Model establishment method and device for rail transit flood prevention, server and medium

Abstract

The application discloses a model establishment method and device for track traffic flood prevention, a server and a medium. The method comprises the following steps: acquiring digital twin data of a track traffic station, determining the number of flood prevention key points and corresponding flood prevention attribute categories according to the digital twin data; when the total number of flood prevention input control key points, flood prevention transitivity key points, flood prevention decisive key points and flood prevention control key points is greater than a first threshold value, a dimensionality reduction improved feedforward neural network is established; and when the number of one fourth of the flood prevention input control key points, the number of the flood prevention transitivity key points, the number of the flood prevention decisive key points and the number of the flood prevention control key points is less than the first threshold value, a GNN model is established. The complexity of the current station can be determined, and a suitable model can be selected according to the complexity. By using the suitable model, the calculation amount can be reduced, and the overfitting phenomenon caused by the excessive complexity of the model can be avoided.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of AI model, and particularly relates to a model establishing method and device for track traffic flood prevention, a server and a medium. BACKGROUND

[0002] Urban rail transit is an indispensable part of the current urban infrastructure layout in China, and as a reorganized part of urban sustainable development, it can effectively alleviate the contradiction between people and urban traffic congestion. With global climate change, extreme rainfall events occur frequently, and the flood prevention system has become an important part of urban resilience construction. Due to the particularity of urban rail transit, once waterlogging occurs, it will cause light traffic suspension, affect urban traffic and people's travel, and cause heavy property losses. Therefore, during the planning and design of the pre-construction and post-reconstruction, the flood prevention performance of the rail transit station needs to be effectively evaluated, and then the flood prevention equipment is reasonably planned to ensure that the flood prevention performance of the rail transit station meets the flood prevention requirements.

[0003] In the prior art, a neural network model can be used to effectively evaluate the flood prevention performance of the rail transit station. By inputting the data of each station and the collected waterlogging situation, the flooding situation of each key part in the station is obtained. In the process of implementing the present application, the inventors have found the following technical problems: Due to the existence of elevated, ordinary and complex transfer stations of various types of stations, and the data types of the stations are various and complex. If the AI model is relatively simple, although it can produce good prediction results for elevated and ordinary stations, the accuracy will be greatly reduced for complex transfer stations. If the AI model is relatively complex, a large amount of default data will be generated for elevated and ordinary stations, and overfitting is easy to occur, and the prediction accuracy will be greatly reduced for this kind of scene. SUMMARY

[0004] The purpose of the present application is to overcome the deficiencies in the prior art, solve the technical problem of low accuracy in the prior art when a single model is used to evaluate and predict the flood prevention performance of the rail transit station, and provide a model establishing method, device, server and medium for track traffic flood prevention.

[0005] The purpose of the present application is achieved by the following technical solutions:

[0006] A model establishing method for track traffic flood prevention, comprising:

[0007] obtaining digital twin data of a rail transit station, determining the number of flood prevention key points and corresponding flood prevention attribute categories according to the digital twin data, the flood prevention attribute categories including flood prevention input control key points, flood prevention transfer key points, flood prevention decision key points and flood prevention control key points;

[0008] when the total number of the anti-flooding input control key points, the anti-flooding transitivity key points, the anti-flooding decisive key points and the anti-flooding control key points is greater than the first threshold value, a dimensionality reduction improved feedforward neural network is established, which takes the data of the anti-flooding input control key points and the anti-flooding control key points as inputs and outputs the predicted data of the anti-flooding transitivity key points and the anti-flooding decisive key points;

[0009] when the total number of the anti-flooding input control key points, the anti-flooding transitivity key points, the anti-flooding decisive key points and the anti-flooding control key points is less than the first threshold value, a GNN model is established, which establishes the connection relationship between the updated hierarchical nodes according to the hierarchical relationship of the clustering nodes of the anti-flooding input control key points, the anti-flooding transitivity key points, the anti-flooding decisive key points and the anti-flooding control key points, and constantly updates each hierarchical node according to the data of the anti-flooding input control key points, and obtains the predicted classification of the anti-flooding transitivity key points and the anti-flooding decisive key points according to the updated hierarchical nodes.

[0010] Further, the GNN model comprises:

[0011] an aggregation layer, configured to establish an input node graph according to the positional relationship of the anti-flooding input control key points, and generate anti-flooding input aggregation nodes by using an aggregation function;

[0012] a graph establishment layer, configured to generate a directed graph by using the positional relationship and the water flow sequence relationship of the anti-flooding input aggregation nodes, the anti-flooding transitivity key points, the anti-flooding decisive key points and the anti-flooding control key points, wherein the anti-flooding control key points are negative;

[0013] a fully connected layer, configured to output the anti-flooding possibility probability of the anti-flooding transitivity key points and the anti-flooding decisive key points by using the attributes of the anti-flooding transitivity key points and the anti-flooding decisive key points output by the graph establishment layer.

[0014] Further, the obtaining of the digital twin data of the rail transit station comprises:

[0015] obtaining virtual precipitation data predicted by a meteorological large model;

[0016] determining a corresponding anti-flooding management strategy according to the virtual precipitation data, and updating the digital twin data based on the anti-flooding management strategy.

[0017] Further, the determining of the corresponding anti-flooding management strategy according to the virtual precipitation data and the updating of the digital twin data based on the anti-flooding management strategy comprise:

[0018] adjusting the height of the water retaining facility and the drainage power of the peripheral drainage facility according to the precipitation data and a water retaining facility and peripheral drainage facility operation table;

[0019] According to the adjusted water retaining facilities and peripheral drainage facilities, the digital twin data is updated.

[0020] Further, the number of flood prevention key points and the corresponding flood prevention attribute categories are determined according to the digital twin data, including:

[0021] According to the flood prevention requirements, determine each flood prevention attribute category, and according to the digital twin data, select a plurality of flood prevention points according to the flood prevention transfer key points, the flood prevention decisive key points and the flood prevention control key points;

[0022] According to the plurality of flood prevention points, determine the redundancy index and the hot standby index of each flood prevention point;

[0023] According to the redundancy index and the hot standby index, calculate the vulnerability of each flood prevention point, and calculate the stability of each flood prevention point;

[0024] According to the vulnerability and stability, select the flood prevention key points from the flood prevention points.

[0025] Further, according to the digital twin data, a plurality of flood prevention points are selected according to each category, including:

[0026] According to the site digital twin data, the stairs in the site are divided according to the grid unit;

[0027] The grid unit is selected as the flood prevention transfer key point from the bottom of the stairs, the side wall of the middle layer and the position of the middle layer;

[0028] Using the turbulence model, the water flow velocity of each flood prevention transfer key point is calculated according to the data of the flood prevention input control key point.

[0029] Further, the dimensionality reduction improved feedforward neural network includes:

[0030] The input layer is used to convert the location of the flood prevention input control key point and the flood prevention control key point, the predicted precipitation, the time of the predicted precipitation, the spatial position data, the redundancy index and the hot standby index into flood prevention point multidimensional data;

[0031] The dimensionality reduction layer is used to perform dimensionality reduction processing on the flood prevention point multidimensional data to obtain dimensionality reduction data of the flood prevention point multidimensional data;

[0032] A plurality of hidden layers are used to extract abstract features through linear transformation;

[0033] Two output layers, one of which is used to output the flood prevention prediction result according to the abstract features using the activation function, and the other is used to output the flow velocity prediction result of the flood prevention transfer control point according to the abstract features output by the partial hidden layer;

[0034] An adjusting unit is configured to adjust a bias matrix corresponding to the partial hidden layer according to a result obtained by the turbulent flow model.

[0035] The application further provides a model establishing device for rail transit flood prevention.

[0036] An obtaining module is configured to obtain digital twin data of a rail transit station, and determine a number of flood prevention key points and corresponding flood prevention attribute categories according to the digital twin data, wherein the flood prevention attribute categories include flood prevention input control key points, flood prevention transfer key points, flood prevention decision key points and flood prevention control key points.

[0037] A first establishing module is configured to establish a dimensionality reduction improved feedforward neural network when a total number of the flood prevention input control key points, the flood prevention transfer key points, the flood prevention decision key points and the flood prevention control key points is greater than a first threshold value, wherein the dimensionality reduction improved feedforward neural network takes data of the flood prevention input control key points and the flood prevention control key points as input, and outputs predicted data of the flood prevention transfer key points and the flood prevention decision key points.

[0038] A second establishing module is configured to establish a GNN model when a number of one fourth of the flood prevention input control key points, a sum of the flood prevention transfer key points, the flood prevention decision key points and the flood prevention control key points is less than the first threshold value, wherein the GNN model establishes a connection relationship between updated hierarchical nodes according to a clustering node of the flood prevention input control key points, a hierarchical relationship of the flood prevention transfer key points, the flood prevention decision key points and the flood prevention control key points, constantly updates each hierarchical node according to data of the flood prevention input control key points, and obtains predicted classification of the flood prevention transfer key points and the flood prevention decision key points according to the updated hierarchical nodes.

[0039] The application further provides a server, comprising:

[0040] One or more processors;

[0041] A storage device configured to store one or more programs;

[0042] When the one or more programs are executed by the one or more processors, the one or more processors implement the model establishing method for rail transit flood prevention.

[0043] The application further provides a storage medium containing computer executable instructions, which are used to execute the model establishing method for rail transit flood prevention when executed by a computer processor.

[0044] Compared with the prior art, the technical scheme of the application has the beneficial effects that:

[0045] 1. The application can accurately capture the actual situation of the station and make complexity judgment: the application obtains digital twin data of rail transit station, digitizes and virtualizes information such as physical space, meteorological data and flood prevention measures, so that the flood prevention evaluation is established on the basis of real and dynamic data. This data-driven mode can reflect the structure and operation state of the station in real time.

[0046] Using digital twin data, multiple flood prevention points are set in the station according to the flood prevention requirements, and through redundancy, hot standby and vulnerability and stability calculation, the key points are accurately identified from multiple potential flood prevention points, and they are divided into "flood prevention input control", "flood prevention transfer", "flood prevention decisiveness" and "flood prevention control" categories, so as to capture the influence of different key factors on the flood prevention effect. This multi-dimensional data extraction and analysis effectively reduces the risk of missing key factors and provides comprehensive and detailed data support for subsequent model establishment.

[0047] 2. The application can realize the flexibility selection of the model: according to the total number of key points in the station and the proportional relationship between the categories (for example, when the number of flood prevention input control key points is sufficient and the total number is higher than the first threshold value, the dimensionality reduction improved feedforward neural network is adopted; when the number of key points is small and the proportion is lower than the threshold value, the graph neural network GNN is adopted), the adaptive selection of model is realized.

[0048] The dimensionality reduction improved feedforward neural network network eliminates data redundancy through the dimensionality reduction layer after inputting the data of flood prevention input control key points and flood prevention control key points, and then extracts abstract features through multiple hidden layers, and finally outputs the prediction results of flood prevention transfer and flood prevention decisiveness key points. This design not only reduces the data dimension and computational complexity, but also avoids the overfitting problem caused by too complex model.

[0049] The graph neural network (GNN) model is suitable for complex station structure and small amount of data scene. The GNN model establishes the connection relationship between each level node (including aggregation node, update level node, etc.), and transmits, fuses and classifies information according to the natural law of space position and water flow direction, so as to ensure the classification accuracy of prediction. This model is especially suitable for flood prevention risk assessment in the case of non-homogeneous, multi-structure and small sample.

[0050] 3. The application combines virtual precipitation data and flood prevention management strategy: the application obtains virtual precipitation data by using meteorological big model prediction, and dynamically adjusts the physical facility parameters (such as the height of water retaining facility and the drainage power of drainage facility) according to the precipitation data and the operation table of water retaining facility and peripheral drainage facility, so that the digital twin data is updated in real time.

[0051] And based on the closed-loop feedback mechanism, this cycle update based on actual rainfall prediction data and facility adjusted digital twin data not only makes the model prediction have real-time correction function, but also provides timely feedback and improvement basis for flood prevention management strategy, thereby effectively preventing risks under extreme rainfall events.

[0052] 4. The application can reduce the operation burden and improve the prediction accuracy: by accurately judging the complexity of the site data, the application adaptively selects a simplified or complex model, which not only effectively reduces the calculation amount, but also avoids overfitting problem caused by too little data, and finally realizes higher prediction accuracy.

[0053] The dimension reduction layer is used to pre-process the input data, remove redundant information, and ensure that the subsequent hidden layer can more efficiently extract key abstract features, thereby effectively improving the accuracy and robustness of the prediction results.

[0054] 5. The application provides a scientific basis for flood prevention measure planning: the key point data acquisition, key model selection and real-time closed-loop update of the application make the site flood prevention performance evaluation more accurate, which provides a quantitative and verifiable scientific basis for the planning and improvement of urban rail transit flood prevention measures, thereby facilitating the rational allocation of flood control resources, reducing accident risks, and reducing economic losses and social impacts caused by extreme rainfall events.

[0055] In addition, different evaluation strategies are adopted for different types of stations such as elevated, ordinary and complex transfer stations, and the application provides a technical solution with wide adaptability and fine risk management, which can effectively predict flood prevention risks in different scenarios and improve overall efficiency.

[0056] In summary, through real-time acquisition and fine processing of digital twin data, multi-angle classification of key points, adaptive selection of models and dynamic strategy regulation, the application can significantly improve the accuracy of rail transit station flood prevention performance evaluation, reduce model calculation complexity and risk overfitting, thereby providing reliable, effective and scientific decision-making basis for reasonable planning of flood prevention measures and urban traffic safety. BRIEF DESCRIPTION OF DRAWINGS

[0057] Figure 1 is a flowchart of the model establishment method for rail transit flood prevention provided by embodiment one of the application;

[0058] Figure 2 is a flowchart of the model establishment method for rail transit flood prevention provided by embodiment two of the application;

[0059] Figure 3 is a structural schematic diagram of the model establishment device for rail transit flood prevention provided by embodiment three of the application;

[0060] Figure 4 is a structural schematic diagram of a server provided by an embodiment four of the present application. DETAILED DESCRIPTION

[0061] The present application will be further described below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely intended for the purpose of interpretation of the present application and are not limiting of the present application. In addition, it should be noted that only the parts related to the present application are shown in the accompanying drawings for the purpose of description.

[0062] Embodiment one

[0063] Figure 1 is a flowchart of a model establishment method for rail transit flood prevention provided by an embodiment one of the present application. The embodiment can be applicable to the case of model establishment for rail transit flood prevention, in particular, station flood prevention. The method can be executed by a model establishment device for rail transit flood prevention, and specifically includes the following steps:

[0064] In step 110, digital twin data of the rail transit station is obtained, and the number of flood prevention key points and corresponding flood prevention attribute categories are determined according to the digital twin data. The flood prevention attribute categories include flood prevention input control key points, flood prevention transitivity key points, flood prevention decision key points, and flood prevention control key points.

[0065] In the planning of the rail transit station, a high-precision three-dimensional model can usually be constructed based on building information modeling (BIM) technology to truly reproduce the spatial structure and equipment distribution of the station, such as gates, elevators, ventilation systems, etc. The digital twin data of the rail transit station can be obtained from the BIM, and the digital twin data can be obtained according to the data currently collected by the sensors.

[0066] In the present embodiment, the flood prevention key points can be divided into flood prevention input control key points, flood prevention transitivity key points, flood prevention decision key points, and flood prevention control key points. The flood prevention input control key points can be points for entering external rainwater in the station. Exemplarily, they can be corresponding keys such as entrances and exits, construction interfaces, etc. The flood prevention transitivity key points can be points that may generate various risks in the process of external rainwater flowing from the flood prevention input control key points. The flood prevention decision key points can be key points that will generate risks once flooded, such as power supply, communication, environmental control, elevators, and ticketing systems, etc. The flood prevention control key points can be points where various drainage equipment is located, which can control the degree of flooding to a certain extent.

[0067] Further, the obtaining of the digital twin data of the rail transit station can include: obtaining virtual precipitation data predicted by a meteorological big model; determining a corresponding flood prevention management strategy according to the virtual precipitation data, and updating the digital twin data based on the flood prevention management strategy. The determining of the corresponding flood prevention management strategy according to the virtual precipitation data and the updating of the digital twin data based on the flood prevention management strategy can include: adjusting the height of the water retaining facility and the drainage power of the peripheral drainage facility according to the precipitation data and a water retaining facility and peripheral drainage facility operation table; and updating the digital twin data according to the adjusted water retaining facility and peripheral drainage facility. Generally, the station can use the water retaining facility and the peripheral drainage facility according to the corresponding waterlogging situation of the predicted rainfall intensity, so as to reduce the inflow waterlogging of the station. In this case, the inflow rainwater of the station will also change, and therefore, the digital twin data can be updated according to the adjusted water retaining facility and peripheral drainage facility, so as to better simulate the flooding situation.

[0068] Further, since the power supply, communication, environment control, elevator and ticketing systems in the station correspond to a plurality of devices respectively, if each device is set as a key point, the data amount is increased, and then the calculation amount is increased. Therefore, in the embodiment, representative devices can be selected as the flood prevention key points from the above-mentioned devices.

[0069] For example, the method can include: determining each flood prevention attribute category according to the flood prevention requirement, and selecting a plurality of flood prevention points according to the flood prevention transitivity key point, the flood prevention decisiveness key point and the flood prevention control key point according to the digital twin data; determining the redundancy index and the hot standby index of each flood prevention point according to the plurality of flood prevention points; calculating the vulnerability of each flood prevention point according to the redundancy index and the hot standby index, and calculating the stability of each flood prevention point; and selecting the flood prevention key point from the flood prevention points according to the vulnerability and the stability. By using the above-mentioned method, the station can be set in the lowest reliability state, and the devices capable of normal operation can be determined as the flood prevention key points.

[0070] In step 120, when the total number of the flood prevention input control key points, the flood prevention transitivity key points, the flood prevention decisiveness key points and the flood prevention control key points is greater than a first threshold value, a dimensionality reduction improved feedforward neural network is established, which takes the data of the flood prevention input control key points and the flood prevention control key points as inputs, and outputs the predicted data of the flood prevention transitivity key points and the flood prevention decisiveness key points.

[0071] In the embodiment, the first threshold value can be set according to the performance of the server and the processing performance of the model. The total number of the anti-flooding input control key points, the anti-flooding transitivity key points, the anti-flooding decisive key points and the anti-flooding control key points is calculated. When the total number exceeds the first threshold value, it can be determined that the current site is a complex site, and the data processing amount is large, so a more complex model needs to be used to obtain better processing results.

[0072] Optionally, the model can use a dimension reduction improved feedforward neural network. Through dimension reduction processing, complex data can be processed to reduce the dimension of irrelevant data. The operation amount is compressed, and at the same time, the rainwater flow is a linear relationship. By using the linear processing advantage of the feedforward neural network, better processing results can be obtained for complex sites. The dimension reduction improved feedforward neural network takes the data of the anti-flooding input control key points and the anti-flooding control key points as input, and outputs the predicted data of the anti-flooding transitivity key points and the anti-flooding decisive key points.

[0073] In step 130, when the number of one-fourth of the anti-flooding input control key points, the sum of the number of the anti-flooding transitivity key points, the anti-flooding decisive key points and the anti-flooding control key points is less than the first threshold value, a GNN model is established. The GNN model establishes the connection relationship between the updated hierarchical nodes according to the hierarchical relationship of the clustering nodes of the anti-flooding input control key points, the anti-flooding transitivity key points, the anti-flooding decisive key points and the anti-flooding control key points, and constantly updates each hierarchical node according to the data of the anti-flooding input control key points, and obtains the predicted classification of the anti-flooding transitivity key points and the anti-flooding decisive key points according to the updated hierarchical nodes.

[0074] In the embodiment, if the number of key points is small, it can be considered as a normal site. Based on this situation, the GNN model can be selected. The graph neural network (GNN) is a deep learning model based on graph structure data, which realizes feature learning and reasoning by aggregating nodes, edges and their neighborhood information. Information transmission and feature updating are performed through graph structure (topological relationship of nodes and edges). It is particularly suitable for the scene of water flow.

[0075] Although the GNN model performs excellently in the scene provided in the embodiment, too many node numbers will seriously affect the calculation efficiency and require more computing resources. Therefore, in the embodiment, the number of nodes, i.e. the number of key points, needs to be considered first.

[0076] Exemplarily, the sum of the number of the quarter of the number of the flood control input control key points, the flood control transitive key points, the flood control decisive key points and the flood control control key points can be calculated, and then the sum is compared with the first threshold value, and when the sum is less than the first threshold value, it can be determined that the number of the key points is suitable for using the GNN model. The reason for considering the quarter of the number of the flood control input control key points is that the flood control input control key points can be compressed by aggregation, but too much aggregation may affect the prediction accuracy. Therefore, in the embodiment, a hierarchical clustering method can be used to perform two times of clustering to obtain the clustered flood control input control key points as the node input into the GNN model. If the number of the flood control input control key points is odd, the number of the nodes output by the hierarchical clustering can be used as a constraint to ensure that the number of the clustered nodes is the quarter of the number of the flood control input control key points. The GNN model establishes the connection relationship between the hierarchical nodes according to the hierarchical relationship of the clustered nodes of the flood control input control key points, the flood control transitive key points, the flood control decisive key points and the flood control control key points, and constantly updates each hierarchical node according to the data of the flood control input control key points, and obtains the prediction classification of the flood control transitive key points and the flood control decisive key points according to the updated hierarchical nodes.

[0077] In the embodiment, the GNN model comprises: a layer combination, which is used to establish an input node graph according to the position relationship of the flood control input control key points, and generate flood control input aggregation nodes by using an aggregation function; a graph establishment layer, which is used to generate a directed graph by using the position and water flow sequence relationship of the flood control input aggregation nodes, the flood control transitive key points, the flood control decisive key points and the flood control control key points, wherein the flood control control key points are negative; and a fully connected layer, which is used to output the flood control possibility probability of the flood control transitive key points and the flood control decisive key points by using the attributes of the flood control transitive key points and the flood control decisive key points output by the graph establishment layer.

[0078] Exemplarily, various data of the key points and the aggregated key points, such as the position, the predicted precipitation, the spatial position of the site, etc., can be used as the attributes of the nodes, the relationship between the nodes and the edges in the graph is iterated, and then the attributes of the flood control transitive key points and the flood control decisive key points are obtained by using the fully connected layer, and the flood control possibility probability of the flood control transitive key points and the flood control decisive key points is output based on the attributes. Since the flood control control key points can be various drainage devices, and the essence is the reverse flow process of the water flow, the flood control control key points can be set as negative.

[0079] The embodiment obtains digital twin data of a rail transit station, determines the number of flood prevention key points and corresponding flood prevention attribute categories according to the digital twin data, the flood prevention attribute categories include flood prevention input control key points, flood prevention transfer key points, flood prevention decision key points and flood prevention control key points, when the total number of flood prevention input control key points, flood prevention transfer key points, flood prevention decision key points and flood prevention control key points is greater than a first threshold, a dimensionality reduction improved feedforward neural network is established, the dimensionality reduction improved feedforward neural network takes data of the flood prevention input control key points and the flood prevention control key points as input, and outputs predicted data of the flood prevention transfer key points and the flood prevention decision key points, when the number of one fourth of the flood prevention input control key points, the number of the flood prevention transfer key points, the number of the flood prevention decision key points and the number of the flood prevention control key points is less than the first threshold, a GNN model is established, the GNN model establishes a connection relationship between updated hierarchical nodes according to a hierarchical relationship of clustered nodes of the flood prevention input control key points, the flood prevention transfer key points, the flood prevention decision key points and the flood prevention control key points, constantly updates each hierarchical node according to data of the flood prevention input control key points, and obtains predicted classification of the flood prevention transfer key points and the flood prevention decision key points according to the updated hierarchical nodes. By using the digital twin data of the rail transit station, various key points required by the model and corresponding types are determined, and the number of key points of each type is used, so that the accuracy of station flood prediction is further improved, and accurate reference basis is provided for planning and improvement of flood prevention measures.

[0080] Embodiment two

[0081] Figure 2is a flowchart of a model establishment method for rail transit flood prevention provided in Embodiment Two of the present application, and the present embodiment is optimized on the basis of the above-mentioned embodiment. The dimension reduction improved feedforward neural network is specifically optimized as follows: an input layer, which is used to convert the locations of flood prevention input control key points and flood prevention control key points, predicted precipitation, the time of predicted precipitation, spatial position data, redundancy index and hot standby index into flood prevention point multidimensional data; a dimension reduction layer, which is used to perform dimension reduction processing on the multidimensional data to obtain dimension reduction data of the multidimensional data; a plurality of hidden layers, which are used to extract abstract features through linear transformation; two output layers, wherein one output layer is used to output a flood prevention prediction result according to the abstract features using an activation function, and the other output layer is used to output a flow rate prediction result of a flood prevention transfer control point according to the abstract features output by part of the hidden layers; and an adjustment unit, which is used to adjust the bias matrix corresponding to the part of the hidden layers according to the result calculated by the turbulent flow model. Correspondingly, the selecting a plurality of flood prevention points according to each category from the digital twin data is specifically optimized as follows: dividing the stairs in the station according to the grid unit according to the station digital twin data; selecting grid units at the bottom of the stairs, the side wall of the intermediate layer and the intermediate layer position as flood prevention transfer key points; and calculating the water flow rate of each flood prevention transfer key point according to the data of the flood prevention input control key points using the turbulent flow model.

[0082] Referring to Figure 2 , the model establishment method for rail transit flood prevention comprises the following steps.

[0083] In step 210, the digital twin data of the rail transit station is obtained, and the stairs in the station are divided according to the grid unit according to the station digital twin data; grid units at the bottom of the stairs, the side wall of the intermediate layer and the intermediate layer position are selected as flood prevention transfer key points.

[0084] In the rail transit station, the stairs are the main channel for people to go up and down. The stairs have a certain height difference, and if the external rainwater enters the entrance and exit, it will quickly flow into the lower station through the stairs. If the flow rate is too fast, it will cause a greater risk. Therefore, in the present embodiment, the stairs in the station can be divided according to the grid unit according to the station digital twin data; grid units at the bottom of the stairs, the side wall of the intermediate layer and the intermediate layer position are selected as flood prevention transfer key points. The flow rate at the bottom of the stairs is the fastest, and the side wall of the intermediate layer will produce a large flow rate impact due to the impact of the side wall. The intermediate layer needs the flow rate of the position of the intermediate layer because the intermediate layer is provided with a stair buffer flat layer. Therefore, the intermediate layer position is divided into a plurality of units, and part of the points are selected as flood prevention transfer key points.

[0085] In step 220, the water flow rate of each flood prevention transfer key point is calculated according to the data of the flood prevention input control key points using the turbulent flow model.

[0086] In the embodiment, the turbulence model can adopt a k-ε model, and the water flow velocity of each flood control transfer key point is calculated by using the input flood control input control key point, and the rainwater data input by the key point corresponding to the stairs.

[0087] In step 230, when the total number of flood control input control key points, flood control transfer key points, flood control decision key points and flood control control key points is greater than the first threshold value, a dimension reduction improved feedforward neural network is established, the dimension reduction improved feedforward neural network comprising: an input layer for converting the location of the flood control input control key point and the flood control control key point, the predicted precipitation, the time of the predicted precipitation, the spatial position data, the redundancy index and the hot standby index into flood control point multidimensional data; a dimension reduction layer for dimension reduction processing of the multidimensional data to obtain dimension reduction data of the multidimensional data; a plurality of hidden layers for extracting abstract features by linear transformation; two output layers, wherein one output layer is used to output flood control prediction results according to the abstract features by using an activation function, and the other output layer is used to output flow velocity prediction results of the flood control transfer control point according to the abstract features output by the part of the hidden layers; an adjustment unit for adjusting the bias matrix corresponding to the part of the hidden layers according to the result calculated by the turbulence model.

[0088] In the embodiment, the dimension reduction improved feedforward neural network can convert various parameters such as the location of the flood control input control key point and the flood control control key point, the predicted precipitation, the time of the predicted precipitation, the spatial position data, the redundancy index and the hot standby index into flood control point multidimensional data by using the input layer.

[0089] Because the input data has many attribute dimensions, it is not convenient to calculate in the later stage, and because the dimension is high, it is necessary to set multiple layers to extract features, which not only requires a large amount of computing power, but also is prone to overfitting. Therefore, in the embodiment, a dimension reduction layer can be added, which can perform dimension reduction processing on the multidimensional data to obtain dimension reduction data of the multidimensional data. For example, factor analysis can be used to perform dimension reduction processing on the multidimensional data.

[0090] The data processed by dimension reduction can be input into the hidden layer, the abstract features are extracted by linear transformation, and then the flood prevention prediction result is output by using the activation function. In the embodiment, another output layer can also be provided, which uses an activation function to output the flow rate prediction result of the flood prevention transfer control point for the abstract features output by part of the hidden layer. In the above steps, the flow velocity of the flood transfer key point can be calculated by using the turbulent flow model. The abstract features obtained by the hidden layer are corrected by using the comparison result between the two. Since the hidden layer uses linear transformation, each hidden layer has a bias matrix, and the bias matrix can be optimized and adjusted by using the different results obtained by the two, so as to further improve the prediction result of the dimension reduction improved feedforward neural network.

[0091] In step 240, the dimension reduction improved feedforward neural network takes the data of the flood prevention input control key point and the flood prevention control key point as input, and outputs the prediction data of the flood prevention transfer key point and the flood prevention decisive key point.

[0092] In step 250, when the number of the quarter of the flood prevention input control key point, the number of the flood prevention transfer key point, the number of the flood prevention decisive key point and the flood prevention control key point is less than the first threshold value, a GNN model is established. The GNN model establishes the connection relationship between the updated hierarchical nodes according to the hierarchical relationship of the cluster nodes of the flood prevention input control key point, the flood prevention transfer key point, the flood prevention decisive key point and the flood prevention control key point, and constantly updates each hierarchical node according to the data of the flood prevention input control key point, and obtains the prediction classification of the flood prevention transfer key point and the flood prevention decisive key point according to the updated hierarchical nodes.

[0093] The embodiment improves the feedforward neural network by dimension reduction, and specifically optimizes the input layer for converting the locations of the flood prevention input control key points and the flood prevention control key points, the predicted precipitation, the time of the predicted precipitation, the spatial position data, the redundancy index and the hot standby index into flood prevention point multidimensional data; the dimension reduction layer for performing dimension reduction processing on the multidimensional data to obtain dimension reduction data of the multidimensional data; the plurality of hidden layers for extracting abstract features through linear transformation; the two output layers, wherein one output layer is used to output the flood prevention prediction result according to the abstract features using the activation function, and the other output layer is used to output the flow rate prediction result of the flood prevention transfer control point according to the abstract features output by the partial hidden layer; and the adjusting unit is used to adjust the bias matrix corresponding to the partial hidden layer according to the result calculated by the turbulence model. Correspondingly, the flood prevention points are selected according to each category based on the digital twin data, and specifically optimized as follows: the site is divided into grid units according to the site digital twin data; the grid units at the bottom of the stairs, the side wall of the middle layer and the middle layer position are selected as the flood prevention transfer key points; and the flow velocity of each flood prevention transfer key point is calculated by the turbulence model according to the data of the flood prevention input control key point. The dimension reduction feature of the dimension reduction improved feedforward neural network can effectively compress the massive data, effectively reduce the complexity of the model, and the bias matrix of the hidden layer can be optimized and adjusted by the flood prevention transfer control point flow rate calculation result calculated by the turbulence model, thereby reducing the problem of calculation accuracy reduction caused by dimension reduction.

[0094] Embodiment three

[0095] Figure 3 is a structural schematic diagram of a model establishment device for rail transit flood prevention provided by the third embodiment of the present application, referring to Figure 3 , the model establishment device for rail transit flood prevention comprises:

[0096] The acquisition module 310 is configured to acquire digital twin data of a rail transit station, determine the number of flood prevention key points and corresponding flood prevention attribute categories according to the digital twin data, and the flood prevention attribute categories include flood prevention input control key points, flood prevention transfer key points, flood prevention decision key points and flood prevention control key points.

[0097] The first establishment module 320 is configured to establish a dimension reduction improved feedforward neural network when the total number of the flood prevention input control key points, the flood prevention transfer key points, the flood prevention decision key points and the flood prevention control key points is greater than a first threshold value, and the dimension reduction improved feedforward neural network takes the data of the flood prevention input control key points and the flood prevention control key points as input and outputs the prediction data of the flood prevention transfer key points and the flood prevention decision key points.

[0098] The second establishing module 330 is used to establish a GNN model when the sum of the number of one-quarter of the flood prevention input control key points, the number of flood prevention transitive key points, the number of flood prevention decisive key points and the number of flood prevention control key points is less than a first threshold. The GNN model establishes and updates the connection relationship between hierarchical nodes according to the hierarchical relationship of the cluster nodes, flood prevention transitive key points, flood prevention decisive key points and flood prevention control key points of the flood prevention input control key points, and continuously updates each hierarchical node according to the data of the flood prevention input control key points, and obtains the predicted classification of the flood prevention transitive key points and the flood prevention decisive key points according to the updated hierarchical nodes.

[0099] The model building device for rail transit flood prevention provided in this embodiment obtains digital twin data of rail transit stations, determines the number of flood prevention key points and corresponding flood prevention attribute categories based on the digital twin data, and the flood prevention attribute categories include flood prevention input control key points, flood prevention transitive key points, flood prevention decisive key points and flood prevention control key points; when the total number of flood prevention input control key points, flood prevention transitive key points, flood prevention decisive key points and flood prevention control key points is greater than a first threshold, a dimensionality reduction improved feedforward neural network is established, and the dimensionality reduction improved feedforward neural network takes the data of the flood prevention input control key points and the flood prevention control key points as input, and outputs the flood prevention attribute categories. Prediction data for flooding transitive key points and flood prevention decisive key points is collected. When the sum of the number of one-quarter of the flood prevention input control key points, the number of flood prevention transitive key points, the number of flood prevention decisive key points, and the number of flood prevention control key points is less than a first threshold, a GNN model is established. The GNN model updates the connection relationships between hierarchical nodes based on the hierarchical relationship of the cluster nodes, flood prevention transitive key points, flood prevention decisive key points, and flood prevention control key points of the flood prevention input control key points. Each hierarchical node is continuously updated based on the data of the flood prevention input control key points, and the predicted classification of flood prevention transitive key points and flood prevention decisive key points is obtained based on the updated hierarchical nodes. By utilizing the digital twin data of rail transit stations, the various key points and corresponding types required for the model are determined. The number of key points of each type is used to determine the complexity of the current station, and an appropriate model is selected based on the complexity. Using an appropriate model not only reduces the computational workload but also avoids overfitting caused by overly complex models, further improving the accuracy of station flooding predictions and providing an accurate reference for planning and improving flood prevention measures.

[0100] Based on the above embodiments, the GNN model includes:

[0101] An aggregation layer, for establishing an input node graph according to the positional relationship of the flood prevention input control key points, and generating flood prevention input aggregation nodes using an aggregation function;

[0102] a graph establishing layer, configured to generate a directed graph by using the positions of the flood control input aggregation node, the flood control transitive key point, the flood control decisive key point and the flood control control key point and the water flow sequence relationship, wherein the flood control control key point is negative;

[0103] a full connection layer, configured to output the flood control possibility probability of the flood control transitive key point and the flood control decisive key point by using the attributes of the flood control transitive key point and the flood control decisive key point output by the graph establishing layer.

[0104] On the basis of the above embodiments, the acquisition module comprises:

[0105] an acquisition unit, configured to acquire virtual precipitation data predicted by a meteorological big model;

[0106] an updating unit, configured to determine a corresponding flood control management strategy according to the virtual precipitation data, and update the digital twin data based on the flood control management strategy.

[0107] On the basis of the above embodiments, the updating unit comprises:

[0108] an adjusting subunit, configured to adjust the height of the water retaining facility and the drainage power of the peripheral drainage facility according to the precipitation data and a water retaining facility and peripheral drainage facility operation table;

[0109] an updating subunit, configured to update the digital twin data according to the adjusted water retaining facility and peripheral drainage facility.

[0110] On the basis of the above embodiments, the acquisition module comprises:

[0111] a selection unit, configured to determine each flood control attribute category according to a flood control requirement, and select a plurality of flood control points according to the flood control transitive key point, the flood control decisive key point and the flood control control key point according to the digital twin data;

[0112] a determination unit, configured to determine the redundancy index and the hot standby index of each flood control point according to the plurality of flood control points;

[0113] a calculation unit, configured to calculate the vulnerability of each flood control point according to the redundancy index and the hot standby index, and calculate the stability of each flood control point;

[0114] a flood control key point selection unit, configured to select a flood control key point from the flood control points according to the vulnerability and the stability.

[0115] On the basis of the above embodiments, the selection unit comprises:

[0116] a division subunit, configured to divide the stairs in the site according to the site digital twin data according to a grid unit;

[0117] The selecting unit is used for selecting a grid unit as a flood prevention transmission key point from the bottom of the stairs, the side wall of the intermediate layer and the intermediate layer position;

[0118] The speed calculating unit is used for calculating the water flow speed of each flood prevention transmission key point according to the data of the flood prevention input control key point by using a turbulence model.

[0119] On the basis of the above-mentioned embodiments, the dimension reduction improved feedforward neural network comprises:

[0120] The input layer is used for converting the location of the flood prevention input control key point and the flood prevention control key point, the predicted precipitation, the time of the predicted precipitation, the spatial position data, the redundancy index and the hot standby index into flood prevention point multidimensional data;

[0121] The dimension reduction layer is used for performing dimension reduction processing on the multidimensional data to obtain dimension reduction data of the multidimensional data;

[0122] The plurality of hidden layers are used for extracting abstract features through linear transformation;

[0123] The two output layers, wherein one output layer is used for outputting a flood prevention prediction result according to the abstract features by using an activation function, and the other output layer is used for outputting a flow speed prediction result of a flood prevention transmission control point according to the abstract features output by the partial hidden layer;

[0124] The adjusting unit is used for adjusting the bias matrix corresponding to the partial hidden layer according to the result calculated by the turbulence model.

[0125] The model establishment device for rail transit flood prevention provided by the embodiment of the application can execute the model establishment method for rail transit flood prevention provided by any embodiment of the application, has the function modules and beneficial effects corresponding to the execution method.

[0126] Embodiment four

[0127] Figure 4 A structural schematic diagram of a server provided for the fourth embodiment of the application. Figure 4 A block diagram of an exemplary server 12 suitable for use in implementing embodiments of the application is shown. Figure 4 The server 12 shown is merely one example and should not be taken as limiting the scope of the functionality or use of the fourth embodiment of the application.

[0128] As shown in Figure 4 The server 12 is shown as a general-purpose computing device. The components of the server 12 can include, but are not limited to, one or more processors or processing units 16, a system memory 28, and a bus 18 that couples various system components including the system memory 28 and the processing unit 16.

[0129] Bus 18 represents one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processor, or a local bus using any of a variety of bus architectures. Examples of these architectures include, but are not limited to, an Industry Standard Architecture (ISA) bus, a Micro Channel Architecture (MAC) bus, an Enhanced ISA bus, a Video Electronics Standards Association (VESA) local bus, and a Peripheral Component Interconnect (PCI) bus.

[0130] The server 12 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by the server 12, including volatile and non-volatile media, removable and non-removable media.

[0131] System memory 28 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) 30 and / or cache memory 32. Server 12 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, storage system 34 may be configured to read and write non-removable, non-volatile magnetic media ( Figure 4 Not shown, usually called a "hard drive"). Although Figure 4 Although not shown, a magnetic disk drive for reading and writing to a removable non-volatile magnetic disk (e.g., a "floppy disk"), as well as an optical disk drive for reading and writing to a removable non-volatile optical disk (e.g., a CD-ROM, DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to bus 18 via one or more data media interfaces. Memory 28 may include at least one program product having a set (e.g., at least one) of program modules configured to perform the functions of various embodiments of the present invention.

[0132] A program / utility 40 having a set (at least one) of program modules 42 may be stored, for example, in memory 28. Such program modules 42 include, but are not limited to, an operating system, one or more application programs, other program modules, and program data, each of which, or some combination thereof, may include an implementation of a network environment. Program modules 42 generally implement the functions and / or methodologies of the embodiments described herein.

[0133] The server 12 may also communicate with one or more external devices 14 (e.g., a keyboard, pointing device, display 24, etc.), one or more devices that enable a user to interact with the server 12, and / or any device that enables the server 12 to communicate with one or more other computing devices (e.g., a network card, a modem, etc.). This communication may occur via an input / output (I / O) interface 22. Furthermore, the server 12 may communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and / or a public network such as the Internet) via a network adapter 20. As shown, the network adapter 20 communicates with the other modules of the server 12 via the bus 18. It should be understood that, although not shown, other hardware and / or software modules may be used in conjunction with the server 12, including but not limited to microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0134] The processing unit 16 executes various functional applications and data processing by running programs stored in the system memory 28 , such as implementing the model building method for rail transit flood prevention provided in an embodiment of the present invention.

[0135] Example 5

[0136] The fifth embodiment of the present invention further provides a storage medium comprising computer-executable instructions, which, when executed by a computer processor, are used to execute any of the model building methods for rail transit flooding prevention provided in the above embodiments.

[0137] The computer storage medium of the embodiments of the present invention may adopt any combination of one or more computer-readable media. The computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples (a non-exhaustive list) of computer-readable storage media include: an electrical connection having one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In this document, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0138] A computer readable signal medium can include a propagated data signal with computer executable code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated signal can take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium can be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate or transport programming code.

[0139] Program code embodied on a computer readable medium can be transmitted using any appropriate medium, including but not limited to wireless, wire line, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

[0140] Computer program code for carrying out operations for aspects of the present application can be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C++ or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computer (for example, through the Internet using an Internet Service Provider).

[0141] Note that the foregoing examples have been provided merely for the purpose of explanation and are in no way to be construed as limiting of the present application. While the application has been described with reference to preferred embodiments and specific mountings, it will be apparent to one skilled in the art that various changes and modifications can be made therein without departing from the spirit and scope of the application. Accordingly, the scope of the application is to be limited only by the appended claims.

Claims

1. A model building method for rail transit flood prevention, characterized in that: include: Obtain digital twin data of rail transit stations, and determine the number of flood prevention key points and corresponding flood prevention attribute categories based on the digital twin data. The flood prevention attribute categories include flood prevention input control key points, flood prevention transitive key points, flood prevention decisive key points, and flood prevention control key points. When the total number of flood prevention input control key points, flood prevention transitive key points, flood prevention decisive key points and flood prevention control key points is greater than a first threshold, establishing a dimensionality reduction improved feedforward neural network, wherein the dimensionality reduction improved feedforward neural network takes the data of the flood prevention input control key points and the flood prevention control key points as input, and outputs predicted data of the flood prevention transitive key points and the flood prevention decisive key points; When the sum of one-quarter of the number of the flood prevention input control key points, the number of the flood prevention transitive key points, the flood prevention decisive key points, and the flood prevention control key points is less than a first threshold, a GNN model is established, wherein the GNN model establishes and updates the connection relationship between the hierarchical nodes according to the hierarchical relationship of the cluster nodes of the flood prevention input control key points, the flood prevention transitive key points, the flood prevention decisive key points, and the flood prevention control key points, and continuously updates each hierarchical node according to the data of the flood prevention input control key points, and obtains the prediction classification of the flood prevention transitive key points and the flood prevention decisive key points according to the updated hierarchical nodes; The GNN model includes: An aggregation layer, for establishing an input node graph according to the positional relationship of the flood prevention input control key points, and generating flood prevention input aggregation nodes using an aggregation function; A graph building layer is used to generate a directed graph using the positions of the flood prevention input aggregation nodes, flood prevention transitive key points, flood prevention decisive key points, and flood prevention control key points and the water flow sequence relationship, wherein the flood prevention control key points are negative; The fully connected layer is used to use the attributes of the flood prevention transitive key points and the flood prevention decisive key points output by the graph building layer to output the flood prevention possibility probability of the flood prevention transitive key points and the flood prevention decisive key points; Determining the number of flood prevention key points and corresponding flood prevention attribute categories based on digital twin data includes: Determine each flood prevention attribute category according to flood prevention requirements, and select several flood prevention points based on digital twin data according to flood prevention transfer key points, flood prevention decisive key points, and flood prevention control key points; Determine the redundancy index and hot standby index of each flood prevention point based on a number of flood prevention points; Calculating the vulnerability of each flood prevention point according to the redundancy index and the hot standby index, and calculating the stability of each flood prevention point; Selecting key flood control points from among the flood control points based on the fragility and stability; Based on the digital twin data, several flood prevention points were selected for each category, including: Divide the stairs within the site into grid units based on the site digital twin data; Grid cells were selected from the bottom of the stairs, the side walls of the middle layer, and the middle layer as the key points of flood prevention and transferability. The flow velocity at the bottom of the stairs is the fastest, and the side walls of the middle layer produce a high flow velocity impact due to the side wall impact. Therefore, a stair buffer leveling layer is set at the middle layer. The water velocity of each flood control transfer key point is calculated based on the data of flood control input control key points using the turbulence model; The dimension reduction improved feedforward neural network includes: The input layer is used to convert flood control input control key points and locations of flood control key points, predicted precipitation, time of predicted precipitation, spatial location data, redundancy index and thermal readiness index into multidimensional data of flood control points; A dimensionality reduction layer is used to perform dimensionality reduction processing on the multidimensional data of the flood prevention points to obtain dimensionality-reduced data of the multidimensional data of the flood prevention points; Multiple hidden layers for extracting abstract features through linear transformation; Two output layers, one of which is used to output flood prevention prediction results using an activation function based on abstract features, and the other is used to output the flow velocity prediction results of flood prevention transfer control points based on the abstract features output by some hidden layers; The adjustment unit is used to adjust the bias matrix corresponding to the part of the hidden layer according to the comparison result of the result calculated by the turbulence model and the flow velocity prediction of the flood prevention transfer control point output by another output layer, so as to improve the prediction result of the dimensionality reduction improved feedforward neural network.

2. The model building method for rail transit flood prevention according to claim 1, characterized in that: The obtaining of digital twin data of rail transit stations includes: Obtain virtual precipitation data predicted by large meteorological models; Determine the corresponding flood prevention management strategy based on the virtual precipitation data, and update the digital twin data based on the flood prevention management strategy.

3. The model building method for rail transit flood prevention according to claim 2, characterized in that: Determining a corresponding flood prevention management strategy based on virtual precipitation data, and updating digital twin data based on the flood prevention management strategy, includes: Adjust the height of water retaining facilities and the drainage capacity of peripheral drainage facilities according to precipitation data and the comparison table of water retaining facilities and peripheral drainage facilities; Update the digital twin data based on the adjusted water retaining facilities and peripheral drainage facilities.

4. A model building device for rail transit flood prevention, characterized in that: include: An acquisition module is used to obtain digital twin data of rail transit stations and determine the number of flood prevention key points and corresponding flood prevention attribute categories based on the digital twin data. The flood prevention attribute categories include flood prevention input control key points, flood prevention transitive key points, flood prevention decisive key points, and flood prevention control key points. A first establishing module is configured to establish a dimensionality reduction improved feedforward neural network when the total number of flood prevention input control key points, flood prevention transitive key points, flood prevention decisive key points, and flood prevention control key points is greater than a first threshold value. The dimensionality reduction improved feedforward neural network takes the data of the flood prevention input control key points and the flood prevention control key points as input and outputs predicted data of the flood prevention transitive key points and the flood prevention decisive key points; The second establishment module is used to establish a GNN model when the sum of the number of one-quarter of the flood prevention input control key points, the number of flood prevention transitive key points, the number of flood prevention decisive key points and the number of flood prevention control key points is less than a first threshold value. The GNN model establishes and updates the connection relationship between hierarchical nodes according to the hierarchical relationship of the cluster nodes, flood prevention transitive key points, flood prevention decisive key points and flood prevention control key points of the flood prevention input control key points, and continuously updates each hierarchical node according to the data of the flood prevention input control key points, and obtains the prediction classification of the flood prevention transitive key points and the flood prevention decisive key points according to the updated hierarchical nodes; The GNN model includes: An aggregation layer, for establishing an input node graph according to the positional relationship of the flood prevention input control key points, and generating flood prevention input aggregation nodes using an aggregation function; A graph building layer is used to generate a directed graph using the positions of the flood prevention input aggregation nodes, flood prevention transitive key points, flood prevention decisive key points, and flood prevention control key points and the water flow sequence relationship, wherein the flood prevention control key points are negative; The fully connected layer is used to use the attributes of the flood prevention transitive key points and the flood prevention decisive key points output by the graph building layer to output the flood prevention possibility probability of the flood prevention transitive key points and the flood prevention decisive key points; The acquisition module includes: The selection unit is used to determine each flood prevention attribute category according to the flood prevention requirements, and select several flood prevention points according to the flood prevention transfer key points, flood prevention decisive key points and flood prevention control key points based on the digital twin data; A determination unit, used for determining a redundancy index and a hot standby index of each flood prevention point based on a plurality of flood prevention points; a calculation unit, configured to calculate the vulnerability of each flood prevention point according to the redundancy index and the hot standby index, and calculate the stability of each flood prevention point; a stability calculation unit, configured to calculate the vulnerability of each flood prevention point according to the redundancy index and the hot standby index, and calculate the stability of each flood prevention point; a flood prevention key point selection unit, configured to select a flood prevention key point from the flood prevention points according to the vulnerability and stability; The selection unit includes: Divide sub-units, which are used to divide the stairs in the site into grid units based on the site digital twin data; The selection unit is used to select grid units from the bottom of the stairs, the side walls of the middle layer, and the middle layer as the key points of flood prevention transferability. The flow velocity at the bottom of the stairs is the fastest, and the side walls of the middle layer produce a high flow velocity impact due to the side wall impact. The middle layer is equipped with a stair buffer leveling layer. A velocity calculation unit is used to calculate the water flow velocity of each flood control transfer key point based on the data of the flood control input control key point using a turbulence model; Dimensionality reduction to improve feedforward neural networks, including: The input layer is used to convert flood control input control key points and locations of flood control key points, predicted precipitation, time of predicted precipitation, spatial location data, redundancy index and thermal readiness index into multidimensional data of flood control points; A dimensionality reduction layer is used to perform dimensionality reduction processing on the multidimensional data of the flood prevention points to obtain dimensionality-reduced data of the multidimensional data of the flood prevention points; Multiple hidden layers for extracting abstract features through linear transformation; Two output layers, one of which is used to output flood prevention prediction results using an activation function based on abstract features, and the other is used to output the flow velocity prediction results of flood prevention transfer control points based on the abstract features output by some hidden layers; The adjustment unit is used to adjust the bias matrix corresponding to the part of the hidden layer according to the comparison result of the result calculated by the turbulence model and the flow velocity prediction of the flood prevention transfer control point output by another output layer, so as to improve the prediction result of the dimensionality reduction improved feedforward neural network.

5. A server, characterized in that: The server includes: one or more processors; a storage device for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the model building method for rail transit flooding prevention as described in any one of claims 1 to 3.

6. A storage medium containing computer-executable instructions, characterized in that: When executed by a computer processor, the computer executable instructions are used to execute the model building method for rail transit flooding prevention according to any one of claims 1 to 3.

Images