Vehicle trajectory prediction method and device for airport operation, electronic equipment and storage medium

Abstract

According to the vehicle trajectory prediction method and device for airport operation, the electronic equipment and the storage medium, the method comprises the steps that features are extracted through a multi-layer network, the precision and adaptability of trajectory prediction are remarkably improved, path planning is optimized, and the safety of special vehicles in airport operation is enhanced. Through a reasonable feature extraction and modeling strategy, the method effectively reduces the calculation complexity, comprehensively considers the multi-dimensional features of the vehicle, and achieves the more comprehensive and accurate track prediction. The intelligent scheduling platform is constructed, various information of different vehicles is collected and stored, real-time decision making and intelligent scheduling are supported, the intelligent level of airport special vehicle scheduling management is improved, and the airport operation efficiency and safety are improved on the whole.

Description

Technical Field

[0001] The present invention relates to the field of vehicle trajectories, and in particular to a method, device, electronic equipment and storage medium for predicting vehicle trajectories in airport operations. Background Art

[0002] With the rapid development of autonomous driving technology, airport special vehicles are very suitable for adopting autonomous driving technology in the future to improve efficiency and safety because they operate in a clear, low-speed, and highly repeatable operating environment, and have high safety requirements and the potential for intelligent scheduling. However, the trajectory planning of autonomous airport operation special vehicles faces unique challenges and needs to be further optimized and improved. Special vehicles need to understand the motion status of surrounding vehicles in real time and accurately predict their future behavior to optimize path planning, improve operating efficiency, and ensure safety.

[0003] Existing trajectory prediction methods are mainly divided into knowledge-driven and data-driven methods. However, knowledge-driven methods have poor adaptability in complex and dynamic airport environments, while data-driven methods rely on a large amount of high-quality data and have high computational complexity. In addition, the frequent interactions between vehicles increase the complexity of prediction, making it difficult for traditional methods to effectively cope with it, resulting in low efficiency and safety of airport operations. Summary of the invention

[0004] The present application provides a vehicle trajectory prediction method, device, electronic device and storage medium for airport operations to solve the above-mentioned technical problems.

[0005] An embodiment of the present application provides a vehicle trajectory prediction method for airport operations, the method comprising: obtaining a sample data set, the sample data set comprising historical driving data of airport operation vehicles; building an initial vehicle trajectory prediction model, the backbone network of the initial trajectory prediction model comprising a long short-term memory network and a graph attention network; inputting the sample data set into the initial vehicle trajectory prediction model, performing model training on the initial vehicle trajectory prediction model, and obtaining a vehicle trajectory prediction model; inputting information of airport operation vehicles to be identified into the vehicle trajectory prediction model, and obtaining a vehicle trajectory prediction result, so as to complete vehicle trajectory prediction.

[0006] In one embodiment of the present application, obtaining a sample data set includes: within a preset time period, detecting the initial position and the final position of the airport operation vehicle by a position sensor on the airport operation vehicle; detecting the vehicle speed of the airport operation vehicle by a speed sensor on the airport operation vehicle; detecting the vehicle speed acceleration of the airport operation vehicle by an acceleration sensor on the airport operation vehicle; integrating the initial position and the final position of the airport operation vehicle, the vehicle speed and the vehicle acceleration into historical driving data, and storing them in the sample data set.

[0007] In one embodiment of the present application, the sample data set is input into the initial vehicle trajectory prediction model, and the initial vehicle trajectory prediction model is trained to obtain a vehicle trajectory prediction model, including: identifying the historical driving data through a long short-term memory network to capture the vehicle motion position characteristics of airport operation vehicles; identifying the historical driving data through the graph attention network to capture the vehicle trajectory interaction characteristics between airport operation vehicles; and performing batch normalization processing on the airport operation vehicle motion position characteristics and the airport operation vehicle trajectory interaction characteristics through the graph attention network decoder to obtain a vehicle trajectory prediction model.

[0008] In one embodiment of the present application, the historical driving data is identified through a long short-term memory network to capture the vehicle motion position characteristics of the airport operation vehicle, including: analyzing the initial position and the terminal position of the airport operation vehicle through the position encoder in the long short-term memory network to extract the position data of the airport operation vehicle; analyzing the vehicle speed and the vehicle acceleration through the motion encoder in the long short-term memory network to extract the motion state of the airport operation vehicle; connecting the position data of the airport operation vehicle and the motion state of the airport operation vehicle to obtain the vehicle motion position characteristics.

[0009] In one embodiment of the present application, the historical driving data is identified through the graph attention network to capture vehicle trajectory interaction features between airport operation vehicles, including: inputting node features and edge features through the position interaction encoder of the graph attention network, and adopting a multi-head attention mechanism to obtain initial vehicle trajectory interaction features; parsing the time series data in the initial vehicle trajectory interaction features through the motion interaction encoder of the graph attention network to obtain vehicle trajectory interaction features, and the vehicle trajectory interaction features include the motion trajectories of airport operation vehicles.

[0010] In one embodiment of the present application, the airport operation vehicle motion position features and the airport operation vehicle trajectory interaction features are batch normalized to obtain a vehicle trajectory prediction model, including: using a residual network to extract abstract feature representations from the vehicle motion position features; connecting the abstract feature representations and the vehicle trajectory interaction features to obtain combined vehicle trajectory interaction features; normalizing the combined vehicle trajectory interaction features to obtain a vehicle trajectory prediction model.

[0011] In one embodiment of the present application, after obtaining the vehicle trajectory prediction result, it also includes: sending the vehicle trajectory prediction result to an airport data monitoring platform, extracting key features from the vehicle trajectory prediction result through the airport data monitoring platform; performing a safety assessment on airport operating vehicles based on the key features, and immediately triggering an early warning mechanism when there is an abnormal driving condition in the safety assessment result.

[0012] The vehicle trajectory prediction device for airport operations provided in an embodiment of the present application includes: an acquisition module for acquiring a sample data set, wherein the sample data set includes historical driving data of airport operation vehicles; an initial model building module for building an initial vehicle trajectory prediction model, wherein the backbone network of the initial trajectory prediction model includes a long short-term memory network and a graph attention network; a training module for inputting the sample data set into the initial vehicle trajectory prediction model, performing model training on the initial vehicle trajectory prediction model, and obtaining a vehicle trajectory prediction model; a prediction module for inputting information of airport operation vehicles to be identified into the vehicle trajectory prediction model, and obtaining a vehicle trajectory prediction result to complete vehicle trajectory prediction.

[0013] An embodiment of the present application provides an electronic device, including a processor, a memory and a communication bus; the communication bus is used to connect the processor and the memory; the processor is used to execute a computer program stored in the memory to implement the above-mentioned vehicle trajectory prediction method for airport operations.

[0014] The beneficial effects of this application are as follows: by adopting a multi-layer network to extract features, the accuracy and adaptability of trajectory prediction are significantly improved, path planning is optimized, and the safety of special vehicles in airport operations is enhanced. Through reasonable feature extraction and modeling strategies, the present invention effectively reduces the computational complexity, while comprehensively considering the multi-dimensional characteristics of the vehicle to achieve a more comprehensive and accurate trajectory prediction. The construction of an intelligent dispatching platform collects and stores various types of information of different vehicles, supports real-time decision-making and intelligent dispatching, improves the intelligence level of airport special vehicle dispatching management, and improves the overall efficiency and safety of airport operations.

[0015] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The drawings herein are incorporated into the specification and constitute a part of the specification, showing embodiments consistent with the present application, and together with the specification, are used to explain the principles of the present application. Obviously, the drawings described below are only some embodiments of the present application, and for those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative work. In the drawings:

[0017] Figure 1 is a flow chart of a vehicle trajectory prediction method for airport operations shown in an exemplary embodiment of the present application;

[0018] Figure 2 is a block diagram of a vehicle trajectory prediction device for airport operations shown in an exemplary embodiment of the present application;

[0019] Figure 3 It is a structural schematic diagram of an electronic device shown in an exemplary embodiment of the present application. DETAILED DESCRIPTION

[0020] The following will describe the implementation methods of the present application with reference to the accompanying drawings and preferred embodiments. Those skilled in the art can easily understand other advantages and effects of the present application from the contents disclosed in this specification. The present application can also be implemented or applied through other different specific implementation methods, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present application. It should be understood that the preferred embodiments are only for illustrating the present application, not for limiting the scope of protection of the present application.

[0021] It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present application, and thus the drawings only show components related to the present application rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed at will, and the component layout may also be more complicated.

[0022] In the following description, a large number of details are discussed to provide a more thorough explanation of the embodiments of the present application. However, it is obvious to those skilled in the art that the embodiments of the present application can be implemented without these specific details. In other embodiments, well-known structures and devices are shown in the form of block diagrams rather than in detail to avoid making the embodiments of the present application difficult to understand.

[0023] The embodiments of the present application respectively propose a vehicle trajectory prediction method for airport operations, a vehicle trajectory prediction device for airport operations, and an electronic device, and these embodiments will be described in detail below.

[0024] See also Figure 1 , Figure 1 FIG. 1 is a flow chart of a method for predicting vehicle trajectories for airport operations according to an embodiment of the present application. Figure 1 As shown, the method at least includes steps S110 to S140, which are described in detail as follows:

[0025] Step S110, obtaining a sample data set, where the sample data set includes historical driving data of airport operation vehicles.

[0026] Step S120, building an initial vehicle trajectory prediction model, the backbone network of the initial trajectory prediction model includes a long short-term memory network and a graph attention network.

[0027] In one embodiment of the present application, the long short-term memory network is an LSTM (Long Short-Term Memory) network, and the graph attention network is a GAT network (Graph Attention Networks).

[0028] Step S130: input the sample data set into the initial vehicle trajectory prediction model, perform model training on the initial vehicle trajectory prediction model, and obtain a vehicle trajectory prediction model.

[0029] Step S140, inputting the information of the airport operation vehicle to be identified into the vehicle trajectory prediction model to obtain the vehicle trajectory prediction result to complete the vehicle trajectory prediction.

[0030] exist Figure 1 In the technical solution shown, by adopting a multi-layer network to extract features, the accuracy and adaptability of trajectory prediction are significantly improved, path planning is optimized, and the safety of special vehicles in airport operations is enhanced. Through reasonable feature extraction and modeling strategies, the present invention effectively reduces the computational complexity, while comprehensively considering the multi-dimensional characteristics of the vehicle, and achieving a more comprehensive and accurate trajectory prediction. The construction of an intelligent scheduling platform collects and stores various types of information of different vehicles, supports real-time decision-making and intelligent scheduling, improves the intelligence level of airport special vehicle scheduling management, and improves the overall efficiency and safety of airport operations.

[0031] In one embodiment of the present application, obtaining a sample data set includes: within a preset time period, detecting an initial position and a terminal position of an airport operation vehicle by means of a position sensor on the airport operation vehicle; detecting a vehicle speed of the airport operation vehicle by means of a vehicle speed sensor on the airport operation vehicle; detecting a vehicle speed acceleration of the airport operation vehicle by means of an acceleration sensor on the airport operation vehicle; integrating the initial position and terminal position of the airport operation vehicle, the vehicle speed, and the vehicle acceleration into historical driving data, and storing them in the sample data set.

[0032] In one embodiment of the present application, a sample data set is input into an initial vehicle trajectory prediction model, and the initial vehicle trajectory prediction model is trained to obtain a vehicle trajectory prediction model, including: identifying historical driving data through a long short-term memory network to capture vehicle motion position characteristics of airport operation vehicles; identifying historical driving data through a graph attention network to capture vehicle trajectory interaction characteristics between airport operation vehicles; and performing batch normalization processing on the airport operation vehicle motion position characteristics and the airport operation vehicle trajectory interaction characteristics through a graph attention network decoder to obtain a vehicle trajectory prediction model.

[0033] In one embodiment of the present application, historical driving data is identified through a long short-term memory network to capture vehicle motion position characteristics of airport operation vehicles, including: analyzing the initial position and terminal position of the airport operation vehicle through the position encoder in the long short-term memory network to extract the position data of the airport operation vehicle; analyzing the vehicle speed and vehicle acceleration through the motion encoder in the long short-term memory network to extract the motion state of the airport operation vehicle; connecting the position data of the airport operation vehicle and the motion state of the airport operation vehicle to obtain the vehicle motion position characteristics.

[0034] As an example, the position encoder in the long short-term memory network is an LSTM encoder. There are two types of LSTM, namely LSTM position encoder and LSTM motion encoder. There are n such LSTM position encoders in the model, and the position data of each vehicle is processed by an LSTM module with the same set of model parameters. The hidden size of each LSTM position encoder model is 32. Through such an encoder, the position feature of LSTM is obtained.

[0035]

[0036] Among them, LSTM is a position encoder cluster; it consists of a network of n LSTM encoding units. Since the position data is highly correlated with the vehicle trajectory. Therefore, it is necessary to consider another LSTM encoder to process position, velocity, and acceleration. This method can comprehensively model the vehicle motion and use multi-dimensional feature representation. By comprehensively considering the relationship between position, velocity, and acceleration, the vehicle motion behavior can be more accurately analyzed and future trajectory can be predicted. In order to achieve good convergence of the model. Similar to the above-mentioned LSTM position encoder, although the motion data of each vehicle is processed by an LSTM encoder, the parameters of each LSTM motion encoder are shared. Each LSTM motion encoder model has a hidden layer of size 32. Through the LSTM motion encoding network, the motion encoding features can be obtained:

[0037]

[0038] Among them, LSTM' is a motion encoder cluster, which consists of n LSTM encoding unit networks. In addition, the vehicle position is the basis of the vehicle trajectory, representing the spatial coordinates of the vehicle at a certain moment. Considering the speed, acceleration can also provide dynamic information of airport operation vehicles, so the input of the initial vehicle trajectory prediction model is the historical driving data X of all vehicles in the entire airport within the preset time, that is, the coordinates, speed and acceleration of n airport operation vehicles at time t. Specifically, it is obtained by the following method:

[0039]

[0040] Among them, X is the vehicle motion position feature,

[0041]

[0042] Among them, t is the time, x t is the position data of n airport operation vehicles in the airport at time t, and p is the motion state of the airport operation vehicles;

[0043]

[0044] in, is the horizontal coordinate of the kth vehicle at time t; is the vertical coordinate of the kth vehicle at time t; is the speed of the kth vehicle at time t; is the acceleration of the kth vehicle at time t.

[0045] In one embodiment of the present application, historical driving data is identified through a graph attention network to capture vehicle trajectory interaction features between airport operation vehicles, including: inputting node features and edge features through a position interaction encoder of the graph attention network, and using a multi-head attention mechanism to obtain initial vehicle trajectory interaction features; parsing time series data in the initial vehicle trajectory interaction features through a motion interaction encoder of the graph attention network to obtain vehicle trajectory interaction features, and the vehicle trajectory interaction features include motion trajectories of airport operation vehicles.

[0046] As an example, the GAT encoder is divided into two categories: GAT position interaction encoder and GAT motion interaction encoder, where the GAT position interaction encoder requires input node features Γ and edge features Λ, and the LSTM position encoding features are used as node features of the GAT position interaction encoder. It is convenient to use location information to model and analyze complex relationships in the graph structure, and Λ usually represents the connection information between node i and node j. It is necessary to calculate the distance between airport operation vehicle i and airport operation vehicle j to represent the edge connection information σ between node i and node j at the current time step t. i,j :

[0047]

[0048] Among them, σ i,j is the edge connection information, i is the airport operation vehicle i, j is the airport operation vehicle j, Dx is the horizontal distance threshold between vehicles, and Dy is the vertical distance threshold between vehicles, x i t is the horizontal coordinate of the i-th vehicle at time t; y i t is the vertical coordinate of the kth vehicle at time t.

[0049] The graph edge input of GAT is:

[0050] Λ={σ i,i |i,j∈{1,2,…,n}}

[0051] Among them, Λ is the edge feature, σ i,j is the edge connection information, i is the airport operation vehicle i, j is the airport operation vehicle j, and n is a natural number. The GAT position interaction encoder adopts a multi-head attention mechanism with 4 heads, and the final output GAT feature encoding feature can be obtained:

[0052]

[0053] Among them, F GAT is the output of the GAT position interaction encoder, whose dimension is (n×32), Γ is the node feature, and Λ is the edge feature.

[0054] The combined vehicle trajectory interaction features are obtained by connecting the abstract feature representation and the vehicle trajectory interaction features, that is, combining the LSTM motion encoding features with the GAT position interaction features, and then inputting the GAT motion interaction encoder with a multi-head attention mechanism. Finally, the GAT outputs a (n×32)-dimensional feature representation. Specifically, the combined vehicle trajectory interaction features are normalized, specifically, the output Y is the future trajectory of all observed vehicles from time step (t+1) to (t+tf), and its mathematical expression is:

[0055]

[0056] Among them, Y is the future trajectory of all observed vehicles and t is the time.

[0057] In one embodiment of the present application, the motion position features of airport operation vehicles and the trajectory interaction features of airport operation vehicles are batch normalized to obtain a vehicle trajectory prediction model, including: using a residual network to extract abstract feature representations from the vehicle motion position features; connecting the abstract feature representations and the vehicle trajectory interaction features to obtain combined vehicle trajectory interaction features; normalizing the combined vehicle trajectory interaction features to obtain a vehicle trajectory prediction model.

[0058] As an example, a residual network is used to process motion data X, further extract abstract feature representations, and then concatenate with GAT motion interaction features. The merged feature matrix is ​​batch normalized. The normalized matrix is ​​used as input to an LSTM decoder with a hidden layer size of 25. The LSTM decoder aggregates multi-dimensional features and performs temporal modeling to predict future trajectories Y.

[0059] In one embodiment of the present application, after obtaining the vehicle trajectory prediction results, it also includes: sending the vehicle trajectory prediction results to an airport data monitoring platform, extracting key features from the vehicle trajectory prediction results through the airport data monitoring platform; performing a safety assessment on airport operation vehicles based on the key features, and immediately triggering an early warning mechanism when there is an abnormal driving condition in the safety assessment results.

[0060] Figure 2 FIG. 1 is a block diagram of a vehicle trajectory prediction device for airport operations according to an exemplary embodiment of the present application. Figure 2As shown, according to an embodiment of the present application, a vehicle trajectory prediction device 200 for airport operations includes: an acquisition module 210, an initial model building module 220, a training module 230 and a prediction module 240, wherein the acquisition module is used to acquire a sample data set, and the sample data set includes historical driving data of airport operation vehicles; the initial model building module is used to build an initial vehicle trajectory prediction model, and the backbone network of the initial trajectory prediction model includes a long short-term memory network and a graph attention network; the training module is used to input the sample data set into the initial vehicle trajectory prediction model, and perform model training on the initial vehicle trajectory prediction model to obtain a vehicle trajectory prediction model; the prediction module is used to input the information of the airport operation vehicle to be identified into the vehicle trajectory prediction model, and obtain the vehicle trajectory prediction result to complete the vehicle trajectory prediction. By adopting a multi-layer network to extract features, the accuracy and adaptability of trajectory prediction are significantly improved, the path planning is optimized, and the safety of special vehicles in airport operations is enhanced. Through reasonable feature extraction and modeling strategies, the present invention effectively reduces the computational complexity, and at the same time comprehensively considers the multi-dimensional characteristics of the vehicle, and realizes a more comprehensive and accurate trajectory prediction. The construction of an intelligent dispatching platform collects and stores various types of information on different vehicles, supports real-time decision-making and intelligent dispatching, improves the intelligence level of airport special vehicle dispatching management, and improves the overall airport operating efficiency and safety.

[0061] In one embodiment of the present application, the acquisition module is used to detect the initial position and terminal position of the airport operation vehicle through the position sensor on the airport operation vehicle within a preset time period; detect the vehicle speed of the airport operation vehicle through the speed sensor on the airport operation vehicle; detect the vehicle speed acceleration of the airport operation vehicle through the acceleration sensor on the airport operation vehicle; integrate the initial position and terminal position of the airport operation vehicle, the vehicle speed and the vehicle acceleration into historical driving data, and store them in a sample data set.

[0062] In one embodiment of the present application, the training module is used to identify historical driving data through a long short-term memory network to capture the vehicle motion position characteristics of airport operation vehicles; identify historical driving data through a graph attention network to capture the vehicle trajectory interaction characteristics between airport operation vehicles; and perform batch normalization processing on the airport operation vehicle motion position characteristics and the airport operation vehicle trajectory interaction characteristics through a graph attention network decoder to obtain a vehicle trajectory prediction model.

[0063] In one embodiment of the present application, the training module is used to analyze the initial position and the terminal position of the airport operation vehicle through the position encoder in the long short-term memory network to extract the position data of the airport operation vehicle; analyze the vehicle speed and the vehicle acceleration through the motion encoder in the long short-term memory network to extract the motion state of the airport operation vehicle; connect the position data of the airport operation vehicle and the motion state of the airport operation vehicle to obtain the vehicle motion position characteristics.

[0064] In one embodiment of the present application, the training module is used to input node features and edge features through the position interaction encoder of the graph attention network, and adopt a multi-head attention mechanism to obtain initial vehicle trajectory interaction features; through the motion interaction encoder of the graph attention network, the time series data in the initial vehicle trajectory interaction features is parsed to obtain vehicle trajectory interaction features, and the vehicle trajectory interaction features include the motion trajectory of airport operating vehicles.

[0065] In one embodiment of the present application, the training module is used to extract abstract feature representations from vehicle motion position features using a residual network; connect the abstract feature representations and vehicle trajectory interaction features to obtain combined vehicle trajectory interaction features; normalize the combined vehicle trajectory interaction features to obtain a vehicle trajectory prediction model.

[0066] In one embodiment of the present application, the prediction module is used to send the vehicle trajectory prediction results to the airport data monitoring platform, and extract key features from the vehicle trajectory prediction results through the airport data monitoring platform; perform a safety assessment on the airport operation vehicles based on the key features, and immediately trigger an early warning mechanism when there is an abnormal driving condition in the safety assessment results.

[0067] It should be noted that the device provided in the above embodiment and the method provided in the above embodiment belong to the same concept, wherein the specific manner in which each module and unit performs the operation has been described in detail in the method embodiment, and will not be repeated here. In practical applications, the device provided in the above embodiment can distribute the above functions to different functional modules as needed, that is, divide the internal structure of the device into different functional modules to complete all or part of the functions described above, and this is not limited here.

[0068] See also Figure 3 , Figure 3 is a schematic diagram of a structure of an electronic device shown in an exemplary embodiment of the present application. It should be noted that: Figure 3 The electronic device 300 shown is only an example and should not bring any limitation to the functions and usage area of ​​the embodiments of the present application.

[0069] like Figure 3 As shown, the electronic device 300 includes a processor 301, a memory 302 and a communication bus 303; the communication bus 303 is used to connect the processor 301 and the memory 302; the processor 301 is used to execute the computer program stored in the memory 302 to implement one or more methods in the above embodiments.

[0070] The electronic device provided in the present application includes a processor, a memory, a transceiver and a communication interface. The memory and the communication interface are connected to the processor and the transceiver and complete communication with each other. The memory is used to store computer programs, the communication interface is used to communicate, and the processor and the transceiver are used to run the computer program so that the electronic device executes each step of the above method.

[0071] In this embodiment, the memory may include a random access memory (RAM), and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0072] The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0073] The above embodiments are merely illustrative of the principles and effects of the present application, and are not intended to limit the present application. Anyone familiar with the technology may modify or change the above embodiments without violating the spirit and scope of the present application. Therefore, all equivalent modifications or changes made by a person of ordinary skill in the art without departing from the spirit and technical ideas disclosed in the present application shall still be covered by the claims of the present application.

Claims

1. A vehicle trajectory prediction method for airport operations, characterized in that: The method comprises: Acquire a sample data set, wherein the sample data set includes historical driving data of airport operation vehicles; Building an initial vehicle trajectory prediction model, wherein the backbone network of the initial trajectory prediction model includes a long short-term memory network and a graph attention network; Inputting the sample data set into the initial vehicle trajectory prediction model, performing model training on the initial vehicle trajectory prediction model, and obtaining a vehicle trajectory prediction model; The information of the airport operation vehicle to be identified is input into the vehicle trajectory prediction model to obtain the vehicle trajectory prediction result to complete the vehicle trajectory prediction.

2. The vehicle trajectory prediction method for airport operations according to claim 1, characterized in that: Get a sample dataset, including: Within a preset time period, detecting the initial position and the terminal position of the airport operation vehicle by means of a position sensor on the airport operation vehicle; Detecting the vehicle speed of the airport operation vehicle by means of a vehicle speed sensor on the airport operation vehicle; Detecting the vehicle speed of the airport operation vehicle by an acceleration sensor on the airport operation vehicle; The initial position and terminal position of the airport operation vehicle, the vehicle speed and the vehicle acceleration are integrated into historical driving data and stored in the sample data set.

3. The vehicle trajectory prediction method for airport operations according to claim 2, characterized in that: Inputting the sample data set into the initial vehicle trajectory prediction model, performing model training on the initial vehicle trajectory prediction model, and obtaining a vehicle trajectory prediction model, including: Recognize the historical driving data through a long short-term memory network to capture the vehicle movement position characteristics of the airport operation vehicles; The historical driving data is identified by the graph attention network to capture the vehicle trajectory interaction characteristics between airport operation vehicles; The graph attention network decoder is used to perform batch normalization processing on the motion position features of the airport operation vehicles and the trajectory interaction features of the airport operation vehicles to obtain a vehicle trajectory prediction model.

4. The vehicle trajectory prediction method for airport operations according to claim 3, characterized in that: The historical driving data is recognized by a long short-term memory network to capture the vehicle motion position characteristics of the airport operation vehicles, including: Analyzing the initial position and the terminal position of the airport operation vehicle by the position encoder in the long short-term memory network to extract the position data of the airport operation vehicle; Analyzing the vehicle speed and the vehicle acceleration by means of a motion encoder in the long short-term memory network to extract the motion state of the airport operation vehicle; The position data of the airport operation vehicle and the motion state of the airport operation vehicle are connected to obtain the vehicle motion position feature.

5. The vehicle trajectory prediction method for airport operations according to claim 3, characterized in that: The historical driving data is identified by the graph attention network to capture the vehicle trajectory interaction features between airport operation vehicles, including: The node features and edge features are input through the position interaction encoder of the graph attention network, and the initial vehicle trajectory interaction features are obtained by using a multi-head attention mechanism; The time series data in the initial vehicle trajectory interaction features are parsed through the motion interaction encoder of the graph attention network to obtain vehicle trajectory interaction features, and the vehicle trajectory interaction features include the motion trajectories of airport operating vehicles.

6. The vehicle trajectory prediction method for airport operations according to claim 3, characterized in that: The motion position features of the airport operation vehicles and the trajectory interaction features of the airport operation vehicles are batch normalized to obtain a vehicle trajectory prediction model, including: A residual network is used to extract abstract feature representation from the vehicle motion position features; Connecting the abstract feature representation and the vehicle trajectory interaction feature to obtain a combined vehicle trajectory interaction feature; The combined vehicle trajectory interaction features are normalized to obtain a vehicle trajectory prediction model.

7. The vehicle trajectory prediction method for airport operations according to claim 5, characterized in that: After obtaining the vehicle trajectory prediction result, it also includes: Sending the vehicle trajectory prediction result to an airport data monitoring platform, and extracting key features from the vehicle trajectory prediction result through the airport data monitoring platform; A safety assessment is performed on airport operation vehicles based on the key features, and when an abnormal driving condition is found in the safety assessment result, an early warning mechanism is immediately triggered.

8. A vehicle trajectory prediction device for airport operations, characterized in that: The device comprises: An acquisition module, used for acquiring a sample data set, wherein the sample data set includes historical driving data of airport operation vehicles; An initial model building module, used to build an initial vehicle trajectory prediction model, wherein the backbone network of the initial trajectory prediction model includes a long short-term memory network and a graph attention network; A training module, used for inputting the sample data set into the initial vehicle trajectory prediction model, performing model training on the initial vehicle trajectory prediction model, and obtaining a vehicle trajectory prediction model; The prediction module is used to input the information of the airport operation vehicle to be identified into the vehicle trajectory prediction model to obtain the vehicle trajectory prediction result to complete the vehicle trajectory prediction.

9. An electronic device, characterized in that: The electronic device comprises: one or more processors; A storage device for storing one or more programs, which, when executed by the one or more processors, enables the electronic device to implement the vehicle trajectory prediction method for airport operations as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and the computer program is used to make a computer execute the method as described in any one of claims 1 to 7.

Images