A bus route space-time feature-based weekly vehicle behavior cognition method and system

Abstract

This invention belongs to the field of vehicle control technology and provides a method and system for recognizing weekly vehicle behavior based on the spatiotemporal characteristics of bus routes. The method includes: acquiring multi-source information data fused from vehicle, road, and cloud sources; extracting spatiotemporal features of dynamic bus route scenarios based on the acquired multi-source information data, and constructing a spatiotemporal scene feature vector for the bus route; predicting the vehicle queuing status at intersections based on the constructed spatiotemporal scene feature vector; and generating a vehicle trajectory sequence by combining the predicted intersection vehicle queuing status with a preset vehicle trajectory prediction model, thereby completing the recognition of weekly vehicle behavior based on the spatiotemporal characteristics of bus routes.

Description

Technical Field

[0001] This invention belongs to the field of vehicle control technology, specifically relating to a method and system for recognizing the weekly vehicle behavior based on the spatiotemporal characteristics of bus routes. Background Technology

[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.

[0003] In complex urban road environments, the operation of fixed-route buses is simultaneously affected by multiple factors, including route structure, signal timing, stop locations, and surrounding traffic flow, resulting in significant spatiotemporal dynamics and uncertainties in overall operation. Achieving accurate modeling of the spatiotemporal characteristics of routes and high-precision prediction of traffic conditions, while ensuring bus safety and punctuality, is a significant technical challenge for current intelligent bus scheduling and energy consumption optimization. Based on a vehicle-road-cloud collaborative architecture, integrating vehicle-side operational information, roadside detection data, and global traffic status from the cloud, new possibilities are provided for constructing spatiotemporal characteristic models and dynamic prediction mechanisms for public transport scenarios.

[0004] Existing technologies can use spatial attention-based vehicle trajectory prediction to model short-term interaction relationships between vehicles, thereby improving trajectory prediction accuracy. This is suitable for dense trajectory sequences obtained based on vehicle-mounted or visual perception. However, these technologies only focus on improving the trajectory model structure itself, without systematically integrating data fusion from vehicle-cloud collaboration, the periodic spatiotemporal characteristics of fixed routes, and the probabilistic representation of intersection queuing into the trajectory prediction process. Therefore, in the scenario of intersection queuing-trajectory coupling prediction for fixed-route buses, there are still shortcomings in adaptability and overall cognitive ability.

[0005] However, existing methods still mostly focus on individual vehicle trajectory learning or local traffic state estimation, making it difficult to simultaneously characterize the periodic patterns of route operation, the spatiotemporal propagation characteristics of intersection queuing, and the dynamic constraints of the road network. This fails to meet the needs of fixed-route buses for refined prediction in complex traffic environments. Summary of the Invention

[0006] To address the aforementioned issues, this invention proposes a method and system for recognizing weekly vehicle behavior based on the spatiotemporal characteristics of bus routes. It utilizes multi-source information, including vehicle status, road traffic flow, and signal timing, collected through vehicle-road-cloud collaboration. This information is processed using spatiotemporal synchronization and sliding window techniques to form a structured sequence. By integrating temporal context, route spatial topology, and station queuing characteristics, a dynamic spatiotemporal feature vector is constructed to characterize bus operation patterns. An online temporal model is used to predict intersection queue length, dissipation time, and the number of vehicles that can pass, generating prior traffic conditions. The prediction results, along with historical vehicle states, are input into a Transformer-based trajectory prediction network, outputting multiple future trajectories with probability distributions. This approach improves the accuracy and stability of queue prediction for fixed-route buses in complex urban traffic environments, thereby enhancing operational safety and traffic efficiency.

[0007] According to some embodiments, the first solution of the present invention provides a method for recognizing weekly vehicle behavior based on the spatiotemporal characteristics of bus routes, employing the following technical solution: A method for recognizing weekly vehicle behavior based on the spatiotemporal characteristics of bus routes, comprising: Acquire multi-source information data from vehicle-road-cloud fusion; Based on the acquired multi-source information data, the spatiotemporal features of the dynamic scene of the bus route are extracted, and the spatiotemporal scene feature vector of the bus route is constructed. Based on the constructed spatiotemporal scene feature vector of the bus route, predict the vehicle queuing status at the intersection; By combining the predicted vehicle queuing status at the intersection with the pre-set vehicle trajectory prediction model, a vehicle trajectory sequence is generated to complete the weekly vehicle behavior cognition based on the spatiotemporal characteristics of bus routes.

[0008] As a further technical limitation, the acquired multi-source information data from vehicle-road-cloud fusion is a structured spatiotemporal fusion dataset, which includes vehicle status data, roadside information, and cloud data; the vehicle status data includes speed. acceleration Energy consumption Braking state Gear position (g), positioning information (x, y), and heading angle The roadside information includes road traffic flow. Lane occupancy Signal timing period Phase status and remaining green time The cloud-based data includes road network geometry, road attributes, historical traffic flow data, and road segment saturation capacity. and signal control information.

[0009] As a further technical limitation, the extracted spatiotemporal features of the dynamic scene of the bus route include the time features of the route scene. Line scene spatial characteristics and route traffic conditions ;in, ; Based on the time period type, morning peak = 1; off-peak = 2; evening peak = 3. , This serves as a benchmark for the operational level during the same period, while also incorporating sine and cosine functions for the current time. Perform continuous encoding; , , The current road segment type. This represents the remaining distance to the nearest critical intersection downstream. This is a topological attribute representing the number of lanes in the current road segment. , and They are respectively , This refers to the historical statistics of road section capacity utilization. This represents the average operating speed of vehicles on the line within the current window.

[0010] Furthermore, based on the spatiotemporal scene features of the railway line, a unified multidimensional spatiotemporal feature vector at the railway line level is constructed through nonlinear fusion using a feature mapping model. ,Right now .

[0011] As a further technical constraint, a traffic flow saturation index is calculated based on the real-time traffic flow and saturation capacity of the intersection approach lanes. Combining the obtained traffic flow saturation index, different features in the spatiotemporal features are adaptively weighted using a spatiotemporal context weight vector to construct weighted input features. Based on the obtained weighted input features and an online recursive temporal network, the vehicle queuing status of the intersection is predicted, resulting in the probability distribution of vehicle queue length, vehicle queue clearance time, and number of passable vehicles at the intersection.

[0012] As a further technical limitation, the obtained intersection vehicle queuing state prediction results are encoded to obtain traffic state feature vectors; based on the obtained traffic state feature vectors and vehicle historical state sequences, vehicle state code sets and weekly vehicle state code sets are obtained; combined with the Transformer-based interactive model, vehicle trajectory deduction is performed under intersection queuing constraints to obtain multiple weekly vehicle probability trajectories and their confidence levels under intersection queuing constraints, generating vehicle trajectory sequences.

[0013] According to some embodiments, a second aspect of the present invention provides a weekly vehicle behavior cognition system based on the spatiotemporal characteristics of bus routes, employing the following technical solution: A weekly vehicle behavior cognition system based on the spatiotemporal characteristics of bus routes, comprising: The acquisition module is configured to acquire multi-source information data from vehicle-road-cloud fusion. The module is configured to extract the spatiotemporal features of the dynamic scene of the bus route based on the acquired multi-source information data, and construct the spatiotemporal scene feature vector of the bus route. The prediction module is configured to predict the vehicle queuing status at intersections based on the constructed spatiotemporal scene feature vector of the bus route. The cognition module is configured to combine the predicted vehicle queuing status at the intersection with a preset vehicle trajectory prediction model to generate a vehicle trajectory sequence, thereby completing the weekly vehicle behavior cognition based on the spatiotemporal characteristics of the bus route.

[0014] According to some embodiments, a third aspect of the present invention provides a computer-readable storage medium, employing the following technical solution: A computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the steps of the weekly vehicle behavior cognition method based on the spatiotemporal characteristics of bus routes as described in the first aspect of the present invention.

[0015] According to some embodiments, the fourth aspect of the present invention provides an electronic device, which adopts the following technical solution: An electronic device includes a memory, a processor, and a program stored in the memory and running on the processor. When the processor executes the program, it implements the steps in the weekly vehicle behavior cognition method based on the spatiotemporal characteristics of bus routes as described in the first aspect of the present invention.

[0016] According to some embodiments, the fifth aspect of the present invention provides a computer program product, which adopts the following technical solution: A computer program product includes software code, wherein the program in the software code performs the steps of the weekly vehicle behavior cognition method based on the spatiotemporal characteristics of bus routes as described in the first aspect of the present invention.

[0017] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention addresses the issues of packet loss and time delay in vehicle-side data under vehicle-road-cloud collaborative environments. By introducing a hierarchical completion mechanism based on sampling continuity detection, and combining vehicle kinematic constraints, roadside traffic conditions, and historical operating patterns, it adaptively fuses missing data, effectively ensuring the integrity and consistency of time-series data. This provides a stable data foundation for subsequent spatiotemporal modeling and prediction, and improves the robustness of prediction results.

[0018] This invention constructs a spatiotemporal scene modeling method for bus routes, which uniformly depicts the time patterns, spatial structure, and downstream intersection and station constraints of the routes. This enables the model to represent the traffic operation status from the overall perspective of the routes, thereby significantly improving the accuracy and stability of intersection queuing and surrounding vehicle trajectory prediction in complex scenarios.

[0019] This invention decomposes the prediction process into two levels: intersection queue prediction and surrounding vehicle trajectory prediction. It utilizes the spatiotemporal characteristics of the route and the queue prediction results to transmit information between model levels, thereby achieving hierarchical modeling and dynamic fusion of traffic conditions. This improves the model's adaptability to time-varying traffic conditions and its prediction robustness.

[0020] This invention can be adapted to different routes and intersection operating scenarios, providing reliable environmental awareness support for bus operation scheduling, priority traffic decisions and energy-saving control, which helps to improve operating efficiency and reduce energy consumption. Attached Figure Description

[0021] The accompanying drawings, which form part of this embodiment, are used to provide a further understanding of this embodiment. The illustrative embodiments and their descriptions are used to explain this embodiment and do not constitute an improper limitation of this embodiment.

[0022] Figure 1 This is a flowchart of the weekly vehicle behavior cognition method based on the spatiotemporal characteristics of bus routes in Embodiment 1 of the present invention; Figure 2 This is a schematic diagram illustrating the steps of the weekly vehicle behavior cognition method based on the spatiotemporal characteristics of bus routes in Embodiment 1 of the present invention; Figure 3 This is a schematic diagram of intersection queuing prediction in Embodiment 1 of the present invention; Figure 4 This is a schematic diagram illustrating the spatiotemporal correlation between bus route traffic flow and intersection queue length in Embodiment 1 of the present invention. Figure 5 This is a schematic diagram of the vehicle trajectory prediction in Embodiment 1 of the present invention; Figure 6 This is a structural block diagram of the weekly vehicle behavior cognition system based on the spatiotemporal characteristics of bus routes in Embodiment 2 of the present invention. Detailed Implementation

[0023] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0024] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0025] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0026] In this invention, terms such as "upper," "lower," "left," "right," "front," "back," "vertical," "horizontal," "side," and "bottom" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. These terms are used only to facilitate the description of the structural relationships of the various components or elements of this invention and do not specifically refer to any component or element in this invention. They should not be construed as limiting the invention.

[0027] In this invention, terms such as "fixed connection," "connected," and "linked" should be interpreted broadly, indicating a fixed connection, an integral connection, or a detachable connection; a direct connection or an indirect connection through an intermediate medium. Those skilled in the art can determine the specific meaning of these terms in this invention based on the specific circumstances, and they should not be construed as limitations on the invention.

[0028] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0029] Example 1 Embodiment 1 of this invention introduces a method for recognizing weekly vehicle behavior based on the spatiotemporal characteristics of bus routes.

[0030] like Figure 1 The method for recognizing weekly vehicle behavior based on the spatiotemporal characteristics of bus routes, as shown, includes: Acquire multi-source information data from vehicle-road-cloud fusion; Based on the acquired multi-source information data, the spatiotemporal features of the dynamic scene of the bus route are extracted, and the spatiotemporal scene feature vector of the bus route is constructed. Based on the constructed spatiotemporal scene feature vector of the bus route, predict the vehicle queuing status at the intersection; By combining the predicted vehicle queuing status at the intersection with the pre-set vehicle trajectory prediction model, a vehicle trajectory sequence is generated to complete the weekly vehicle behavior cognition based on the spatiotemporal characteristics of bus routes.

[0031] Below, in conjunction with Figure 2 Detailed description of the unfolding method in this embodiment: As one or more implementation methods, this embodiment is based on the collection of multi-source information data through vehicle-road-cloud fusion.

[0032] Based on the vehicle-road-cloud collaborative architecture, the vehicle terminal collects vehicle operating status parameters, including speed. acceleration Energy consumption Braking state Gear position (g), positioning information (x, y), heading angle Roadside traffic flow data collection Lane occupancy Signal timing period Phase status and remaining green time Cloud-based integration of road network geometry, road attributes, historical traffic flow data, and road segment saturation capacity. The system includes signal control information; the road network geometry provides road topology and lane centerlines, road attributes include the number of lanes, road segment length, and speed limits, and signal control information includes static information such as timing strategies. The saturation capacity of the road segment is determined by extracting the maximum traffic flow rate sequence during periods of smooth traffic flow in historical data when vehicle speeds consistently exceed 50 km / h, and taking the 95th percentile.

[0033] This embodiment performs unified preprocessing on the collected vehicle, road, and cloud data, including data cleaning, data completion, and preliminary structuring. Data cleaning removes GPS jump points, abnormal speed changes, and traffic noise through median filtering and moving average algorithms, and performs rule verification on abnormal signal state changes. In the data completion stage of this embodiment, to address potential packet loss and latency issues during the wireless transmission of vehicle-side data to the cloud, the first step is based on the sampling period. The arrival status of vehicle-side data is continuously monitored. If data with the corresponding timestamp is not received within multiple consecutive sampling periods or the data arrival delay exceeds a preset threshold, it is determined to be a communication anomaly and a hierarchical completion mechanism is triggered.

[0034] For short-term data loss, based on vehicle kinematic continuity constraints, historical speed and acceleration states are used to predict and complete the missing position and speed. For missing vehicle-end data over continuous time periods, traffic flow, lane occupancy, and signal phase information collected from the roadside are combined with historical trajectory patterns of the same route and time period stored in the cloud to inversely estimate the vehicle's operating state within that time period, thereby generating a virtual vehicle-end state sequence consistent with the actual traffic environment. For data with significant time delays, it is backfilled onto a unified timeline based on its original timestamp upon arrival, and consistency verification and weighted fusion are performed with the completed data. Subsequently, all data is organized into a preliminary time series form according to the collection timestamp.

[0035] This embodiment can achieve time alignment and spatial correction of vehicle, road and cloud data; time alignment uses a unified clock based on the network time protocol and performs linear interpolation on sampling points that cannot be aligned; spatial correction uses hidden Markov model map matching to project vehicle positioning onto the road centerline and verify speed and heading angle.

[0036] In this embodiment, the window length is set to... The sliding step size is Continuous observation segments are constructed and a unified time series is formed. Based on the above process, a structured spatiotemporal fusion dataset containing vehicle status, roadside traffic, signal control, and road network attributes is generated, and sampled according to a unified period. The output can be directly used for spatiotemporal feature modeling.

[0037] As one or more implementation methods, this embodiment performs spatiotemporal feature modeling of dynamic scenes of bus routes.

[0038] This embodiment uses a unified sampling period after spatiotemporal alignment and fusion. The structured spatiotemporal dataset (containing multi-source information such as vehicle operating status, roadside traffic flow, signal phase information, and route spatial location) is used as input. Based on historical traffic flow patterns, the route operation is divided into typical time periods of morning peak, off-peak, and evening peak. Within each time period, the route is expanded along the spatial dimension to identify potential operational constraints such as downstream intersections, stations, and bottleneck sections corresponding to different spatial locations on the route.

[0039] Data sequences based on a unified timeline, utilizing a sliding time window Construct continuous samples across multiple time periods and extract the following three types of features: (1) Temporal characteristics of the line scene Its mathematical expression is: ; in, Based on the time period type, morning peak = 1; off-peak = 2; evening peak = 3. , Simultaneous operational level indicators, while also introducing sine and cosine functions for the current moment. Perform continuous encoding.

[0040] (2) Spatial characteristics of the line scene Its mathematical expression is: ; in, , Current road segment type The remaining distance to the nearest critical intersection downstream. Topological attributes such as the number of lanes in the current road segment.

[0041] (3) Traffic conditions of the line Its mathematical expression is: ; in, and for , This refers to the historical statistics of road section capacity utilization. The average operating speed of vehicles on the line within the current window.

[0042] Based on the aforementioned spatiotemporal features of the railway lines, a unified multidimensional spatiotemporal feature vector at the railway line level is constructed through nonlinear fusion using a feature mapping model. ,Right now ; in, For macroscopic time context features, For a set of spatial topological features, It is a set of characteristics of the traffic operation status of the line.

[0043] This embodiment completes the construction of feature vectors, obtaining route-level spatiotemporal scene feature vectors that reflect the operational characteristics of bus routes at different times and spatial locations. .

[0044] As one or more implementation methods, this embodiment is based on intersection queuing prediction based on the spatiotemporal characteristics of the line.

[0045] This embodiment utilizes the constructed line-level multidimensional spatiotemporal feature vector The complex line operation environment is decomposed into several prediction sub-models under specific spatiotemporal scenarios to achieve high-precision queuing status prediction.

[0046] Dynamic spatiotemporal feature vectors output from previous steps Road saturation capacity Traffic flow Lane occupancy Traffic signal timing cycle For input, use Time context in and spatial topology The current moment is divided into several typical prediction scenario subsets through clustering or rule logic. ,like: Sub-model High congestion periods (morning and evening rush hours) + intersection approach lane constraints scenario.

[0047] Sub-model Off-peak hours + station queuing interference scenario.

[0048] Sub-model Transitional periods + traffic convergence scenarios at bottleneck sections.

[0049] Based on the above inputs, the traffic flow saturation index is calculated according to the real-time traffic flow and saturation capacity of the intersection approach lanes. ; Constructing weighted input features Through the spatiotemporal context weight vector Adaptive weighting of different features highlights key factors that significantly influence queue formation under different spatiotemporal conditions. in, , This represents element-wise multiplication. This represents the historical queue length.

[0050] The queuing prediction model employs an online recursive temporal network, with its core structure consisting of LSTM units. The model's input is the spatiotemporal feature vector output within a rolling time window. Internally constructed sequence For different spatiotemporal scenarios The model calls the corresponding parameter branch. ,Right now ; ; in, For the future Predicted queue length at any given time For model parameters, For the prediction loss function, This represents the actual queue length observation at a future time.

[0051] The model output includes a probability distribution of queue length. Queue dissipation time and the number of vehicles that can pass ; in, By predicting values The upper injection follows a mean of 0 and a variance of . Gaussian noise The probability distribution of queue length is obtained by sampling multiple times.

[0052] The queue will dissipate in time as follows: ; in, Let be the expected value of the queue length. For traffic flow rate, This is the empirical delay coefficient.

[0053] The number of vehicles that can pass is: ; in, Indicates the saturation capacity of the road (unit: vehicles / hour). The remaining time for the green light (in seconds). This indicates the maximum number of vehicles that can pass through during the current signal cycle.

[0054] As one or more implementation methods, this embodiment predicts the weekly vehicle trajectory based on intersection queuing information.

[0055] In this embodiment, the prediction object is the bus and the set of all surrounding vehicles that may interact with the bus in an intersection environment.

[0056] This embodiment uses, as follows Figure 3 The intersection queuing prediction result shown is the input, i.e. , , and the output such as Figure 4 The dynamic spatiotemporal feature vector shown .

[0057] The queuing prediction results are encoded into a traffic state feature vector, i.e.: ; in, and This represents the expected value and variance of the predicted queue length under constraints of time period and route spatial location. and These represent the dissipation time and the number of vehicles that can pass, calculated based on the queuing conditions at the intersection. Used to quantify traffic operation status and capacity at intersections.

[0058] Traffic state feature vector The fusion of historical state sequences of buses and other vehicles involves first extracting the historical dynamic features of the vehicles through a time-series coding network to obtain the state codes of the buses. and the status coding of the car Set; in which, and These represent the historical state sequences of passenger cars and weekly cars, respectively, and their specific components are as follows: ; in, Let be a set of the weekly car states, the composition of each element of which is... similar.

[0059] This embodiment will , and The input is a Transformer-based interactive module that simultaneously injects traffic state features into the queries of all participants. ,key ,value The matrix-guided model performs trajectory deduction under intersection queuing constraints, and its definition is as follows: , ; , guest ; , ; in, Code the historical status of the bus. For the set of historical state codes of the Zhouche, This represents vector concatenation. , , This is a learnable weight matrix.

[0060] In this embodiment, the attention weight matrix is: ; in, Let be the dimension constant of the key vector. Represents the normalized exponential function, Each element Representing the The degree of influence of weekly vehicle characteristics on bus trajectory.

[0061] The decoder outputs multiple probabilistic trajectories of the target bus and the surrounding vehicles at H future time points, i.e. ; in, It reflects multiple possible trajectories and their uncertainties in the form of a probability distribution. This is the characteristic matrix of the chariot.

[0062] Therefore, the output of this embodiment is as follows Figure 5 The diagram shows multiple probabilistic trajectories of vehicles in each cycle under queuing constraints at an intersection, along with their confidence levels.

[0063] This embodiment addresses the issues of packet loss and time delay in vehicle-side data under vehicle-road-cloud collaborative environments. By introducing a hierarchical completion mechanism based on sampling continuity detection, and combining vehicle kinematic constraints, roadside traffic conditions, and historical operating patterns, it adaptively fuses missing data, effectively ensuring the integrity and consistency of time-series data. This provides a stable data foundation for subsequent spatiotemporal modeling and prediction, and improves the robustness of prediction results.

[0064] This embodiment constructs a spatiotemporal scene modeling method for bus routes, which uniformly depicts the time patterns, spatial structure, and downstream intersection and station constraints of the routes. This enables the model to represent the traffic operation status from the overall perspective of the routes, thereby significantly improving the accuracy and stability of intersection queuing and surrounding vehicle trajectory prediction in complex scenarios.

[0065] This embodiment decomposes the prediction process into two levels: intersection queue prediction and surrounding vehicle trajectory prediction. It utilizes the spatiotemporal characteristics of the route and the queue prediction results to transmit information between model levels, thereby achieving hierarchical modeling and dynamic fusion of traffic conditions. This improves the model's adaptability to time-varying traffic conditions and its prediction robustness.

[0066] This embodiment can be adapted to different routes and intersection operating scenarios, providing reliable environmental awareness support for bus operation scheduling, priority traffic decisions and energy-saving control, which helps to improve operating efficiency and reduce energy consumption.

[0067] Example 2 Embodiment 2 of the present invention introduces a weekly vehicle behavior cognition system based on the spatiotemporal characteristics of bus routes.

[0068] like Figure 6 The illustrated system for recognizing bus behavior based on the spatiotemporal characteristics of bus routes includes: The acquisition module is configured to acquire multi-source information data from vehicle-road-cloud fusion. The module is configured to extract the spatiotemporal features of the dynamic scene of the bus route based on the acquired multi-source information data, and construct the spatiotemporal scene feature vector of the bus route. The prediction module is configured to predict the vehicle queuing status at intersections based on the constructed spatiotemporal scene feature vector of the bus route. The cognition module is configured to combine the predicted vehicle queuing status at the intersection with a preset vehicle trajectory prediction model to generate a vehicle trajectory sequence, thereby completing the weekly vehicle behavior cognition based on the spatiotemporal characteristics of the bus route.

[0069] The detailed steps are the same as those of the weekly vehicle behavior cognition method based on the spatiotemporal characteristics of bus routes provided in Example 1, and will not be repeated here.

[0070] Example 3 Embodiment 3 of the present invention provides a computer-readable storage medium.

[0071] A computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the steps of the weekly vehicle behavior cognition method based on the spatiotemporal characteristics of bus routes as described in Embodiment 1 of the present invention.

[0072] The detailed steps are the same as those of the weekly vehicle behavior cognition method based on the spatiotemporal characteristics of bus routes provided in Example 1, and will not be repeated here.

[0073] Example 4 Embodiment 4 of the present invention provides an electronic device.

[0074] An electronic device includes a memory, a processor, and a program stored in the memory and running on the processor. When the processor executes the program, it implements the steps in the weekly vehicle behavior cognition method based on the spatiotemporal characteristics of bus routes as described in Embodiment 1 of the present invention.

[0075] The detailed steps are the same as those of the weekly vehicle behavior cognition method based on the spatiotemporal characteristics of bus routes provided in Example 1, and will not be repeated here.

[0076] Example 5 Embodiment 5 of the present invention provides a computer program product.

[0077] A computer program product includes software code, wherein the program in the software code performs the steps of the weekly vehicle behavior cognition method based on the spatiotemporal characteristics of bus routes as described in Embodiment 1 of the present invention.

[0078] The detailed steps are the same as those of the weekly vehicle behavior cognition method based on the spatiotemporal characteristics of bus routes provided in Example 1, and will not be repeated here.

[0079] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of the present invention can be implemented using various computer languages, such as the object-oriented programming language Java and the interpreted scripting language JavaScript.

[0080] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0081] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0082] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0083] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0084] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

[0085] The above description is merely a preferred embodiment of this practice and is not intended to limit the scope of this practice. Various modifications and variations can be made to this practice by those skilled in the art. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of this practice should be included within the protection scope of this practice.

Claims

1. A method for recognizing weekly vehicle behavior based on the spatiotemporal characteristics of bus routes, characterized in that, include: Acquire multi-source information data from vehicle-road-cloud fusion; Based on the acquired multi-source information data, the spatiotemporal features of the dynamic scene of the bus route are extracted, and the spatiotemporal scene feature vector of the bus route is constructed. Based on the constructed spatiotemporal scene feature vector of the bus route, predict the vehicle queuing status at the intersection; By combining the predicted vehicle queuing status at the intersection with the pre-set vehicle trajectory prediction model, a vehicle trajectory sequence is generated to complete the weekly vehicle behavior cognition based on the spatiotemporal characteristics of bus routes.

2. The method for recognizing weekly vehicle behavior based on the spatiotemporal characteristics of bus routes as described in claim 1, characterized in that, The acquired multi-source information data from vehicle-road-cloud fusion is a structured spatiotemporal fusion dataset, which includes vehicle status data, roadside information, and cloud data; the vehicle status data includes speed. acceleration Energy consumption Braking state Gear position (g), positioning information (x, y), and heading angle The roadside information includes road traffic flow. Lane occupancy Signal timing period Phase status and remaining green time The cloud-based data includes road network geometry, road attributes, historical traffic flow data, and road segment saturation capacity. and signal control information.

3. The method for recognizing weekly vehicle behavior based on the spatiotemporal characteristics of bus routes as described in claim 1, characterized in that, The extracted spatiotemporal features of the dynamic scenes of bus routes include the time features of the route scenes. Line scene spatial characteristics and route traffic conditions ;in, ; Based on the time period type, morning peak = 1; off-peak = 2; evening peak = 3. , This serves as a benchmark for the operational level during the same period, while also incorporating sine and cosine functions for the current time. Perform continuous encoding; , , The current road segment type. This represents the remaining distance to the nearest critical intersection downstream. This is a topological attribute representing the number of lanes in the current road segment. , and They are respectively , This refers to the historical statistics of road section capacity utilization. This represents the average operating speed of vehicles on the line within the current window.

4. The method for recognizing weekly vehicle behavior based on the spatiotemporal characteristics of bus routes as described in claim 3, characterized in that, Based on the spatiotemporal features of the railway line, a unified multidimensional spatiotemporal feature vector at the railway line level is constructed through nonlinear fusion using a feature mapping model. ,Right now .

5. The method for recognizing weekly vehicle behavior based on the spatiotemporal characteristics of bus routes as described in claim 1, characterized in that, Based on the real-time traffic flow and saturation capacity of the intersection approach lanes, a traffic flow saturation index is calculated. Combining the obtained traffic flow saturation index, different features in the spatiotemporal features are adaptively weighted using a spatiotemporal context weight vector to construct a weighted input feature. Based on the obtained weighted input features and online recursive temporal network, the vehicle queuing status at the intersection is predicted, and the probability distribution of the vehicle queue length, the vehicle queuing time, and the number of vehicles that can pass through the intersection are obtained.

6. The method for recognizing weekly vehicle behavior based on the spatiotemporal characteristics of bus routes as described in claim 1, characterized in that, The obtained intersection vehicle queuing state prediction results are encoded to obtain traffic state feature vectors. Based on the obtained traffic state feature vectors and vehicle historical state sequences, vehicle state code sets and weekly vehicle state code sets are obtained. Combined with the Transformer-based interactive model, vehicle trajectories are extrapolated under intersection queuing constraints to obtain multiple weekly vehicle probability trajectories and their confidence levels under intersection queuing constraints, thus generating vehicle trajectory sequences.

7. A weekly vehicle behavior cognition system based on the spatiotemporal characteristics of bus routes, characterized in that, include: The acquisition module is configured to acquire multi-source information data from vehicle-road-cloud fusion. The module is configured to extract the spatiotemporal features of the dynamic scene of the bus route based on the acquired multi-source information data, and construct the spatiotemporal scene feature vector of the bus route. The prediction module is configured to predict the vehicle queuing status at intersections based on the constructed spatiotemporal scene feature vector of the bus route. The cognition module is configured to combine the predicted vehicle queuing status at the intersection with a preset vehicle trajectory prediction model to generate a vehicle trajectory sequence, thereby completing the weekly vehicle behavior cognition based on the spatiotemporal characteristics of the bus route.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of the weekly vehicle behavior cognition method based on the spatiotemporal characteristics of bus routes as described in any one of claims 1-6.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the program, it implements the steps of the weekly vehicle behavior cognition method based on the spatiotemporal characteristics of bus routes as described in any one of claims 1-6.

10. A computer program product, comprising software code, characterized in that, The program in the software code executes the steps of the weekly vehicle behavior cognition method based on the spatiotemporal characteristics of bus routes as described in any one of claims 1-6.

Images