A method, device and storage medium for identifying key control nodes of an intelligent connected vehicle platoon

By constructing dynamic parameter matrix and similarity matrix, performing spectral clustering, identifying key control nodes in intelligent connected vehicle queues, solving the problem of inaccurate identification in the prior art and improving the stability and safety of traffic flow .

CN117576905BActive Publication Date: 2024-11-26SOUTHEAST UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202311538852.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-11-17
Publication Date
2024-11-26
Estimated Expiration
2043-11-17

AI Technical Summary

Technical Problem

In an intelligent connected environment, it is difficult for the existing technology to accurately identify key control nodes in the vehicle queue, resulting in poor control effect, ignoring the driving style and dynamic parameter heterogeneity between vehicles, and being unable to realize the intelligent connected vehicle queue Optimization control.

Method used

By obtaining the vehicle's motion trajectory information, constructing a dynamic parameter matrix and a similarity matrix, establishing a graph Laplace matrix, performing spectral clustering, identifying key control nodes in the vehicle queue, and comprehensively considering the vehicle itself and relative motion parameters.

Benefits of technology

Accurate identification of key control nodes for intelligent connected vehicle queues is achieved, the stability and safety of traffic flow are improved, and the vehicle speed fluctuations are reduced, which is significantly better than other control methods.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117576905B_ABST
    Figure CN117576905B_ABST
Patent Text Reader

Abstract

The present invention discloses a method, device and storage medium for identifying key control nodes of an intelligent connected vehicle queue, and belongs to the field of intelligent traffic control technology. The method includes: obtaining the motion trajectory information of the intelligent connected vehicle queue, extracting the vehicle self-motion parameters and the relative motion parameters of the workshop of all vehicles in the queue; establishing a graph structure representation according to the distribution of intelligent connected vehicles, calculating the characteristic parameters according to the vehicle dynamics parameters and constructing a similarity matrix, and deriving a Laplace matrix based on the graph structure and the characteristic matrix of the intelligent connected vehicle queue; and based on the Laplace matrix, performing characteristic decomposition, determining the optimal number of clustering clusters, and then performing a spectral clustering operation of the intelligent connected vehicle queue; according to the clustering results, extracting the key vehicles in each clustering cluster, and outputting the key control node identification results of the intelligent connected vehicle queue to the intelligent connected vehicle queue. The present invention can provide a decision-making basis and guarantee for the safe and efficient control of the intelligent connected vehicle queue.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to a method, device and storage medium for identifying key control nodes of an intelligent networked vehicle queue, and belongs to the technical field of intelligent traffic control and vehicle-road cooperative automatic driving. Background Art

[0002] With the cross-integration of transportation, automobile industry, information and communication industries, the research and application of traffic management and control in the intelligent network environment has begun to accelerate. Supported by advanced perception, computing and communication technologies, intelligent networked vehicles can obtain dynamic and static information in traffic flow in real time and perform computational analysis to achieve real-time data-driven traffic management and control, overcoming the poor effect and long time of model-driven traffic management and control methods in traditional traffic flow, and becoming one of the key technologies to ensure traffic operation efficiency and safety in the future.

[0003] The development of intelligent networked transportation technology is a long-term accumulation process. It is still in the early stages of development. Technologies such as perception, computing and communication need to be further improved. There are still many technical problems to be solved to achieve a complete intelligent networked environment. For a long time, mixed traffic flows composed of intelligent networked vehicles and manually driven vehicles, as well as low-level intelligent networked environments, will be widely present in the transportation system.

[0004] In the operation of the traffic system, the implementation of real-time single-vehicle control for all vehicles is subject to technical limitations, which easily leads to the waste of computing and communication resources, and the stability and safety of control cannot be guaranteed, which lacks feasibility. In the intelligent network environment, a feasible method for real-time intelligent traffic control of traffic flow is to analyze the changing trend of motion parameters of the intelligent networked vehicle queue during its motion process, thereby identifying the potential vehicles in the vehicle queue that are prone to continuous fluctuations in motion state, and impose targeted control on the vehicles that introduce traffic interference to the traffic flow and cause the traffic operation efficiency or traffic safety to decrease, so as to achieve the coordinated optimization control of the intelligent networked vehicle queue. The key problem of this control method is to accurately identify the key control nodes in the vehicle queue. In existing studies, this problem is usually simplified, and the uniform distribution method is used to identify the key control nodes, or only a single factor (such as distance) is considered to complete the identification of the key control nodes, ignoring the different driving styles of different vehicles in the vehicle queue during operation and the heterogeneity of the changes in dynamic parameters, resulting in the identification of the key control nodes cannot match the operation of the vehicle queue, which is not conducive to the implementation of optimization control. Summary of the invention

[0005] The purpose of the present invention is to provide a method, device and storage medium for identifying key control nodes of an intelligent connected vehicle queue, comprehensively considering the vehicle's own motion parameters and the relative motion parameters of the vehicle in the intelligent connected vehicle queue during the movement process, identifying the key control nodes of the intelligent connected vehicle queue, and providing a basis for the optimized control of the intelligent connected vehicle queue.

[0006] In order to solve the above technical problems, the present invention is implemented by adopting the following technical solutions:

[0007] In one aspect, the present invention provides a method for identifying key control nodes of an intelligent connected vehicle fleet, comprising:

[0008] Acquire the motion trajectory information of the intelligent networked vehicle fleet, and process it to obtain the dynamic parameters of all vehicles in the fleet during the motion process, wherein the dynamic parameters include the vehicle's own motion parameters and the relative motion parameters of the vehicle;

[0009] Establish a graph structure representation of the queue based on the distribution of intelligent connected vehicles;

[0010] Constructing a characteristic parameter matrix and a similarity matrix of vehicle dynamics parameters according to the vehicle's own motion parameters and the vehicle-to-vehicle relative motion parameters;

[0011] Based on the graph structure representation, the characteristic parameter matrix and the similarity matrix, a graph Laplacian matrix of the intelligent connected vehicle platoon is constructed;

[0012] Based on the Laplace matrix of the intelligent connected vehicle fleet, eigendecomposition is performed, and a spectral clustering operation of the intelligent connected vehicle fleet is performed according to a preset optimal number of clustering clusters;

[0013] According to the clustering results, the key control vehicles in each cluster are extracted as the key control node identification results of the intelligent connected vehicle queue and output to the intelligent connected vehicle queue.

[0014] Optionally, among the dynamic parameters, the vehicle's own motion parameters include the position of the i-th vehicle at the t-th time speed and acceleration

[0015] The inter-vehicle relative motion parameters include the vehicle distance between the i-th vehicle and its preceding vehicle (i-1-th vehicle) at time t. Speed ​​Difference and acceleration difference

[0016] Optionally, the graph structure representation of the queue is established according to the distribution of the intelligent networked vehicles, including: taking each vehicle in the intelligent networked vehicles as a node, and constructing an adjacency matrix according to the communication connection relationship between the vehicles, which is expressed as:

[0017]

[0018] Among them, w ij represents the adjacent relationship between the node i where the i-th vehicle is located and the node j where the j-th vehicle is located. If there is a communication connection between node i and node j, then w ij =1, otherwise w ij =0;

[0019] Construct the degree matrix, expressed as: in, Indicates the number of other nodes that have a communication connection relationship with node i.

[0020] Optionally, the characteristic parameter matrix of the vehicle dynamics parameters is expressed as:

[0021]

[0022] Among them, p ij (k m ) represents the dynamic parameter k between the node i where the i-th vehicle is located and the node j where the j-th vehicle is located m The eigenvalue of k m represents the value of the mth kinetic parameter;

[0023] For the vehicle's own motion parameters, the characteristic values ​​of the dynamic parameters are calculated according to the following formula:

[0024]

[0025] For the relative motion parameters of the workshop, the characteristic values ​​of the dynamic parameters are calculated according to the following formula:

[0026]

[0027] The similarity matrix is ​​expressed as:

[0028]

[0029] in,

[0030]

[0031] s ij (k m ) represents the dynamic parameter k between the node i where the i-th vehicle is located and the node j where the j-th vehicle is located m The similarity, P i (k m ) represents the kinetic parameter k m The i-th row of the characteristic parameter matrix of , σ represents the Gaussian kernel function parameters.

[0032] Optionally, constructing a graph Laplacian matrix of the intelligent connected vehicle platoon based on the graph structure representation and the feature parameter matrix and the similarity matrix includes:

[0033] Based on the feature parameter matrix and the similarity matrix, a comprehensive similarity matrix is ​​constructed, which is expressed as:

[0034]

[0035] Among them, ω m (k) represents the characteristic parameter k m The weight of similarity can be set according to experience or needs. M represents the feature parameter k m The number of types;

[0036] According to the graph structure representation and the comprehensive similarity matrix, a graph Laplacian matrix L is constructed, which is expressed as:

[0037]

[0038] In the formula, f represents the feature vector, f i 、f j denote the eigenvectors represented by eigenvalue i and eigenvalue j respectively, N denotes the number of real eigenvalues ​​of the graph Laplacian matrix L, Represents the element in the i-th row and j-th column in the comprehensive similarity matrix.

[0039] Optionally, the Laplace matrix based on the intelligent connected vehicle fleet is used to perform eigendecomposition, and a spectral clustering operation of the intelligent connected vehicle fleet is performed according to a preset optimal number of clustering clusters, including:

[0040] Construct the standardized Laplace matrix D -1 / 2 LD 1 / 2 ;

[0041] Calculate D -1 / 2 LD 1 / 2 The eigenvectors f corresponding to the smallest l eigenvalues ​​λ in , and standardize the l eigenvectors f to form an l×N dimensional feature matrix T;

[0042] Divide the feature matrix T into N l-dimensional samples according to rows;

[0043] Using the clustering method, set the number of clusters to g∈[g min ,g max ], take each cluster number and perform clustering respectively, and get the clustering result corresponding to each cluster number Each clustering result CR i There are several clusters C iComposition, calculate the clustering index value of the clustering result corresponding to each cluster number;

[0044] Select the number of clusters g corresponding to the optimal clustering index value b And the corresponding clustering result CR b is the optimal number of clusters and the best clustering result.

[0045] Optionally, the above-mentioned calculation of the clustering index value of the clustering result corresponding to each clustering cluster number is specifically calculating the Davies-Bouldin index value of the clustering result corresponding to each clustering cluster number;

[0046] The best clustering result is the clustering result with the smallest Davies-Bouldin index value.

[0047] Optionally, extracting key control vehicles in each cluster according to the clustering result as the key control node identification result of the intelligent connected vehicle fleet includes:

[0048] Each cluster C in the clustering result b The lead car V i Key Control Vehicle Key Control Vehicle Set V key c ontrol vehicles =(V i );

[0049] The key control node identification results of the intelligent connected vehicle platoon are expressed as the key control node set where ν i Indicates vehicle V i The corresponding control node in the intelligent connected vehicle fleet.

[0050] In a second aspect, the present invention provides a key control node identification device for an intelligent connected vehicle fleet, comprising:

[0051] A motion trajectory information acquisition module is configured to acquire the motion trajectory information of the intelligent networked vehicle queue, and process the information to obtain the dynamic parameters of all vehicles in the queue during the motion process, wherein the dynamic parameters include the vehicle's own motion parameters and the relative motion parameters of the vehicle.

[0052] A graph structure representation building module is configured to build a graph structure representation of the queue according to the distribution of intelligent connected vehicles;

[0053] A feature similarity analysis module is configured to construct a feature parameter matrix and a similarity matrix of vehicle dynamics parameters according to the vehicle's own motion parameters and the vehicle-to-vehicle relative motion parameters;

[0054] A graph Laplacian matrix construction module is configured to construct a graph Laplacian matrix of the intelligent connected vehicle platoon based on the graph structure representation and the feature parameter matrix and the similarity matrix;

[0055] A feature clustering module is configured to perform feature decomposition based on the Laplace matrix of the intelligent connected vehicle fleet, and perform a spectral clustering operation on the intelligent connected vehicle fleet according to a preset optimal number of clustering clusters;

[0056] And, the key control node identification module is configured to extract the key control vehicles in each cluster according to the clustering results, as the key control node identification results of the intelligent connected vehicle queue, and output them to the intelligent connected vehicle queue.

[0057] In a third aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps of the method for identifying key control nodes of a fleet of intelligent connected vehicles as described in the first aspect are performed.

[0058] Beneficial Effects

[0059] Compared with the prior art, the present invention has the following beneficial effects:

[0060] (1) The key control node identification method of the present invention comprehensively considers the motion parameters of each vehicle in the queue and the relative motion parameters between adjacent vehicles during the operation of the intelligent connected vehicle queue, covering multiple factors that need to be met under the consistency goal of vehicle queue operation, and can accurately identify the key nodes that need to be controlled in the intelligent connected vehicle queue, providing a guarantee for improving traffic safety and operation efficiency;

[0061] Experiments have shown that applying control to the intelligent networked vehicle queue based on the key control node identification results of the present invention can reduce the fluctuation of vehicle speed in the intelligent networked vehicle queue and improve the stability of the vehicle queue. The control effect is significantly better than the control effect based on other key control nodes.

[0062] (2) Based on the fact that the present invention can effectively utilize the perception, computing, and communication technologies of intelligent connected vehicles and intelligent connected roads in an intelligent connected environment, the key control nodes of the intelligent connected vehicle queue can be identified, which provides an efficient method for the application of intelligent connected traffic management and control technology in the current traffic system;

[0063] (3) The present invention is of great significance for future research on traffic management and control methods in complex scenarios under intelligent network environments, and can provide a theoretical and methodological basis for identifying key control nodes in groups of intelligent networked vehicles. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1The figure is a flow chart of an embodiment of a method for identifying key control nodes of a smart connected vehicle platoon according to the present invention;

[0065] Figure 2 The figure is a schematic diagram of a detailed process of identifying key control nodes of a smart connected vehicle platoon in one embodiment of the present invention;

[0066] Figure 3 The figure shows a schematic diagram of the intelligent connected vehicle platoon simulation experiment scenario;

[0067] Figure 4 The figure shows the speed variation without applying control to the key control nodes;

[0068] Figure 5 The speed variation diagram under control of the key control node results based on the forward distribution method is shown;

[0069] Figure 6 Shown is the speed variation diagram under control of the key control node results based on the betweenness centrality method;

[0070] Figure 7 Shown is the speed variation diagram under control of the key control node results based on the control rank method;

[0071] Figure 8 Shown is the speed variation diagram under control of the key control node results based on the uniform distribution method;

[0072] Fig. 9 Shown is a speed variation diagram under control based on the key control node identification result of the present invention. DETAILED DESCRIPTION

[0073] The present invention is further described below with reference to the accompanying drawings and specific embodiments.

[0074] Technical concept

[0075] Before the realization of fully intelligent networked traffic management and control, the implementation of real-time single-vehicle control of all vehicles is subject to technical limitations. In order to solve the problem that it is difficult to achieve traffic flow optimization control through a full control method in the current intelligent networked environment, the present invention aims to propose a key control node identification method for an intelligent networked vehicle queue, taking into account the different driving styles of different vehicles in the intelligent networked vehicle queue during operation and the heterogeneity of dynamic parameter changes, overcoming the shortcomings of existing studies that use a uniform distribution method to identify key control nodes, or only consider a single factor (such as distance) to complete the identification of key control nodes and ignore driving heterogeneity. For the intelligent networked vehicle queue, the change trend of the motion parameters of the motion process is analyzed, thereby identifying the potential vehicles in the vehicle queue that are prone to continuous fluctuations in motion state, introducing traffic interference to the traffic flow, and causing a decrease in traffic operation efficiency or traffic safety, and applying targeted control to them, so as to achieve collaborative optimization control of the intelligent networked vehicle queue.

[0076] Example 1

[0077] refer to Figure 1 This embodiment introduces a method for identifying key control nodes of an intelligent connected vehicle fleet, which includes:

[0078] Acquire the motion trajectory information of the intelligent networked vehicle fleet, and process it to obtain the dynamic parameters of all vehicles in the fleet during the motion process, wherein the dynamic parameters include the vehicle's own motion parameters and the relative motion parameters of the vehicle;

[0079] Establish a graph structure representation of the queue based on the distribution of intelligent connected vehicles;

[0080] Constructing a characteristic parameter matrix and a similarity matrix of vehicle dynamics parameters according to the vehicle's own motion parameters and the vehicle-to-vehicle relative motion parameters;

[0081] Based on the graph structure representation, the characteristic parameter matrix and the similarity matrix, a graph Laplacian matrix of the intelligent connected vehicle platoon is constructed;

[0082] Based on the Laplace matrix of the intelligent connected vehicle fleet, eigendecomposition is performed, and a spectral clustering operation of the intelligent connected vehicle fleet is performed according to a preset optimal number of clustering clusters;

[0083] According to the clustering results, the key control vehicles in each cluster are extracted as the key control node identification results of the intelligent connected vehicle queue and output to the intelligent connected vehicle queue.

[0084] This embodiment considers the vehicle's own motion parameters and the relative motion parameters of the vehicle, and the identification of key control nodes in the intelligent connected vehicle queue helps to ensure the improvement of road traffic operation efficiency, safety and stability under intelligent connected transportation.

[0085] Example 2

[0086] refer to Figure 2 This embodiment specifically introduces the implementation process of a key control node identification method for an intelligent connected vehicle queue. The method of this embodiment can be implemented by an intelligent connected road service end or any terminal, which obtains the key control node identification result and outputs it to a node with control authority in the intelligent connected vehicle queue, or performs subsequent control strategy analysis and execution according to the key control node identification result.

[0087] The implementation process of the method specifically includes the following contents.

[0088] 1. Obtain the motion trajectory information of the intelligent connected vehicle queue, process the trajectory information and extract the dynamic parameters of all vehicles in the queue during the motion process.

[0089] In this embodiment, the extracted vehicle dynamics parameters include vehicle self-motion parameters and vehicle-to-vehicle relative motion parameters; wherein the vehicle self-motion parameters include the position of the i-th vehicle at the t-th time speed and acceleration The inter-vehicle relative motion parameters include the vehicle distance between the i-th vehicle and its preceding vehicle (i-1-th vehicle) at time t. Speed ​​Difference and acceleration difference

[0090] 2. Establish a graph structure representation of the queue based on the distribution of intelligent connected vehicles.

[0091] In this embodiment, the graph structure is represented by an adjacency matrix, which is expressed as:

[0092]

[0093] Among them, w ij represents the adjacent relationship between the node i where the i-th vehicle is located and the node j where the j-th vehicle is located. If there is a communication connection between node i and node j, then w ij =1, otherwise w ij =0;

[0094] The degree matrix is ​​constructed as:

[0095]

[0096] in,

[0097] 3. Based on the vehicle's own motion parameters and the workshop interaction motion parameters, the queue characteristic parameters are calculated and the similarity matrix is ​​constructed.

[0098] In this part, we first calculate the characteristic parameters of the dynamic parameters based on the vehicle dynamic parameters. Specifically, we use k m Represents the value of the mth kinetic parameter. For each kinetic parameter, a characteristic parameter matrix can be obtained:

[0099] Taking the vehicle speed and acceleration as examples, p ij (k m ) is calculated as:

[0100]

[0101] Taking the three vehicle relative motion parameters of vehicle spacing, speed difference, and acceleration difference as examples, p ij (k m ) is calculated as:

[0102]

[0103] Then the characteristic parameter matrix is ​​expressed as:

[0104]

[0105] Among them, p ij (k m ) represents the dynamic parameter k between the node i where the i-th vehicle is located and the node j where the j-th vehicle is located m For the vehicle's own motion parameters, when i≠j, p ij (k m )=0; For the relative motion parameters of the workshop, when i=j, p ij (k m )=0;

[0106] The similarity matrix based on the feature parameter matrix is ​​expressed as:

[0107]

[0108] Among them, s ij (k m ) represents the dynamic parameter k between the node i where the i-th vehicle is located and the node j where the j-th vehicle is located m The similarity of P i (k m ) represents the kinetic parameter k m The i-th row of the characteristic parameter matrix of , σ is the Gaussian kernel function parameter.

[0109] 4. Derive the Laplace matrix based on the intelligent connected vehicle platoon graph structure and characteristic matrix.

[0110] The comprehensive similarity matrix based on different feature similarity matrices is expressed as:

[0111]

[0112] where ω m (k) represents the characteristic parameter k m weight in terms of similarity;

[0113] The Laplace matrix of the graph of the intelligent connected vehicle platoon based on vehicle dynamics is derived as follows:

[0114]

[0115] 5. According to the Laplace matrix of the intelligent connected vehicle fleet, perform eigendecomposition, determine the optimal number of clusters, and perform spectral clustering operation of the intelligent connected vehicle fleet.

[0116] Based on the above figure Laplace matrix, construct the standardized Laplace matrix D -1 / 2 LD 1 / 2 ;

[0117] Calculate D -1 / 2 LD 1 / 2 The eigenvectors f corresponding to the smallest l eigenvalues ​​λ are normalized to form an l×N-dimensional feature matrix T;

[0118] The feature matrix T is divided into N l-dimensional samples according to the rows, and the clustering method is used to set the number of clusters to g∈[g min ,g max ], take each possible number of clusters for clustering, and get the clustering result corresponding to each number of clusters Each clustering result CR i There are several clusters C i Composition, calculate the clustering index value of the clustering result corresponding to each cluster number;

[0119] Select the number of clusters g corresponding to the optimal clustering index value b And the corresponding clustering result CR b is the optimal number of clusters and the best clustering result.

[0120] When selecting the best clustering number and the best clustering result, this embodiment calculates the Davies-Bouldin index value of the clustering result corresponding to each clustering number, and takes the clustering result with the smallest Davies-Bouldin index value as the best clustering result.

[0121] 6. According to the clustering results, extract the key vehicles in each cluster, and output the key control node identification results of the intelligent connected vehicle queue to the intelligent connected vehicle queue.

[0122] Specifically, this part is: identifying the best clustering result CR b Each cluster C b The lead car V i For key control vehicles, forming the key control vehicle set V key c ontrol vehicles =(V i );

[0123] Output the key control node set of intelligent connected vehicle platoon

[0124] In order to verify that the key control node identification method of the intelligent connected vehicle queue provided by the present invention can accurately identify the key control nodes in the intelligent connected vehicle queue and achieve excellent control effect, the following experiment is conducted to identify the key control nodes of the intelligent connected vehicle queue in the set simulation scenario and control the key control nodes to observe the control effect, specifically:

[0125] A set of vehicle platoon trajectory data obtained under real-world scenarios is selected. The vehicle platoon consists of 25 intelligent connected vehicles and is used as the initial state of the simulation experiment. A single-lane highway vehicle platoon following scenario is set as the experimental scenario. The simulation duration is 300s and the simulation time step is 0.1s. The basic schematic diagram of the experimental scenario is as follows: Figure 3 As shown. Based on the initial state of the vehicle queue trajectory data, the vehicle's own motion parameters and the relative motion parameters of the workshop during the vehicle movement process are extracted; according to the distribution of the intelligent networked vehicles, the graph structure representation of the queue is established, and the calculation of the queue characteristic parameters and the construction of the similarity matrix are completed according to the vehicle's own motion parameters and the workshop interactive motion parameters, and the Laplace matrix based on the graph structure and characteristic matrix of the intelligent networked vehicle queue is derived; according to the Laplace matrix of the intelligent networked vehicle queue, the characteristic decomposition is performed, the optimal number of clustering clusters is determined, and the spectral clustering operation of the intelligent networked vehicle queue is performed; according to the clustering results, the key vehicles in each cluster are extracted, and the key control node identification results of the intelligent networked vehicle queue are output. The control based on the PID controller is implemented for the vehicles at the output key nodes, and the control based on the following model is applied to other vehicles according to their motion characteristics. In order to verify that the key control node identification method for the intelligent networked vehicle queue provided by the present invention can accurately identify the key control nodes in the intelligent networked vehicle queue, the control experiments based on 4 commonly used key node identification methods are selected as comparisons, and the 4 commonly used key node identification methods are respectively the front distribution method, the betweenness centrality method, the control rank method and the uniform distribution method.

[0126] Figure 4 The speed variation without applying control to the key control nodes is given;

[0127] Figure 5 The speed variation under control of the key control node results based on the forward distribution method is given;

[0128] Figure 6 The speed variation under control is given based on the key control node results of the betweenness centrality method;

[0129] Figure 7 The speed variation under control is given based on the key control node results of the control rank method;

[0130] Figure 8 The speed variation under control of the key control node results based on the uniform distribution method is given;

[0131] Fig. 9 The speed variation under control based on the key control node results of this patent is given.

[0132] Depend on Figure 4-9 It can be seen that by using the key control node identification method of the present invention to control the intelligent connected vehicle queue, the fluctuation of vehicle speed in the intelligent connected vehicle queue can be reduced, and the stability of the vehicle queue can be improved. It is an efficient optimization control method. The control effect based on the key control node of this patent is better than the control effect based on other key control nodes, and the effect of improving the stability and efficiency of the intelligent connected vehicle queue is more obvious, which proves that the key control node identification method for the intelligent connected vehicle queue provided by the present invention can accurately identify the key control nodes in the intelligent connected vehicle queue.

[0133] The present invention selects a highway single-lane vehicle platoon following driving scenario for analysis, and other single-lane following driving scenarios may have the same key node identification method for following driving as this scenario.

[0134] Example 3

[0135] refer to Figure 2 This embodiment introduces a key control node identification device for an intelligent connected vehicle platoon, which includes:

[0136] A motion trajectory information acquisition module is configured to acquire the motion trajectory information of the intelligent networked vehicle queue, and process the information to obtain the dynamic parameters of all vehicles in the queue during the motion process, wherein the dynamic parameters include the vehicle's own motion parameters and the relative motion parameters of the vehicle.

[0137] A graph structure representation building module is configured to build a graph structure representation of the queue according to the distribution of intelligent connected vehicles;

[0138] A feature similarity analysis module is configured to construct a feature parameter matrix and a similarity matrix of vehicle dynamics parameters according to the vehicle's own motion parameters and the vehicle-to-vehicle relative motion parameters;

[0139] A graph Laplacian matrix construction module is configured to construct a graph Laplacian matrix of the intelligent connected vehicle platoon based on the graph structure representation and the feature parameter matrix and the similarity matrix;

[0140] A feature clustering module is configured to perform feature decomposition based on the Laplace matrix of the intelligent connected vehicle fleet, and perform a spectral clustering operation on the intelligent connected vehicle fleet according to a preset optimal number of clustering clusters;

[0141] And, the key control node identification module is configured to extract the key control vehicles in each cluster according to the clustering results, as the key control node identification results of the intelligent connected vehicle queue, and output them to the intelligent connected vehicle queue.

[0142] The specific functional implementation of the above functional modules refers to the relevant content in the method of Example 2 and will not be repeated here.

[0143] Example 4

[0144] This embodiment introduces a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the steps of the method described in Embodiment 1 or 2 are executed.

[0145] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may 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.

[0146] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0147] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0148] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0149] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for identifying key control nodes of an intelligent connected vehicle platoon, characterized in that: include: Obtaining motion trajectory information of a fleet of intelligent connected vehicles, and processing to obtain dynamic parameters of all vehicles in the fleet during motion, wherein the dynamic parameters include the vehicle's own motion parameters and the relative motion parameters of the vehicle; Establish a graph structure representation of the queue based on the distribution of intelligent connected vehicles; Constructing a characteristic parameter matrix and a similarity matrix of vehicle dynamics parameters according to the vehicle's own motion parameters and the vehicle-to-vehicle relative motion parameters; Based on the graph structure representation, the characteristic parameter matrix and the similarity matrix, a graph Laplacian matrix of the intelligent connected vehicle platoon is constructed; Based on the Laplace matrix of the intelligent connected vehicle fleet, eigendecomposition is performed, and a spectral clustering operation of the intelligent connected vehicle fleet is performed according to a preset optimal number of clustering clusters; According to the clustering results, the key control vehicles in each cluster are extracted as the key control node identification results of the intelligent connected vehicle queue, and output to the intelligent connected vehicle queue; in: Among the dynamic parameters, the vehicle's own motion parameters include The car is in Location at the moment ,speed and acceleration The relative motion parameters of the workshop include The vehicle and the vehicle in front of it are The distance between vehicles at the time , speed difference and acceleration difference ; The graph structure representation of the queue is established according to the distribution of the intelligent networked vehicles, including: taking each vehicle in the intelligent networked vehicles as a node, and constructing an adjacency matrix according to the communication connection relationship between the vehicles, which is expressed as: , in, , is the number of nodes, Representative The node where the vehicle is located With The node where the vehicle is located The adjacent relationship between nodes With Node If there is a communication connection between ,otherwise ; Construct the degree matrix, expressed as: ,in, , indicating that the node The number of other nodes with which communication connections exist; The characteristic parameter matrix of the vehicle dynamics parameters is expressed as: , in, Representative The node where the vehicle is located With The node where the vehicle is located The kinetic parameters between The characteristic value of represents the value of the mth kinetic parameter; For the vehicle's own motion parameters, the characteristic values ​​of the dynamic parameters are calculated according to the following formula: , For the relative motion parameters of the workshop, the characteristic values ​​of the dynamic parameters are calculated according to the following formula: , The similarity matrix is ​​expressed as: , in, , Representative The node where the vehicle is located With The node where the vehicle is located The kinetic parameters between The similarity of Representative kinetic parameters The characteristic parameter matrix of OK, Represents the Gaussian kernel function parameters; The method of constructing a graph Laplacian matrix of the intelligent connected vehicle platoon based on the graph structure representation and the characteristic parameter matrix and the similarity matrix includes: Based on the feature parameter matrix and the similarity matrix, a comprehensive similarity matrix is ​​constructed, which is expressed as: , in, Representative characteristic parameters The weight in terms of similarity, M Represents characteristic parameters The number of types; According to the graph structure representation and the comprehensive similarity matrix, a graph Laplacian matrix is ​​constructed , expressed as: , In the formula, represents the feature vector, , denote the eigenvectors represented by eigenvalue i and eigenvalue j respectively, N denotes the number of real eigenvalues ​​of the graph Laplacian matrix L, Represents the element in the i-th row and j-th column in the comprehensive similarity matrix.

2. The method according to claim 1, characterized in that: The Laplace matrix based on the intelligent networked vehicle queue is used to perform eigendecomposition, and a spectral clustering operation of the intelligent networked vehicle queue is performed according to a preset optimal clustering cluster number, including: Construct the normalized Laplacian matrix ; calculate The smallest Eigenvalues The corresponding eigenvector , and feature vector Standardization, forming a The characteristic matrix of ; The feature matrix Divide into rows indivual Dimensional sample; Using the clustering method, set the number of clusters to , take each cluster number and cluster them separately, and get the clustering result corresponding to each cluster number , each clustering result Several clusters Composition, calculate the clustering index value of the clustering result corresponding to each cluster number; Select the number of clusters corresponding to the optimal clustering index value And the corresponding clustering results is the optimal number of clusters and the best clustering result.

3. The method according to claim 2, characterized in that: The clustering index value of the clustering result corresponding to each clustering cluster number is calculated, specifically, the Davies-Bouldin index value of the clustering result corresponding to each clustering cluster number is calculated; The best clustering result is the clustering result with the smallest Davies-Bouldin index value.

4. The method according to claim 1, characterized in that: The step of extracting the key control vehicles in each cluster according to the clustering results as the key control node identification results of the intelligent connected vehicle fleet includes: Each cluster in the clustering result The lead car Key Control Vehicle Key Control Vehicle Set ; The key control node identification results of the intelligent connected vehicle platoon are expressed as the key control node set ,in Indicates vehicle The corresponding control node in the intelligent connected vehicle fleet.

5. A device for identifying key control nodes of an intelligent networked vehicle platoon using the method for identifying key control nodes of an intelligent networked vehicle platoon according to any one of claims 1 to 4, characterized in that: include: A motion trajectory information acquisition module is configured to acquire motion trajectory information of a fleet of intelligent networked vehicles, and process the information to obtain dynamic parameters of all vehicles in the fleet during motion, wherein the dynamic parameters include the vehicle's own motion parameters and the relative motion parameters of the vehicle; A graph structure representation building module is configured to build a graph structure representation of the queue according to the distribution of intelligent connected vehicles; A feature similarity analysis module is configured to construct a feature parameter matrix and a similarity matrix of vehicle dynamics parameters according to the vehicle's own motion parameters and the vehicle-to-vehicle relative motion parameters; A graph Laplacian matrix construction module is configured to construct a graph Laplacian matrix of the intelligent connected vehicle platoon based on the graph structure representation and the feature parameter matrix and the similarity matrix; A feature clustering module is configured to perform feature decomposition based on the Laplace matrix of the intelligent connected vehicle fleet, and perform a spectral clustering operation on the intelligent connected vehicle fleet according to a preset optimal number of clustering clusters; And, the key control node identification module is configured to extract the key control vehicles in each cluster according to the clustering results, as the key control node identification results of the intelligent connected vehicle queue, and output them to the intelligent connected vehicle queue.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the key control node identification method of the intelligent connected vehicle fleet as described in any one of claims 1 to 4 are performed.