Comprehensive traffic control method and system based on multi-task cooperation

Abstract

The invention relates to the technical field of traffic control, and provides a comprehensive traffic control method and system based on multi-task cooperation. The method comprises the following steps: processing traffic data through a time-space convolution kernel preset for different data modes to generate a state tensor so as to determine task viscosity parameters and sequence parameters between traffic control tasks, generating a task vector formed by the traffic control tasks and a self-adaptive constraint coefficient, constructing and solving a collaborative optimization function, and performing collaborative optimization on the task vector and the self-adaptive constraint coefficient. Determining control parameters for traffic control; and generating a to-be-executed control instruction set based on the control parameters, and distributing control instructions in the control instruction set to corresponding execution units. Through multi-source data fusion and spatio-temporal feature extraction, comprehensive perception of a traffic state is realized, a collaborative optimization model is constructed based on task association analysis, a multi-task target is dynamically balanced, global optimal control parameters are generated, the collaboration and control efficiency of traffic control are improved, the adaptability to a complex scene is enhanced, and the traffic control efficiency is improved. And intelligent and refined traffic control is realized.

Description

Technical Field

[0001] The present application relates to the field of traffic control technology, and more specifically, to a comprehensive traffic control method and system based on multi-task collaboration. Background Art

[0002] With the acceleration of urbanization and the continued growth of motor vehicle ownership, urban traffic congestion is becoming increasingly severe. Traditional single-task traffic control models are no longer able to meet the complex and ever-changing traffic needs. Existing technologies generally use real-time control to regulate traffic lights, but different traffic control tasks may have conflicting objectives. For example, signal optimization aims to minimize delays, while bus priority may require extended green light durations. As a result, traffic control is both less coordinated and less efficient. Summary of the Invention

[0003] The present application provides a comprehensive traffic control method and system based on multi-task collaboration, which can at least to some extent solve the problem of low coordination and efficiency of traffic control.

[0004] Other features and advantages of the present application will become apparent from the following detailed description, or may be learned in part by practice of the present application.

[0005] According to one aspect of the present application, a comprehensive traffic control method based on multi-task collaboration is provided, including: acquiring traffic data from multiple sources through intelligent sensors; processing the traffic data through spatiotemporal convolution kernels preset for different data modalities to generate a state tensor; determining the task viscosity parameters and sequence parameters between each traffic control task based on the state tensor, and generating a task vector composed of the traffic control tasks according to the sequence parameters of the traffic control tasks; constructing a collaborative optimization function based on the adaptive constraint coefficients generated by the task vectors and the task viscosity parameters, and determining the control parameters for traffic control by solving the collaborative optimization function; generating a control instruction set to be executed based on the control parameters, and distributing the control instructions in the control instruction set to corresponding execution units.

[0006] In the present application, based on the aforementioned scheme, the traffic data is processed by a spatiotemporal convolution kernel preset for different data modalities to generate a state tensor, including: obtaining a spatiotemporal convolution kernel preset for different data modalities; generating an original feature matrix corresponding to the traffic data at a set spatiotemporal position, performing convolution linear processing on the original feature matrix through the spatiotemporal convolution kernel to generate first data; processing the first data through an activation function to generate second data; extracting a feature coding matrix from the traffic data through principal component analysis, and generating the state tensor through an outer product operation between the feature coding matrix and the second data.

[0007] In the present application, based on the aforementioned solution, the generating of the original feature matrix corresponding to the traffic data at the set spatiotemporal position, performing convolution linear processing on the original feature matrix through the spatiotemporal convolution kernel to generate the first data includes: based on the traffic data of the controlled area, extracting a feature coding matrix from the traffic data through principal component analysis; performing convolution linear processing on the original feature matrix through the spatiotemporal convolution kernel to generate the first data as follows: ; Where i and N represent the type identifier and total number of types of traffic data respectively. represents the credibility weight of the i-th type of traffic data, represents the spatiotemporal convolution kernel of the i-th data source, Represents the original feature matrix of the i-th type of traffic data at the (x, y) position at time t, represents the bias parameter obtained through training, Represents the convolution operation.

[0008] In the present application, based on the above-mentioned solution, the feature coding matrix is ​​extracted from the traffic data by principal component analysis, and the state tensor is generated by the outer product operation between the feature coding matrix and the second data, including: extracting the feature coding matrix from the traffic data by principal component analysis ; Generate the state tensor by performing the outer product operation between the feature encoding matrix and the second data for: ;in, represents the activation function, Represents the outer product of tensors.

[0009] In the present application, based on the aforementioned scheme, the task viscosity parameters and sequence parameters between each traffic control task are determined based on the state tensor, and a task vector composed of the traffic control tasks is generated according to the sequence parameters of the traffic control task, including: determining the task viscosity parameters between each traffic control task based on the state tensor; determining the sequence parameters of the traffic control task based on the task viscosity parameters; and sorting and combining the traffic control tasks according to the sequence parameters of the traffic control task to generate a task vector.

[0010] In the present application, based on the aforementioned scheme, the adaptive constraint coefficient generated according to the task vector and the task viscosity parameter is used to construct a collaborative optimization function, and the control parameters for traffic control are determined by solving the collaborative optimization function, including: generating an adaptive constraint coefficient between tasks according to the task vector and the task viscosity parameter; generating constraint terms between tasks through dynamic causal reasoning based on traffic control task parameters; constructing a collaborative optimization function based on the adaptive constraint coefficient and constraint terms between tasks; and determining the control parameters for traffic control by solving the collaborative optimization function.

[0011] In the present application, based on the aforementioned scheme, the control instruction set to be executed is generated based on the control parameters, and the control instructions in the control instruction set are distributed to the corresponding execution units, including: generating the control instruction set to be executed based on the control parameters, and determining the execution units corresponding to the instructions in the control instruction set through a preset database; and distributing the control instructions in the control instruction set to the corresponding execution units.

[0012] According to one aspect of the present application, a comprehensive traffic control system based on multi-task collaboration is provided, comprising:

[0013] An acquisition unit, used to acquire traffic data from multiple sources through smart sensors;

[0014] A state unit, configured to process the traffic data using spatiotemporal convolution kernels preset for different data modalities to generate a state tensor;

[0015] A task unit, configured to determine a task stickiness parameter and a sequence parameter between the traffic control tasks based on the state tensor, and generate a task vector consisting of the traffic control tasks according to the sequence parameters of the traffic control tasks;

[0016] a parameter unit, configured to construct a collaborative optimization function based on the task vector and the adaptive constraint coefficient generated by the task viscosity parameter, and determine control parameters for traffic control by solving the collaborative optimization function;

[0017] The instruction unit is used to generate a control instruction set to be executed based on the control parameters, and distribute the control instructions in the control instruction set to corresponding execution units.

[0018] According to one aspect of the present application, a computer-readable medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the integrated traffic control method based on multi-task collaboration as described in the above embodiment is implemented.

[0019] According to one aspect of the present application, an electronic device is provided, comprising: one or more processors; a storage device for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the comprehensive traffic control method based on multi-task collaboration as described in the above embodiments.

[0020] According to one aspect of the present application, a computer program product or computer program is provided, comprising computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the integrated traffic control method based on multi-task collaboration provided in the various optional implementations described above.

[0021] The technical solution of this application obtains multi-source traffic data through intelligent sensors; processes the traffic data through preset spatiotemporal convolution kernels for different data modalities to generate a state tensor; based on the state tensor, determines the task viscosity parameters and sequence parameters between each traffic control task, and generates a task vector composed of the traffic control task according to the sequence parameters of the traffic control task; constructs a collaborative optimization function based on the adaptive constraint coefficients generated by the task vector and the task viscosity parameters, and determines the control parameters for traffic control by solving the collaborative optimization function; generates a set of control instructions to be executed based on the control parameters, and distributes the control instructions in the control instruction set to the corresponding execution units. Through multi-source data fusion and spatiotemporal feature extraction, a comprehensive perception of traffic status is achieved, a collaborative optimization model is constructed based on task association analysis, multi-task objectives are dynamically balanced, and global optimal control parameters are generated; finally, the instruction distribution mechanism is used to ensure the precise execution of control actions, improve the coordination and efficiency of traffic control, enhance adaptability to complex scenarios, and realize intelligent and refined traffic control.

[0022] 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

[0023] The accompanying drawings are incorporated into and constitute a part of the specification, illustrate 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 those skilled in the art can derive other drawings based on these drawings without inventive effort.

[0024] Figure 1A flowchart of a comprehensive traffic control method based on multi-task collaboration in one embodiment of the present application is schematically shown.

[0025] Figure 2 The flowchart of generating a state tensor in one embodiment of the present application is schematically shown.

[0026] Figure 3 A schematic diagram of an integrated traffic control system based on multi-task collaboration in one embodiment of the present application is shown schematically.

[0027] Figure 4 A schematic diagram of the structure of a computer system suitable for implementing an electronic device according to an embodiment of the present application is shown. DETAILED DESCRIPTION

[0028] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that this application will be thorough and complete and will fully convey the concepts of the example embodiments to those skilled in the art.

[0029] In addition, described feature, structure or characteristic can be combined in one or more embodiments in any suitable manner.In the following description, many specific details are provided so as to provide a full understanding of the embodiments of the present application. However, it will be appreciated by those skilled in the art that the technical scheme of the present application can be put into practice without one or more of the specific details, or other methods, components, devices, steps etc. can be adopted. In other cases, known methods, devices, implementations or operations are not shown or described in detail to avoid blurring the various aspects of the application.

[0030] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically separate entities. That is, these functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different networks and / or processor devices and / or microcontroller devices.

[0031] The flowcharts shown in the accompanying drawings are for illustrative purposes only and do not necessarily include all contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may be decomposed, while others may be combined or partially combined. Therefore, the actual execution order may vary depending on the actual situation.

[0032] The implementation details of the technical solution of this application are described in detail below:

[0033] Figure 1FIG2 shows a flow chart of a comprehensive traffic control method based on multi-task collaboration according to an embodiment of the present application. Figure 1 As shown, the integrated traffic control method based on multi-task collaboration includes at least steps S110 to S150, which are described in detail as follows:

[0034] S110, obtains traffic data from multiple sources through smart sensors.

[0035] In one embodiment of this application, a smart sensor array deployed in a traffic network collects multi-source, heterogeneous traffic data in real time. These sensors include, but are not limited to: roadside cameras for capturing vehicle trajectories and pedestrian movements; a combination of radar and lidar for object detection and speed measurement in three-dimensional space; geomagnetic sensors embedded in the road surface for sensing vehicle presence and passing events; and weather stations for monitoring environmental impacts on traffic flow.

[0036] Optionally, various sensors transmit raw data to edge computing nodes through wired or wireless communication links for preprocessing, including data cleaning, time synchronization and format standardization, to provide high-quality input for subsequent fusion analysis.

[0037] The multi-source data fusion mechanism creates a three-dimensional perception of traffic scenes through spatiotemporal alignment and feature correlation. For example, vehicle type information identified by cameras and speed data measured by radar are fused in a spatiotemporal coordinate system to form a complete trajectory of a single target. Vehicle arrival events detected by geomagnetic sensors are combined with signal cycle data to infer intersection queue lengths and delay indicators. This multimodal data complementation mechanism not only overcomes the limitations of a single data source but also enhances the reliability of detecting anomalies (such as traffic accidents and illegal parking) through cross-validation, providing real-time, accurate environmental perception input for traffic control systems.

[0038] The above process uses intelligent sensors to acquire multi-source traffic data, ensuring comprehensive awareness of traffic conditions and providing a solid data foundation for subsequent processing. The fusion of multi-source data improves perception accuracy and robustness, enabling it to cope with complex and changing traffic environments.

[0039] S120 , processing the traffic data using spatiotemporal convolution kernels preset for different data modalities to generate a state tensor.

[0040] In one embodiment of the present application, multi-source traffic data is deeply processed to generate a state tensor by using spatiotemporal convolution kernels preset for different data modalities. For example, for visual data captured by a camera, a spatiotemporal convolution kernel is used to simultaneously capture the spatial continuity and temporal correlation of vehicle trajectories; for radar point cloud data, a sparse convolution kernel is used to extract the spatiotemporal features of moving targets; and for geomagnetic sensor data, a one-dimensional temporal convolution kernel is used to analyze the arrival pattern of traffic. These dedicated convolution kernels extract local patterns in the spatiotemporal dimension through a sliding window mechanism, which are then aggregated into global features through a pooling layer, and finally fused into a four-dimensional tensor containing multi-dimensional traffic state information such as speed field, density field, and queue length, providing real-time and refined traffic scene representation for upper-level decision-making.

[0041] like Figure 2 As shown, in one embodiment of the present application, the traffic data is processed by using spatiotemporal convolution kernels preset for different data modalities to generate a state tensor, including:

[0042] S210, obtaining spatiotemporal convolution kernels preset for different data modalities;

[0043] S220, generating an original feature matrix corresponding to the traffic data at a set spatiotemporal position, and performing convolution linear processing on the original feature matrix using the spatiotemporal convolution kernel to generate first data;

[0044] S230, processing the first data through an activation function to generate second data;

[0045] S240 , extracting a feature coding matrix from the traffic data through principal component analysis, and generating the state tensor through an outer product operation between the feature coding matrix and the second data.

[0046] In one embodiment of this application, based on machine learning and deep learning, a spatiotemporal convolution kernel is designed for each data modality. The spatiotemporal convolution kernel can include spatial weights for adjusting spatial parameters and temporal weights for adjusting temporal parameters. For example, a 3D convolution kernel with a 2×2 spatial dimension and 3 frames of temporal dimension is used for positioning data to capture the continuity of vehicle trajectories. A causal convolution kernel is used for checkpoint data to prevent future data leaks.

[0047] In one embodiment of the present application, based on a set matrix format, traffic data is written into a matrix to generate an original feature matrix corresponding to the traffic data at a set spatiotemporal position. The original feature matrix is ​​convolved linearly using the spatiotemporal convolution kernel to generate first data.

[0048] Afterwards, the first data is processed by an activation function to generate second data;

[0049] Afterwards, a feature coding matrix is ​​extracted from the traffic data through principal component analysis, and a state tensor is generated through an outer product operation between the feature coding matrix and the second data.

[0050] Specifically, based on the above process, the state tensor is generated for:

[0051]

[0052] in, Represents the traffic state at a specific location (x, y), a specific time t, and a specific feature f; represents the activation function, i and N represent the type identification and total number of types of traffic data respectively, represents the credibility weight of the i-th type of traffic data, represents the spatiotemporal convolution kernel of the i-th data source, Represents the original feature matrix of the i-th type of traffic data at the (x, y) position at time t, represents the bias parameter obtained through training, represents the feature encoding matrix generated by the traffic area map, represents the convolution operation, Represents the outer product of tensors.

[0053] The above process processes traffic data using pre-defined spatiotemporal convolution kernels tailored to different data modalities to generate a state tensor. These kernels capture the spatiotemporal characteristics of traffic data, generating a refined, multi-dimensional state representation. The state tensor provides comprehensive state information for traffic control, facilitating a more accurate understanding of traffic dynamics.

[0054] S130 , determining task viscosity parameters and sequence parameters between the traffic control tasks based on the state tensor, and generating a task vector consisting of the traffic control tasks according to the sequence parameters of the traffic control tasks.

[0055] In one embodiment of the present application, based on the spatiotemporal evolution of traffic flow implied by the state tensor, a graph attention network dynamically calculates the viscosity parameters and sequence parameters between traffic control tasks. The task viscosity parameters reflect the strength of state coupling between tasks, such as the mutual influence between signal control and ramp regulation. The sequence parameters capture the timing dependencies of task execution through a temporal convolutional network, such as the need to prioritize accident response over conventional optimization. Finally, tasks are topologically sorted according to the sequence parameter weights to generate task vectors that reflect priority and correlation relationships, providing structured input for subsequent collaborative optimization.

[0056] In one embodiment of the present application, based on the state tensor, determining the task stickiness parameter and sequence parameter between each traffic control task, and generating a task vector consisting of the traffic control tasks according to the sequence parameters of the traffic control tasks, including:

[0057] Determining task stickiness parameters between traffic control tasks based on the state tensor;

[0058] Determining a sequence parameter of the traffic control task based on the task stickiness parameter;

[0059] The traffic control tasks are sorted and combined according to the sequence parameters of the traffic control tasks to generate a task vector.

[0060] In one embodiment of the present application, based on the state tensor of the four-dimensional space-time generated in the above steps, the current moment is extracted. The data slice T is used as input to determine the task stickiness parameter between traffic control tasks m and n. for:

[0061]

[0062] in, They represent the mth and nth traffic control task parameters, such as signal cycle, variable lane direction, etc. The larger the value, the more coupled the impact of tasks m and n on the traffic state is, and they need to be optimized collaboratively; the smaller the value, the tasks can be processed independently.

[0063] After calculating the task stickiness parameter, the task stickiness parameter can be thresholded. Traffic control tasks are regarded as strongly related task pairs, such as signal control and ramp control.

[0064] In one embodiment of the present application, the sequence parameter of the traffic control task is determined based on the task stickiness parameter. for:

[0065]

[0066] in, It represents the congestion coefficient, which is dynamically adjusted with the congestion index and ranges from [0.1, 5]. For example, =0.1 indicates a smooth flow state, =5 indicates a serious congestion state; k and n represent the identifiers of traffic control tasks, represents the task stickiness parameter between traffic control tasks k and n, Indicates the current benefit of the nth task, Indicates the increment, Indicates the potential benefit improvement from executing the nth task.

[0067] Optionally, the sequence parameters are recalculated every preset time period (e.g., 5 seconds). When a sudden congestion is detected, the current benefit of the mth task is When the congestion threshold is greater than the set one, the nonlinear enhancement function , ( =5), improve the sequence parameters of related tasks to achieve dynamic control.

[0068] The above process determines the task stickiness and sequence parameters between traffic control tasks based on the state tensor. The task stickiness parameter reflects the degree of correlation between tasks, while the sequence parameter determines the order in which tasks are executed. These two parameters enable dynamic adjustment of task priorities and correlations, ensuring the orderly execution of traffic control tasks.

[0069] In this embodiment, the task benefit increment and task stickiness parameter are combined through an exponential function, and the congestion coefficient is used to adjust the allocation strategy. The calculated sequence parameters of the traffic control tasks are used to indicate the priority or importance of the traffic control tasks. The gradient of the task parameters with respect to the state tensor is calculated to achieve a quantitative assessment of the task impact.

[0070] After determining the sequence parameters for each traffic control task, the tasks are sorted and combined in descending order based on the sequence parameters to generate a task vector. This approach adaptively adjusts the priority allocation strategy based on real-time traffic conditions. Compared to fixed-priority schemes, the dynamic priority mechanism shortens system response time, especially in emergency scenarios (such as traffic accidents). Dynamic priority significantly improves average speeds during peak hours.

[0071] S140 , constructing a collaborative optimization function based on the task vector and the adaptive constraint coefficient generated by the task viscosity parameter, and determining control parameters for traffic control by solving the collaborative optimization function.

[0072] In one embodiment of the present application, adaptive constraint weights are generated based on the priority sequence and association relationships inherent in the task vectors, combined with the task coupling strength represented by the task stickiness parameter. In this embodiment, the adaptive constraint weights can adjust the constraints between tasks in real time as traffic conditions evolve, thereby constructing a multi-task collaborative optimization function. Subsequently, an ant colony algorithm is employed to simulate the task scheduling process. While satisfying physical constraints such as signal timing and lane functionality, a global optimal solution is iteratively searched through pheromone volatilization and reinforcement mechanisms. Ultimately, a traffic control parameter combination that balances efficiency and fairness is output, achieving a dynamic balance among multiple objectives.

[0073] In one embodiment of the present application, a collaborative optimization function is constructed based on the adaptive constraint coefficient generated by the task vector and the task stickiness parameter, and control parameters for traffic control are determined by solving the collaborative optimization function, including:

[0074] generating an adaptive constraint coefficient between tasks according to the task vector and the task stickiness parameter;

[0075] Based on the traffic control task parameters, constraints between tasks are generated through dynamic causal reasoning;

[0076] Construct collaborative optimization functions based on adaptive constraint coefficients and constraint terms between tasks;

[0077] By solving the collaborative optimization function, control parameters for traffic control are determined.

[0078] In one embodiment of the present application, an adaptive constraint coefficient between tasks m and n is generated based on the task vector and the task stickiness parameter. for:

[0079]

[0080] in, Indicates the preset steep slope coefficient, which can be 10; represents the task stickiness parameter between the mth and nth tasks, The stickiness threshold can be set to 0.5. The adaptive constraint coefficient allows the system to automatically adjust the constraint strength when task relevance changes, reducing task switching latency.

[0081] In one embodiment of the present application, based on the traffic control task parameters, the constraint terms between tasks m and n are generated by dynamic causal reasoning. for:

[0082]

[0083] in, represents the loss function of the mth task, Indicates based on The expected value calculated is Denote the parameters of the mth and nth traffic control tasks, respectively. The constraints calculated in the above process are used to enforce causal consistency between tasks and avoid parameter conflicts.

[0084] In one embodiment of the present application, a collaborative optimization function is constructed based on the sequence parameters, the adaptive constraint coefficient and the constraint term. for:

[0085]

[0086] During the above operation, the sequence parameters As a weight, high priority tasks have a greater impact on the optimization goal. >τ, If it approaches 1, the constraint is activated; otherwise, if it approaches 0, the tasks are optimized independently.

[0087] In one embodiment of the present application, the control parameters for traffic control, i.e., traffic control task parameters, are determined by solving the collaborative optimization function. Specifically, in the solving process, the collaborative optimization function is first calculated. The gradient of the optimal solution is determined by linear programming, and the parameters are updated along the search direction to ensure the convergence of the optimization process within the feasible domain, and finally the control parameters for traffic control are obtained.

[0088] The above process generates adaptive constraint coefficients based on the task vector and task stickiness parameters, and constructs a collaborative optimization function. This comprehensively considers the objectives and constraints of multiple tasks, achieving dynamic balancing and optimization across them through the collaborative optimization function. The introduction of adaptive constraint coefficients enables the system to adjust the constraints between tasks based on varying traffic conditions, thereby generating more optimized control parameters.

[0089] S150: Generate a control instruction set to be executed based on the control parameters, and distribute the control instructions in the control instruction set to corresponding execution units.

[0090] In one embodiment of the present application, generating a control instruction set to be executed based on the control parameters, and distributing the control instructions in the control instruction set to corresponding execution units includes:

[0091] generating a control instruction set to be executed based on the control parameters, and determining, through a preset database, execution units corresponding to the instructions in the control instruction set;

[0092] Distribute the control instructions in the control instruction set to corresponding execution units.

[0093] In one embodiment of the present application, after generating control parameters, a pre-defined rule engine and logic conversion module generate an executable set of control instructions. This process first maps the numerical control parameters to specific traffic management actions, such as signal phase switching and variable lane direction adjustment. Subsequently, by querying a pre-defined configuration database, the physical execution unit corresponding to each instruction is determined, such as the traffic light at a specific intersection or the variable information sign at a road section, ensuring a precise association between the instructions and the hardware devices.

[0094] The generated control instruction set is distributed to the corresponding execution unit via the distributed message middleware. Instruction routing is performed based on the identification code of the execution unit, and a lightweight communication protocol is used to ensure real-time transmission.

[0095] Optionally, during the distribution process, instructions are serialized into a device-compatible format, and a checksum is embedded to ensure transmission integrity. After receiving the instructions, the execution unit needs to feedback a confirmation signal to form a closed-loop control.

[0096] The above process generates a set of control instructions to be executed based on the control parameters and distributes them to the corresponding execution units. The optimization results are converted into actual control actions, ensuring that traffic control instructions reach the execution units accurately and promptly. Through a pre-defined database and instruction distribution mechanism, the system ensures precise association of instructions with hardware devices, achieving closed-loop traffic control.

[0097] The technical solution of this application obtains multi-source traffic data through intelligent sensors; processes the traffic data through preset spatiotemporal convolution kernels for different data modalities to generate a state tensor; based on the state tensor, determines the task viscosity parameters and sequence parameters between each traffic control task, and generates a task vector composed of the traffic control task according to the sequence parameters of the traffic control task; constructs a collaborative optimization function based on the adaptive constraint coefficients generated by the task vector and the task viscosity parameters, and determines the control parameters for traffic control by solving the collaborative optimization function; generates a set of control instructions to be executed based on the control parameters, and distributes the control instructions in the control instruction set to the corresponding execution units. Through multi-source data fusion and spatiotemporal feature extraction, a comprehensive perception of traffic status is achieved, a collaborative optimization model is constructed based on task association analysis, multi-task objectives are dynamically balanced, and global optimal control parameters are generated; finally, the instruction distribution mechanism is used to ensure the precise execution of control actions, improve the coordination and efficiency of traffic control, enhance adaptability to complex scenarios, and realize intelligent and refined traffic control.

[0098] The following introduces an embodiment of the integrated traffic control system based on multi-task collaboration of the present application, which can be used to execute the integrated traffic control method based on multi-task collaboration in the above-mentioned embodiment of the present application. It can be understood that the integrated traffic control system based on multi-task collaboration can be a computer program (including program code) running on a computer device, for example, the integrated traffic control system based on multi-task collaboration is an application software; the integrated traffic control system based on multi-task collaboration can be used to execute the corresponding steps in the method provided in the embodiment of the present application. For details not disclosed in the embodiment of the integrated traffic control system based on multi-task collaboration of the present application, please refer to the embodiment of the integrated traffic control method based on multi-task collaboration mentioned above in the present application.

[0099] Figure 3 A block diagram of an integrated traffic control system based on multi-task collaboration according to an embodiment of the present application is shown.

[0100] Reference Figure 3 As shown, according to an embodiment of the present application, a comprehensive traffic control system based on multi-task collaboration includes:

[0101] An acquisition unit 310 is configured to acquire traffic data from multiple sources through smart sensors;

[0102] A state unit 320 is configured to process the traffic data using a spatiotemporal convolution kernel preset for different data modalities to generate a state tensor;

[0103] A task unit 330 is configured to determine a task stickiness parameter and a sequence parameter between the traffic control tasks based on the state tensor, and generate a task vector consisting of the traffic control tasks according to the sequence parameters of the traffic control tasks;

[0104] a parameter unit 340 for constructing a collaborative optimization function based on the task vector and the adaptive constraint coefficient generated by the task viscosity parameter, and determining control parameters for traffic control by solving the collaborative optimization function;

[0105] The instruction unit 350 is configured to generate a control instruction set to be executed based on the control parameters, and distribute the control instructions in the control instruction set to corresponding execution units.

[0106] In the present application, based on the aforementioned scheme, the traffic data is processed by a spatiotemporal convolution kernel preset for different data modalities to generate a state tensor, including: obtaining a spatiotemporal convolution kernel preset for different data modalities; generating an original feature matrix corresponding to the traffic data at a set spatiotemporal position, performing convolution linear processing on the original feature matrix through the spatiotemporal convolution kernel to generate first data; processing the first data through an activation function to generate second data; extracting a feature coding matrix from the traffic data through principal component analysis, and generating the state tensor through an outer product operation between the feature coding matrix and the second data.

[0107] In the present application, based on the aforementioned solution, the generating of the original feature matrix corresponding to the traffic data at the set spatiotemporal position, performing convolution linear processing on the original feature matrix through the spatiotemporal convolution kernel to generate the first data includes: based on the traffic data of the controlled area, extracting a feature coding matrix from the traffic data through principal component analysis; performing convolution linear processing on the original feature matrix through the spatiotemporal convolution kernel to generate the first data as follows: ; Where i and N represent the type identifier and total number of types of traffic data respectively. represents the credibility weight of the i-th type of traffic data, represents the spatiotemporal convolution kernel of the i-th data source, Represents the original feature matrix of the i-th type of traffic data at the (x, y) position at time t, represents the bias parameter obtained through training, Represents the convolution operation.

[0108] In the present application, based on the above-mentioned solution, the feature coding matrix is ​​extracted from the traffic data by principal component analysis, and the state tensor is generated by the outer product operation between the feature coding matrix and the second data, including: extracting the feature coding matrix from the traffic data by principal component analysis ; Generate the state tensor by performing the outer product operation between the feature encoding matrix and the second data for: ;in, represents the activation function, Represents the outer product of tensors.

[0109] In the present application, based on the aforementioned scheme, the task viscosity parameters and sequence parameters between each traffic control task are determined based on the state tensor, and a task vector composed of the traffic control tasks is generated according to the sequence parameters of the traffic control task, including: determining the task viscosity parameters between each traffic control task based on the state tensor; determining the sequence parameters of the traffic control task based on the task viscosity parameters; and sorting and combining the traffic control tasks according to the sequence parameters of the traffic control task to generate a task vector.

[0110] In the present application, based on the aforementioned scheme, the adaptive constraint coefficient generated according to the task vector and the task viscosity parameter is used to construct a collaborative optimization function, and the control parameters for traffic control are determined by solving the collaborative optimization function, including: generating an adaptive constraint coefficient between tasks according to the task vector and the task viscosity parameter; generating constraint terms between tasks through dynamic causal reasoning based on traffic control task parameters; constructing a collaborative optimization function based on the adaptive constraint coefficient and constraint terms between tasks; and determining the control parameters for traffic control by solving the collaborative optimization function.

[0111] In the present application, based on the aforementioned scheme, the control instruction set to be executed is generated based on the control parameters, and the control instructions in the control instruction set are distributed to the corresponding execution units, including: generating the control instruction set to be executed based on the control parameters, and determining the execution units corresponding to the instructions in the control instruction set through a preset database; and distributing the control instructions in the control instruction set to the corresponding execution units.

[0112] The technical solution of this application obtains multi-source traffic data through intelligent sensors; processes the traffic data through preset spatiotemporal convolution kernels for different data modalities to generate a state tensor; based on the state tensor, determines the task viscosity parameters and sequence parameters between each traffic control task, and generates a task vector composed of the traffic control task according to the sequence parameters of the traffic control task; constructs a collaborative optimization function based on the adaptive constraint coefficients generated by the task vector and the task viscosity parameters, and determines the control parameters for traffic control by solving the collaborative optimization function; generates a set of control instructions to be executed based on the control parameters, and distributes the control instructions in the control instruction set to the corresponding execution units. Through multi-source data fusion and spatiotemporal feature extraction, a comprehensive perception of traffic status is achieved, a collaborative optimization model is constructed based on task association analysis, multi-task objectives are dynamically balanced, and global optimal control parameters are generated; finally, the instruction distribution mechanism is used to ensure the precise execution of control actions, improve the coordination and efficiency of traffic control, enhance adaptability to complex scenarios, and realize intelligent and refined traffic control.

[0113] Figure 4 A schematic diagram of the structure of a computer system suitable for implementing an electronic device according to an embodiment of the present application is shown.

[0114] It should be noted that the computer system of the electronic device in this embodiment is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0115] In this embodiment, the computer system includes a central processing unit 401, which can perform various appropriate actions and processes based on programs stored in a read-only memory 402 or programs loaded from a storage unit 408 into a random access memory 403, such as executing the integrated traffic control method based on multi-task collaboration described in the above embodiment. The random access memory 403 also stores various programs and data required for system operation. The central processing unit 401, the read-only memory 402, and the random access memory 403 are connected to each other via a bus 404. An input / output interface 405 is also connected to the bus 404.

[0116] The following components are connected to the input / output interface 405: an input section 406 including a keyboard, mouse, and the like; an output section 407 including devices such as a cathode ray tube (CRT), a liquid crystal display (LCD), and speakers; a storage section 408 including devices such as a hard disk; and a communication section 409 including a network interface card such as a LAN (Local Area Network) card or a modem. The communication section 409 performs communication processing via a network such as the Internet. A drive 410 is also connected to the input / output interface 405 as needed. Removable media 411, such as a magnetic disk, an optical disk, a magneto-optical disk, or a semiconductor memory, is installed in the drive 410 as needed, so that computer programs read from the removable media can be installed in the storage section 408 as needed.

[0117] In particular, according to an embodiment of the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present application includes a computer program product that includes a computer program carried on a computer-readable medium, the computer program including a computer program for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network via the communication section 409 and / or installed from a removable medium 411. When the computer program is executed by the central processing unit 401, the various functions defined in the system of the present application are performed.

[0118] It should be noted that the computer-readable medium described in the embodiments of the present application may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to, an electrical connection having one or more conductors, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), flash memory, optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present application, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In the present application, a computer-readable signal medium may include a data signal transmitted in baseband or as part of a carrier wave, which carries a computer-readable computer program. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. A computer program embodied on a computer-readable medium may be transmitted using any suitable medium, including but not limited to wireless, wired, or any suitable combination thereof.

[0119] The flowcharts and block diagrams in the accompanying drawings illustrate the possible implementation architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present application. Among them, each box in the flowchart or block diagram can represent a module, program segment, or part of the code, and the above-mentioned module, program segment, or part of the code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flowchart, and the combination of boxes in the block diagram or flowchart, can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0120] The units involved in the embodiments described in this application may be implemented by software or hardware, and the units described may also be set in a processor. In some cases, the names of these units do not constitute limitations on the units themselves.

[0121] According to one aspect of the present application, a computer program product or computer program is provided, the computer program product or computer program including computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the methods provided in the various optional implementations described above.

[0122] As another aspect, the present application also provides a computer-readable medium, which may be included in the electronic device described in the above embodiments; or may exist independently without being incorporated into the electronic device. The computer-readable medium carries one or more programs, and when the one or more programs are executed by the electronic device, the electronic device implements the integrated traffic control method based on multi-task collaboration described in the above embodiments.

[0123] It should be noted that, although several modules or units of the device for action execution are mentioned in the above detailed description, this division is not mandatory. In fact, according to the embodiment of the application, the features and functions of two or more modules or units described above can be concretized in one module or unit. On the contrary, the features and functions of one module or unit described above can be further divided into multiple modules or units to be concretized.

[0124] Through the description of the above embodiments, it is easy for those skilled in the art to understand that the example embodiments described herein can be implemented by software or by combining software with necessary hardware. Therefore, the technical solution according to the embodiments of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, and includes a number of instructions to enable a computing device (which can be a personal computer, a server, a touch terminal, or a network device, etc.) to execute the method according to the embodiments of the present application.

[0125] Those skilled in the art will readily conceive of other embodiments of the present application after considering the specification and practicing the embodiments disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present application that follow the general principles of this application and include common knowledge or customary techniques in the art that are not disclosed herein.

[0126] It should be understood that the present application is not limited to the exact structures described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.

Claims

1. A comprehensive traffic control method based on multi-task collaboration, characterized in that: include: Acquire traffic data from multiple sources through smart sensors; Processing the traffic data using spatiotemporal convolution kernels preset for different data modalities to generate a state tensor; Determining task stickiness parameters and sequence parameters between the traffic control tasks based on the state tensor, and generating a task vector consisting of the traffic control tasks according to the sequence parameters of the traffic control tasks; constructing a collaborative optimization function based on the task vector and the adaptive constraint coefficient generated by the task stickiness parameter, and determining control parameters for traffic control by solving the collaborative optimization function; A control instruction set to be executed is generated based on the control parameters, and the control instructions in the control instruction set are distributed to corresponding execution units.

2. The integrated traffic control method based on multi-task collaboration according to claim 1 is characterized in that: The traffic data is processed using spatiotemporal convolution kernels preset for different data modalities to generate a state tensor, including: Get preset spatiotemporal convolution kernels for different data modalities; Generating an original feature matrix corresponding to the traffic data at a set spatiotemporal position, and performing convolution linear processing on the original feature matrix using the spatiotemporal convolution kernel to generate first data; Processing the first data through an activation function to generate second data; A feature coding matrix is ​​extracted from the traffic data through principal component analysis, and the state tensor is generated through an outer product operation between the feature coding matrix and the second data.

3. The integrated traffic control method based on multi-task collaboration according to claim 2 is characterized in that: Generating an original feature matrix corresponding to the traffic data at a set spatiotemporal position, and performing convolution linear processing on the original feature matrix using the spatiotemporal convolution kernel to generate first data, including: Based on the traffic data of the controlled area, extracting a feature coding matrix from the traffic data through principal component analysis; The original feature matrix is ​​convolved linearly using the spatiotemporal convolution kernel to generate the first data: ; Among them, i and N represent the type identification and total number of types of traffic data respectively. represents the credibility weight of the i-th type of traffic data, represents the spatiotemporal convolution kernel of the i-th data source, Represents the original feature matrix of the i-th type of traffic data at the (x, y) position at time t, represents the bias parameter obtained through training, Represents the convolution operation.

4. The integrated traffic control method based on multi-task collaboration according to claim 3 is characterized in that: Extracting a feature coding matrix from the traffic data through principal component analysis, and generating the state tensor through an outer product operation between the feature coding matrix and the second data, including: Extracting Feature Coding Matrix from Traffic Data via Principal Component Analysis ; Generate the state tensor by performing an outer product operation between the feature encoding matrix and the second data for: ； in, represents the activation function, represents the outer product of tensors, represents the feature encoding matrix.

5. The integrated traffic control method based on multi-task collaboration according to claim 1 is characterized in that: Determining task stickiness parameters and sequence parameters between the traffic control tasks based on the state tensor, and generating a task vector consisting of the traffic control tasks according to the sequence parameters of the traffic control tasks, including: Determining task stickiness parameters between traffic control tasks based on the state tensor; Determining a sequence parameter of the traffic control task based on the task stickiness parameter; The traffic control tasks are sorted and combined according to the sequence parameters of the traffic control tasks to generate a task vector.

6. The integrated traffic control method based on multi-task collaboration according to claim 1 is characterized in that: Constructing a collaborative optimization function based on the task vector and the adaptive constraint coefficient generated by the task stickiness parameter, and determining control parameters for traffic control by solving the collaborative optimization function, including: generating an adaptive constraint coefficient between tasks according to the task vector and the task stickiness parameter; Based on the traffic control task parameters, constraints between tasks are generated through dynamic causal reasoning; Construct collaborative optimization functions based on adaptive constraint coefficients and constraint terms between tasks; By solving the collaborative optimization function, control parameters for traffic control are determined.

7. The integrated traffic control method based on multi-task collaboration according to claim 1 is characterized in that: Generating a control instruction set to be executed based on the control parameters, and distributing the control instructions in the control instruction set to corresponding execution units, including: generating a control instruction set to be executed based on the control parameters, and determining, through a preset database, execution units corresponding to the instructions in the control instruction set; Distribute the control instructions in the control instruction set to corresponding execution units.

8. A comprehensive traffic control system based on multi-task collaboration, characterized in that: include: An acquisition unit, used to acquire traffic data from multiple sources through smart sensors; A state unit, configured to process the traffic data using spatiotemporal convolution kernels preset for different data modalities to generate a state tensor; A task unit, configured to determine a task stickiness parameter and a sequence parameter between the traffic control tasks based on the state tensor, and generate a task vector consisting of the traffic control tasks according to the sequence parameters of the traffic control tasks; a parameter unit, configured to construct a collaborative optimization function based on the task vector and the adaptive constraint coefficient generated by the task viscosity parameter, and determine control parameters for traffic control by solving the collaborative optimization function; The instruction unit is used to generate a control instruction set to be executed based on the control parameters, and distribute the control instructions in the control instruction set to corresponding execution units.

9. The integrated traffic control system based on multi-task collaboration according to claim 8 is characterized in that: The traffic data is processed using spatiotemporal convolution kernels preset for different data modalities to generate a state tensor, including: Get preset spatiotemporal convolution kernels for different data modalities; Generating an original feature matrix corresponding to the traffic data at a set spatiotemporal position, and performing convolution linear processing on the original feature matrix using the spatiotemporal convolution kernel to generate first data; Processing the first data through an activation function to generate second data; A feature coding matrix is ​​extracted from the traffic data through principal component analysis, and the state tensor is generated through an outer product operation between the feature coding matrix and the second data.

10. The integrated traffic control system based on multi-task collaboration according to claim 9 is characterized in that: Generating an original feature matrix corresponding to the traffic data at a set spatiotemporal position, and performing convolution linear processing on the original feature matrix using the spatiotemporal convolution kernel to generate first data, including: Based on the traffic data of the controlled area, extracting a feature coding matrix from the traffic data through principal component analysis; The original feature matrix is ​​convolved linearly using the spatiotemporal convolution kernel to generate the first data: ; Among them, i and N represent the type identification and total number of types of traffic data respectively. represents the credibility weight of the i-th type of traffic data, represents the spatiotemporal convolution kernel of the i-th data source, Represents the original feature matrix of the i-th type of traffic data at the (x, y) position at time t, represents the bias parameter obtained through training, Represents the convolution operation.

Images