A longitudinal control method for vehicle platooning under any communication link failure

Abstract

The present invention relates to a longitudinal control method for vehicle platooning under any communication link failure, comprising the following steps: Step S1: Analyze possible communication link failure situations under the Bidirectional-leader communication topology structure, and classify them into three major types according to the failure location and handling methods; Step S2: Utilize the Bidirectional-leader topology structure characteristics, and combine with the distributed model predictive control algorithm to construct corresponding active communication topologies and controller transformation schemes for each failure type. The present invention can autonomously adapt to the occurrence of any communication link failure under the topology structure defined by Bidirectional-leader, avoid collisions between vehicles, realize the normal operation of the platoon, and provide a safe, feasible and practical corresponding method for communication failure situations in the actual implementation process of vehicle platooning.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent transportation, and particularly to a longitudinal control method for vehicle platooning under any communication link failure. Background Art

[0002] With the accelerating urbanization process and the explosive growth of the number of automobiles in various countries around the world, the problem of urban congestion has become increasingly serious. Accompanied by this are frequent traffic accidents and air pollution, which pose a severe challenge to the urban carrying capacity and social operation efficiency. Since vehicle platooning has great potential in alleviating road congestion, improving traffic safety, and reducing fuel consumption, it has attracted extensive attention in recent years.

[0003] The problem of vehicle platooning longitudinal control can be regarded as an optimization problem with multiple inputs, multiple parameters, and multiple constraints. Therefore, the distributed model predictive control method can effectively solve this problem. The distributed model predictive control framework for vehicle platooning longitudinal control mainly includes four parts: vehicle kinematic or dynamic model, communication topology, distributed controller, and platoon geometry. Among them, the communication topology plays a crucial role in the stability of vehicle platooning. Each vehicle in the platoon needs to rely on the data flow direction defined by the communication topology, obtain the data of neighboring vehicles through vehicle-to-vehicle (V2V) communication, and use this data to construct its local optimal control law. Therefore, once the communication link fails, it will bring huge potential safety hazards to the platoon. Therefore, it is of great practical significance to study a longitudinal control method for vehicle platooning based on distributed model predictive control under any communication link failure.

[0004] Distributed model predictive control is widely used to solve the vehicle platooning control problem. Since the controllers of vehicle platooning are distributed on each vehicle, and the vehicle platooning longitudinal control problem is an optimization problem with multiple inputs, multiple parameters, and multiple constraints, we can intuitively use the distributed model predictive control method to solve the vehicle platooning longitudinal control problem under any communication link failure. However, for different communication link failure situations, how to design a distributed model predictive controller algorithm applicable to all failure cases is one of the cores of solving the vehicle platooning longitudinal control problem under any communication link failure.

[0005] Currently, most vehicle platoon control methods mostly adopt fixed information flow topologies, such as: (a) Predecessor-following, (b) Predecessor-leader Following, (c) Bidirectional, and (d) Bidirectional-leader adopted under normal communication conditions in this paper. However, fixed information flow topologies cannot cope with the unpredictable platoon driving environment, which requires us to actively transform the communication topology of the platoon when necessary to deal with different driving conditions. For example, when a certain communication link fails, we need to re-establish a new communication link to ensure the normal driving of the vehicle platoon. Therefore, how to design an active transformation scheme for the vehicle platoon communication topology is one of the cores to solve the longitudinal control problem of vehicle platoons under any communication link failure.

[0006] Currently, certain achievements have been made in the research on the longitudinal control problem of vehicle platoons under communication failures. However, the current methods still have many drawbacks. For example, some vehicle platoon control methods are only applicable to the case of partial communication link failures and cannot cover all cases of any communication link failure; or they can be applicable to vehicle platoon control under any vehicle communication failure, but cannot ensure that the performance of the vehicle platoon is not affected after the failure. This greatly limits the practical engineering application of vehicle platoons. Summary of the Invention

[0007] In view of this, the purpose of the present invention is to provide a longitudinal control method for vehicle platoons under any communication link failure, aiming to solve the above problems.

[0008] To achieve the above purpose, the present invention adopts the following technical solutions:

[0009] A longitudinal control method for vehicle platoons under any communication link failure, comprising the following steps:

[0010] Step S1: Analyze the possible communication link failure situations under the Bidirectional-leader communication topology structure, and divide them into three major types according to the failure location and treatment methods;

[0011] Step S2: Utilize the Bidirectional-leader topology structure characteristics, and combine with the distributed model predictive control algorithm to construct corresponding active communication topologies and controller transformation schemes for each failure type.

[0012] Further, the specific content of step S1 is as follows:

[0013] In the Bidirectional-leader communication topology structure, for a vehicle platoon composed of N + 1 vehicles, where 1 vehicle is the leader and N vehicles are the followers,

[0014] The leader numbered 0 provides the desired trajectory for all followers and the emergency braking signal, where y0(t) = [p0(t), v0(t)] T is the position and speed information of the leader at time t, s exp,i (t) = [d exp,i (t), 0] T represents the vector of the desired position spacing and speed difference between vehicle i and its preceding vehicle;

[0015] In addition, the follower provides the assumed output trajectory to the adjacent downstream follower where adjacent vehicles transmit the emergency braking signal to each other;

[0016] According to the specific fault location and treatment method, the communication fault situation is divided into three major types: (1) Loss of connection between the leader and any follower; (2) Loss of connection between any follower m and follower m + 1; (3) Loss of connection between the leader and follower 1.

[0017] Furthermore, for normal communication and communication fault situation (1), a controller based on distributed model predictive control is designed as follows:

[0018] The vehicle formation control objective is to maintain a certain distance between vehicles and the whole vehicle formation moves at a constant speed:

[0019]

[0020]

[0021] where p tr,n (t) represents the actual position of vehicle n, p exp,n (t) represents its desired position; v n (t) represents the actual speed of vehicle n,

[0022] Let

[0023] d exp,n represents the desired position spacing between vehicle i and its preceding vehicle, d tr,n represents the actual position spacing between vehicle i and its preceding vehicle, d adj,n is used to adjust the distance between vehicle i and its preceding vehicle, d delay represents the distance traveled within the maximum communication delay for the vehicle to receive the emergency braking signal, d brake represents the conservative emergency braking distance, τ max represents the maximum communication delay, v max represents the maximum speed, a max represents the maximum braking acceleration;

[0024] Using the vehicle position p n (t), speed v n (t), acceleration a n (t), establish the state - space expression of the following linear system:

[0025]

[0026] where

[0027] By performing Euler approximation on the above formula, we get:

[0028]

[0029] where E is the identity matrix, T is the sampling time, x(k|t) represents the system state value at the k - th sampling time after time t, u(k|t) represents the system input value at the k - th sampling time after time t, and y(k + 1|t) represents the system output value at the k - th sampling time after time t;

[0030] For any vehicle i ∈ {1, 2,..., N} at time t, there is the following optimal control problem:

[0031]

[0032] s.t.

[0033]

[0034]

[0035]

[0036]

[0037]

[0038] where represents the optimal control sequence J to be solved at the k - th sampling time after time t i represents the cost function of vehicle i, N p represents the prediction horizon, l i represents the cost function of vehicle i at any time within k = 0 ~ Np - 1, represents the predicted output trajectory at the k - th sampling time after time t, represents the assumed output trajectory at the k - th sampling time after time t, that is, the expected output trajectory at the current time calculated using the optimal control input calculated at the previous time, Denotes the system predicted state at the k-th sampling time after time t, x i (t) represents the system state value at time t, Denotes the predicted speed at the k-th sampling time after time t, s exp,i (k|t) represents the vector of the desired position spacing and speed difference between vehicle i and its leading vehicle at the k-th sampling time after time t, y exp,i (k|t) represents that the leader numbered 0 provides the desired trajectory for all followers at the k-th sampling time after time t;

[0039] The specific expression of the cost function is as follows:

[0040]

[0041] Where Q i ≥0, R i ≥0, F i ≥0, G i ≥0 are all symmetric weight matrices; the first term of the cost function (11) Represents the error between the predicted output trajectory and the desired trajectory; the second term of the cost function (11) Represents the error between the predicted input quantity and the input quantity 0 during the uniform formation operation; the third term of the cost function (11) Represents the error between the predicted output trajectory of vehicle i and its assumed output trajectory; the fourth term of the cost function (11) Represents the error between the predicted output trajectory of vehicle i and the assumed output trajectory of its leading vehicle.

[0042] Furthermore, for communication failure cases (2) and (3), the active communication topology transformation scheme is as follows:

[0043] When the disconnection occurs between follower m and follower m + 1, the active communication topology transformation scheme is to establish a new communication connection between follower m + 1 and follower m - 1. At this time, the delay for follower m + 1 to obtain the emergency braking signal of follower m increases, that is, resulting in d exp,m+1 = d adj,m+1 + 2*τ max v max + d brake , d exp,i Represents the desired distance between vehicle i and its leading vehicle, d adj,i Represents the adjustable spacing between vehicle i and its leading vehicle, which is used to adjust the density of the vehicle formation; the active controller transformation scheme actively adjusts the distributed model predictive controller under normal communication according to the transformed communication topology structure. For vehicle i = m + 1, the constraints (9) of its optimal control problem and the last term of the cost function are changed as follows:

[0044]

[0045]

[0046] When the disconnection occurs between the leader and the follower 1, the active communication topology transformation scheme is to establish a new communication connection between the follower 2 and the leader, and change the information flow direction between the follower 2 and the follower 1. At this time, the delay for the follower 1 to obtain the emergency braking signal of the leader increases, that is, it causes d exp,1 = d adj,1 + 2*τ max v max + d brake ; The active controller transformation scheme actively adjusts the distributed model predictive controller under normal communication according to the transformed communication topology structure; for vehicle i = 2, the constraints (9) of its optimal control problem and the end term of the cost function also have the changes shown in equations (12) and (13) respectively; for vehicle i = 1, the constraints (9) of its optimal control problem and the end term of the cost function have the following changes respectively:

[0047]

[0048]

[0049] A longitudinal control system for vehicle formation under any communication link failure, including a processor, a memory, and a computer program stored on the memory. When the processor executes the computer program, it specifically executes the steps in the above-mentioned longitudinal control method for vehicle formation under any communication link failure.

[0050] The present invention has the following beneficial effects compared with the prior art:

[0051] The present invention can effectively solve the problem of longitudinal control of vehicle formation under any communication link failure, that is, it can still maintain the normal driving of the vehicle formation after any communication link failure, and can effectively prevent collisions between any vehicles in the formation, providing a strong guarantee for the safety of vehicle formation operation. Compared with the existing research on longitudinal control of vehicle formation under dynamic communication topology, this method first proposes a longitudinal control method for vehicle formation based on distributed model predictive control that can effectively solve any communication link failure in vehicle formation, and has good scalability. Under the assumption that there is an upper bound on the communication delay between any two vehicles, it can theoretically be extended to a formation composed of an infinite number of vehicles. The proposed method promotes the further advantages of vehicle formation in reducing road congestion, improving traffic safety, and reducing fuel consumption to a certain extent. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 It is a diagram of the Bidirectional - leader communication topology structure under normal communication based on the prior art of the present invention and the active communication topology transformation scheme in the case of different communication link failures;

[0053] Figure 2 It is a schematic diagram of the actual positions and desired positions of each vehicle in the formation based on the prior art of the present invention;

[0054] Figure 3 It is a specific disposal scheme diagram for communication link failure situations (2) and (3) based on the prior art of the present invention;

[0055] Figure 4 It is a simulation schematic diagram for communication link failure situation (1) based on the embodiment;

[0056] Figure 5 It is a simulation schematic diagram for communication link failure situation (2) based on the embodiment;

[0057] Figure 6 It is a simulation schematic diagram for communication link failure situation (3) based on the embodiment;

[0058] Figure 7 It is a comparative simulation schematic diagram based on the embodiment. Detailed implementation manners

[0059] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0060] Please refer to Figure 1 , the present invention provides a longitudinal control method for vehicle formation under any communication link failure. Its topological structure under normal communication, various communication link failure situations, and the active communication topology transformation scheme are as Figure 1 shown, and the design of the controller based on distributed model predictive control will be introduced in detail in step two.

[0061] Step one: Analyze the possible communication link failure situations under the Bidirectional - leader communication topology structure

[0062] a. Information flow content and information flow direction design

[0063] For a vehicle formation composed of N + 1 vehicles, where 1 vehicle is the leader and N vehicles are followers. We conduct research based on the very common Bidirectional - leader communication topology structure.

[0064] First, as Figure 1 shown, define the information flow content and information flow direction in the communication topology structure of the vehicle formation:

[0065] The leader numbered 0 provides the desired trajectory for all followers (numbered 1 to N). and an emergency braking signal (used to avoid emergency collision avoidance once a vehicle accident occurs), where y0(t) = [p0(t), v0(t)] T is the position and speed information of the leader at time t, s exp,i (t) = [d exp,i (t), 0] T represents the vector of the desired position spacing and speed difference between vehicle i and its preceding vehicle.

[0066] In addition, the followers (numbered 1 to N - 1) provide the assumed output trajectory to the adjacent downstream followers (numbered 2 to N). where adjacent vehicles transmit the emergency braking signal to each other.

[0067] b. Analysis of communication link failure cases

[0068] According to the specific failure location and handling method, it is divided into three major types: (1) loss of connection between the leader and any follower (except follower 1), (2) loss of connection between any follower m and follower m + 1, (3) loss of connection between the leader and follower 1;

[0069] Step 2: Utilize the Bidirectional - leader topological structure characteristics, and combine the distributed model predictive control algorithm to construct the corresponding active communication topology and controller transformation scheme for each failure type.

[0070] (1) Design a controller based on distributed model predictive control

[0071] a. Vehicle formation control objective

[0072] The control objective is to keep a certain distance between vehicles and the whole vehicle formation moves at a constant speed:

[0073]

[0074]

[0075] where p tr,n (t) represents the actual position of vehicle n, p exp,n (t) represents its desired position; v n (t) represents the actual speed of vehicle n, The schematic diagram is as Figure 2 shown. Figure 2 in d exp,n represents the desired position spacing between vehicle i and its preceding vehicle, d tr,n represents the actual position spacing between vehicle i and its preceding vehicle, d adj,nFor adjusting the distance, d, between vehicle i and its preceding vehicle delay Indicates the distance, d, traveled within the maximum communication delay for the vehicle to receive an emergency braking signal brake Indicates the conservative emergency braking distance, τ max Indicates the maximum communication delay, v max Indicates the maximum speed, a max Indicates the maximum braking acceleration.

[0076] b. Vehicle mathematical model

[0077] In model predictive control, a complex model is not a good choice. Reasonably simplifying the model and designing constraint conditions that meet the driving conditions are the focus of the research. Here, we use the vehicle position p n (t), speed v n (t), acceleration a n (t) to establish the state-space expression of the following linear system:

[0078]

[0079] where

[0080] By performing Euler approximation on the above equation, we can obtain:

[0081]

[0082] where E is the identity matrix, T is the sampling time, x(k|t) represents the system state value at k sampling times after time t, u(k|t) represents the system input value at the kth sampling time after time t, and y(k + 1|t) represents the system output value at the kth sampling time after time t.

[0083] c. Controller design

[0084] For any vehicle i ∈ {1, 2,..., N} at time t, there is the following optimal control problem:

[0085]

[0086] s.t.

[0087]

[0088]

[0089]

[0090]

[0091]

[0092] Among them represents the optimal control sequence J to be solved at the k-th sampling time after time t i represents the cost function of vehicle i, N p represents the prediction horizon, l i represents the cost function of vehicle i at any time within k = 0 to Np - 1 represents the predicted output trajectory at the k-th sampling time after time t represents the assumed output trajectory at the k-th sampling time after time t, that is, the expected output trajectory at the current time calculated using the optimal control input calculated at the previous time represents the system predicted state at the k-th sampling time after time t, x i (t) represents the system state value at time t represents the predicted speed at the k-th sampling time after time t, s exp,i (k|t) represents the vector of the expected position spacing and speed difference between vehicle i and its leading vehicle at the k-th sampling time after time t, y exp,i (k|t) represents that the leader numbered 0 provides the expected trajectory for all followers (numbered 1 to N) at the k-th sampling time after time t. I = 2. Specifically, when there is no direct communication between vehicle i and the leader, I = 1. Constraint (6) is a model constraint; Constraint (7) is used to initialize the current state variables; Constraint (8) is an input quantity constraint; the role of the terminal constraint (9) is to make vehicle i, its leading vehicle, and the leader maintain good consistency; the terminal constraint (0) forces vehicle i to maintain a constant speed at the end of the prediction horizon

[0093] The specific expression of the cost function is as follows

[0094]

[0095] Among them, Q i ≥0, R i ≥0, F i ≥0, G i ≥0 are all symmetric weight matrices. Specifically, when there is no direct communication between vehicle i and the leader, Q i = 0; the first term of the cost function (11) represents the error between the predicted output trajectory and the expected trajectory; the second term of the cost function (11) represents the error between the predicted input quantity and the input quantity 0 during the formation's uniform motion; the third term of the cost function (11) represents the error between the predicted output trajectory of vehicle i and its own assumed output trajectory; the fourth term of the cost function (11) Denote the error between the predicted output trajectory of vehicle i and the assumed output trajectory of the leading vehicle in front. The design idea of this cost function is that it is desired that the followers in the vehicle formation maintain good consistency with the leader and its leading vehicle in front, and try to maintain a constant speed operation.

[0096] This algorithm is applicable to normal communication topology situations and communication link failure situations. (1) Loss of connection between the leader and any follower (except follower 1).

[0097] (2) Design of the active communication topology transformation scheme and the active controller transformation scheme

[0098] For the communication link failure situations (2) loss of connection between any follower m and follower m + 1 and (3) loss of connection between the leader and follower 1, the active communication topology transformation scheme is as Figure 3 shown.

[0099] When the loss of connection between follower m and follower m + 1 occurs, the active communication topology transformation scheme is to establish a new communication connection between follower m + 1 and follower m - 1 (follower 0 is the leader). At this time, the delay for follower m + 1 to obtain the emergency braking signal of follower m increases, that is, it leads to d exp,m+1 = d adj,m+1 + 2*τ max v max + d brake where d exp,i represents the desired distance between vehicle i and the leading vehicle in front, and d adj,i represents the adjustable spacing between vehicle i and the leading vehicle in front, which is used to adjust the density of the vehicle formation; the active controller transformation scheme actively adjusts the distributed model predictive controller under normal communication conditions according to the transformed communication topology structure. For vehicle i = m + 1, the constraints (9) and the last term of the cost function of its optimal control problem are changed as follows:

[0100]

[0101]

[0102] When the loss of connection between the leader and follower 1 occurs, the active communication topology transformation scheme is to establish a new communication connection between follower 2 and the leader, and change the information flow direction between follower 2 and follower 1. At this time, the delay for follower 1 to obtain the emergency braking signal of the leader increases, that is, it leads to d exp,1 = d adj,1 + 2*τ max v max + d brake; The active controller transformation scheme actively adjusts the distributed model predictive controller under normal communication conditions according to the transformed communication topology. For vehicle i = 2, the constraints (9) of its optimal control problem and the last term of the cost function are changed as shown in Eqs. (12) and (13) respectively; for vehicle i = 1, the constraints (9) of its optimal control problem and the last term of the cost function are changed as follows:

[0103]

[0104]

[0105] Example 1:

[0106] In order to introduce the present invention in detail, a specific example is given below to demonstrate the effectiveness and superiority of the proposed longitudinal control method for vehicle formation under any communication link failure.

[0107] In the given simulation case, the initial position of the leader is defined as 0 m, the initial speed is defined as 10 m / s, the maximum speed V of each vehicle max is limited to the maximum communication delay τ max is defined as 1 / the maximum acceleration |a max | is defined as 6 / s 2 , the adjustable spacing d adj,i is defined as 2 m. When the communication link fails at 1 s, and in order to verify whether the vehicle formation has a certain anti-interference ability after the failure, we set that the leader performs an acceleration motion with an acceleration of 2 / s for 1 s at 5 s 2 and a deceleration motion with an acceleration of -2 / s for 1 s at 10 s 2 . Before the failure and when the communication link failure situation (1) occurs, the spacing between adjacent vehicles is d exp,i = d adj,i + τ max v max + d brake = 21 (m); when the communication link failure situations (2) and (3) occur, the spacing between vehicle m + 1 and the vehicle in front becomes d exp,m+1 = d adj,m+1 + 2*τ max v max + d brake = 22 (m).

[0108] As Figure 4 shown is the simulation diagram under the communication link failure situation (1), and as Figure 5 shown is the simulation diagram under the communication link failure situation (2), and as Figure 6The figure shows a simulation diagram under the communication link failure condition (3), where (a), (b), (c), and (d) respectively represent the acceleration, actual position, deviation from the expected position given by the leader, and actual distance from the vehicle in front in the vehicle formation. From Figure 4 , 5 , and 6, it can be seen that the proposed longitudinal control method for vehicle formation under any communication link failure can effectively handle any communication link failure situation and has a certain anti-interference ability, verifying the effectiveness of this method.

[0109] In the given comparative simulation cases, a longitudinal control method for vehicle formation under any communication link failure proposed in this paper is compared with a model predictive control method for interconnected vehicle formation under a traditional switched communication topology. This method adopts the strategy of keeping the leader at a constant speed after the leader's communication fails, which leads to the inability of the vehicle formation to perform acceleration and deceleration operations after the failure, thus causing the vehicle formation to lose certain performance. As Figure 7 shown, where (a), (b), (c), and (d) respectively represent the speed information of the comparative method and the method proposed in this paper under three communication link failure classification situations. The simulation results show that the proposed longitudinal control method for vehicle formation under any communication link failure can ensure that the original performance of the vehicle formation remains unchanged, verifying the superiority of this method.

[0110] The above are only the preferred embodiments of the present invention, and all equivalent changes and modifications made according to the scope of the patent application of the present invention shall fall within the scope covered by the present invention.

Claims

1. A longitudinal control method for vehicle platoons under any communication link failure, characterized in that, It includes the following steps: Step S1: Analyze the possible communication link failure situations under the Bidirectional-leader communication topology structure, and classify them into three major types according to the failure location and handling method; Step S2: Utilize the Bidirectional-leader topology structure characteristics, and combine with the distributed model predictive control algorithm to construct corresponding active communication topologies and controller transformation schemes for each failure type; The specific content of step S1 is as follows: In the Bidirectional-leader communication topology structure, for a vehicle formation composed of N + 1 vehicles, where 1 vehicle is the leader and N vehicles are followers, The leader numbered 0 provides the desired trajectory for all followers and an emergency braking signal, where y0(t) = [p0(t), v0(t)] T is the position and velocity information of the leader at time t, s exp,i (t) = [d exp,i (t), 0] T represents the vector of the desired position spacing and speed difference between vehicle i and its preceding vehicle; In addition, the follower provides a hypothetical output trajectory to an adjacent downstream follower wherein adjacent vehicles transmit an emergency braking signal to each other; According to the specific failure location and handling method, the communication failure situations are divided into three major types: (1) Loss of connection between the leader and any follower; (2) Loss of connection between any follower m and follower m + 1; (3) Loss of connection between the leader and follower 1; For normal communication and communication failure situation (1), design a controller based on distributed model predictive control, specifically as follows: The vehicle formation control objective is to keep a certain distance between vehicles and the whole vehicle formation moves at a constant speed: where p tr,n (t) represents the actual position of vehicle n, and p exp,n (t) represents its desired position; v n (t) represents the actual speed of vehicle n, Set d exp,n represents the desired position spacing between vehicle i and its preceding vehicle, d tr,n represents the actual position spacing between vehicle i and its preceding vehicle, d adj,n is used to adjust the distance between vehicle i and its preceding vehicle, d delay represents the distance traveled by the vehicle within the maximum communication delay for receiving an emergency braking signal, d brake represents the conservative emergency braking distance, τ max represents the maximum communication delay, v max represents the maximum speed, a max represents the maximum braking acceleration; Using the vehicle position p n (t), speed v n (t), acceleration a n (t), establish the state - space expression of the following linear system: y(t) = Cx(t) (3) Among them By performing Euler approximation on the above formula, we get: Among them E is the identity matrix, T is the sampling time, x(k|t) represents the system state value at k sampling times after time t, u(k|t) represents the system input value at the kth sampling time after time t, and y(k + 1|t) represents the system output value at the kth sampling time after time t; For any vehicle i ∈ {1, 2,..., N} at time t, there is the following optimal control problem: Among them represents the optimal control sequence J to be solved at the k-th sampling time after time t i represents the cost function of vehicle i, N p represents the prediction horizon, l i represents the cost function of vehicle i at any time within k = 0 to Np - 1 represents the predicted output trajectory at the k-th sampling time after time t represents the assumed output trajectory at the k-th sampling time after time t, that is, the expected output trajectory at the current time calculated using the optimal control input calculated at the previous time represents the system predicted state at the k-th sampling time after time t, x i (t) represents the system state value at time t represents the predicted speed at the k-th sampling time after time t, s exp,i (k|t) represents the vector of the expected position spacing and speed difference between vehicle i and its leading vehicle at the k-th sampling time after time t, y exp,i (k|t) represents that the leader numbered 0 provides the expected trajectory for all followers at the k-th sampling time after time t The specific expression of the cost function is as follows: where Q i ≥0, R i ≥0, F i ≥0, G i ≥0 are all symmetric weight matrices; the first term of the cost function (11) represents the error between the predicted output trajectory and the desired trajectory; the second term of the cost function (11) represents the error between the predicted input quantity and the input quantity 0 during the formation's uniform motion; the third term of the cost function (11) represents the error between the predicted output trajectory of vehicle i and its assumed output trajectory; the fourth term of the cost function (11) represents the error between the predicted output trajectory of vehicle i and the assumed output trajectory of the preceding vehicle; For communication failure situations (2) and (3), the active communication topology transformation scheme is specifically as follows: When the communication loss occurs between follower m and follower m + 1, the active communication topology transformation scheme is to establish a new communication connection between follower m + 1 and follower m - 1. At this time, the delay for follower m + 1 to obtain the emergency braking signal of follower m increases, that is, it leads to d exp,m+1 = d adj,m+1 + 2*τ max v max + d brake , d exp,i represents the desired distance between vehicle i and the vehicle in front, and d adj,i represents the adjustable spacing between vehicle i and the vehicle in front, which is used to adjust the density of the vehicle formation; The active controller transformation scheme actively adjusts the distributed model predictive controller in the case of normal communication according to the transformed communication topology structure. For vehicle i = m + 1, the constraints (9) of its optimal control problem and the last term of the cost function are changed as follows: When the loss of communication occurs between the leader and follower 1, the active communication topology transformation scheme is to establish a new communication connection between follower 2 and the leader and change the information flow direction between follower 2 and follower 1. At this time, the delay for follower 1 to obtain the emergency braking signal from the leader increases, that is, it leads to d exp,1 = d adj,1 + 2*τ max v max + d brake ; The active controller transformation scheme actively adjusts the distributed model predictive controller under normal communication according to the transformed communication topology structure; For vehicle i = 2, the constraints (9) of its optimal control problem and the end term of the cost function are also changed as shown in Eqs. (12) and (13) respectively; For vehicle i = 1, the constraints (9) of its optimal control problem and the end term of the cost function are changed as follows:

2. A longitudinal control system for vehicle formation under any communication link failure, characterized in that, It includes a processor, a memory, and a computer program stored on the memory. When the processor executes the computer program, it specifically executes the steps in a vehicle formation longitudinal control method under any communication link failure as claimed in claim 1.

Images