Zero-trust intelligent connected vehicle platoon control method under communication interruption based on DMPC

Abstract

The present invention specifically relates to a DMPC-based zero-trust intelligent connected vehicle platoon control method under communication interruption, comprising: defining a hypothetical output state sequence and a predicted output state sequence in an initial state; establishing an optimization objective cost function; for a single vehicle: solving the optimization objective cost function based on the hypothetical output state sequence, the predicted output state sequence, and the hypothetical output state sequences of its neighbors at the current moment to obtain the optimal predicted control input sequence at the current moment; calculating the optimal input state sequence and combining it with the optimal predicted control input sequence to calculate the optimal input state at the next moment; and calculating the hypothetical output state sequence at the next moment and distributing it to neighboring vehicles in communication with the vehicle. The present invention combines the current states of multiple parties to predict the future motion state of a vehicle and can adjust the driving state of each vehicle based on the communication topology of the vehicle platoon and the states of neighboring vehicles, thereby improving the stability of the longitudinal control of the intelligent connected vehicle platoon.

Description

Technical Field

[0001] The present invention relates to the fields of Internet of Vehicles and big data, and in particular to a DMPC-based zero-trust intelligent connected vehicle platoon control method under communication interruption. Background Art

[0002] In the connected vehicle (IoV) environment, vehicle-to-vehicle communication (VVCC) technology enables information sharing and interaction between vehicles, promoting the development of intelligent transportation. Connected vehicles can obtain information about surrounding vehicles (location, speed, acceleration, etc.), thereby alleviating traffic congestion and reducing traffic accidents, and has become a hot topic in current research.

[0003] Vehicle platooning couples vehicles in a platoon by introducing technologies like wireless communications. This, coupled with timely data sharing, expands vehicles' beyond-visual-range awareness of their surroundings, enabling earlier and more precise platoon control. This has resulted in a range of benefits, including improved overall traffic efficiency, reduced fuel consumption, and reduced exhaust emissions. Consequently, it has attracted worldwide attention.

[0004] Zero Trust (ZT) was first proposed in 2010 by then-Forrester analyst John Kindervag. Its core concept is that by default, no one, anything, or person inside or outside the network is trustworthy. Anyone, anything, or anything attempting to access the network and its resources must be verified before authorization. The key is to break the default trust, adopting the principle of "continuous verification, never trust." Zero Trust advocates a shift in security architecture from a network-centric to an identity-centric one, requiring all access behaviors to be subject to fine-grained, adaptive, identity-centric access control. Zero Trust represents a new generation of network security protection and has garnered increasing attention in recent years.

[0005] In a platoon of intelligent connected vehicles, the local area network connecting the vehicles operates in a zero-trust environment, changing randomly as the vehicles move. In essence, data obtained from other vehicle nodes is not completely reliable. For untrusted vehicle nodes, other vehicle nodes in the LAN employ corresponding information confirmation and filtering strategies to obtain reliable information. Therefore, in a zero-trust environment, intelligent connected vehicle platoons can effectively reach group consensus, thereby improving vehicle driving safety and enhancing road traffic capacity.

[0006] However, due to the time-varying traffic environment constraints of wireless communication systems, not only communication delays are included, but also practical application scenarios may also face environmental limitations, such as tunnels and mountainous areas, leading to temporary interruptions and restorations in inter-vehicle communication. To address this issue, the prior art proposes a distributed model predictive control (DMPC) method for vehicle platoon control systems under switching communication topologies. Specifically, a DMPC algorithm is designed for multi-vehicle systems switching communication topologies, and the effectiveness of the model predictive control algorithm under switching topologies is demonstrated through numerical simulations. Distributed model predictive control is a distributed form of model predictive control, suitable for large-scale system control problems. When the system is large and complex, with multiple inputs and outputs, a centralized model predictive controller must solve a large-scale optimization problem, which is computationally burdensome and inefficient. Instead, the original system can be decomposed into multiple coupled subsystems. A local MPC controller is designed for each subsystem to solve the local optimization problem and obtain a local optimal solution, fully utilizing the performance of each processor.

[0007] With the continuous advancement of wireless communication technologies (i.e., DSRC, LTE-V, and 5G), the types of communication topologies for vehicle platoons are becoming increasingly diverse. However, existing solutions for controlling intelligent connected vehicle platoons during communication interruptions mostly use a single communication topology. This fails to effectively adapt to diverse communication topologies and control intelligent connected vehicle platoons under the influence of communication anomalies, resulting in poor longitudinal control stability for intelligent connected vehicle platoons. Therefore, designing a method for achieving longitudinal stability control for intelligent connected vehicle platoons is an urgent technical challenge. Summary of the Invention

[0008] In response to the shortcomings of the above-mentioned existing technologies, the technical problem to be solved by the present invention is: how to provide a DMPC-based zero-trust intelligent connected vehicle platoon control method under communication interruption? This method can combine the current states of multiple parties to predict the future motion state of vehicles, and can adjust the driving state of each vehicle based on the communication topology of the vehicle platoon and the status of neighboring vehicles. This method can improve the stability of the longitudinal control of the intelligent connected vehicle platoon, greatly enhance driving safety and improve fuel economy, which is of great significance to alleviating traffic congestion, road safety and environmental pollution.

[0009] In order to solve the above technical problems, the present invention adopts the following technical solutions:

[0010] A DMPC-based zero-trust intelligent connected vehicle platoon control method under communication interruption includes:

[0011] S1: Define the hypothetical output state sequence and predicted output state sequence of each vehicle in the vehicle queue under the initial state;

[0012] S2: Establishing the corresponding optimization objective cost function based on the communication topology of each vehicle in the vehicle queue;

[0013] S3: For a single vehicle in the vehicle queue: obtain the neighbor hypothetical output state sequence of the neighbor vehicles communicating with the vehicle at the current moment, and solve its optimization objective cost function based on the vehicle's current hypothetical output state sequence, predicted output state sequence, and neighbor hypothetical output state sequence to obtain the optimal predicted control input sequence for the vehicle at the current moment;

[0014] S4: Calculating a corresponding optimal input state sequence based on the optimal predicted control input sequence of the vehicle at the current moment, and calculating the optimal input state at the next moment in combination with the optimal predicted control input sequence of the vehicle at the current moment, so as to control the driving state of the vehicle in the vehicle queue at the next moment;

[0015] S5: Calculate a hypothetical control input sequence for the next moment based on the optimal predicted control input sequence for the vehicle at the current moment, and then calculate a hypothetical output state sequence for the next moment in combination with a preset vehicle dynamics model. Distribute the hypothetical output state sequence for the vehicle at the next moment to neighboring vehicles with which it communicates;

[0016] S6: Repeat steps S3 to S5 until the vehicle queue control is completed.

[0017] Preferably, the hypothetical output state sequence of the vehicle is calculated by the following steps:

[0018] S101: defining a hypothetical control input sequence of the vehicle at time t=0;

[0019] The formula is described as:

[0020]

[0021] Where: represents the hypothetical control input sequence of vehicle i at time t = 0; Np represents the prediction time domain; k represents the kth time domain in the prediction time domain; v i (0) represents the speed of vehicle i at time t = 0; f represents the hypothetical control input function;

[0022] S102: Calculate the corresponding predicted input state sequence based on the assumed control input sequence of the vehicle at time t=0;

[0023] The formula is described as:

[0024]

[0025]

[0026]

[0027]

[0028] Where: represents the assumed control input sequence The corresponding predicted input state sequence; and x i (0) represents the predicted input state sequence and predicted input state of vehicle i at time t = 0; δt represents the discrete step length; p i (t) represents the vehicle position; v i (t) represents the vehicle speed; m i represents the vehicle mass; η T,i Indicates transmission efficiency; T i (t) represents the driving or braking torque required by the vehicle; C A,i represents the air resistance coefficient; g represents the gravity coefficient; f i represents the rolling resistance coefficient; τ i represents the inertial delay;

[0029] S103: Calculating a hypothetical output state sequence of the vehicle at time t=0 based on the predicted input state sequence;

[0030] The formula is described as:

[0031]

[0032]

[0033]

[0034] Where: represents the hypothetical output state sequence of vehicle i at time t = 0; Represents the predicted input state sequence The corresponding predicted output state sequence; Represents the output matrix.

[0035] Preferably, the predicted output state sequence is calculated using the following formula:

[0036]

[0037]

[0038] Where: represents the predicted output state sequence of vehicle i at time t = 0; and x i (0) represents the predicted input state sequence and predicted input state of vehicle i at time t = 0; Represents the output matrix.

[0039] Preferably, the optimization objective cost function of the vehicle is expressed by the following formula:

[0040]

[0041]

[0042]

[0043]

[0044]

[0045]

[0046]

[0047] in, represents the control variable to be optimized;

[0048] Representing a collection The number of elements in ;

[0049]

[0050]

[0051]

[0052] represents the deviation of the position information between vehicles i and j;

[0053] Where: represents the optimization target cost of vehicle i at time t; They represent the predicted output state, assumed output state, and neighbor assumed output state of vehicle i at time t respectively; represents the predicted input state of vehicle i at time t; They represent the predicted output state sequence, hypothetical output state sequence, and neighbor hypothetical output state sequence of vehicle i at time t respectively; represents the predicted input state sequence of vehicle i at time t; represents the predicted input state sequence of vehicle i at time t; represents the predictive control input sequence of vehicle i at time t; represents the neighborhood set of vehicle i at time t; represents the in-degree matrix following vehicle i at time t; represents the traction set of vehicle i at time t, p i(t) = 1 means that the following vehicle i can obtain the state information of the leading vehicle node at time t, i (t) = 0 means that the following vehicle i cannot obtain the status information of the pilot vehicle node at time t; represents the predictive control input; u min,i Indicates the minimum value of the predictive control input; u max,i Indicates the maximum value of the predictive control input; Represents the hypothetical output sequence in the prediction time domain; Represents the predicted output torque in the prediction time domain; Indicates the prediction speed within the prediction time domain.

[0054] Preferably, the vehicle also obtains the expected output sequence of the leading vehicle at the current moment, and then solves its optimization target cost function based on the expected output sequence of the vehicle at the current moment, the assumed output state sequence, the predicted output state sequence and the neighbor's assumed output state sequence to obtain the optimal predicted control input sequence of the vehicle at the current moment.

[0055] Preferably, the optimization objective cost function of the vehicle is solved by the following formula:

[0056]

[0057] Where: y i,des (k|t) represents the predicted output state sequence, hypothetical output state sequence, neighbor hypothetical output state sequence, and expected output sequence of vehicle i at time t respectively; represents the predicted input state sequence of vehicle i at time t; represents the deviation of the position information between vehicles i and j; when vehicle i can receive the state information of the pilot vehicle, Q i >0, otherwise, Q i =0; R i ≥0 reflects that the vehicle queue tends to travel at a constant speed; F i Reflects the assumed reference trajectory that vehicle i makes for itself; G i represents the average cooperative cost function weight matrix of vehicle i and its neighboring vehicles, describing the cooperative control target of the vehicle queue; Indicates the prediction speed; represents the hypothetical output sequence of vehicle j.

[0058] Preferably, the optimal input state at the next moment is calculated by the following steps:

[0059] S401: Calculating the corresponding optimal input state sequence based on the optimal predicted control input sequence of the vehicle at the current moment;

[0060] The formula is described as follows:

[0061]

[0062] Where: represents the optimal predictive control input sequence of vehicle i at time t; represents the optimal input state sequence of vehicle i at time t;

[0063] S402: Selecting the first input state from the optimal input state sequence of the vehicle at the current moment as the corresponding input state;

[0064] The formula is described as follows:

[0065]

[0066] Where: represents the optimal input state sequence The first input state in x i (t) represents the input state of vehicle i at time t;

[0067] S403: Selecting the first predictive control input from the optimal predictive control input sequence of the vehicle at the current moment as the controller;

[0068] The formula is described as follows:

[0069]

[0070] Where: Represents a controller; represents the optimal predictive control input sequence The first predictive control input in ;

[0071] S404: Calculating the optimal input state at the next moment based on the optimal input state information at the current moment and the controller;

[0072] The formula is described as follows:

[0073]

[0074] Where: x i (t+1) represents the input state of vehicle i at time t+1.

[0075] Preferably, the optimal input state includes the vehicle position and vehicle speed, that is, the vehicle position and vehicle speed of the vehicle in the vehicle queue at the next moment are predicted through the optimal input state.

[0076] Preferably, the vehicle dynamics model is expressed by the following formula:

[0077] For a single vehicle:

[0078] x i (t+1)=φ i (x i )+ψ i ·u i (t),i∈N;

[0079] y i (t) = ηx i (t);

[0080]

[0081]

[0082]

[0083] Where: x i (t+1) represents the input state of vehicle i at time t+1; u i (t) Control input of vehicle i at time t; x i (t) represents the input state of vehicle i at time t; represents the output matrix; δt represents the discrete step length; p i (t) represents the vehicle position; v i (t) represents the vehicle speed; m i represents the vehicle mass; η T,i Indicates transmission efficiency; T i (t) represents the driving or braking torque required by the vehicle; C A,i represents the air resistance coefficient; g represents the gravity coefficient; f i represents the rolling resistance coefficient; τ i Indicates inertial delay.

[0084] Preferably, the hypothetical output state sequence of the vehicle at the next moment is calculated by the following steps:

[0085] S501: Calculating a hypothetical control input sequence at the next moment based on the optimal predicted control input sequence of the vehicle at the current moment;

[0086] The formula is described as:

[0087]

[0088] Where: represents the hypothetical control input sequence of vehicle i at time t+1; represents the optimal predictive control input sequence of vehicle i at time t;

[0089] S502: Calculating hypothetical input state information at the next moment based on the hypothetical control input sequence of the vehicle at the next moment;

[0090] The formula is described as:

[0091]

[0092] Where: represents the hypothetical input state information of vehicle i at time t+1; represents the optimal predictive control input sequence of vehicle i at time t;

[0093] S503: Calculating a hypothetical output state sequence at the next moment based on the hypothetical input state information of the vehicle at the next moment;

[0094] The formula is described as:

[0095]

[0096] Where: represents the hypothetical output state sequence of vehicle i at time t+1.

[0097] Compared with the prior art, the DMPC-based zero-trust intelligent connected vehicle platoon control method under communication interruption has the following beneficial effects:

[0098] The present invention combines the vehicle's current hypothetical output state sequence and predicted output state sequence to solve its optimization objective cost function and obtain the optimal predicted control input sequence. The optimal input state sequence is then calculated based on the vehicle's current optimal predicted control input sequence. The optimal input state at the next moment is then calculated in combination with the vehicle's current optimal predicted control input sequence to control the vehicle's driving state in the vehicle queue at the next moment. This method, in other words, is able to predict the vehicle's future motion state by combining multiple current states, including the vehicle's own state, predicted state, and neighboring vehicle states, by solving an open-loop optimal control problem within a finite time domain. This improves the effectiveness of longitudinal control of intelligent connected vehicle queues.

[0099] The present invention establishes a corresponding optimization objective cost function based on the communication topology of vehicles in a vehicle platoon, obtains the neighbor hypothetical output state sequence of neighboring vehicles communicating with the vehicle, and solves the optimization objective cost function to obtain the optimal predicted control input sequence for the vehicle at the current moment. The optimal input state of the vehicle at the next moment is further calculated to control the driving state of the vehicle in the vehicle platoon at the next moment. That is, each vehicle can adaptively adjust its own driving state based on the communication topology of the vehicle platoon and the state of neighboring vehicles. It is applicable to different communication topologies, thereby improving the stability of the longitudinal control of the intelligent connected vehicle platoon, and can greatly enhance driving safety and improve fuel economy. It is of great significance to alleviate traffic congestion, road safety and environmental pollution. BRIEF DESCRIPTION OF THE DRAWINGS

[0100] In order to make the purpose, technical solutions and advantages of the invention more clear, the present invention will be further described in detail below with reference to the accompanying drawings, in which:

[0101] Figure 1 The flowchart of the DMPC-based zero-trust intelligent connected vehicle platoon control method under communication interruption;

[0102] Figure 2 Predict control flow graph for the model;

[0103] Figure 3 An example of a switching communication topology;

[0104] Figure 4 The figure shows the communication topology switching diagram. DETAILED DESCRIPTION

[0105] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. The components of the embodiments of the present invention generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the invention claimed for protection, but only represents selected embodiments of the present invention. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0106] It should be noted that similar reference numerals and letters denote similar items in the following figures. Therefore, once an item is defined in one figure, it does not require further definition or explanation in subsequent figures. In the description of the present invention, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer" indicate positions or relationships based on the positions or relationships shown in the figures, or the positions or relationships in which the inventive product is typically placed when in use. These terms are intended solely to facilitate the description of the present invention and simplify the description. They do not indicate or imply that the devices or components referred to must have a specific orientation, be constructed, or operate in a specific orientation, and are therefore not to be construed as limiting the present invention. Furthermore, the terms "first," "second," and "third," etc., are used solely to distinguish descriptions and are not to be construed as indicating or implying relative importance. Furthermore, terms such as "horizontal" and "vertical" do not imply that a component is absolutely horizontal or overhanging, but rather may be slightly tilted. For example, "horizontal" simply means that its direction is more horizontal than "vertical," and does not mean that the structure must be completely horizontal, but rather may be slightly tilted. In the description of the present invention, it should also be noted that, unless otherwise expressly specified or limited, the terms "disposed," "installed," "connected," and "connected" should be understood in a broad sense. For example, they may refer to fixed connections, detachable connections, or integral connections; they may refer to mechanical connections or electrical connections; they may refer to direct connections or indirect connections through an intermediate medium; and they may refer to internal communication between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on the specific circumstances.

[0107] The following is a further detailed description through specific implementation methods:

[0108] Example:

[0109] This embodiment discloses a DMPC-based zero-trust intelligent connected vehicle platoon control method under communication interruption.

[0110] like Figure 1 and Figure 2 As shown in FIG, a DMPC-based zero-trust intelligent connected vehicle platoon control method under communication interruption includes:

[0111] S1: Define the hypothetical output state sequence and predicted output state sequence of each vehicle in the vehicle queue under the initial state;

[0112] S2: Establishing the corresponding optimization objective cost function based on the communication topology of each vehicle in the vehicle queue;

[0113] In this embodiment, the vehicle platoon switches between four different communication topologies: PF, LPF, TPF, and LTPF. Figure 4 shown.

[0114] S3: For a single vehicle in the vehicle queue: obtain the neighbor hypothetical output state sequence of the neighbor vehicles communicating with the vehicle at the current moment, and solve its optimization objective cost function based on the vehicle's current hypothetical output state sequence, predicted output state sequence, and neighbor hypothetical output state sequence to obtain the optimal predicted control input sequence for the vehicle at the current moment;

[0115] S4: Calculating a corresponding optimal input state sequence based on the optimal predicted control input sequence of the vehicle at the current moment, and then calculating the optimal input state of the vehicle at the next moment based on the optimal predicted control input sequence and the optimal input state sequence of the vehicle at the current moment, so as to control the driving state of the vehicle in the vehicle queue at the next moment;

[0116] In this embodiment, the driving state of the vehicle includes the vehicle position and the vehicle speed.

[0117] S5: Calculate a hypothetical control input sequence for the next moment based on the optimal predicted control input sequence for the vehicle at the current moment, and then calculate a hypothetical output state sequence for the next moment in combination with a preset vehicle dynamics model. Distribute the hypothetical output state sequence for the vehicle at the next moment to neighboring vehicles with which it communicates;

[0118] In this embodiment, a vehicle distributes its hypothetical output state sequence to neighboring vehicles with which it communicates through V2V wireless communication technology.

[0119] S6: Repeat steps S3 to S5 until the vehicle queue control ends (the preset control time is reached).

[0120] The present invention combines the vehicle's current hypothetical output state sequence and predicted output state sequence to solve its optimization objective cost function and obtain the optimal predicted control input sequence. The optimal input state sequence is then calculated based on the vehicle's current optimal predicted control input sequence. The optimal input state at the next moment is then calculated in combination with the vehicle's current optimal predicted control input sequence to control the vehicle's driving state in the vehicle queue at the next moment. This method, in other words, is able to predict the vehicle's future motion state by combining multiple current states, including the vehicle's own state, predicted state, and neighboring vehicle states, by solving an open-loop optimal control problem within a finite time domain. This improves the effectiveness of longitudinal control of intelligent connected vehicle queues.

[0121] The present invention establishes a corresponding optimization objective cost function based on the communication topology of vehicles in a vehicle platoon, obtains the neighbor hypothetical output state sequence of neighboring vehicles communicating with the vehicle, and solves the optimization objective cost function to obtain the optimal predicted control input sequence for the vehicle at the current moment. The optimal input state of the vehicle at the next moment is further calculated to control the driving state of the vehicle in the vehicle platoon at the next moment. That is, each vehicle can adaptively adjust its own driving state based on the communication topology of the vehicle platoon and the state of neighboring vehicles. It is applicable to different communication topologies, thereby improving the stability of the longitudinal control of the intelligent connected vehicle platoon, and can greatly enhance driving safety and improve fuel economy. It is of great significance to alleviate traffic congestion, road safety and environmental pollution.

[0122] In this example, consider a platoon of N+1 intelligent connected vehicles in a zero-trust environment, traveling longitudinally on a flat road with a constant spacing, ignoring the pitch angles of the vehicles. Vehicles sense and transmit vehicle status information, including absolute position, velocity, and acceleration, through V2V wireless communication, GPS positioning, and sensors such as lidar. In a zero-trust environment, a platoon should not implicitly trust any person, device, system, or signal within or outside the network. Instead, each vehicle node should first verify trust in the node before transmitting signals. Only after confirming the authenticity of the transmitted signal will vehicle status information be transmitted. Figure 3 A schematic diagram of communication topology switching is shown: in a zero-trust environment, due to the interruption of communication between vehicles due to the untrustworthiness of some nodes, the communication topology of the vehicle queue is switched from the LPF communication topology to the TPF communication topology.

[0123] In this embodiment, the four different communication topologies of PF, LPF, TPF and LTPF are switched as follows: Figure 4 shown.

[0124] The communication topology of the intelligent connected vehicle platoon at time t can be Characterization. In a dynamic directed graph middle, represents the set of N following vehicle nodes, represents a time-varying set of edges, Represents a set of time-varying adjacency matrices. Similarly, In, a ij (t) = 1 means that at time t, vehicle node i can receive the status information of vehicle node j, ij (t) = 0 means that at time t, vehicle node i cannot receive the status information of vehicle node j, and it is assumed that there is no self-loop in this paper, that is, for any time, a ij (t)≠1. If the connected graph A subgraph of The tree of all nodes of , then the subgraph is called The spanning tree.

[0125] Assumptions in Represents the number of candidate communication topologies. and They represent the in-degree matrix and out-degree matrix following vehicle i at time t, namely:

[0126]

[0127]

[0128] Notice, It only contains the following vehicle number.

[0129] In order to characterize the information interaction relationship between the leading vehicle node and the following vehicle node, the following traction set is defined:

[0130]

[0131] Among them, p i (t) = 1 means that the following vehicle node i can obtain the status information of the leading vehicle node at time t, i (t) = 0 means that the following vehicle node i cannot obtain the state information of the leading vehicle node at time t. In order to obtain all the in-degree vehicle information of vehicle node i, the neighborhood set of vehicle node i in the entire intelligent connected vehicle queue is defined as:

[0132]

[0133] In practical applications, vehicle dynamics models are often nonlinear, and the dynamic characteristics of each vehicle are also different. The nonlinear dynamics prediction model under discrete state is designed as follows:

[0134] The state vector of a single vehicle is denoted as x i (t)=[p i (t),v i (t),T i (t)] T , the control input is recorded as u i (t), the discrete step length is recorded as δt

[0135]

[0136] Furthermore, the above formula can be written as:

[0137] x i (t+1)=φ i (xi )+ψ i ·u i (t),i∈N;

[0138] in, The detailed expression is:

[0139]

[0140] The state output equation of intelligent connected vehicle i is:

[0141] y i (t)=[p i (t)v i (t)] T =ηx i (t);

[0142] in, is the output matrix, that is:

[0143]

[0144] At the same time, in order to characterize the lumped discrete nonlinear prediction model of the entire intelligent connected vehicle fleet, let: and denote the lumped state vector, state output vector, and control input vector of the vehicle platoon respectively. Then, the overall dynamic prediction model of the vehicle platoon can be expressed as:

[0145] X(t+1)=Φ(X(t))+Ψ·U(t);

[0146] Y(t)=ΓX(t).

[0147] Where Φ=[φ1(x1) T ,φ2(x2) T ,…,φ N (x N ) T ] T , Ψ=diag{ψ1,ψ2,…,ψ N}, Γ=diag{η, η,..., η}.

[0148] Therefore, the vehicle dynamics model is expressed as follows:

[0149] For a single vehicle:

[0150] x i (t+1)=φ i (x i )+Ψ i ·u i (t),i∈N;

[0151] Here, x i represents the input of the i-th vehicle.

[0152] y i (t) = ηx i (t);

[0153]

[0154]

[0155]

[0156] Where: x i (t+1) represents the input state of vehicle i at time t+1; u i (t) Control input of vehicle i at time t; x i (t) represents the input state of vehicle i at time t; represents the output matrix; δt represents the discrete step length; p i (t) represents the vehicle position; v i (t) represents the vehicle speed; m i represents the vehicle mass; η T,i Indicates transmission efficiency; T i (t) represents the driving or braking torque required by the vehicle; C A,i represents the air resistance coefficient; g represents the gravity coefficient; f i represents the rolling resistance coefficient; τ i Indicates inertial delay.

[0157] For vehicle queues:

[0158] X(t+1)=Φ(X(t))+Ψ·U(t);

[0159] Y(t) = ΓX(t);

[0160]

[0161]

[0162]

[0163] Φ=[φ1(x1) T ,φ2(x2) T ,…,φ N (x N ) T ] T ;

[0164] Ψ=diag{ψ1,ψ2,…,ψ N};

[0165] Γ = diag{η,η,…,η};

[0166] Where: N represents the number of vehicles in the vehicle queue.

[0167] The vehicle dynamics model constructed in the present invention is used to describe the vehicle's ability to move on a two-dimensional plane. The change in the vehicle's motion state is due to the resultant force acting on its tires, but the microscopic force process of the tires is extremely complex, requiring a large number of complex equations and data-calibrated parameters to characterize its laws. It is not practical to build a microscopic physical model to study the control method of intelligent connected vehicle platoons. This is because even if the complex model is simplified to a considerable extent, it can still accurately describe the vehicle's motion laws, and the simplified model is more convenient for theoretical analysis. Since the present invention is aimed at the control of intelligent connected vehicle platoons, it only needs to consider the vehicle's ability to move on a one-dimensional road, that is, the vehicle's longitudinal dynamics model. The longitudinal force of an intelligent connected vehicle is mainly determined by the engine, transmission system, braking system, air resistance, tire friction, rolling resistance and gravity. Due to the complex and changeable road environment in actual applications, existing methods often use simple dynamic models. The present invention designs a more accurate vehicle dynamics model for the DMPC algorithm.

[0168] In the specific implementation process, the following steps are used to calculate the hypothetical output state sequence of the vehicle:

[0169] S101: defining a hypothetical control input sequence of the vehicle at time t=0;

[0170] The formula is described as:

[0171]

[0172] Where: represents the hypothetical control input sequence of vehicle i at time t = 0; Np represents the prediction time domain; k represents the kth time domain in the prediction time domain; v i (0) represents the speed of vehicle i at time t = 0; f represents the hypothetical control input function;

[0173] S102: Calculate the corresponding predicted input state sequence based on the assumed control input sequence of the vehicle at time t=0;

[0174]

[0175]

[0176]

[0177]

[0178] In this embodiment, They represent the optimal input state sequences from time 0 to k and from time 0 to k+1 respectively.

[0179] Where: represents the assumed control input sequence The corresponding predicted input state sequence; and x i (0) represents the predicted input state sequence and predicted input state of vehicle i at time t = 0; δt represents the discrete step length; p i (t) represents the vehicle position; v i (t) represents the vehicle speed; m i represents the vehicle mass; η T,i Indicates transmission efficiency; T i (t) represents the driving or braking torque required by the vehicle; C A,i represents the air resistance coefficient; g represents the gravity coefficient; f i represents the rolling resistance coefficient; τ i represents the inertial delay;

[0180] S103: Calculating a hypothetical output state sequence of the vehicle at time t=0 based on the predicted input state sequence;

[0181] The formula is described as:

[0182]

[0183]

[0184]

[0185] Where: represents the hypothetical control input sequence of vehicle i at time t = 0; Represents the predicted input state sequence The corresponding predicted output state sequence; Represents the output matrix.

[0186] The predicted output state sequence is calculated using the following formula:

[0187]

[0188]

[0189] Where: represents the predicted output state sequence of vehicle i at time t = 0; and x i (0) represents the predicted input state sequence and predicted input state of vehicle i at time t = 0; Represents the output matrix.

[0190] In the specific implementation process, the discrete state space model is assumed to be as follows:

[0191] x(k+1)=f(x(k),u(k)),x(0)=x0;

[0192] y(k)=h(x(k),u(k));

[0193] Where: x(k) represents the state quantity of the controlled system, u(k) represents the control input quantity of the system, y(k) represents the control output quantity of the system, and x0 represents the initial state quantity of the system.

[0194] Based on the above discrete state space prediction model, the output of the next Np time domains is predicted and the control quantity in the control time domain Nc is expressed as follows:

[0195]

[0196]

[0197] At the same time, it is necessary to optimize the deviation between the system output and the reference quantity and the system control quantity to ensure that the target cost function can enable the controlled system to quickly and smoothly track the desired reference trajectory. The target function is often in the following form:

[0198]

[0199] Where: Q and R are weight matrices, and r represents the reference trajectory.

[0200] Taking into account the optimization objectives and constraints, the specific optimization problem can be described as follows:

[0201]

[0202] The above optimization problem can be properly processed at the same time, and the quadratic programming problem of control constraints and outputs can be solved to obtain the optimal control input sequence, and the first component can be selected as the actual control quantity at the current moment.

[0203] Combine the member vehicle i in the queue and the neighbor set The state of the neighboring vehicles in the local rolling horizon optimization objective cost function is established. The model predictive control algorithm of all vehicles in the intelligent connected vehicle queue adopts the same prediction horizon Np. In each prediction horizon [t, t+Np], three control input sequences are defined and are the predicted control input sequence, the optimal predicted control input sequence and the hypothetical control input sequence, k∈[0,1,2,…,Np-1,Np]. Correspondingly, and They are the predicted output sequence, optimal predicted output sequence, and hypothetical output sequence obtained by applying the above control variables to the intelligent connected vehicle platoon system. The hypothetical output sequence will be sent to neighboring vehicles as one of the parameters of the target cost function on the vehicle node.

[0204] Therefore, the following optimization objective cost function can be obtained:

[0205]

[0206]

[0207]

[0208] Indicates that the first input state of the predicted input state sequence is used as the predicted input state;

[0209]

[0210]

[0211]

[0212] in, represents the control variable to be optimized;

[0213] Representing a collection The number of elements in ;

[0214]

[0215]

[0216]

[0217] represents the deviation of the position information between vehicles i and j;

[0218] because Parameters such as are calculated based on the communication topology of the vehicle platoon, so the optimization objective cost function of the vehicle can be considered to be constructed based on the communication topology of the vehicle.

[0219] Where: represents the optimization target cost of vehicle i at time t, J i represents the vehicle platoon cooperative tracking cost function; They represent the predicted output state, assumed output state, and neighbor assumed output state of vehicle i at time t respectively; represents the predicted input state of vehicle i at time t; They represent the predicted output state sequence, hypothetical output state sequence, and neighbor hypothetical output state sequence of vehicle i at time t respectively; represents the predicted input state sequence of vehicle i at time t; represents the predicted input state sequence of vehicle i at time t; represents the predictive control input sequence of vehicle i at time t; represents the neighborhood set of vehicle i at time t; represents the in-degree matrix following vehicle i at time t; represents the traction set of vehicle i at time t, p i (t) = 1 means that the following vehicle i can obtain the state information of the leading vehicle node at time t, i (t) = 0 means that the following vehicle i cannot obtain the status information of the pilot vehicle node at time t; represents the predictive control input; u min,i Indicates the minimum value of the predictive control input; u max,i Indicates the maximum value of the predictive control input; represents the hypothetical output sequence in the prediction time domain; T i p )Np|t) represents the predicted output torque in the prediction time domain; Indicates the prediction speed within the prediction time domain.

[0220] During the specific implementation process, the vehicle also obtains the expected output sequence of the leading vehicle at the current moment, and then solves its optimization target cost function based on the vehicle's expected output sequence at the current moment, the assumed output state sequence, the predicted output state sequence and the neighbor's assumed output state sequence to obtain the optimal predicted control input sequence of the vehicle at the current moment.

[0221] The optimized objective cost function of the vehicle is solved by the following formula:

[0222]

[0223] Where: y i,des (k|t_represents the predicted output state sequence, hypothetical output state sequence, neighbor hypothetical output state sequence, and expected output sequence of vehicle i at time t respectively; represents the predicted input state sequence of vehicle i at time t; represents the deviation of the position information between vehicles i and j; Q i 、R i 、F i , G i Represents the weight matrix, which is a symmetric non-negative definite matrix. Its calculation method is defined as Where: When vehicle i can receive the state information of the pilot vehicle, Qi >0, otherwise, Q i =0; R i ≥0 reflects that the vehicle queue tends to travel at a constant speed; F i Reflects the assumed reference trajectory that vehicle i makes for itself; G i represents the average cooperative cost function weight matrix of vehicle i and its neighboring vehicles, describing the cooperative control target of the vehicle queue; Indicates the prediction speed; represents the hypothetical output sequence of vehicle j.

[0224] It is worth noting that the target optimization problem is a nonlinear and non-convex programming problem. If it is assumed that it has a feasible solution, the existing numerical algorithms such as SQP and active set method can be applied to solve the optimal solution.

[0225] The present invention constructs and solves the aforementioned optimization objective cost function. By solving an open-loop optimal control problem within a finite time domain, the present invention can predict the vehicle's future motion state by combining its own current state, predicted state, and neighboring vehicle states, thereby improving the effectiveness of longitudinal control of intelligent connected vehicle platoons.

[0226] In the specific implementation process, the optimal input state at the next moment is calculated through the following steps:

[0227] S401: Calculating the corresponding optimal input state sequence based on the optimal predicted control input sequence of the vehicle at the current moment;

[0228] The formula is described as follows:

[0229]

[0230] Where: u i * (k|t) represents the optimal predictive control input sequence of vehicle i at time t; represents the optimal input state sequence of vehicle i at time t; They represent the optimal input state sequences from time t to k and from time t to k+1 respectively.

[0231] S402: Selecting the first input state from the optimal input state sequence of the vehicle at the current moment as the corresponding input state;

[0232] The formula is described as follows:

[0233]

[0234] Where: represents the optimal input state sequence The first input state in x i(t) represents the input state of vehicle i at time t;

[0235] S403: Selecting the first predictive control input from the optimal predictive control input sequence of the vehicle at the current moment as the controller;

[0236] The formula is described as follows:

[0237] make is the optimal solution of the target cost function at time t. According to the rolling optimization time domain control principle, the optimal solution at time t is The first component in serves as the controller for the local model predictive control of vehicle i, namely:

[0238]

[0239] Where: Represents a controller; represents the optimal predictive control input sequence The first predictive control input in ;

[0240] S404: Calculating the optimal input state at the next moment based on the optimal input state information at the current moment and the controller;

[0241] The formula is described as follows:

[0242]

[0243] Where: x i (t+1) represents the input state of vehicle i at time t+1 (i.e., the next time).

[0244] Specifically, the optimal input state includes the vehicle position and vehicle speed, that is, the vehicle position and vehicle speed of the vehicle in the vehicle queue at the next moment are controlled by the optimal input state.

[0245] In the specific implementation process, the following steps are used to calculate the hypothetical output state sequence of the vehicle at the next moment:

[0246] S501: Calculating a hypothetical control input sequence at the next moment based on the optimal predicted control input sequence of the vehicle at the current moment;

[0247] The formula is described as:

[0248]

[0249] Where: represents the hypothetical control input sequence of vehicle i at time t+1; represents the optimal predictive control input sequence of vehicle i at time t;

[0250] S502: Calculating hypothetical input state information at the next moment based on the hypothetical control input sequence of the vehicle at the next moment;

[0251] The formula is described as:

[0252]

[0253] Where: represents the hypothetical input state information of vehicle i at time t+1; represents the optimal predictive control input sequence of vehicle i at time t; Represent the hypothetical input state information from time t+1 to k and from time t+1 to k+1 respectively.

[0254] S503: Calculating a hypothetical output state sequence at the next moment based on the hypothetical input state information of the vehicle at the next moment;

[0255] The formula is described as:

[0256]

[0257] Where: represents the hypothetical output state sequence of vehicle i at time t+1.

[0258] The present invention aims to solve the problem of zero-trust intelligent connected vehicle platoon control under communication interruption, and proposes a distributed model prediction longitudinal vehicle platoon control method. The present invention uses the state trajectory information of adjacent vehicles to construct the target cost function, which can be applied to different communication topologies. First, the rolling domain control method is adopted to replace the global optimal solution with the local optimal solution, and the actual state of the system is used for feedback correction. Corresponding to each sampling moment, the state of the system at the future moment is predicted according to the prediction model, and the open-loop optimal control problem in the finite time domain is solved. Then, the first component in the solved optimal control sequence is used to control the system at the current moment. In addition, the prediction model is feedback-corrected according to the actual output of the system to realize the closed loop of the control optimization process. At the next sampling moment, the above optimization process is repeated to achieve optimal control of the system. This method will greatly improve driving safety and improve fuel economy, and is of great significance to alleviating traffic congestion, road safety and environmental pollution.

[0259] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit the technical solutions. Those skilled in the art should understand that modifications or equivalent replacements of the technical solutions of the present invention that do not depart from the purpose and scope of the technical solutions of the present invention should be included in the scope of the claims of the present invention.

Claims

1. A zero-trust intelligent connected vehicle platoon control method under communication interruption based on DMPC, characterized in that: include: S1: Define the hypothetical output state sequence and predicted output state sequence of each vehicle in the vehicle queue under the initial state; S2: Establishing the corresponding optimization objective cost function based on the communication topology of each vehicle in the vehicle queue; In step S2, the optimized target cost function of the vehicle is expressed by the following formula: s.t.1) 2) 3) 4) in, represents the control variable to be optimized; Representing a collection The number of elements in ; represents the deviation of the position information between vehicles i and j; Where: represents the optimization target cost of vehicle i at time t; They represent the predicted output state, assumed output state, and neighbor assumed output state of vehicle i at time t respectively; represents the predicted input state of vehicle i at time t; They represent the predicted output state sequence, hypothetical output state sequence, and neighbor hypothetical output state sequence of vehicle i at time t respectively; represents the predicted input state sequence of vehicle i at time t; represents the predicted input state sequence of vehicle i at time t; represents the predictive control input sequence of vehicle i at time t; represents the neighborhood set of vehicle i at time t; represents the in-degree matrix following vehicle i at time t; represents the traction set of vehicle i at time t, p i (t) = 1 means that the following vehicle i can obtain the state information of the leading vehicle node at time t, i (t) = 0 means that the following vehicle i cannot obtain the status information of the pilot vehicle node at time t; represents the predictive control input; u min,i Indicates the minimum value of the predictive control input; u max,i Indicates the maximum value of the predictive control input; Represents the hypothetical output sequence in the prediction time domain; Represents the predicted output torque in the prediction time domain; Indicates the prediction speed within the prediction time domain; S3: For a single vehicle in the vehicle queue: obtain the neighbor hypothetical output state sequence of the neighbor vehicles communicating with the vehicle at the current moment, and solve its optimization objective cost function based on the vehicle's current hypothetical output state sequence, predicted output state sequence, and neighbor hypothetical output state sequence to obtain the optimal predicted control input sequence for the vehicle at the current moment; S4: Calculating a corresponding optimal input state sequence based on the optimal predicted control input sequence of the vehicle at the current moment, and calculating the optimal input state at the next moment in combination with the optimal predicted control input sequence of the vehicle at the current moment, so as to control the driving state of the vehicle in the vehicle queue at the next moment; S5: Calculate a hypothetical control input sequence for the next moment based on the optimal predicted control input sequence for the vehicle at the current moment, and then calculate a hypothetical output state sequence for the next moment in combination with a preset vehicle dynamics model. Distribute the hypothetical output state sequence for the vehicle at the next moment to neighboring vehicles with which it communicates; S6: Repeat steps S3 to S5 until the vehicle queue control is completed.

2. The DMPC-based zero-trust intelligent connected vehicle platoon control method under communication interruption as claimed in claim 1, characterized in that: In step S1, the hypothetical output state sequence of the vehicle is calculated by the following steps: S101: defining a hypothetical control input sequence of the vehicle at time t=0; The formula is described as: Where: represents the hypothetical control input sequence of vehicle i at time t = 0; Np represents the prediction time domain; k represents the kth time domain in the prediction time domain; v i (0) represents the speed of vehicle i at time t = 0; f represents the hypothetical control input function; S102: Calculate the corresponding predicted input state sequence based on the assumed control input sequence of the vehicle at time t=0; The formula is described as: Where: represents the assumed control input sequence The corresponding predicted input state sequence; and x i (0) represents the predicted input state sequence and predicted input state of vehicle i at time t = 0; δt represents the discrete step length; p i (t) represents the vehicle position; v i (t) represents the vehicle speed; m i Indicates the vehicle mass; η T,i Indicates transmission efficiency; T i (t) represents the driving or braking torque required by the vehicle; C A,i represents the air resistance coefficient; g represents the gravity coefficient; f i represents the rolling resistance coefficient; τ i represents the inertial delay; S103: Calculating a hypothetical output state sequence of the vehicle at time t=0 based on the predicted input state sequence; The formula is described as: Where: represents the hypothetical output state sequence of vehicle i at time t = 0; Represents the predicted input state sequence The corresponding predicted output state sequence; Represents the output matrix.

3. The DMPC-based zero-trust intelligent connected vehicle platoon control method under communication interruption as claimed in claim 1, characterized in that: In step S1, the predicted output state sequence is calculated using the following formula: Where: represents the predicted output state sequence of vehicle i at time t = 0; and x i (0) represents the predicted input state sequence and predicted input state of vehicle i at time t = 0; Represents the output matrix.

4. The DMPC-based zero-trust intelligent connected vehicle platoon control method under communication interruption as claimed in claim 1, characterized in that: In step S3, the vehicle also obtains the expected output sequence of the leading vehicle at the current moment, and then solves its optimization target cost function based on the expected output sequence, assumed output state sequence, predicted output state sequence and neighbor assumed output state sequence of the vehicle at the current moment to obtain the optimal predicted control input sequence of the vehicle at the current moment.

5. The DMPC-based zero-trust intelligent connected vehicle platoon control method under communication interruption as claimed in claim 4, characterized in that: The optimized objective cost function of the vehicle is solved by the following formula: Where: y i,des (k|t) represents the predicted output state sequence, hypothetical output state sequence, neighbor hypothetical output state sequence, and expected output sequence of vehicle i at time t respectively; represents the predicted input state sequence of vehicle i at time t; represents the deviation of the position information between vehicles i and j; when vehicle i can receive the state information of the pilot vehicle, Q i >0, otherwise, Q i =0; R i ≥0 reflects that the vehicle queue tends to travel at a constant speed; F i Reflects the assumed reference trajectory that vehicle i makes for itself; G i represents the average cooperative cost function weight matrix of vehicle i and its neighboring vehicles, describing the cooperative control target of the vehicle queue; Indicates the prediction speed; represents the hypothetical output sequence of vehicle j.

6. The DMPC-based zero-trust intelligent connected vehicle platoon control method under communication interruption as claimed in claim 1, characterized in that: In step S4, the optimal input state at the next moment is calculated by the following steps: S401: Calculating the corresponding optimal input state sequence based on the optimal predicted control input sequence of the vehicle at the current moment; The formula is described as follows: Where: represents the optimal predictive control input sequence of vehicle i at time t; represents the optimal input state sequence of vehicle i at time t; S402: Selecting the first input state from the optimal input state sequence of the vehicle at the current moment as the corresponding input state; The formula is described as follows: Where: represents the optimal input state sequence The first input state in x i (t) represents the input state of vehicle i at time t; S403: Selecting the first predictive control input from the optimal predictive control input sequence of the vehicle at the current moment as the controller; The formula is described as follows: Where: Represents a controller; represents the optimal predictive control input sequence The first predictive control input in ; S404: Calculating the optimal input state at the next moment based on the optimal input state information at the current moment and the controller; The formula is described as follows: Where: x i (t+1) represents the input state of vehicle i at time t+1.

7. The DMPC-based zero-trust intelligent connected vehicle platoon control method under communication interruption as claimed in claim 1, characterized in that: In step S4, the optimal input state includes the vehicle position and vehicle speed, that is, the vehicle position and vehicle speed of the vehicle in the vehicle queue at the next moment are predicted through the optimal input state.

8. The DMPC-based zero-trust intelligent connected vehicle platoon control method under communication interruption as claimed in claim 1, characterized in that: In step S5, the vehicle dynamics model is expressed by the following formula: For a single vehicle: x i (t+1)=φ i (x i )+ψ i ·u i (t),i∈N; y i (t)=ηx i (t); Where: x i (t+1) represents the input state of vehicle i at time t+1; u i (t) Control input of vehicle i at time t; x i (t) represents the input state of vehicle i at time t; represents the output matrix; δt represents the discrete step length; p i (t) represents the vehicle position; v i (t) represents the vehicle speed; m i Indicates the vehicle mass; η T,i Indicates transmission efficiency; T i (t) represents the driving or braking torque required by the vehicle; C A,i represents the air resistance coefficient; g represents the gravity coefficient; f i represents the rolling resistance coefficient; τ i Indicates inertial delay.

9. The DMPC-based zero-trust intelligent connected vehicle platoon control method under communication interruption as claimed in claim 1, characterized in that: In step S5, the hypothetical output state sequence of the vehicle at the next moment is calculated by the following steps: S501: Calculating a hypothetical control input sequence at the next moment based on the optimal predicted control input sequence of the vehicle at the current moment; The formula is described as: Where: represents the hypothetical control input sequence of vehicle i at time t+1; represents the optimal predictive control input sequence of vehicle i at time t; S502: Calculating hypothetical input state information at the next moment based on the hypothetical control input sequence of the vehicle at the next moment; The formula is described as: Where: represents the hypothetical input state information of vehicle i at time t+1; represents the optimal predictive control input sequence of vehicle i at time t; S503: Calculating a hypothetical output state sequence at the next moment based on the hypothetical input state information of the vehicle at the next moment; The formula is described as: Where: represents the hypothetical output state sequence of vehicle i at time t+1.

Images