A design method for collaborative control strategy of vehicle platoons in merging areas based on IVCPS

Abstract

The present invention relates to a method for designing a collaborative control strategy for vehicle queues in a merging area based on IVCPS, and belongs to the field of intelligent traffic management and control. The method includes: for vehicles in the lane merging area, dividing them into several vehicle queues according to the initial spacing between vehicles; mapping the fleet in the ramp to the main road to form a virtual queue based on a virtual mapping method, and numbering the fleet in the virtual queue according to the distance from the fleet to the merging point; allocating the merging order of the fleet through mixed integer linear programming; from the perspective of IVCPS, the main road fleet plays a game according to the game strategy set, and controls the main road fleet to reserve a merging gap for the ramp fleet that needs to merge according to the game result. The present invention can effectively deal with the problems of low traffic efficiency of vehicles in the merging area under high traffic flow and increased energy consumption caused by frequent starting and stopping of vehicles, and minimize the impact of the merging of the ramp fleet on the main road.

Description

Technical Field

[0001] The present invention belongs to the field of intelligent traffic management and control, and relates to a design method for a collaborative control strategy of vehicle queues in a merging area based on IVCPS. Background Art

[0002] According to statistics, merging areas, as key nodes where vehicles merge onto the main line of urban expressways or highways, are high-incidence areas for traffic accidents. On the one hand, the conflicts and potential dangers caused by vehicles changing lanes, accelerating, and decelerating in merging areas are a major challenge to traffic safety. On the other hand, the mismatch between traffic demand and road network capacity in merging areas can cause temporary traffic jams, and frequent starting and stopping can also lead to increased energy consumption. Therefore, appropriate control of vehicles in merging areas is necessary to ensure safe and smooth traffic.

[0003] With the advancement of computing, control, and communication technologies, traditional single-point technologies are no longer able to meet the demands of the new generation of informatization and networking. This has led to the emergence of cyber-physical systems (CPS). The deep integration of CPS with transportation systems connects physical entities such as people, vehicles, and roads with information and applications. This empowers traditional transportation systems with the capabilities of perception, judgment, control, and decision-making, promoting comprehensive improvements in transportation safety and efficiency.

[0004] Against the backdrop of the rapid development of intelligent connected vehicles (ICVs), which can exchange information with roadside equipment and the cloud through V2X technology, a vehicle collaborative control framework built under the Intelligent Vehicles CyberPhysical Systems (IVCPS) is designed to design collaborative control strategies for vehicle platoons. By enabling the emergence of functions through collaborative vehicle decision-making, it is expected to address the low efficiency of vehicles in merging areas under high traffic flows, as well as the increased energy consumption caused by frequent vehicle starts and stops.

[0005] When existing vehicles make collaborative decisions, there are still the following deficiencies:

[0006] (1) Existing research does not fully utilize the characteristics of IVCPS, but instead focuses on platoon stability or merging zone efficiency in isolation. This results in the lack of a collaborative design framework that comprehensively considers multiple scales, making it impossible to fully optimize the overall performance of traffic flow.

[0007] (2) The merging order of vehicles is the core issue of coordinated merging. However, most existing studies are based on the first-in-first-out (FIFO) approach and do not consider the speed difference between ramps and main roads. Research on merging order is mainly divided into optimal strategies and suboptimal strategies. The computational complexity of the optimal strategy increases sharply when the traffic volume is high. The solution of the suboptimal strategy is only better for the current optimization object and may not be beneficial to the overall traffic operation.

[0008] (3) The actual driving environment of vehicles is complex and changeable. When ramp vehicles merge, vehicles on the outer side of the main road are not only affected by the ramp vehicles, but also need to consider the overall traffic situation and the vehicle conditions in other lanes. In most studies on coordinated decision-making between two vehicles in merging areas, the dynamic changes of surrounding traffic are not taken into account, and the possibility of lateral coordination is not considered. Summary of the Invention

[0009] In view of this, the purpose of this invention is to provide a design method for a coordinated control strategy for vehicle queues in merging areas based on ICVPS, which reduces computational complexity, optimizes traffic efficiency, and improves fuel economy.

[0010] In order to achieve the above object, the present invention provides the following technical solutions:

[0011] A design method for a coordinated control strategy for vehicle platoons in a merging area based on IVCPS is proposed. Specifically:

[0012] For vehicles in the lane merging area, they are divided into several vehicle queues according to the initial distance between the vehicles;

[0013] Based on the distance the convoy reaches the merging point, the convoys on the ramp are mapped to the main road to form a virtual queue based on virtual mapping. The convoys in the virtual queue are numbered according to the distance the convoy reaches the merging point.

[0014] The merging order of the merging area is planned based on the vehicle queue. The merging order is assigned through mixed integer linear programming. Each fleet passes through the merging point in sequence according to the merging order.

[0015] From the perspective of IVCPS, if there is a ramp team between two teams with adjacent merging orders, the main road team will play a game according to the game strategy set. According to the game results, the main road team is controlled to reserve a merging gap for the ramp team that needs to merge. Among them, if the two teams with adjacent merging orders are both main road teams, no game is played.

[0016] Furthermore, the vehicle queue is divided according to the vehicle spacing threshold, and the vehicle spacing threshold is expressed as:

[0017]

[0018] Where, v max 、amax Represent the maximum speed and maximum acceleration of the vehicle, v i represents the speed of vehicle i, L p Represents the length of the vehicle formation area; if the initial vehicle distance d of the i-th vehicle i,i-1 (t0)>d T When , the i-th vehicle is regarded as the leader vehicle of the newly formed vehicle queue; otherwise, it is classified as the following vehicle of the front vehicle queue.

[0019] Furthermore, the merging order of the merging area is planned based on the vehicle queue. The merging order is described as a MILP problem, and the objective function of the problem is defined as:

[0020]

[0021] Where, Represents each fleet LP in the virtual control area i The time allocated to pass through the area and the minimum passing time; the objective function represents minimizing the delay time of all fleets passing through the merging point.

[0022] Based on the objective function, the MILP problem is constructed by combining speed, safety, and merging constraints:

[0023]

[0024]

[0025]

[0026]

[0027] C4:s i ={0,1}

[0028] Where, are the longest and shortest time for queue LPi to leave the merging area, si is a binary judgment parameter, and s i =1 indicates fleet LP i For the main road fleet, otherwise s i =0;v i ′ represents the fleet LP i The velocity, v′ max 、v′ min Represents fleet LP i The maximum and minimum speeds, a′ max , a′ min Represents fleet LP i The maximum and minimum acceleration, L c Indicates the distance between the convoy and the merging point, Li Indicates the length of the current fleet, p i represents the current position of the pilot car in the convoy; the travel time obtained by solving the MILP problem Determine the confluence order.

[0029] Furthermore, the game strategy set is shown in the following table:

[0030]

[0031] Among them, LP O Indicates the convoy in the lane on the main road close to the ramp side, LP I Represents LP O The convoy in the adjacent lane; k Indicates fleet LP O Strategy, k=1,2; I l Indicates fleet LP I strategy, l = 1, 2, 3;

[0032] Indicates fleet LP O The revenue function of vehicle i in , ε i 、 δ i Represents fleet LP O The safety benefit, time benefit, and stability benefit of vehicle i, α i , β i ,λ i They represent safety benefit factor, time benefit factor, and stability benefit factor, respectively. i +β i +λ i =1;

[0033] Indicates fleet LP I The revenue function of vehicle j in is, ε j 、 Represents fleet LP I The safety benefit and time benefit of vehicle j, α j , β j They represent the safety benefit factor and time benefit factor respectively, and they represent the weights of safety benefit and time benefit, and α j +β j =1.

[0034] According to the game strategy set, calculate the fleet LP O The benefits of vehicle i before and after changing lanes Compare and The size of selects the execution strategy of vehicle i; the platoon LP of the target lane of lane change I The vehicle j in the game calculates the profit based on the strategy of vehicle i. For fleet LP I and vehicle i, and the Stackelberg game is used to solve the longitudinal control strategies of the two.

[0035] Among them, Stackelberg game is used to solve the fleet LP I The process of the longitudinal control strategy of vehicle i includes:

[0036] 1) Determine the profit function of the leader and followers: the leader is the LP of the team O The vehicle i in the example has a profit function of U I (I l ,I n ), indicating fleet LP I Selection Strategy I n When vehicle i selects strategy I l The profit of vehicle i at this time; the follower is the fleet LP I , the profit function is U I (I l ,I n ), indicating fleet LP O Vehicle i selection strategy O k When the fleet LP I Vehicle j in the selection strategy I l The revenue of vehicle j at time ;

[0037] 2) Design the follower's optimal response strategy as follows:

[0038]

[0039] 3) Design the leader's optimization function to select a strategy that maximizes benefits under the follower's optimal response. The leader's optimization function is expressed as:

[0040]

[0041] 4) Set the follower's optimal response strategy I n I in l Replaced with the leader's optimization function The final optimal response strategy of the follower is obtained, that is, thereby and The composed strategies constitute the optimal vertical control decisions of leaders and followers.

[0042] The beneficial effects of the present invention are:

[0043] (1) This paper designs a vehicle collaborative control architecture based on the multi-scale perspective of IVCPS, targeting the characteristics of merging area scenarios and the dynamic and emergent characteristics of the cyber-physical system of intelligent connected vehicles. It also proposes corresponding design solutions at different scales and architectural layers. Through the collaborative perception and decision-making of different structural layers under IVCPS, combined with the multi-scale empowerment of vehicle-road-cloud integrated technology, it can effectively address the low traffic efficiency of vehicles in merging areas under high traffic flow and the increased energy consumption caused by frequent vehicle starts and stops.

[0044] (2) Starting from the system-level scale, the present invention addresses the problem of difficulty in achieving a balance between optimal decision-making and computational efficiency when determining the merging order under high traffic flow. Based on the idea of ​​vehicle platooning, multiple vehicle queues are constructed to make collaborative decisions to narrow the solution scope, and mixed integer linear programming is used to optimize the fleet traffic, thereby reducing the complexity of decision-making calculations.

[0045] (3) Starting from the unit-level scale, the present invention addresses the problem of vehicle competition in vehicle collaborative control under dynamic traffic environments. Based on the optimized merging order, the present invention considers the impact of surrounding traffic participants and traffic environment on the fleet decision-making in real time, and considers the dynamic nature of the traffic environment in the ramp merging area. With the goal of optimizing traffic efficiency, the present invention proposes a collaborative decision-making game design scheme for the fleet, so that each vehicle in the fleet adopts a corresponding strategy to achieve an improvement in traffic capacity, and the merging of the ramp fleet minimizes the impact on the main road.

[0046] Other advantages, objects, and features of the present invention will be described in part in the following description and, in part, will be apparent to those skilled in the art upon examination of the following description or may be learned from practice of the present invention. The objects and other advantages of the present invention may be realized and obtained through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be described in detail below with reference to the accompanying drawings, in which:

[0048] Figure 1 Schematic diagram of the vehicle cooperative control framework under IVCPS;

[0049] Figure 2 A schematic diagram of the vehicle queue formation;

[0050] Figure 3 Schematic diagram of virtual queue strategy;

[0051] Figure 4 Schematic diagram of vehicle game participants;

[0052] Figure 5 Flowchart of the method for designing a coordinated control strategy for vehicle platoons in merging areas. DETAILED DESCRIPTION

[0053] The following describes the embodiments of the present invention by means of specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present invention, and the following embodiments and features in the embodiments can be combined with each other without conflict.

[0054] To address the issues of high decision-making complexity and reduced traffic efficiency caused by competition between vehicles under high traffic flow, this paper proposes a vehicle platoon collaborative control strategy design method based on the IVCPS multi-scale framework, aiming to reduce computational complexity, optimize traffic efficiency, and effectively improve fuel economy.

[0055] The following content will describe the technical solution of the present invention in detail in parts:

[0056] 1. Design of collaborative control strategy based on IVCPS

[0057] From the perspective of IVCPS, traffic operation is divided into five scales: node level, unit level, subsystem level, system level and regional level, and a detailed division is made from micro to macro, as shown in Table 1.

[0058] Table 1 Multi-scale division of intelligent vehicle cyber-physical system

[0059]

[0060]

[0061] The multi-scale collaborative control architecture for merging areas based on IVCPS is a system design that tightly integrates cyberspace and physical space. Within this collaborative control framework, intelligent connected vehicles can send and receive real-time status information about themselves and surrounding vehicles. This information is transmitted to the cloud and other vehicles via wireless communication technology. The cloud analyzes both vehicle-side and road-section data to develop an optimized merging strategy based on current traffic conditions and sends instructions back to the vehicles. Upon receiving these instructions, the vehicles automatically adjust their speed and direction to smoothly and safely complete the merging process.

[0062] The multi-scale vehicle cooperative confluence architecture based on IVCPS designed in this paper can be divided into two main layers: the cyber layer and the physical layer, which are interconnected through an efficient communication mechanism. Figure 1 shown.

[0063] The IVCPS architecture is integrated into the collaborative decision-making of platoons in merging areas. At the information layer, decisions on the merging order of vehicles are made in the cloud based on vehicle information. The dynamic interaction between vehicles on the main road and surrounding vehicles affects the collaborative decision-making. At the physical layer, the safety and stability of the vehicle platoon during driving is particularly critical.

[0064] (1) At the information level, decisions need to be made on vehicle coordination. The traffic environment in the merging area is real-time and complex. Faced with the time-varying traffic environment in the merging area, intelligent connected vehicles should be able to respond in real time and make decisions on their own traffic order based on the current traffic conditions. In addition, in order to coordinate with the vehicles merging from the ramp, vehicles on the main road can choose the optimal decision between the two strategies of deceleration and lane changing. However, when changing lanes, the vehicle will inevitably have a conflict of interest with the vehicle behind it in the target lane, and the deceleration operation will affect the traffic efficiency of the lane. Therefore, game theory is introduced into the decision-making model, and the impact of vehicle lane changing and deceleration behavior is considered to establish a vehicle queue coordination game model. This requires analyzing the traffic situation at the system level and performing merging control through vehicle group coordination.

[0065] (2) At the physical layer, vehicle platoons must be controlled to ensure stable and efficient platoon operation. In the longitudinal direction, following vehicles within a platoon must closely follow the leading vehicle, and control between vehicles in the platoon must be implemented based on decision-making results. This requires efficient communication and information sharing between vehicles at the unit level, down to the control strategy for each vehicle.

[0066] Through the collaborative perception and decision-making of different structural layers under IVCPS, combined with the multi-scale empowerment of vehicle-road-cloud integrated technology, it is possible to effectively address the problems of low traffic efficiency of vehicles in merging areas under high traffic flow and increased energy consumption caused by frequent vehicle starts and stops.

[0067] 2. Optimization of confluence order

[0068] The merging order of vehicles is a key factor in ramp-to-ramp coordinated merging, determining the order in which vehicles on the main road and on the ramp coordinate their movements. Existing technologies often default to a first-in, first-out (FIFO) merging order, ignoring the speed differences between the main road and the ramp. Using an optimal strategy, the order in which two vehicles pass can be formulated as a mixed-integer linear program. However, the computational complexity of optimizing the optimal strategy increases exponentially with increasing traffic volume. Furthermore, the results obtained with suboptimal strategies may not necessarily be beneficial to overall traffic flow.

[0069] Therefore, this embodiment, based on the vehicle collaborative merging architecture, considers the system-level scale and designs a fleet collaborative decision-making model at the IVCPS information layer. Specifically, based on the concept of vehicle platooning, ICVs entering the merging area are divided into multiple queues according to the vehicle queue division strategy. Decisions are made on a queue-by-queue basis, enabling sequential planning of the merging area at the information layer. This improves traffic efficiency by leveraging the close spacing within the vehicle queues; it also optimizes decision-making time by making decisions on a queue-by-queue basis.

[0070] In this embodiment, the collaborative decision-making process is as follows:

[0071] First, according to the initial vehicle spacing, that is, the vehicle spacing d when the vehicle enters the vehicle formation area i,i-1 (t0), the vehicles on the main road and ramp are divided into multiple sub-queues, and coordinated control is performed in the form of sub-queues. In order to clarify the division of vehicle sub-queues, the vehicle spacing threshold is defined, that is,

[0072]

[0073] Among them, v max 、a max Represent the maximum speed and maximum acceleration of the vehicle, v i represents the speed of vehicle i, L p is the length of the vehicle formation area. T The threshold value of the vehicle spacing is used to prevent the speed and acceleration fluctuations caused by the large spacing between following vehicles when a queue is formed. i,i-1 (t0)>d T When , the i-th vehicle is regarded as the leader of the newly formed vehicle queue, otherwise, it is classified as the following vehicle in the front queue. Figure 2 The diagram shows the formation process of a vehicle queue.

[0074] Based on the vehicle formation, in the queue mapping area, the ramp fleet is mapped to the outer lane of the main road to form a virtual queue. The inner lane of the main road is prohibited from changing lanes, so no virtual mapping operation is performed. The virtual queue method is centered on the merging point, and the ramp vehicles are projected onto the main road according to their position from the merging point. The vehicle queues entering the queue mapping area are numbered and indexed according to their distance from the merging point. The index is LP. k Indicates that, according to this index, the ramp vehicle queue is mapped to the corresponding position of the main road, such as Figure 3 When a vehicle queue passes the merging point, the coordinated merging operation is completed, and the indexes of all sub-queues are reduced by 1. The vehicle queue with index 0 indicates that it has left the current virtual queue.

[0075] Secondly, to improve the traffic efficiency of the merging area of ​​the expressway, the traffic order of the main road fleet and the ramp fleet is optimized with reference to the optimal strategy. The merging order is described as a mixed integer linear programming (MILP) problem, and the objective function is defined as follows:

[0076]

[0077] LP for each fleet in the virtual control area i The time allocated to pass through the area and the minimum passing time, the objective function is defined as minimizing the delay time of all platoons passing through the merging point.

[0078] Referring to the suboptimal strategy, multiple vehicles are considered as a queue and decisions are made in a queue-like manner to reduce computational complexity. Therefore, the MILP constraint function is designed based on speed conditions (C1, C2) and safety and merging conditions (C3, C4) in the form of a queue:

[0079]

[0080]

[0081]

[0082]

[0083] C4:s i ={0,1}

[0084] in, queue LP i The maximum and minimum time to exit the merging area, s i is a binary judgment parameter, s i =1 indicates fleet LP i For the main road fleet, otherwise s i =0. v i ′ represents the fleet LP i The velocity, v′ max 、v′ min Represents fleet LP i The maximum and minimum speeds, a′ max , a′ min Represents fleet LP i The maximum and minimum acceleration, L c Indicates the distance between the convoy and the merging point, L i Indicates the length of the current fleet, p i Indicates the current position of the convoy leader.

[0085] Through the si The optimal merging order of vehicles can be obtained by judging i = 0, indicating that the current convoy is on the ramp and must follow the merge constraint to merge onto the main road. This means the ramp convoy follows the previous convoy in the virtual space. Conversely, it follows the safety constraint, meaning the main road convoy travels at a preset speed (possibly following vehicles on the main road or controlling its speed to reserve space for the ramp convoy to merge).

[0086] 3. Fleet Collaborative Decision-Making

[0087] When a vehicle chooses a decision that maximizes its profit, it will proceed accordingly. However, when a convoy makes a lane change, this often impacts traffic conditions in the target lane. Especially during high traffic conditions, lane changes can disrupt normal driving in the target lane, and in severe cases, can cause congestion. This creates a clear conflict of interest between vehicles on the two lanes, necessitating a detailed study of the game-playing relationship between them, starting at the IVCPS unit level.

[0088] According to the results of the merging order optimization, two measures can be taken for the main road convoy whose merging order is behind the ramp convoy. The first is to use longitudinal control to reserve a gap, and the second is to change lanes to the outer lane, such as Figure 4 Show.

[0089] Based on the idea of ​​game theory, the dynamic interest relationship of the teams is described. The game strategy set and symbolic representation between the teams are shown in Table 2:

[0090] Table 2 Fleet game strategy set

[0091]

[0092] According to the strategy sets and benefits between the teams, the game benefit matrix between the teams is shown in Table 3.

[0093] Table 3 Fleet game payoff matrix

[0094]

[0095] Indicates fleet LP O Vehicle i selection strategy O k When the fleet LP I Vehicle j in the selection strategy I l The revenue of vehicle i is, k=1,2,l=1,2,3; Indicates fleet LP O Vehicle i selection strategy O k When the fleet LP I Vehicle j in the selection strategy I lThe benefit of vehicle j at this time. Vehicle j will participate in the decision-making of vehicle i only when it is behind vehicle i in the longitudinal position.

[0096] Fleet LP O The final profit function of the vehicle in is expressed as:

[0097]

[0098] Among them, ε i 、 δ i Represents fleet LP O The safety benefit, time benefit, and stability benefit of vehicle i, α i , β i ,λ i They represent the safety benefit factor, time benefit factor, and stability benefit factor, respectively, and represent the weight of each benefit. In addition, there is a mutual constraint relationship between these parameters, namely, α i +β i +λ i =1.

[0099] Fleet LP I The final profit function of the vehicle in is expressed as:

[0100]

[0101] Among them, ε j 、 Represents fleet LP I The safety benefit and time benefit of vehicle j, α j , β j They represent the safety benefit factor and time benefit factor respectively, and represent the weight of each benefit, and α j +β j =1.

[0102] Based on the construction of strategy sets and payoff functions, the two-lane scenario during lane changing has obvious leader-follower characteristics. The Stackelberg equilibrium solution strategy is adopted to solve the model. By combining the optimal response of the follower and the optimal strategy of the leader, the equilibrium point of the Stackelberg game is obtained, and finally the optimal strategy combination of the leader and follower is obtained.

[0103] The specific steps are as follows:

[0104] (1) Determine the profit function of the participants: the leader is the LP of the team O The vehicle i in the example has a profit function of U I (I l ,I n ), indicating fleet LP ISelection Strategy I n When vehicle i selects strategy I l The profit of vehicle i at time t; the follower is the fleet LP I , the profit function is Indicates fleet LP O Vehicle i selection strategy O k When the fleet LP I Vehicle j in the selection strategy I l The revenue of vehicle j at time .

[0105] (2) Follower optimal response function: For a given leader strategy, formulate the follower's optimal response strategy.

[0106]

[0107] (3) Leader’s optimization function: After obtaining the follower’s optimal response function, the leader’s goal is to update the strategy to maximize the profit under the follower’s optimal response.

[0108]

[0109] The follower's optimal response function I n I in l Optimal strategy for replacing the leader Get the final optimal strategy of the follower and The composed strategies constitute the optimal decisions of followers and leaders, that is, each vehicle in the convoy can adopt the corresponding strategy to achieve the improvement of traffic capacity and minimize the impact of the ramp convoy on the main road.

[0110] 4. Decision-making Steps

[0111] Combining the contents of the first, second, and third parts above, the following steps are given for designing a coordinated control strategy for vehicle queues in merging areas:

[0112] (1) Based on the system-level scale under IVCPS, the merging order is first optimized. According to the initial distance between vehicles, the vehicles are divided into several vehicle queues. The merging control and decision-making are carried out in the form of vehicle queues at the unit level, which can improve the overall traffic efficiency and narrow the solution scope. Subsequently, the ramp fleet is mapped to the main road based on the virtual mapping method, and the fleet is numbered according to the distance from the merging point. However, due to the difference in speed between the main road and the ramp vehicles, the ramp vehicles merging into the main road also need to maintain a safe distance. Therefore, it is necessary to comprehensively consider the speed constraints, the safety constraints within the queue, and the safety constraints of the ramp vehicles merging into the queue, and redefine the passage order of the queue. The present invention uses mixed integer linear programming to constrain various conditions and assigns the merging order according to the results.

[0113] Assume that there are two fleets P1 and P2 on the main road in the merging area, and one fleet P3 on the ramp needs to merge. According to the dynamic model, speed constraints for each fleet are established, and safety, merging constraints, and lane constraints are constructed. Based on the constraints, a mixed integer linear programming model is established. The goal is to minimize the total travel time of all fleets in the merging area. The Gurobi solver is used to solve the optimization problem of the above constraints to obtain the optimal travel time and merging order for each fleet. The merging order is determined based on the calculated time interval. For example, if the passing time of fleet P3 is greater than the passing time of fleet P1 plus the time interval d 1,3 , then P1's passing order is greater than P3. P2 and P3 determine their passing order according to the same principle.

[0114] (2) Under IVCPS, fleets engage in a coordinated game based on the unit scale. After the merging order is assigned, the game can begin. First, the game subject is identified. According to the assigned merging order, if the adjacent fleet is a main road fleet, there is no need to engage in a game, and the follow-up operation is directly executed according to the strategy. If there is a ramp fleet among the adjacent fleets, the main road fleet needs to make a game decision to reserve a gap for the ramp fleet to merge into the main road.

[0115] For the convoy in the lane near the ramp on the main road, it makes decisions based on the convoy game strategy set. If there is a convoy C in this lane, O Want to change lanes to enter the target lane, the target lane has another team C I , C O Considering the safety benefit and speed benefit after lane change, vehicle i in C I The vehicle j in the equation considers safety benefit, time benefit and stability benefit. Set the benefit function parameters respectively and calculate the fleet C according to the initial fleet state. O The benefits of vehicle i before and after changing lanes If the total revenue Greater than the gain before changing lanes Then execute lane change decision O1, from team C O At this time, Team C I The vehicle j in the calculation is based on the optimal benefit (i.e. ) Select Strategy I l , while Team C O Vehicle i in

[15] updates the optimization function according to the strategy selected by vehicle j, so that vehicle i obtains its longitudinal control decision when changing lanes.

[0116] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not limiting. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention can be modified or replaced by equivalents without departing from the purpose and scope of the technical solutions, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for designing a coordinated control strategy for vehicle queues in a merging area based on IVCPS, characterized by: For vehicles in the lane merging area, they are divided into several vehicle queues according to the initial distance between the vehicles; Based on the distance the convoy reaches the merging point, the convoys on the ramp are mapped to the main road to form a virtual queue based on virtual mapping. The convoys in the virtual queue are numbered according to the distance the convoy reaches the merging point. The merging order of the merging area is planned based on the vehicle queue. The merging order is assigned through mixed integer linear programming. Each fleet passes through the merging point in sequence according to the merging order. From the perspective of IVCPS, if there is a ramp team between two adjacent convoys with a merge order, the main road team will play a game based on the game strategy set. Based on the game results, the main road team will be controlled to reserve a merging gap for the ramp team that needs to merge. If both convoys with adjacent merging orders are on the main road, no game will be played. The game strategy set is shown in the following table: Among them, LP O Indicates the convoy in the lane on the main road close to the ramp side, LP I Represents LP O The convoy in the adjacent lane; k Indicates fleet LP O Strategy, k=1,2; I l represents the strategy of fleet LP1, l = 1, 2, 3; Indicates fleet LP O The revenue function of vehicle i in , ε i 、 δ i Represents fleet LP O The safety benefit, time benefit, and stability benefit of vehicle i, α i , β i ,λ i They represent safety benefit factor, time benefit factor, and stability benefit factor, respectively. i +β i +λ i =1; Indicates fleet LP I The revenue function of vehicle j in is, ε j 、 Represents fleet LP I The safety benefit and time benefit of vehicle j, α j , β j Represent the safety benefit factor and time benefit factor respectively, and α j +β j =1; According to the game strategy set, calculate the fleet LP O The benefits of vehicle i before and after changing lanes Compare and The size of selects the execution strategy of vehicle i; the platoon LP of the target lane of lane change I The vehicle j in the game calculates the profit based on the strategy of vehicle i. For fleet LP I and vehicle i, the Stackelberg game is used to solve the longitudinal control strategies of the two; Using Stackelberg game to solve fleet LP I The process of the longitudinal control strategy of vehicle i includes: 1) Determine the profit function of the leader and followers: the leader is the LP of the team O The vehicle i in the example has a profit function of U I (I l ,I n ), indicating fleet LP I Selection Strategy I n When vehicle i selects strategy I l The profit of vehicle i at this time; the follower is the fleet LP I , the profit function is Indicates fleet LP O Vehicle i selection strategy O k When the fleet LP I Vehicle j in the selection strategy I l The revenue of vehicle j at time ; 2) Design the follower's optimal response strategy as follows: 3) Design the leader’s optimization function to update the strategy of maximizing benefits under the follower’s optimal response. The leader’s optimization function is expressed as: 4) Set the follower's optimal response strategy I n I in l Replaced with the leader's optimization function The final optimal response strategy of the follower is obtained, that is, thereby and The composed strategies constitute the optimal vertical control decisions of followers and leaders.

2. The method according to claim 1, wherein: The vehicle queue is divided according to the vehicle spacing threshold, which is expressed as: Where, v max 、a max Represent the maximum speed and maximum acceleration of the vehicle, v i represents the speed of vehicle i, L p Represents the length of the vehicle formation area; if the initial vehicle distance d of the i-th vehicle i,i-1 (t0)>d T When , the i-th vehicle is regarded as the leader vehicle of the newly formed vehicle queue; otherwise, it is classified as the following vehicle of the front vehicle queue.

3. The method according to claim 1, wherein: The merging order in the merging area is planned based on the vehicle queue. The merging order is described as a MILP problem, and the objective function of the problem is defined as: Where, Represents each fleet LP in the virtual control area i The allocated time to pass through the merging area and the minimum passing time; the objective function represents minimizing the delay time of all fleets passing through the merging point.

4. The method according to claim 3, wherein: Based on the objective function, the MILP problem is constructed by combining speed, safety, and merging constraints: C4:s i ={0,1} Where, Team LP i The maximum and minimum time to exit the merging area, s i is a binary judgment parameter, s i =1 indicates fleet LP i For the main road fleet, otherwise s i =0;v′ i Indicates fleet LP i The velocity, v′ max 、v′ min Represents fleet LP i The maximum and minimum speeds, a′ max , a′ min Represents fleet LP i The maximum and minimum acceleration, L c Indicates the distance between the convoy and the merging point, L i Indicates the length of the current fleet, p i represents the current position of the pilot car in the convoy; the travel time obtained by solving the MILP problem Determine the confluence order.

Images