A Dual-Trigger Protocol Control Method for Unmanned Vehicle Formations

Abstract

The present invention belongs to the field of driverless technology and relates to a dual-trigger protocol control method for driverless vehicle formation, including: Step 1, based on a positive multi-agent system, collect the driving state information of driverless vehicles in formation driving, and establish a state space model of the driverless vehicle formation system; Step 2, preset and based on an event-triggered mechanism, through linear programming combined with the Lyapunov function method, construct a dual-trigger protocol model for the observer and controller of the driverless vehicles in the formation system, and finally establish a closed-loop augmented system for the formation driving of driverless vehicles. The present invention uses a positive multi-agent system to model the driverless vehicle formation system, designs a dual-trigger observation protocol through linear programming combined with the Lyapunov function method, improves the system calculation efficiency, can observe the state of driverless vehicles in a timely manner, and conducts effective control to avoid abnormal formation driving of driverless vehicles.

Description

Technical Field

[0001] The present invention belongs to the technical field of driverless technology, and relates to a dual-trigger protocol control method for driverless vehicle formation. Background Art

[0002] With the rapid development of Internet technology, artificial intelligence, and big data, intelligentization has become a trend and a fashion, and it is an inevitable trend in the development of the automotive industry. The formation driving of driverless vehicles provides a perfect humanized operation management solution for the development of contemporary transportation. It can not only reduce the demand of transportation enterprises for drivers, but also reduce the labor intensity of fleet drivers, vehicle fuel consumption, safety accidents, and transportation costs, greatly improving the profit space of the fleet. This technology has high social and economic value. Among them, the formation of driverless vehicles is a key problem worthy of research, which plays an important role in reducing fuel consumption, improving the operation efficiency of the fleet, and reducing traffic congestion. However, there are still many problems with the formation tasks on the road: First, the dynamic vehicle motion states on the road are complex, and the vehicle formation coordination is difficult; Second, the vehicle perception is limited, and the formation system stability is poor; Third, the fixed formation mode makes the system flexibility insufficient and has a greater impact on surrounding vehicles.

[0003] The existing formation methods based on traditional control require complex controller design. The system-level control method has high requirements for the stability of individual vehicles. Facing complex and changeable road scenarios, the fixed control mode will also lose system flexibility and adaptability to environmental changes. Moreover, due to the difficulty of vehicle formation coordination and inaccurate state observation during driving, the driverless vehicle formation control system is prone to failure and is extremely likely to cause the system to get out of control, thus affecting the driving safety of driverless vehicles. Therefore, in order to ensure the driving safety of driverless vehicles, it is of great significance to observe and control the vehicle driving control system in a timely and accurate manner. Summary of the Invention

[0004] In order to solve the above technical problems existing in the prior art, the present invention proposes a dual-trigger protocol control method for driverless vehicle formation. Based on positive multi-agent system modeling, observer protocol, and controller protocol, it can effectively detect vehicle anomalies and perform timely control to avoid abnormal formation driving of driverless vehicles. The specific technical solution is as follows:

[0005] A dual-trigger protocol method for driverless vehicle formation includes the following steps:

[0006] Step 1, based on the positive multi-agent system, collect the driving state information of the driverless vehicles in formation driving, and establish a state space model of the driverless vehicle formation system;

[0007] Step 2: Preset and based on the event-triggering mechanism, construct a dual-trigger protocol model for the observer and controller of the driverless vehicles in the formation system through linear programming combined with the Lyapunov function method, and finally establish a closed-loop augmented system for the platoon driving of driverless vehicles.

[0008] Further, the specific content of Step 1 is as follows: According to the non-negativity of the driving speed of the driverless vehicles, model the driverless vehicle formation system as a positive multi-agent system. Then, take the driving state of the driving speed of the driverless vehicles as the state of the positive multi-agent system, the signal generated by the controller of the driverless vehicles as the control input, and the output of the sensors of the driverless vehicles as the system output to establish a state space model of the driverless vehicle formation system:

[0009] ,

[0010] ;

[0011] Among them, is the operating state of each driverless vehicle during platoon driving at time t, represents the operating state of the i-th driverless vehicle during platoon driving at time t, is the control protocol of the i-th driverless vehicle during platoon driving at time t, that is, the signal generated by the controller, represents the operating state of the i-th driverless vehicle collected by the sensor at time t, , represents the number of driverless vehicles in the driverless vehicle platoon system; is the system matrix, respectively represent an n-dimensional vector, a positive integer, an n×n-dimensional, an n×r-dimensional, and a q×n-dimensional matrix.

[0012] Further, the event-triggering mechanism refers to the adaptive event-triggering condition during the information interaction between the platoon-driving driverless vehicles, and the expression is as follows:

[0013] ,

[0014] ,

[0015] Among them, , represents the observer error, , represents the controller error, represents the state sampling signal of the observer, represents the state sampling signal of the observer at the event trigger point, represents the state sampling signal of the controller, Represents the state sampling signal of the controller at the event trigger point, and represents a constant that changes over time;

[0016] When the event trigger condition is not satisfied; therefore, according to the event trigger condition, we have:

[0017] ,

[0018] ;

[0019] wherein, is the maximum value of the observer error during event triggering, is the maximum value of the controller error during event triggering.

[0020] Furthermore, the dual-trigger protocol model of the observer and controller of the driverless vehicle for constructing the formation system is specifically as follows:

[0021] Construct the dual-trigger protocol model, and the expression is as follows:

[0022] ,

[0023] ,

[0024] ,

[0025] ;

[0026] wherein, is the controller protocol, represents the state sampling signal of the controller at the event trigger point, is the observer output, represents the communication situation between agents, i.e., driverless vehicles i and j, represents the system error, represents the observer input state of the driverless vehicle, is the observer output state of the driverless vehicle; L and K are the observer gain matrix and the controller feedback matrix respectively; represents the case of no leader, represents the neighboring vehicle of driverless vehicle i; F is the gain matrix of the formation system, and the specific expression is:

[0027] ;

[0028] wherein, represents an n-dimensional vector with all elements being 1, represents an r-dimensional vector with all elements being 1, An n-dimensional vector in which the i-th element is 1 and the remaining elements are 0, the vector .

[0029] Furthermore, the closed-loop augmented system for establishing the formation driving of driverless vehicles is specifically expressed as follows:

[0030] ,

[0031] where, , is the augmented state vector, represents the operating state of the i-th driverless vehicle during formation driving at time t, represents the system operation error of the i-th driverless vehicle, is transpose, is transpose;

[0032] Then the dynamics of the closed-loop state system and the closed-loop error system are respectively expressed as:

[0033] ,

[0034] ;

[0035] The following system matrices are calculated:

[0036] ,

[0037] where, is the system matrix of the closed-loop augmented system, is the Laplacian matrix.

[0038] Furthermore, the method of combining linear programming with the Lyapunov function is specifically as follows:

[0039] Design a constant vector such that the following conditional inequalities hold:

[0040] Condition 1: ,

[0041] Condition 2: ,

[0042] Condition 3: ,

[0043] Condition 4: ,

[0044] Condition 5: ,

[0045] Condition 6: ,

[0046] Condition 7: ;

[0047] Among them, Conditions 1 to 3 are used to limit that the matrix is greater than or equal to zero, Conditions 4 and 5 are used to limit that the Lyapunov function is less than zero, and Conditions 6 and 7 are used to transform the conditions into a linear form;

[0048] Furthermore, according to the set event-triggering conditions, the constructed closed-loop augmented system, the designed system gain matrix, and the said Conditions 1 to 3, the system is transformed into:

[0049] ,

[0050] Among them, , represents the initial state of the system at , and , the system matrix is: ;

[0051] Next, from Condition 1, it can be deduced that:

[0052] ,

[0053] The deduced formula represents the inequality after substituting the gain matrix F into Condition 1, making in the closed-loop augmented system a Metzler matrix;

[0054] From Condition 2, it can be obtained that:

[0055] ,

[0056] The deduced formula represents the inequality after substituting the gain matrix F into Condition 1, making in the closed-loop augmented system a Metzler matrix;

[0057] With the help of Condition 3, it can be known that:

[0058] ,

[0059] The deduced formula represents the inequality after substituting the gain matrix F into Condition 1, making in the closed-loop augmented system a Metzler matrix;

[0060] Therefore, is a Metzler matrix;

[0061] Then, define the index: , where is the i-th element of , where , , is the element at the i-th row and j-th column of is a Metzler matrix, then for the matrix, we get ; so for at any time, there is always , therefore, for any initial state , we get .

[0062] Furthermore, the Lyapunov function is , where ,

[0063] , taking the derivative of the Lyapunov function gives , where ,

[0064] ;

[0065] From the gain matrix F and condition 6, we get:

[0066] ,

[0067] ;

[0068] The derived formula represents the conversion of the gain matrix from the matrix factorization form to the vector product form, turning the multiplication of multiple variables into a single form;

[0069] Thus, we have:

[0070] ,

[0071] ,

[0072] The derived formula represents the simplification condition of the gain matrix inequality, turning the condition into the form of linear programming;

[0073] From condition 4 and condition 7, we get:

[0074] ;

[0075] The derived formula represents the stability condition of the Lyapunov function;

[0076] From Condition 5 and Condition 7, we get:

[0077] ;

[0078] The derived formula represents the stability condition of the Lyapunov function;

[0079] From the above conclusion, we obtain:

[0080] ;

[0081] Therefore, , indicating the stability of the unmanned vehicle formation system.

[0082] Beneficial effects:

[0083] 1. The present invention uses a positive multi-agent system to model the unmanned vehicle formation, avoiding system redundancy, improving the model accuracy, providing a new solution idea for the control of unmanned vehicle formation, and reducing the risks existing in the system operation, thereby enhancing the stability, reliability, and safety of the system;

[0084] 2. The present invention observes the state of the vehicle control system in real time through an observer, and makes timely and effective control of vehicle anomalies or failures through a controller, thereby reducing the occurrence of vehicle accidents to a certain extent and ensuring the safe operation of the unmanned vehicle formation;

[0085] 3. The present invention designs a dual-trigger protocol through linear programming combined with the linear Lyapunov function method, reducing the use of computing resources of the vehicle system and improving the computing efficiency of the system. Description of the drawings

[0086] Figure 1 is a schematic flowchart of a dual-trigger protocol control method for an unmanned vehicle formation according to the present invention;

[0087] Figure 2 is a schematic diagram of an unmanned vehicle formation according to an embodiment of the present invention;

[0088] Figure 3 is a structure diagram of a dual-trigger observer and controller of an unmanned vehicle according to an embodiment of the present invention. Detailed implementation manners

[0089] In order to make the objectives, technical solutions, and technical effects of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to the accompanying drawings of the specification and embodiments.

[0090] As Figure 1 shown, a dual-trigger protocol control method for an unmanned vehicle formation according to an embodiment of the present invention includes the following steps:

[0091] Step 1: Based on the positive multi-agent system, collect the driving state information of the driverless vehicles in formation driving, and establish the state space model of the driverless vehicle formation system.

[0092] As Figure 2 shown, in the embodiment of the present invention, several driverless vehicles in formation driving are taken as the research object. During the actual driving process of several vehicles, it is necessary to ensure that each vehicle does not collide with its adjacent vehicle and the distance is not too far apart, so as to ensure the safety of multiple vehicles during the actual driving process.

[0093] Considering that the driving speed of the driverless vehicles in the system is non-negative, the vehicle formation system is modeled as a positive multi-agent system. The driving states such as the driving speed of the driverless vehicles are used as the states of the positive multi-agent system, the signals generated by the controllers of the driverless vehicles are used as the control inputs, and the outputs of the sensors of the driverless vehicles are used as the system outputs. The state space model of the driverless vehicle formation system is established as follows:

[0094] ,

[0095] ;

[0096] wherein, is the operating state of each driverless vehicle in formation driving at time t, represents the operating state of the i-th driverless vehicle in formation driving at time t, is the control protocol of the i-th driverless vehicle in formation driving at time t, that is, the signal generated by the controller, represents the operating state of the i-th driverless vehicle collected by the sensor at time t, , represents the number of driverless vehicles in the driverless vehicle formation system; is the system matrix, respectively represent n-dimensional vectors, positive integers, n×n-dimensional, n×r-dimensional, and q×n-dimensional matrices; when an observer anomaly occurs in the driverless vehicle, it cannot drive in formation safely.

[0097] Step 2: Preset and based on the event-triggered mechanism, through the method of linear programming combined with the Lyapunov function, construct the dual-trigger protocol model of the observer and controller of the driverless vehicles in formation driving, and finally establish the closed-loop augmented system for the driverless vehicle formation driving.

[0098] Specifically, the event-triggered mechanism refers to the adaptive event-triggered condition during the information interaction process among the driverless vehicles in formation driving, and the expression is as follows:

[0099] ,

[0100] ,

[0101] Among them, , represents the observer error, , represents the controller error, represents the state sampling signal of the observer, represents the state sampling signal of the observer at the event trigger point, represents the state sampling signal of the controller, represents the state sampling signal of the controller at the event trigger point, and represent constants that change with time; when , the event trigger condition is not satisfied; therefore, according to the event trigger condition, we can obtain:

[0102] ,

[0103] ;

[0104] Among them, is the maximum value of the observer error in event triggering, is the maximum value of the controller error in event triggering.

[0105] The dual-trigger protocol model for constructing the observer and controller of the driverless vehicles in the formation system specifically includes:

[0106] The constructed dual-trigger protocol model, and the specific expression is as follows:

[0107] ,

[0108] ,

[0109] ,

[0110] ;

[0111] Among them, is the controller protocol, represents the state sampling signal of the controller at the event trigger point, is the observer output, represents the communication situation between agents, i.e., driverless vehicles i and j, represents the system error, represents the input state of the observer of the driverless vehicle, is the observer output state of the driverless vehicle; L and K are the observer gain matrix and the controller feedback matrix respectively; represents the case without a leader, represents the neighboring vehicles of the driverless vehicle i; F is the gain matrix of the formation system, and the specific expression is:

[0112] ;

[0113] where, represents an n-dimensional vector with all elements being 1, represents an r-dimensional vector with all elements being 1, represents an n-dimensional vector with the i-th element being 1 and the remaining elements being 0, and the vector .

[0114] Construct a closed-loop augmented system for the formation driving of driverless vehicles, and the specific expression is as follows:

[0115] ,

[0116] where, , is the augmented state vector, represents the operating state of the i-th driverless vehicle during formation driving at time t, represents the system operation error of the i-th driverless vehicle, is transpose of, is transpose of;

[0117] Then the dynamics of the closed-loop state system and the closed-loop error system are respectively expressed as:

[0118] ,

[0119] ;

[0120] The following system matrices are calculated:

[0121] ,

[0122] where, is the system matrix of the closed-loop augmented system, is the Laplacian matrix.

[0123] The method of combining linear programming with the Lyapunov function is as follows:

[0124] Design a constant vector , so that the following conditional inequalities hold:

[0125] Condition 1: ,

[0126] Condition 2: ,

[0127] Condition 3: ,

[0128] Condition 4: ,

[0129] Condition 5: ,

[0130] Condition 6: ,

[0131] Condition 7: .

[0132] Among them, Conditions 1 to 3 are used to limit the matrix to be greater than or equal to zero, Conditions 4 and 5 are used to limit the Lyapunov function to be less than zero, and Conditions 6 and 7 are used to transform the conditions into a first-order form;

[0133] According to the set event-triggering conditions, the constructed closed-loop augmented system, the designed system gain matrix, and the said Conditions 1 to 3, the system is transformed into:

[0134] ,

[0135] Among them, , represents the initial state of the system at , and , the system matrix is:

[0136] ,

[0137] It is deduced from Condition 1 that:

[0138] ,

[0139] The deduced formula represents the inequality after substituting the gain matrix F into Condition 1, and the purpose is to make in the closed-loop augmented system a Metzler matrix;

[0140] It can be obtained from Condition 2 that:

[0141] ,

[0142] The deduced formula represents the inequality after substituting the gain matrix F into Condition 1, and the purpose is to make in the closed-loop augmented system a Metzler matrix;

[0143] With the help of Condition 3, it can be known that:

[0144] ,

[0145] The derived formula represents the inequality after substituting the gain matrix F into Condition 1, aiming to make the closed-loop augmented system be a Metzler matrix;

[0146] Therefore, is a Metzler matrix. Define a set of indices: where is the i-th element of . It can be obtained that , where , is the element in the i-th row and j-th column of . If is a Metzler matrix, then for the matrix , it can be obtained that . Therefore, for at any time, there is always . Therefore, for any initial state , it can be obtained that

[0147] . Thus, the positivity of the unmanned vehicle formation system is guaranteed. , where ,

[0148] Taking the derivative of the Lyapunov function can obtain , where ,

[0149] ;

[0150] From the gain matrix F and Condition 6, it can be obtained that:

[0151] ,

[0152] ;

[0153] The derived formula represents the transformation of the gain matrix from the matrix decomposition form to the vector product form, aiming to convert the multiplication of multiple variables into a single form;

[0154] From this, it can be obtained that:

[0155] ,

[0156] ;

[0157] The derived formula represents the simplification condition of the gain matrix inequality, aiming to transform the condition into the form of linear programming;

[0158] From condition 4 and condition 7, it can be obtained that:

[0159] ;

[0160] The derived formula represents the stability condition of the Lyapunov function, aiming to prove the stability of the system.

[0161] From condition 5 and condition 7, it can be obtained that:

[0162] ;

[0163] The derived formula represents the stability condition of the Lyapunov function, aiming to prove the stability of the system.

[0164] From the above conclusions, it can be obtained that:

[0165] ;

[0166] Therefore, , the stability of the unmanned vehicle formation system is guaranteed.

[0167] As Figure 3 shown, it is the structure diagram of the double-trigger observer and controller of the unmanned vehicle. Through the observer arranged on each unmanned vehicle, the running state of each unmanned vehicle can be accurately observed. When an abnormal driving state is found, this state is intervened through the event trigger condition of the observer, and controlled through the event trigger condition of the controller to ensure the safe and stable driving.

[0168] The above is only the preferred embodiment of the present invention and does not impose any form of limitation on the present invention. Although the implementation process of the present invention has been described in detail above, for those familiar with the field, they can still modify the technical solutions recorded in the foregoing examples, or make equivalent replacements for some of the technical features. Any modifications, equivalent replacements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A dual-trigger protocol control method for an unmanned vehicle formation, characterized in that, It includes the following steps: Step 1: Based on the positive multi-agent system, collect the driving state information of the driverless vehicles in formation driving, and establish the state space model of the driverless vehicle formation system; Step 2: Preset and based on the event-triggered mechanism, through the method of linear programming combined with the Lyapunov function, construct the double-trigger protocol model of the observer and controller of the driverless vehicles in the formation system, and finally establish the closed-loop augmented system for the formation driving of driverless vehicles; The construction of the double-trigger protocol model of the observer and controller of the driverless vehicles in the formation system is specifically as follows: Construct the double-trigger protocol model, and the expression is as follows: ； Among them, is the controller protocol, represents the state sampling signal of the controller at the event trigger point, is the observer output, represents the communication situation between agents, i.e., autonomous vehicles i and j, represents the system error, represents the input state of the observer of the autonomous vehicle, is the output state of the observer of the autonomous vehicle, represents the state sampling signal of the observer, represents the state sampling signal of the observer at the event trigger point; represents the operating state of the i-th autonomous vehicle collected by the sensor at time t, , represents the number of unmanned vehicles in the autonomous vehicle formation system; is the system matrix, and L and K are the observer gain matrix and the controller feedback matrix respectively; represents the case without a leader, represents the autonomous vehicle 's neighboring vehicles; F is the gain matrix of the formation system, and the specific expression is: ； Among them, represents an n-dimensional vector with all elements being 1, represents an r-dimensional vector with all elements being 1, represents the n-dimensional vector in which the -th element is 1 and the remaining elements are 0, and the vector 2. The dual-trigger protocol control method for an unmanned vehicle formation according to claim 1, characterized in that, The specific content of Step 1 is as follows: According to the fact that the driving speed of the driverless vehicle is non-negative, model the driverless vehicle formation system as a positive multi-agent system. Then, take the driving state of the driving speed of the driverless vehicle as the state of the positive multi-agent system, the signal generated by the controller of the driverless vehicle as the control input, and the output of the sensor of the driverless vehicle as the system output, and establish the state space model of the driverless vehicle formation system: ， ； Among them, is the operating state of each driverless vehicle during platoon driving at time t, represents the operating state of the th driverless vehicle during platoon driving at time t, is the control protocol of the ith driverless vehicle during platoon driving at time t, that is, the signal generated by the controller, represents the operating state of the ith driverless vehicle collected by the sensor at time t, , represents the number of driverless vehicles in the driverless vehicle platoon system; is the system matrix, respectively represent an n-dimensional vector, a positive integer, an n×n-dimensional, an n×r-dimensional, and a q×n-dimensional matrix.

3. The dual-trigger protocol control method for an unmanned vehicle formation according to claim 2, characterized in that, The event-triggered mechanism refers to the adaptive event-triggered condition during the information interaction process between the driverless vehicles in formation driving, and the expression is as follows: ， ， Among them, , represents the observer error, , represents the controller error, represents the state sampling signal of the observer, represents the state sampling signal of the observer at the event trigger point, represents the state sampling signal of the controller, represents the state sampling signal of the controller at the event trigger point, and represent time-varying constants; When the event trigger condition does not hold; thus, according to the event trigger condition, it can be obtained that: ， ； Among them, is the maximum value of the observer error in event triggering, is the maximum value of the controller error in event triggering.

4. The dual-trigger protocol control method for an unmanned vehicle formation according to claim 1, characterized in that, The specific expression of the established closed-loop augmented system for the formation driving of driverless vehicles is as follows: ， Among them, , is the augmented state vector, indicating the operating state of the th driverless vehicle during platoon driving at time t, indicating the system operation error of the th driverless vehicle, is the transpose, is the transpose; Then the dynamics of the closed-loop state system and the closed-loop error system are respectively expressed as: The following system matrix is calculated: Among them, is the system matrix of the closed-loop augmented system, is the Laplacian matrix.

5. The dual-trigger protocol control method for an unmanned vehicle formation according to claim 4, characterized in that, The method of combining linear programming with the Lyapunov function is specifically as follows: Design constant Vector such that the following conditional inequalities hold: Condition 1: , Condition 2: , Condition 3: , Condition 4: , Condition 5: , Condition 6: , Condition 7: ; Among them, Conditions 1 to 3 are used to define that the matrix is greater than or equal to zero, Conditions 4 and 5 are used to define that the Lyapunov function is less than zero, and Conditions 6 and 7 are used to transform the conditions into a first-order form; Then, according to the set event-triggered condition, the constructed closed-loop augmented system, the designed system gain matrix, and the above Conditions 1 to 3, convert the system into: ， Among them, , represents the initial state of the system at , and and , the system matrix is as follows: ； Next, it is deduced from Condition 1: ， The derived formula represents the inequality after substituting the gain matrix F into Condition 1, making a Metzler matrix; From Condition 2, it can be obtained: ， The derived formula represents the inequality after substituting the gain matrix F into Condition 1, making a Metzler matrix; With the help of Condition 3, it can be known: ， The derived formula represents the inequality after substituting the gain matrix F into Condition 1, making a Metzler matrix; Therefore, is a Metzler matrix; Then, define the metrics: where is the -th element of to obtain , where , is the -th row and -th column element of ; if is a Metzler matrix, then for the matrix of , obtain ; so for at any time, there is always .

6. As for a dual-trigger protocol control method for an unmanned vehicle formation according to claim 5, it is characterized in that, The Lyapunov function is , where ​ For the Lyapunov function Taking the derivative gives , where , ； From the gain matrix F and Condition 6, it is obtained: ； The deduced formula means converting the gain matrix from the matrix decomposition form to the vector product form, and turning the multiplication of multiple variables into a single form; Thus, it is obtained: ， The deduced formula means simplifying the condition of the gain matrix inequality to make the condition into the form of linear programming; From Condition 4 and Condition 7, it is obtained: ； The deduced formula means the stability condition of the Lyapunov function; From Condition 5 and Condition 7, it is obtained: ； The deduced formula means the stability condition of the Lyapunov function; From the above conclusions, it is obtained: ； Therefore, it indicates that the unmanned vehicle formation system is stable.

Images