Method, device and equipment for positioning rescue position of vehicle on expressway and medium

Abstract

The invention provides an expressway vehicle rescue position positioning method and device, equipment and a medium, and the method comprises the steps: querying a license recognition database based on the accident information of an accident vehicle, and determining the information of an ETC door frame where the accident vehicle passes recently; inputting the accident road section and a portal road section in the ETC portal information into a pre-trained graph convolutional network model, performing feature processing on the accident road section and the portal road section based on a traffic knowledge graph structure, and predicting a current vehicle speed of the accident road section and a current vehicle speed of the portal road section; and determining an accident point position of the accident vehicle based on the current vehicle speed of the accident road section, the current vehicle speed of the portal road section, the ETC portal information and the alarm receiving time information. The graph neural network model is adopted to perform feature learning on the accident road section and the portal road section, and the real-time vehicle speed of the current accident related road section and the real-time vehicle speed of the portal road section are predicted, so that the accuracy of accident point location estimation is improved.

Description

Technical Field

[0001] This application relates to the technical field of highway management, and particularly to a positioning method, device, equipment and medium for the location of highway vehicle rescue. Background Art

[0002] Highways have the advantages of large capacity, high operating speed, low transportation cost and significant social benefits. However, once a traffic accident occurs on a highway, it is often shocking, and the accident will occupy the driving lane after it occurs, which is extremely likely to cause road congestion and threaten the driving safety on the highway. Therefore, rapid and accurate accident positioning, timely and effective accident warning, efficient accident handling and rescue are particularly important for ensuring the safe and stable operation of highways, improving the management and service level of highways, and enhancing the ability to handle emergencies. In the prior art, there are some disadvantages in the auxiliary positioning and rescue method for highway traffic accident vehicles based on ETC data: (1) its applicable scope is limited, only for the positioning of ETC accident vehicles, and non-ETC vehicles cannot be covered, resulting in limited rescue scenarios; (2) relying on single ETC gantry data, the positioning accuracy and reliability are relatively low. Therefore, how to improve the accuracy of highway vehicle rescue location positioning has become a technical problem that cannot be underestimated. Summary of the Invention

[0003] In view of this, the purpose of this application is to provide a positioning method, device, equipment and medium for the location of highway vehicle rescue, using a graph neural network model to perform feature learning on the accident section and the gantry section, predicting the real-time vehicle speed of the current accident-related section and the real-time vehicle speed of the gantry section, and thus improving the accuracy of accident point estimation.

[0004] The embodiment of this application provides a positioning method for the location of highway vehicle rescue. The positioning method includes:

[0005] Query the license plate recognition database based on the accident information of the accident vehicle to determine the ETC gantry information that the accident vehicle passed by most recently; wherein, the accident information includes the alarm receiving time information, license plate information and accident section.

[0006] Input the accident section and the gantry section in the ETC gantry information into a pre-trained graph convolutional network model, perform feature processing on the accident section and the gantry section based on the traffic knowledge graph structure, and predict the current vehicle speed of the accident section and the current vehicle speed of the gantry section.

[0007] Determine the accident point location of the accident vehicle based on the current vehicle speed of the accident section, the current vehicle speed of the gantry section, the ETC gantry information and the alarm receiving time information.

[0008] In a possible implementation manner, the traffic knowledge graph structure is determined through the following steps:

[0009] Obtain the attribute information of each highway section and the attribute information of the vehicles traveling on each highway section;

[0010] Take each highway section and each traveling vehicle as nodes of the traffic knowledge graph structure, and take the driving relationship between each traveling vehicle and each highway section and the connectivity relationship between each highway section and other highway sections as edges of the traffic knowledge graph structure;

[0011] Construct the traffic knowledge graph structure based on multiple said nodes, the attribute information corresponding to each said node, and multiple said edges.

[0012] In a possible implementation manner, the feature processing of the accident section and the gantry section based on the traffic knowledge graph structure to predict the current speed of the accident section and the current speed of the gantry section includes:

[0013] Based on the traffic knowledge graph structure, determine the neighbor nodes of the accident section and the neighbor nodes of the gantry section;

[0014] Perform joint feature learning based on the attribute information of the neighbor nodes of the accident section and the attribute information of the accident section to predict the current speed of the accident section;

[0015] Perform joint feature learning based on the attribute information of the neighbor nodes of the gantry section and the attribute information of the gantry section to predict the current speed of the gantry section.

[0016] In a possible implementation manner, the performing joint feature learning based on the attribute information of the neighbor nodes of the accident section and the attribute information of the accident section to predict the current speed of the accident section includes:

[0017] Based on the section historical speed information, traffic flow information, and weather impact factor information in the attribute information of the accident section, determine the first section feature of the accident section;

[0018] Based on the vehicle historical speed information and the recently passed gantry data information in the attribute information of the driving vehicle nodes among the neighbor nodes of the accident section, determine the vehicle features of multiple said driving vehicles;

[0019] Based on the attribute information of other highway section nodes among the neighbor nodes of the accident section, determine the second section feature;

[0020] Based on the activation function, process multiple second section features, vehicle features of multiple traveling vehicles, and the first section feature to determine the current vehicle speed of the accident section.

[0021] In a possible implementation manner, the process of determining the current vehicle speed of the accident section based on the activation function, processing multiple second section features, vehicle features of multiple traveling vehicles, and the first section feature includes:

[0022] Each layer of the graph convolutional network model performs an aggregation process on multiple second section features, vehicle features of multiple traveling vehicles, and the first section feature based on the activation function to determine the aggregated feature output by the last layer of the network;

[0023] Based on the product of the aggregated feature and the weight value, determine the current vehicle speed of the accident section.

[0024] In a possible implementation manner, the process of determining the accident point location of the accident vehicle based on the current vehicle speed of the accident section, the current vehicle speed of the gantry section, the ETC gantry information, and the alarm receiving time information includes:

[0025] Perform a mean process on the current vehicle speed of the accident section and the current vehicle speed of the gantry section to determine the target speed;

[0026] Based on the alarm receiving time information and the time information of passing through the gantry in the ETC gantry information, determine the time difference;

[0027] Based on the product of the time difference, the target speed, and the target value, determine the accident point location of the accident vehicle.

[0028] The embodiment of the present application further provides a positioning device for the rescue location of highway vehicles. The positioning device includes:

[0029] A query module, configured to query the license plate recognition database based on the accident information of the accident vehicle to determine the ETC gantry information that the accident vehicle passed by most recently; wherein, the accident information includes alarm receiving time information, license plate information, and the accident section;

[0030] A vehicle speed determination module, configured to input the accident section and the gantry section in the ETC gantry information into a pre-trained graph convolutional network model, perform feature processing on the accident section and the gantry section based on the traffic knowledge graph structure, and predict the current vehicle speed of the accident section and the current vehicle speed of the gantry section;

[0031] An accident location determination module, configured to determine the accident point location of the accident vehicle based on the current vehicle speed of the accident section, the current vehicle speed of the gantry section, the ETC gantry information, and the alarm receiving time information.

[0032] In a possible implementation manner, the positioning device further includes a knowledge graph structure construction module, and the knowledge graph structure construction module determines the traffic knowledge graph structure through the following steps:

[0033] Obtain the attribute information of each highway section and the attribute information of the vehicles traveling on each highway section;

[0034] Take each highway section and each traveling vehicle as nodes of the traffic knowledge graph structure, and take the driving relationship between each traveling vehicle and each highway section and the connectivity relationship between each highway section and other highway sections as edges of the traffic knowledge graph structure;

[0035] Construct the traffic knowledge graph structure based on multiple nodes, the attribute information corresponding to each node, and multiple edges.

[0036] An embodiment of the present application further provides an electronic device, including: a processor, a memory, and a bus. The memory stores machine-readable instructions executable by the processor. When the electronic device runs, the processor communicates with the memory through the bus. When the machine-readable instructions are executed by the processor, the steps of the positioning method for the rescue position of highway vehicles as described above are executed.

[0037] An embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is run by a processor, the steps of the positioning method for the rescue position of highway vehicles as described above are executed.

[0038] The positioning method, device, equipment and medium for the rescue location of highway vehicles provided by the embodiments of the present application, the positioning method includes: querying the license plate recognition database based on the accident information of the accident vehicle to determine the ETC gantry information that the accident vehicle passed by most recently; wherein, the accident information includes the alarm receiving time information, license plate information and accident section; inputting the accident section and the gantry section in the ETC gantry information into a pre-trained graph convolutional network model, performing feature processing on the accident section and the gantry section based on the traffic knowledge graph structure, and predicting the current vehicle speed of the accident section and the current vehicle speed of the gantry section; based on the current vehicle speed of the accident section, the current vehicle speed of the gantry section, the ETC gantry information and the alarm receiving time information, determining the accident point location of the accident vehicle. The graph neural network model is used to perform feature learning on the accident section and the gantry section, predict the real-time vehicle speed of the current accident-related section and the real-time vehicle speed of the gantry section, thereby improving the accuracy of accident point estimation.

[0039] To make the above objects, features and advantages of the present application more obvious and understandable, the following specifically enumerates preferred embodiments and, in conjunction with the accompanying drawings, makes a detailed description as follows. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings required for use in the embodiments. It should be understood that the following drawings only show some embodiments of the present application, and therefore should not be regarded as limiting the scope. For those of ordinary skill in the art, without creative efforts, other related drawings can also be obtained based on these drawings.

[0041] Figure 1 It is a flowchart of a positioning method for the rescue location of highway vehicles provided by the embodiments of the present application;

[0042] Figure 2 It is one of the structural schematic diagrams of a positioning device for the rescue location of highway vehicles provided by the embodiments of the present application;

[0043] Figure 3 It is the second structural schematic diagram of a positioning device for the rescue location of highway vehicles provided by the embodiments of the present application;

[0044] Figure 4 It is the structural schematic diagram of an electronic device provided by the embodiments of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0045] To make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application usually described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, each other embodiment obtained by those skilled in the art without making creative work belongs to the scope of protection of the present application.

[0046] First, the application scenarios to which the present application is applicable are introduced. The present application can be applied in the field of highway management technology.

[0047] Research has found that highways have the advantages of large capacity, fast speed, low transportation cost and significant social benefits. However, once a traffic accident occurs on a highway, it is often shocking, and after the accident, the lane will be occupied, which is very likely to cause road congestion and threaten the driving safety of the highway. Therefore, fast and accurate accident location, timely and effective accident warning, efficient accident handling and rescue are particularly important to ensure the safe and stable operation of highways, improve the management and service level of highways, and enhance the ability to deal with emergencies. In the existing technology, the auxiliary positioning and rescue method of highway traffic accident vehicles based on ETC data has some shortcomings: (1) Its scope of application is limited, only for ETC accident vehicle positioning, and cannot cover non-ETC vehicles, resulting in limited rescue scenarios; (2) It relies on a single ETC gantry data, and the positioning accuracy and reliability are low. Therefore, how to improve the accuracy of highway vehicle rescue position positioning has become a technical problem that cannot be underestimated.

[0048] Based on this, an embodiment of the present application provides a method for locating the rescue position of a vehicle on a highway, and adopts a graph neural network model to perform feature learning on the accident section and the gantry section, predict the real-time vehicle speed of the current accident-related section and the real-time vehicle speed of the gantry section, thereby improving the accuracy of the accident point estimation.

[0049] See also Figure 1 , Figure 1 The flowchart of a method for locating a rescue position of a vehicle on a highway provided by an embodiment of the present application is as follows. Figure 1 As shown in , the positioning method provided in the embodiment of the present application includes:

[0050] S101: Query the license plate recognition database based on the accident information of the accident vehicle to determine the ETC gantry information that the accident vehicle passed by most recently; wherein, the accident information includes the alarm receiving time information, license plate information, and accident section.

[0051] In this step, query the license plate recognition database according to the accident information of the accident vehicle to determine the ETC gantry information that the accident vehicle passed by most recently.

[0052] Wherein, the accident information includes the alarm receiving time information, license plate information, and accident section.

[0053] Here, starting from when the alarm call is connected, the customer service system receives the alarm information. The customer service system receives the user's alarm call and triggers the positioning process. Collect accident information, including: alarm receiving time T1, license plate information P, and accident section L1. Clarify the basic accident information (T1, P, L1) to provide data support for subsequent positioning.

[0054] Here, according to the license plate information P and the alarm receiving time T1, query the license plate recognition database to obtain the ETC gantry information that the accident vehicle passed by most recently. Input the license plate information P and the alarm receiving time T1 to obtain the ETC gantry information. The ETC gantry information includes the time T0 and the gantry section L0 where the gantry is located. Here, set the gantry position as the origin kilometer number (0km) as the reference point for positioning.

[0055] S102: Input the accident section and the gantry section in the ETC gantry information into a pre-trained graph convolutional network model, perform feature processing on the accident section and the gantry section based on the traffic knowledge graph structure, and predict the current vehicle speed of the accident section and the current vehicle speed of the gantry section.

[0056] In this step, input the accident section and the gantry section in the ETC gantry information into a pre-trained graph convolutional network model, perform feature processing on the accident section and the gantry section according to the traffic knowledge graph structure, and predict the current vehicle speed of the accident section and the current vehicle speed of the gantry section.

[0057] Among them, the identification information of the accident section and the gantry section can be input into the graph convolutional network model so that the attribute information of the accident section and the gantry section can be determined according to the traffic knowledge graph structure.

[0058] Here, the trained Graph Neural Network (GNN) is used to perform feature learning on the historical trajectories of vehicles, traffic flow, ETC gantry data, etc., to predict the real-time vehicle speed of the current accident section, thereby improving the accuracy of accident location estimation. The graph convolutional network model is trained in the following way: The node set in the traffic knowledge graph structure includes all road section nodes and vehicle nodes, and the edge set includes all vehicle-road section edges and road section-road section edges. Each node has an initial feature vector. The core idea of the graph convolutional network is to let each node learn features from its neighbor nodes through a message passing mechanism. The specific steps are as follows: Initialize node features: Assign an initial feature vector to each node. Message passing: Gradually update the node features through multiple layers of graph convolutional operations. Output layer: In the last layer, output the predicted vehicle speed of the road section nodes. The goal of the training process is to minimize the error between the predicted vehicle speed and the real vehicle speed. The commonly used loss function is the mean square error. Optimize the model parameters through the backpropagation algorithm to gradually reduce the loss function and obtain the trained graph convolutional network model.

[0059] In a possible implementation manner, the traffic knowledge graph structure is determined through the following steps:

[0060] A: Obtain the attribute information of each highway section and the attribute information of the vehicles traveling on each highway section.

[0061] Here, the attribute information of the highway section includes historical vehicle speed information, traffic flow, and weather impact factors. The attribute information of the traveling vehicle includes license plate information, vehicle type, ETC passing records, and vehicle historical speed information.

[0062] B: Take each highway section and each traveling vehicle as nodes of the traffic knowledge graph structure, and take the driving relationship between each traveling vehicle and each highway section and the connectivity relationship between each highway section and other highway sections as edges of the traffic knowledge graph structure.

[0063] Here, take each highway section and each traveling vehicle as nodes of the traffic knowledge graph structure, and take the driving relationship between each traveling vehicle and each highway section and the connectivity relationship between each highway section and other highway sections as edges of the traffic knowledge graph structure.

[0064] Among them, the driving relationship means that the vehicle travels on a certain road section at a certain time point, and the connectivity relationship refers to adjacent ETC gantries.

[0065] C: Construct the traffic knowledge graph structure based on multiple said nodes, the attribute information corresponding to each said node, and multiple said edges.

[0066] Here, a traffic knowledge graph structure is constructed based on multiple nodes, the attribute information of the nodes, and the connections between multiple edges.

[0067] In a possible implementation manner, the feature processing of the accident section and the gantry section based on the traffic knowledge graph structure to predict the current vehicle speed of the accident section and the current vehicle speed of the gantry section includes:

[0068] a: Based on the traffic knowledge graph structure, determine the neighbor nodes of the accident section and the neighbor nodes of the gantry section.

[0069] Here, query the neighbor nodes of the accident section and the neighbor nodes of the gantry section according to the traffic knowledge graph structure.

[0070] Among them, the neighbor nodes of the accident section include other highway section nodes that have a connectivity relationship with the accident section and the driving vehicle nodes driving on the accident section.

[0071] b: Perform joint feature learning based on the attribute information of the neighbor nodes of the accident section and the attribute information of the accident section to predict the current vehicle speed of the accident section.

[0072] Here, perform joint feature learning based on the attribute information of the neighbor nodes of the accident section and the attribute information of the accident section to predict the current vehicle speed of the accident section.

[0073] In a possible implementation manner, the performing joint feature learning based on the attribute information of the neighbor nodes of the accident section and the attribute information of the accident section to predict the current vehicle speed of the accident section includes:

[0074] (1): Based on the section historical vehicle speed information, traffic flow information, and weather impact factor information in the attribute information of the accident section, determine the first section feature of the accident section.

[0075] Here, for accident section T i The representation of the first section feature is:

[0076]

[0077] Among them, is the historical vehicle speed of this section, is the traffic flow of this section, is the weather impact factor, is accident section R i is the first section feature of.

[0078] (2): Determine vehicle characteristics of multiple said moving vehicles based on the vehicle historical speed information and the recently passed gantry data information in the attribute information of the moving vehicle nodes among the neighbor nodes of the accident section.

[0079] Here, the representation of the vehicle characteristics of the moving vehicle is:

[0080]

[0081] Among them, is the historical speed of the vehicle, is vehicle V j the ETC gantry data recently passed, is the vehicle characteristic of the moving vehicle.

[0082] (3): Determine the second section characteristic based on the attribute information of other highway section nodes among the neighbor nodes of the accident section.

[0083] Here, determine the second section characteristic according to the attribute information of other highway section nodes among the neighbor nodes of the accident section.

[0084] (4): Determine the current speed of the accident section based on processing multiple said second section characteristics, multiple vehicle characteristics of the moving vehicles, and the first section characteristic by an activation function.

[0085] Here, determine the current speed of the accident section according to processing multiple second section characteristics, multiple vehicle characteristics of the moving vehicles, and the first section characteristic by an activation function.

[0086] In a possible implementation manner, the determining the current speed of the accident section based on processing multiple said second section characteristics, multiple vehicle characteristics of the moving vehicles, and the first section characteristic by an activation function includes:

[0087] I: Each layer of the graph convolutional network model performs an aggregation process on multiple said second section characteristics, multiple vehicle characteristics of the moving vehicles, and the first section characteristic based on the activation function to determine the aggregated characteristic output by the last layer of the network.

[0088] Here, each layer of the graph convolutional network model performs a convolutional aggregation process on multiple second section characteristics, multiple vehicle characteristics of the moving vehicles, and the first section characteristic to determine the aggregated characteristic output by the last layer of the network.

[0089] Among them, for accident section R i , its adjacent section R k and the moving vehicle V driving on it in the accident section jObtain the current vehicle speed information from:

[0090]

[0091] Among them, is the aggregated feature output by the last layer of the network. W1 and W2 are both weight values. is the vehicle feature of the driving vehicle V obtained by convolutional processing of the l-th network layer j of i R i The set of neighbor nodes of R includes accident sections among the adjacent sections in the set of neighbor nodes. is the section feature of other highway sections where the driving vehicle R is located obtained by convolutional processing of the l-th network layer k of

[0092] Here, the vehicle speed information of the section is not only determined by the vehicle speed in the past of this section, but also affected by the speed of the vehicles driving on this section currently, and the vehicle speeds of adjacent sections.

[0093] II: Based on the product of the aggregated feature and the weight value, determine the current vehicle speed of the accident section.

[0094] Here, according to the product of the aggregated feature and the weight value, determine the current vehicle speed of the accident section.

[0095] Among them, the weight value is the weight value of the output layer of the graph convolutional network model.

[0096] c: Based on the attribute information of the neighbor nodes of the gantry section and the attribute information of the gantry section, perform joint feature learning to predict the current vehicle speed of the gantry section.

[0097] Here, according to the attribute information of the neighbor nodes of the gantry section and the attribute information of the gantry section, perform joint feature learning to predict the current vehicle speed of the gantry section.

[0098] Among them, the determination process of the current vehicle speed of the gantry section is the same as the determination process of the current vehicle speed of the above-mentioned accident section, and this part will not be elaborated here.

[0099] S103: Based on the current vehicle speed of the accident section, the current vehicle speed of the gantry section, the ETC gantry information, and the alarm receiving time information, determine the accident point location of the accident vehicle.

[0100] In this step, according to the current vehicle speed of the accident section, the current vehicle speed of the gantry section, the ETC gantry information, and the alarm receiving time information, determine the accident point location of the accident vehicle.

[0101] In a possible implementation manner, determining the accident point location of the accident vehicle based on the current vehicle speed of the accident section, the current vehicle speed of the gantry section, the ETC gantry information, and the alarm receiving time information includes:

[0102] Performing an average processing on the current vehicle speed of the accident section and the current vehicle speed of the gantry section to determine a target speed; determining a time difference based on the alarm receiving time information and the time information of passing through the gantry in the ETC gantry information; determining the accident point location of the accident vehicle based on the product of the time difference, the target speed, and a target value.

[0103] Here, the accident point location of the accident vehicle is determined by the following formula:

[0104] Δt = T1 - T0

[0105]

[0106] Wherein, is the current vehicle speed of the gantry section, is the current vehicle speed of the accident section, X1 is the accident point location of the accident vehicle, T0 is the time information of passing through the gantry, and T1 is the alarm receiving time information.

[0107] In a specific embodiment, the system receives user alarm information, collects the alarm receiving time, license plate information, and accident section to clarify the basic information of the accident. Then, the system queries the license plate recognition database, and according to the license plate information and the alarm receiving time, obtains the information of the gantry that the vehicle passed through recently in the province, including the passing time and the section where the gantry is located. A graph neural network is used to perform feature learning on the historical trajectory, traffic flow, ETC gantry data, etc. of the vehicle, predict the real-time vehicle speed of the relevant accident section, and further improve the accuracy of accident point estimation. It realizes the rapid and accurate positioning of ETC and non-ETC accident vehicles, and provides intelligent and efficient auxiliary support for highway vehicle rescue.

[0108] A positioning method for the rescue location of highway vehicles provided by an embodiment of the present application, the positioning method includes: querying a license plate recognition database based on the accident information of the accident vehicle to determine the ETC gantry information that the accident vehicle passed by most recently; wherein, the accident information includes alarm receiving time information, license plate information, and accident section; inputting the accident section and the gantry section in the ETC gantry information into a pre-trained graph convolutional network model, performing feature processing on the accident section and the gantry section based on the traffic knowledge graph structure, and predicting the current vehicle speed of the accident section and the current vehicle speed of the gantry section; determining the accident point location of the accident vehicle based on the current vehicle speed of the accident section, the current vehicle speed of the gantry section, the ETC gantry information, and the alarm receiving time information. A graph neural network model is used to perform feature learning on the accident section and the gantry section, predict the real-time vehicle speed of the current accident-related section and the real-time vehicle speed of the gantry section, thereby improving the accuracy of accident point estimation.

[0109] Please refer to Figure 2 、 Figure 3 , Figure 2 which is one of the structural schematic diagrams of a positioning device for the rescue location of highway vehicles provided by an embodiment of the present application; Figure 3 which is the second structural schematic diagram of a positioning device for the rescue location of highway vehicles provided by an embodiment of the present application. As Figure 2 shown in

[0110] The positioning device 200 includes:

[0111] A query module 210, configured to query a license plate recognition database based on the accident information of the accident vehicle to determine the ETC gantry information that the accident vehicle passed by most recently; wherein, the accident information includes alarm receiving time information, license plate information, and accident section;

[0112] A vehicle speed determination module 220, configured to input the accident section and the gantry section in the ETC gantry information into a pre-trained graph convolutional network model, perform feature processing on the accident section and the gantry section based on the traffic knowledge graph structure, and predict the current vehicle speed of the accident section and the current vehicle speed of the gantry section;

[0113] An accident location determination module 230, configured to determine the accident point location of the accident vehicle based on the current vehicle speed of the accident section, the current vehicle speed of the gantry section, the ETC gantry information, and the alarm receiving time information. Figure 3 Further, as

[0114] Obtain the attribute information of each highway section and the attribute information of the vehicles traveling on each highway section;

[0115] Take each highway section and each traveling vehicle as nodes of the traffic knowledge graph structure, and take the driving relationship between each traveling vehicle and each highway section and the connectivity relationship between each highway section and other highway sections as edges of the traffic knowledge graph structure;

[0116] Construct the traffic knowledge graph structure based on multiple said nodes, the attribute information corresponding to each said node, and multiple said edges.

[0117] Further, when the vehicle speed determination module 220 is used to perform feature processing on the accident section and the gantry section based on the traffic knowledge graph structure and predict the current vehicle speed of the accident section and the current vehicle speed of the gantry section, the vehicle speed determination module 220 specifically is used for:

[0118] Based on the traffic knowledge graph structure, determine the neighbor nodes of the accident section and the neighbor nodes of the gantry section;

[0119] Perform joint feature learning based on the attribute information of the neighbor nodes of the accident section and the attribute information of the accident section, and predict the current vehicle speed of the accident section;

[0120] Perform joint feature learning based on the attribute information of the neighbor nodes of the gantry section and the attribute information of the gantry section, and predict the current vehicle speed of the gantry section.

[0121] Further, when the vehicle speed determination module 220 is used to perform joint feature learning based on the attribute information of the neighbor nodes of the accident section and the attribute information of the accident section and predict the current vehicle speed of the accident section, the vehicle speed determination module 220 specifically is used for:

[0122] Based on the section historical vehicle speed information, traffic flow information, and weather influence factor information in the attribute information of the accident section, determine the first section feature of the accident section;

[0123] Based on the vehicle historical vehicle speed information and the information of the recently passed gantry data in the attribute information of the traveling vehicle nodes among the neighbor nodes of the accident section, determine the vehicle features of multiple said traveling vehicles;

[0124] Based on the attribute information of other highway section nodes among the neighbor nodes of the accident section, determine the second section feature;

[0125] Based on the activation function, process multiple second-segment features, vehicle features of multiple traveling vehicles, and the first-segment feature to determine the current vehicle speed of the accident section.

[0126] Further, when the vehicle speed determination module 220 is used to process multiple second-segment features, vehicle features of multiple traveling vehicles, and the first-segment feature based on the activation function to determine the current vehicle speed of the accident section, the vehicle speed determination module 220 specifically is used for:

[0127] Each layer of the graph convolutional network model aggregates and processes multiple second-segment features, vehicle features of multiple traveling vehicles, and the first-segment feature based on the activation function to determine the aggregated feature output by the last layer of the network;

[0128] Based on the product of the aggregated feature and the weight value, determine the current vehicle speed of the accident section.

[0129] Further, when the accident location determination module 230 is used to determine the accident point location of the accident vehicle based on the current vehicle speed of the accident section, the current vehicle speed of the gantry section, the ETC gantry information, and the alarm receiving time information, the accident location determination module 230 specifically is used for:

[0130] Perform mean processing on the current vehicle speed of the accident section and the current vehicle speed of the gantry section to determine the target speed;

[0131] Based on the alarm receiving time information and the passing gantry time information in the ETC gantry information, determine the time difference;

[0132] Based on the product of the time difference, the target speed, and the target value, determine the accident point location of the accident vehicle.

[0133] A positioning device for the rescue location of highway vehicles provided by an embodiment of the present application. The positioning device includes: a query module, configured to query a license plate recognition database based on the accident information of an accident vehicle to determine the ETC gantry information that the accident vehicle passed by most recently; wherein, the accident information includes alarm receiving time information, license plate information, and accident section; a vehicle speed determination module, configured to input the accident section and the gantry section in the ETC gantry information into a pre-trained graph convolutional network model, perform feature processing on the accident section and the gantry section based on a traffic knowledge graph structure, and predict the current vehicle speed of the accident section and the current vehicle speed of the gantry section; an accident location determination module, configured to determine the accident point location of the accident vehicle based on the current vehicle speed of the accident section, the current vehicle speed of the gantry section, the ETC gantry information, and the alarm receiving time information. A graph neural network model is used to perform feature learning on the accident section and the gantry section, predict the real-time vehicle speed of the current accident-related section and the real-time vehicle speed of the gantry section, thereby improving the accuracy of accident point estimation.

[0134] Please refer to Figure 4 , Figure 4 which is a schematic structural diagram of an electronic device provided by an embodiment of the present application. As Figure 4 shown in

[0135] the electronic device 400 includes a processor 410, a memory 420, and a bus 430. Figure 1 The memory 420 stores machine-readable instructions executable by the processor 410. When the electronic device 400 runs, the processor 410 communicates with the memory 420 through the bus 430. When the machine-readable instructions are executed by the processor 410, the steps of the positioning method for the rescue location of highway vehicles in the method embodiment as shown above

[0136] can be executed. The specific implementation manner can refer to the method embodiment and will not be elaborated here. Figure 1 An embodiment of the present application further provides a computer-readable storage medium. A computer program is stored on the computer-readable storage medium. When the computer program is run by a processor, the steps of the positioning method for the rescue location of highway vehicles in the method embodiment as shown above

[0137] can be executed. The specific implementation manner can refer to the method embodiment and will not be elaborated here.

[0138] In several embodiments provided in this application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For another example, 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 displayed or discussed couplings or direct couplings or communication connections to each other can be through some communication interfaces. The indirect couplings or communication connections of the devices or units can be in electrical, mechanical, or other forms.

[0139] The units described as separate components may or may not be physically separated. The components displayed as units may or may not be physical units, that is, they can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0140] In addition, in each embodiment of this application, the functional units can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit.

[0141] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a non-volatile computer-readable storage medium executable by a processor. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in each embodiment of this application. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM), random access memories (RAM), magnetic disks, or optical discs that can store program codes.

[0142] Finally, it should be noted that the above-described embodiments are only specific embodiments of the present application, which are used to illustrate the technical solutions of the present application, rather than to limit it. The protection scope of the present application is not limited thereto. Although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: any person skilled in the art within the technical scope disclosed by the present application can still modify the technical solutions described in the foregoing embodiments, or can easily think of changes, or perform equivalent replacements on some of the technical features; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A positioning method for the rescue position of highway vehicles, characterized in that, The positioning method includes: Query the license plate recognition database based on the accident information of the accident vehicle to determine the ETC gantry information that the accident vehicle passed by most recently; wherein, the accident information includes the alarm receiving time information, license plate information, and accident section. Input the accident section and the gantry section in the ETC gantry information into a pre-trained graph convolutional network model, perform feature processing on the accident section and the gantry section based on the traffic knowledge graph structure, and predict the current speed of the accident section and the current speed of the gantry section. Based on the current speed of the accident section, the current speed of the gantry section, the ETC gantry information, and the alarm receiving time information, determine the accident point location of the accident vehicle.

2. The positioning method according to claim 1, characterized in that, Determine the traffic knowledge graph structure through the following steps: Obtain the attribute information of each highway section and the attribute information of the vehicles driving on each highway section. Take each highway section and each driving vehicle as nodes of the traffic knowledge graph structure, and take the driving relationship between each driving vehicle and each highway section and the connectivity relationship between each highway section and other highway sections as edges of the traffic knowledge graph structure. Construct the traffic knowledge graph structure based on multiple nodes, the attribute information corresponding to each node, and multiple edges.

3. The positioning method according to claim 1, wherein The performing feature processing on the accident section and the gantry section based on the traffic knowledge graph structure and predicting the current speed of the accident section and the current speed of the gantry section includes: Based on the traffic knowledge graph structure, determine the neighbor nodes of the accident section and the neighbor nodes of the gantry section. Perform joint feature learning based on the attribute information of the neighbor nodes of the accident section and the attribute information of the accident section, and predict the current speed of the accident section. Perform joint feature learning based on the attribute information of the neighbor nodes of the gantry section and the attribute information of the gantry section, and predict the current speed of the gantry section.

4. The positioning method according to claim 3, wherein The performing joint feature learning based on the attribute information of the neighbor nodes of the accident section and the attribute information of the accident section and predicting the current speed of the accident section includes: Based on the section historical speed information, traffic flow information, and weather influence factor information in the attribute information of the accident section, determine the first section feature of the accident section. Based on the vehicle historical speed information and the information of the most recently passed gantry data in the attribute information of the driving vehicle nodes among the neighbor nodes of the accident section, determine the vehicle features of multiple driving vehicles. Based on the attribute information of other highway section nodes among the neighbor nodes of the accident section, determine the second section feature. Based on an activation function, process multiple second section features, the vehicle features of multiple driving vehicles, and the first section feature to determine the current speed of the accident section.

5. The positioning method according to claim 4, wherein The processing multiple second section features, the vehicle features of multiple driving vehicles, and the first section feature based on an activation function to determine the current speed of the accident section includes: Each layer of the graph convolutional network model aggregates multiple second section features, vehicle features of multiple traveling vehicles, and the first section feature based on the activation function to determine the aggregated feature output by the last layer of the network; Based on the product of the aggregated feature and the weight value, the current vehicle speed of the accident section is determined.

6. The positioning method according to claim 1, characterized in that, The determining of the accident point location of the accident vehicle based on the current vehicle speed of the accident section, the current vehicle speed of the gantry section, the ETC gantry information, and the alarm receiving time information includes: Performing an averaging process on the current vehicle speed of the accident section and the current vehicle speed of the gantry section to determine a target speed; Based on the alarm receiving time information and the time information of passing through the gantry in the ETC gantry information, a time difference is determined; Based on the product of the time difference, the target speed, and a target value, the accident point location of the accident vehicle is determined.

7. A positioning device for the rescue location of highway vehicles, characterized in that, The positioning device includes: A query module for querying the license plate recognition database based on the accident information of the accident vehicle to determine the ETC gantry information that the accident vehicle passed through most recently; wherein, the accident information includes alarm receiving time information, license plate information, and the accident section; A vehicle speed determination module for inputting the accident section and the gantry section in the ETC gantry information into a pre-trained graph convolutional network model, performing feature processing on the accident section and the gantry section based on the traffic knowledge graph structure, and predicting the current vehicle speed of the accident section and the current vehicle speed of the gantry section; An accident location determination module for determining the accident point location of the accident vehicle based on the current vehicle speed of the accident section, the current vehicle speed of the gantry section, the ETC gantry information, and the alarm receiving time information.

8. The positioning device according to claim 7, wherein The positioning device further includes a knowledge graph structure construction module, and the knowledge graph structure construction module determines the traffic knowledge graph structure through the following steps: Obtaining the attribute information of each highway section and the attribute information of the vehicles traveling on each highway section; Taking each highway section and each traveling vehicle as nodes of the traffic knowledge graph structure, and taking the traveling relationship between each traveling vehicle and each highway section and the connectivity relationship between each highway section and other highway sections as edges of the traffic knowledge graph structure; Constructing the traffic knowledge graph structure based on multiple nodes, the attribute information corresponding to each node, and multiple edges.

9. An electronic device, characterized in that, Including: A processor, a memory, and a bus. The memory stores machine-readable instructions executable by the processor. When the electronic device runs, the processor communicates with the memory through the bus. When the machine-readable instructions are run by the processor, the steps of the positioning method for the rescue location of highway vehicles as described in any one of claims 1 to 6 are executed.

10. A computer-readable storage medium, characterized in that, A computer program is stored on the computer-readable storage medium. When the computer program is run by the processor, the steps of the positioning method for the rescue location of highway vehicles as described in any one of claims 1 to 6 are executed.

Images