Multi-unmanned vehicle longitudinal queue control privacy protection method based on bipartite graph

Abstract

The invention provides a privacy protection method for multi-unmanned vehicle longitudinal queue control based on a bipartite graph, and relates to the technical field of intelligent unmanned vehicle queue control. Comprising the following steps: constructing a directed graph comprising a leading unmanned vehicle and N-1 following unmanned vehicles by using a graph theory method; constructing a motion model of the leading unmanned vehicle and the following unmanned vehicle based on the directed graph; introducing a linear system theory, and converting the motion models of the leading unmanned vehicle and the following unmanned vehicle into linear system models; based on the directed graph, introducing a replica to determine a new topological structure and a node set; and based on a linear system model, determining a vehicle longitudinal queue control law according to the new topological structure and the node set so as to perform communication control. According to the invention, the problems of important information leakage in the vehicle longitudinal queue and excessive resource consumption in the information exchange process in the prior art are solved.

Description

Technical Field

[0001] This invention relates to the field of intelligent unmanned vehicle queue control technology, and in particular to a privacy protection method for multi-unmanned vehicle longitudinal queue control based on bipartite graphs. Background Technology

[0002] With the continuous development of intelligent vehicle systems, vehicle longitudinal queuing control has gradually become a research hotspot. As an efficient cooperative driving mode, vehicle longitudinal queuing control has received widespread attention in recent years, and related research results are becoming increasingly abundant. This control method can not only effectively reduce energy consumption and alleviate traffic congestion, but also improve driving safety, thus attracting much attention.

[0003] Vehicle platooning systems typically require vehicles to share real-time location, speed, and direction of travel data to ensure platoon coordination and security. However, this data exchange requirement also introduces privacy risks into vehicle-to-vehicle communication. If this sensitive information is illegally intercepted or leaked during transmission, it could pose a serious threat to individual privacy, citizen safety, and even public safety. Summary of the Invention

[0004] To overcome the shortcomings of existing technologies, the purpose of this invention is to provide a privacy protection method for longitudinal queue control of multiple unmanned vehicles based on bipartite graphs. This invention solves the problems of important information leakage and excessive resource consumption during information exchange in the longitudinal queue of vehicles in existing technologies.

[0005] To achieve the above objectives, the present invention provides the following solution:

[0006] A privacy-preserving method for longitudinal queuing control of multiple unmanned vehicles based on bipartite graphs includes:

[0007] Construct a directed graph containing a leader autonomous vehicle and N-1 following autonomous vehicles using graph theory methods;

[0008] Motion models for the leading unmanned vehicle and the following unmanned vehicle are constructed based on the directed graph;

[0009] By introducing linear system theory, the motion models of the leading unmanned vehicle and the following unmanned vehicle are transformed into linear system models;

[0010] Based on the directed graph, a replica is introduced to determine a new topology and node set;

[0011] Based on a linear system model, a vehicle longitudinal queuing control law is determined according to the new topology and node set for communication control.

[0012] Preferably, it further includes:

[0013] Determine the location and speed information of each vehicle;

[0014] Based on the location and speed information, determine whether the triggering conditions are met. If so, trigger the event and obtain the corresponding vehicle longitudinal queue control law. If not, use the current vehicle longitudinal queue control law for communication control.

[0015] Preferably, the expression for the directed graph of the lead autonomous vehicle and N-1 following autonomous vehicles is:

[0016] ;

[0017] in, It is a collection of autonomous vehicle nodes. ; It represents the set of connecting edges; Represents the neighbor weight matrix, Represents vehicle node Its first Neighbor vehicle nodes Weighting coefficients for communication between them.

[0018] Preferably, the motion model expressions for the lead unmanned vehicle and the following unmanned vehicle are as follows:

[0019] ;

[0020] ;

[0021] Where, p i Indicates the position of driverless car i. Indicates the speed of the driverless car i. This represents the acceleration of driverless car i. This represents the position of the leading driverless car. This indicates that the driverless car is at 0 speed. This represents the acceleration of the driverless car. This represents the inertial lag in longitudinal vehicle dynamics. This indicates the desired acceleration control input that needs to be designed.

[0022] Preferably, the expression for the linear system model is:

[0023] ;

[0024] in, , For state vectors, K is the control gain, A is the first matrix, and B is the second matrix, where...

[0025] .

[0026] Preferably, the expressions for the new topology and the node set are as follows:

[0027] ;

[0028] ;

[0029] in, For the new topology. For the new set of nodes, It represents the set of new edges.

[0030] Preferably, the expression for the vehicle longitudinal queuing control law is:

[0031] ;

[0032] in, This is the vehicle longitudinal queuing control law. For nodes The set of neighboring nodes, The expression is:

[0033]

[0034] in Represents virtual vehicles To real vehicles distance, Represents real vehicles To virtual vehicles The distance.

[0035] Preferably, when the event is triggered, the expression for the corresponding vehicle longitudinal queuing control law is obtained as follows:

[0036] ;

[0037] in, Indicates the first The triggering time sequence of the k-th autonomous vehicle When an event is triggered, the corresponding vehicle longitudinal queue control law is obtained.

[0038] The present invention discloses the following technical effects:

[0039] This invention provides a privacy protection method for longitudinal queue control of multiple unmanned vehicles based on bipartite graphs, comprising: constructing a directed graph containing one leader unmanned vehicle and N-1 following unmanned vehicles using graph theory; constructing motion models of the leader and following unmanned vehicles based on the directed graph; introducing linear system theory to transform the motion models of the leader and following unmanned vehicles into linear system models; introducing a replica to determine a new topology and node set based on the directed graph; and determining the longitudinal queue control law for communication control based on the new topology and node set according to the linear system model. Compared to other privacy protection methods, the bipartite graph-based privacy protection method of this invention is simpler and easier to understand, facilitating the implementation of vehicle privacy protection in vehicle queue control. The distance information for longitudinal queue control under the bipartite graph-based privacy protection method is sent by virtual vehicle information we set, further protecting the privacy of vehicle safety information. By introducing a dynamic event triggering mechanism, vehicle queue control is achieved under the premise of privacy protection. This design not only ensures that vehicles can successfully achieve queue consensus without leaking initial information but also effectively reduces resource consumption. Attached Figure Description

[0040] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0041] Figure 1 A flowchart illustrating a privacy protection method for longitudinal queue control of multiple unmanned vehicles based on a bipartite graph, provided in an embodiment of the present invention;

[0042] Figure 2 This is a structural diagram of a conventional vehicle queue provided in an embodiment of the present invention;

[0043] Figure 3 This is a structural diagram of a vehicle queue based on bipartite graph privacy protection provided in an embodiment of the present invention;

[0044] Figure 4 is a schematic diagram of the information changes of four vehicles under privacy protection and dynamic event triggering provided in the embodiment of the present invention. In Figure 4(a), it is a schematic diagram of position change, Figure 4(b) is a schematic diagram of speed change, and Figure 4(c) is a schematic diagram of acceleration change.

[0045] Figure 5 is a schematic diagram of the time change of the vehicle under dynamic event triggering according to the embodiment of the present invention. In Figure 5(a), the vehicle triggering time period under dynamic event triggering is shown; Figure 5(b) shows the vehicle control under dynamic event triggering. Figure 5(c) shows the changes over time, where the variables in the vehicle event trigger function are displayed. A diagram illustrating how it changes over time. Detailed Implementation

[0046] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0047] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0048] like Figure 1 As shown, this invention provides a privacy protection method for longitudinal queue control of multiple unmanned vehicles based on bipartite graphs, including:

[0049] Step 100: Construct a directed graph containing a leader autonomous vehicle and N-1 following autonomous vehicles using graph theory methods;

[0050] Step 200: Construct motion models for the leading unmanned vehicle and the following unmanned vehicle based on the directed graph;

[0051] Step 300: Introduce linear system theory to convert the motion models of the leading unmanned vehicle and the following unmanned vehicle into linear system models;

[0052] Step 400: Based on the directed graph, introduce a replica to determine a new topology and node set;

[0053] Step 500: Based on the linear system model, determine the vehicle longitudinal queuing control law according to the new topology and node set for communication control.

[0054] Furthermore, it also includes:

[0055] Determine the location and speed information of each vehicle;

[0056] Based on the location and speed information, determine whether the triggering conditions are met. If so, trigger the event and obtain the corresponding vehicle longitudinal queue control law. If not, use the current vehicle longitudinal queue control law for communication control.

[0057] Furthermore, such as Figure 2 As shown in Figure 4, a directed communication topology for the longitudinal vehicle queue is established using graph theory methods, and the leader of the autonomous vehicle in the longitudinal queue is determined:

[0058] Consider a scenario that includes a lead autonomous vehicle and The topology formed by the longitudinal convoy of vehicles following an autonomous vehicle, and the communication structure between these vehicles, can be represented by a directed graph, namely:

[0059] ;

[0060] in, It is a collection of autonomous vehicle nodes. ; It represents the set of connecting edges; Represents the neighbor weight matrix, Represents vehicle node Its first Neighbor vehicle nodes The weighting coefficients for communication between them; if each following vehicle is considered a node, then... Represents Received Information, that is Otherwise it is Hypothetical directed graph There exists a directed spanning tree, that is, a node. If it can reach every other node through a directed path, then we put the node... This is called the root node. We define the set of neighboring nodes of node i as... , .make , , This represents the Laplace matrix.

[0061] If there exists a vehicle node in a directed communication topology such that there are paths from that vehicle node to all other vehicle nodes, then the directed graph communication topology G is said to contain a spanning tree, and that vehicle node is called the root node of the spanning tree.

[0062] In this embodiment, the root node of the vehicle is the vehicle leader, i.e., the vehicle leader node; the turntable that leads the vehicles in this queue is used for... express.

[0063] Other vehicle nodes in the vehicle cluster are divided into leader vehicle nodes or follower vehicle nodes. Leader vehicle nodes can obtain information directly from the queue center, while follower vehicle nodes can only obtain information from the leader vehicle nodes and other follower vehicle nodes.

[0064] The expression for establishing the motion model of the vehicle platoon is:

[0065] ;

[0066] ;

[0067] Where, p i Indicates the position of driverless car i. Indicates the speed of the driverless car i. This represents the acceleration of driverless car i. This represents the position of the leading driverless car. This indicates that the driverless car is at 0 speed. This represents the acceleration of the driverless car. This represents the inertial lag in longitudinal vehicle dynamics. This indicates the desired acceleration control input that needs to be designed.

[0068] Based on the vehicle queuing system model established in step two, transform it into the following linear system (linear system model) form:

[0069] ;

[0070] in, , For state vectors, K is the control gain, A is the first matrix, and B is the second matrix, where...

[0071] .

[0072] It is not difficult to verify that the matrix here , It is controllable.

[0073] Furthermore, the invented privacy-preserving initial value method based on bipartite graphs is applied to vehicle longitudinal queuing control: such as... Figure 3 In such a topology, each vehicle Each has a copy of it. Furthermore, their initial values ​​are different, and the replica's state is propagated through communication. The vehicle updates its controller based on data received from neighboring vehicle nodes, while the replica's state is only affected by the vehicle's... The inherent influence of this property. Under these conditions, we establish a new topology. New set of nodes , This represents the set of new edges, and This indicates a new topology. Next node It is a node The neighboring nodes, among which In the newly expanded diagram It contains 2N vehicle nodes. In this topology and structure, node 1 to node... Represents real vehicle nodes, nodes arrive Each of these represents a copy of the others. The resulting graph is called a bipartite graph, where... Since it reflects the link states between nodes, under such privacy protection, this bipartite graph has the following property:

[0074] ;

[0075] ;

[0076] ;

[0077] like Figure 3 As shown, this graph is an example of a privacy-preserving method based on bipartite graphs to protect the initial values ​​of vehicles. The original directed graph... like Figure 3 As shown in the upper part, the node set is ={1,2,3,4}. For For each real node Introduce a virtual node This way each node Its corresponding virtual node A pair is formed. By adding four virtual nodes, a new directed graph is created. like Figure 3 As shown in the lower half, the new set of nodes is ={1,2,3,4,5,6,7,8}. Real nodes. It can be associated with its corresponding virtual node It can interact and receive information from virtual nodes in the virtual layer. Therefore, the state information of the actual nodes is hidden from neighboring node pairs, ensuring the initial value privacy of the autonomous vehicle system is protected. Under this privacy protection method, the Laplace matrix is ​​as follows:

[0078] ;

[0079] in, A column vector is a column vector whose elements are all zero. This reflects the topological relationship between the leader node and other nodes. It is a non-singular M-matrix.

[0080] like Figure 3 As shown, the control law is a directed graph structure with a lead vehicle, dividing the queue into: leader unmanned vehicle nodes and follower unmanned vehicle nodes. To achieve the longitudinal queue goal, i.e. ; ( (where r and It is a constant greater than 0.

[0081] To further protect vehicle privacy, we will [do something related to] adjacent real vehicles. , Distance between Divided into virtual vehicles With real vehicles distance and virtual vehicles Compared to the real vehicles in front distance The sum of the distances of the two parts, i.e. The control laws we obtain are divided into control laws for the real layer and control laws for the virtual layer:

[0082] To simplify the control law, we define:

[0083]

[0084] in Represents virtual vehicles To real vehicles distance, Represents real vehicles To virtual vehicles The distance.

[0085] The control rules of the virtual layer and the real layer can then be combined to form new control rules, as shown below:

[0086] ;

[0087] It is worth noting that the distance deviation between the cloned vehicle node and its real vehicle node in this invention can be set individually by the designer for each vehicle, thereby further ensuring the security of vehicle privacy.

[0088] Applying dynamic event-triggered design to the longitudinal vehicle queuing, a new event-triggered control law is designed:

[0089] To reduce the consumption of communication resources between vehicles, a dynamic event triggering mechanism is introduced below. For example... Figure 2 As shown in the vehicle structure diagram, each vehicle sends its own information, such as position and speed, to the vehicle communication network. The platform collects and processes this information and then sends it to other vehicles to determine if triggering conditions are met. If the conditions are met, an event is triggered and the controller is notified; if the conditions are not met, the information is directly transmitted to the controller to ensure that vehicles maintain a safe distance and drive in an orderly manner. The event triggering mechanism is introduced below:

[0090] ;

[0091] Indicates the first The first trigger time sequence of an autonomous vehicle. Indicates the first The (k+1)th trigger time sequence of an autonomous vehicle, where the event trigger function... As shown below:

[0092] ;

[0093] Matrix in trigger function satisfy ,matrix The variables below will satisfy certain conditions. The following conditions must be met:

[0094] ;

[0095] parameter It is a constant greater than zero.

[0096] Furthermore, according to the vehicle longitudinal queuing control law, it becomes as follows:

[0097] When the event is triggered, the expression for the corresponding vehicle longitudinal queuing control law is obtained as follows:

[0098] ;

[0099] in Represents the time point when the controller is updated. This represents the time point of the next controller update.

[0100] Define measurement error and tracking error as follows: , in:

[0101] ;

[0102] ;

[0103] Based on the control rate and error term of the above design, as long as the vehicle longitudinal queuing system meets the following conditions, vehicle privacy can be effectively protected from leakage while ensuring that vehicle queuing consensus can be achieved:

[0104] 1. Dynamic event triggering function Parameters in The following formula is satisfied:

[0105] ;

[0106] ;

[0107] in, , .

[0108] 2. Vehicle longitudinal queuing control law Control gain in ,in The matrix is ​​obtained by solving the following Riccati equation:

[0109] ;

[0110] parameter The selection is as follows:

[0111] .

[0112] More specifically, in this embodiment, four unmanned vehicles (one of which is the leader) are performing tasks and coordinating queue control in accordance with the aforementioned control method and privacy protection method.

[0113] Establish a motion model for the longitudinal convoy of autonomous vehicles, and set the model parameters as follows:

[0114] Select matrix With matrix variables in The distance between the follower vehicle and the leader vehicle ,in For virtual vehicle nodes, their distances to their respective real vehicle nodes can be set manually; based on the Riccati equation, the matrix obtained by solving MATLAB... Then control gain The value obtained after calculation is: ; ;

[0115] from Figure 4a As can be seen from the -c diagram, this invention enables vehicles to achieve good vertical queue coordination under our designed bipartite graph-based privacy protection. The dashed lines represent virtual vehicle nodes branching off from the real autonomous vehicle nodes. Ultimately, the virtual nodes do not directly converge on their real vehicles but instead travel in a vertical queue at a certain distance, ensuring that external eavesdroppers cannot locate the actual positions of the real vehicles through the virtual vehicle positions. The smaller diagram illustrates the difference in initial value selection between the real and virtual vehicles, demonstrating that the initial values ​​are effectively protected under our designed privacy protection method. From... Figure 5a As can be seen from -c, the event-triggered design of this invention is successful, which means that the resource consumption of communication is effectively reduced during system operation.

[0116] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0117] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A privacy protection method for longitudinal queuing control of multiple unmanned vehicles based on bipartite graphs, characterized in that, include: Construct a directed graph containing a leader autonomous vehicle and N-1 following autonomous vehicles using graph theory methods; Motion models for the leading unmanned vehicle and the following unmanned vehicle are constructed based on the directed graph; By introducing linear system theory, the motion models of the leading unmanned vehicle and the following unmanned vehicle are transformed into linear system models; Based on the directed graph, a replica is introduced to determine a new topology and node set; Based on a linear system model, a vehicle longitudinal queuing control law is determined according to the new topology and node set for communication control.

2. The privacy protection method for longitudinal queue control of multiple unmanned vehicles based on bipartite graphs according to claim 1, characterized in that, Also includes: Determine the location and speed information of each vehicle; Based on the location and speed information, determine whether the triggering conditions are met. If so, trigger the event and obtain the corresponding vehicle longitudinal queue control law. If not, use the current vehicle longitudinal queue control law for communication control.

3. The privacy protection method for longitudinal queue control of multiple unmanned vehicles based on bipartite graphs according to claim 1, characterized in that, The expression for the directed graph consisting of a lead autonomous vehicle and N-1 following autonomous vehicles is: ； in, It is a collection of autonomous vehicle nodes. ; It represents the set of connecting edges; Represents the neighbor weight matrix. Represents vehicle node Its first Neighbor vehicle nodes Weighting coefficients for communication between them.

4. The privacy protection method for longitudinal queue control of multiple unmanned vehicles based on bipartite graphs according to claim 3, characterized in that, The motion models of the leading and following unmanned vehicles are expressed as follows: ； ； Where, p i Indicates the position of driverless car i. This represents the speed of driverless car i. This represents the acceleration of driverless car i. This represents the position of the leading driverless car. This indicates the speed of the driverless car at speed 0. This represents the acceleration of the driverless car. This represents the inertial lag in longitudinal vehicle dynamics. This indicates the desired acceleration control input that needs to be designed.

5. A privacy protection method for longitudinal queue control of multiple unmanned vehicles based on bipartite graphs according to claim 4, characterized in that, The expression for the linear system model is: ； in, , For state vectors, K is the control gain, A is the first matrix, and B is the second matrix, where... 。 6. A privacy protection method for longitudinal queue control of multiple unmanned vehicles based on bipartite graphs according to claim 5, characterized in that, The expressions for the new topology and node set are as follows: ； ； in, For the new topology. For the new set of nodes, This represents the set of new edges.

7. A privacy protection method for longitudinal queue control of multiple unmanned vehicles based on bipartite graphs according to claim 6, characterized in that, The expression for the vehicle longitudinal queuing control law is: ； in, This is the vehicle longitudinal queuing control law. For nodes The set of neighboring nodes, The expression is: ； in, Represents virtual vehicles To real vehicles distance, Represents real vehicles To virtual vehicles The distance.

8. A privacy protection method for longitudinal queue control of multiple unmanned vehicles based on bipartite graphs according to claim 7, characterized in that, When the event is triggered, the expression for the corresponding vehicle longitudinal queuing control law is obtained as follows: ； in, Indicates the first The triggering time sequence of the k-th autonomous vehicle When an event is triggered, the corresponding vehicle longitudinal queue control law is obtained.

Images