A first module for collecting water surface flow data of the breach area;

[0172] The second module is configured to virtually reconstruct the breach area in a three-dimensional space according to the water surface flow data, and obtain a three-dimensional flow field model;

[0173] The third module is configured to extract target features from the three-dimensional flow field model, and obtain a flow field feature matrix;

[0174] The fourth module is configured to optimize a throwing decision according to the flow field feature matrix, the area throwing state data and the available material data, with a first target function as an optimization target, to obtain an optimal throwing scheme, the first target function representing at least one of minimized flow energy or minimized material loss;

[0175] The fifth module is configured to optimize a transportation decision according to the optimal throwing scheme and transportation resource data, with a second target function as an optimization target, to obtain an optimal transportation scheduling scheme, and to schedule the transportation resources according to the optimal transportation scheduling scheme, the second target function representing maximized transportation efficiency.

[0176] It can be understood that the content in the above-mentioned dike breach plugging method embodiments is applicable to the system embodiments, the system embodiments specifically implement the same functions as the above-mentioned dike breach plugging method embodiments, and achieve the same beneficial effects as the above-mentioned dike breach plugging method embodiments.

[0177] The application also provides an electronic device, which includes a memory, a processor, a program stored in the memory and executable on the processor, and a data bus for realizing connection and communication between the processor and the memory, and the program is executed by the processor to realize the above-mentioned dike breach plugging method. The electronic device can be any smart terminal, such as a tablet computer or a vehicle-mounted computer.

[0178] Please refer to Figure 2 , Figure 2 The hardware structure of the electronic device of another embodiment is illustrated, which includes:

[0179] The processor 901 can be implemented in a general-purpose CPU (Central Processing Unit), a microprocessor, an ASIC (Application Specific Integrated Circuit), or one or more integrated circuits, and is used to execute related programs to implement the technical solutions provided by the embodiments of the application.

[0180] The memory 902 can be implemented in the form of a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 902 can store an operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 902 and is called by the processor 901 to execute the dike breach sealing method of the embodiments of this application.

[0181] Input / output interface 903, used to implement information input and output;

[0182] Communication interface 904, used to implement communication interaction between this device and other devices, which can be achieved through wired means (such as USB, network cable, etc.) or wireless means (such as mobile network, WiFi, Bluetooth, etc.);

[0183] Bus 905 , which transmits information between various components of the device (e.g., processor 901 , memory 902 , input / output interface 903 , and communication interface 904 );

[0184] The processor 901 , the memory 902 , the input / output interface 903 and the communication interface 904 are connected to each other in communication within the device via a bus 905 .

[0185] An embodiment of the present application also provides a storage medium, which is a computer-readable storage medium used for computer-readable storage. The storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the above-mentioned embankment breach sealing method.

[0186] The memory, as a non-transient computer-readable storage medium, can be used to store non-transient software programs and non-transient computer executable programs. In addition, the memory may include a high-speed random access memory and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some embodiments, the memory may optionally include a memory remotely arranged relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0187] The embodiments described in the embodiments of this application are intended to more clearly illustrate the technical solutions of the embodiments of this application and do not constitute a limitation on the technical solutions provided by the embodiments of this application. Those skilled in the art will appreciate that with the evolution of technology and the emergence of new application scenarios, the technical solutions provided in the embodiments of this application are also applicable to similar technical problems.

[0188] Those skilled in the art will understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of the present application, and may include more or fewer steps than shown in the figures, or a combination of certain steps, or different steps.

[0189] The system embodiment described above is merely illustrative. The units described as separate components may or may not be physically separate, i.e., they may be located in one place or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of this embodiment.

[0190] Those skilled in the art will appreciate that all or some of the steps in the methods, systems, and functional modules / units in the devices disclosed above may be implemented as software, firmware, hardware, or appropriate combinations thereof.

[0191] The terms "first", "second", "third", "fourth", etc. (if any) in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0192] It should be understood that in this application, "at least one (item)" means one or more, and "plurality" means two or more. "And / or" is used to describe the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" can mean: only A exists, only B exists, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally indicates that the previous and next associated objects are in an "or" relationship. "At least one of the following items" or similar expressions refers to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, c can be single or multiple.

[0193] In the several embodiments provided in this application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative. For example, the division of the above units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of systems or units, which can be electrical, mechanical or other forms.

[0194] The units described above as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0195] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0196] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes multiple instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of various embodiments of the present application. The aforementioned storage medium includes: various media that can store programs, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0197] The preferred embodiments of the present invention are described above with reference to the accompanying drawings, but are not intended to limit the scope of the present invention. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and essence of the present invention should be within the scope of the present invention.

Claims

1. A dike breach sealing method, characterized in that: The following steps are involved: Collect water surface flow data in the breach area; Performing virtual three-dimensional space reconstruction on the breach area according to the water surface flow data to obtain a three-dimensional flow field model; Performing target feature extraction on the three-dimensional flow field model to obtain a flow field feature matrix; Taking a first objective function as an optimization goal, optimizing the casting decision according to the flow field characteristic matrix, the regional casting state data, and the available material data to obtain an optimal casting solution, wherein the first objective function represents at least one of minimizing flow velocity energy or minimizing material loss; Taking the second objective function as the optimization goal, optimizing the transportation decision according to the optimal casting plan and the transportation resource data to obtain an optimal transportation scheduling plan, and scheduling transportation resources according to the optimal transportation scheduling plan, wherein the second objective function represents maximizing transportation efficiency; The target feature extraction of the three-dimensional flow field model to obtain a flow field feature matrix includes the following steps: Partitioning the three-dimensional flow field model to obtain regional flow field models of different flow velocity areas; Target features are extracted from the flow field models of each region respectively to obtain the flow field feature vectors of different flow velocity regions; Determine the flow field characteristic matrix according to the flow field characteristic vectors in different flow velocity areas; The method takes the first objective function as the optimization target, optimizes the casting decision according to the flow field characteristic matrix, the regional casting state data, and the available material data, and obtains the optimal casting plan, including the following steps: Determining the material throwing constraint for each flow velocity region according to the flow velocities of different flow velocity regions in the flow field characteristic matrix; Determining a casting decision plan for each of the flow rate areas according to the regional casting state data, the available material data, the flow field characteristic vector, and the casting material constraint; Taking the first objective function as the optimization target, optimizing the casting decision plan to obtain the optimal casting plan; The throwing decision plan includes the throwing area coordinates, stone particle size and throwing amount. The stone particle size is calculated by the following formula: ； in, Indicates the stone particle size, represents the maximum flow velocity in the region, is the stability coefficient, is the relative density of stone, is the acceleration due to gravity.

2. The dike breach sealing method according to claim 1, characterized in that: The water surface flow data includes a water surface flow video, and the virtual three-dimensional space reconstruction of the breach area according to the water surface flow data to obtain a three-dimensional flow field model includes the following steps: Preprocessing each frame image in the water surface flow video to obtain preprocessed continuous frame images; Extracting water surface ripple feature points from the continuous frame images using a scale-invariant feature transformation algorithm; Determine the water surface flow velocity distribution data according to the position change of the water surface ripple feature points at each position in the previous and next frames; A three-dimensional flow field model is constructed in a virtual three-dimensional space according to the water surface flow velocity distribution data.

3. The dike breach sealing method according to claim 2, characterized in that: The water surface flow data also includes water flow monitoring data from a water surface sensor, and the virtual three-dimensional space reconstruction of the breach area is performed based on the water surface flow data to obtain a three-dimensional flow field model, further comprising the following steps: Determining the flow velocity distribution of a corresponding area in the water surface flow velocity distribution data according to the collection location of the water flow monitoring data; Correcting the flow velocity distribution in the corresponding area based on the water flow monitoring data; The corrected water surface flow velocity distribution data is obtained according to the corrected flow velocity distribution in each area.

4. The dike breach sealing method according to claim 1, characterized in that: The method of optimizing the transportation decision based on the optimal casting plan and the transportation resource data using the second objective function as the optimization target to obtain the optimal transportation scheduling plan includes the following steps: Determine a material transportation task for each transport resource based on the optimal casting plan and the transport resource data, wherein the material transportation task includes the transport target material, the transport volume, the transport starting point, and the transport destination; Taking the minimum transportation time as the goal, the transportation resources are route-planned according to the material transportation task to obtain a transportation route; Determine the transportation scheduling decision plan based on the material transportation tasks and transportation routes of each transportation resource; Taking the second objective function as the optimization target, the transportation scheduling decision plan is optimized to obtain the optimal transportation scheduling plan.

5. A dike breach sealing system, characterized in that: include: The first module is used to collect water surface flow data in the breach area; The second module is used to reconstruct the breach area in a virtual three-dimensional space according to the water surface flow data to obtain a three-dimensional flow field model; The third module is used to extract target features from the three-dimensional flow field model to obtain a flow field feature matrix; A fourth module is configured to optimize the casting decision based on the flow field characteristic matrix, the regional casting state data, and the available material data, using the first objective function as an optimization target to obtain an optimal casting solution, wherein the first objective function represents at least one of minimizing flow velocity energy or minimizing material loss; a fifth module, configured to optimize transportation decisions based on the optimal drop plan and transportation resource data using a second objective function as an optimization objective, to obtain an optimal transportation scheduling plan, and to schedule transportation resources based on the optimal transportation scheduling plan, wherein the second objective function represents maximizing transportation efficiency; The third module is specifically configured to perform the following steps: Partitioning the three-dimensional flow field model to obtain regional flow field models of different flow velocity areas; Target features are extracted from the flow field models of each region respectively to obtain the flow field feature vectors of different flow velocity regions; Determine the flow field characteristic matrix according to the flow field characteristic vectors in different flow velocity areas; The fourth module is specifically configured to perform the following steps: Determining the material throwing constraint for each flow velocity region according to the flow velocities of different flow velocity regions in the flow field characteristic matrix; Determining a casting decision plan for each of the flow rate areas according to the regional casting state data, the available material data, the flow field characteristic vector, and the casting material constraint; Taking the first objective function as the optimization target, optimizing the casting decision plan to obtain the optimal casting plan; The throwing decision plan includes the throwing area coordinates, stone particle size and throwing amount. The stone particle size is calculated by the following formula: ； in, Indicates the stone particle size, represents the maximum flow velocity in the region, is the stability coefficient, is the relative density of stone, is the acceleration due to gravity.

6. An electronic device, characterized in that: The electronic device includes a memory, a processor, a program stored in the memory and executable on the processor, and a data bus for implementing connection and communication between the processor and the memory. When the program is executed by the processor, the steps of the method described in any one of claims 1 to 4 are implemented.

7. A storage medium, wherein the storage medium is a computer-readable storage medium and is used for computer-readable storage, characterized in that: The storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the steps of any one of claims 1 to 4.

Images