A macro-micro integrated alliance decision-making method based on intelligent connected vehicle clusters

Abstract

The present invention relates to the field of intelligent connected vehicle cluster decision-making technology, and in particular to a macro-micro integrated alliance decision-making method for intelligent connected vehicle clusters. The present invention constructs an initial path for an intelligent connected vehicle cluster, taking into account safety, comfort, efficiency, and energy consumption goals, to construct a cluster alliance decision cost function. Under the conditions of satisfying the kinematic constraints and comfort of the intelligent connected vehicles, an adaptive co-evolutionary algorithm is used to solve the problem. The adaptive co-evolutionary algorithm adopts a diversity-based search allocation mechanism to perform population diversity allocation and traffic flow guidance variation, and optimizes the global exploration and local optimization modules respectively. Periodic global optimization and other operations are performed to generate the optimal intelligent connected vehicle cluster decision, thereby enabling the vehicle to reduce energy consumption and improve comfort while maximizing safety and improving efficiency.

Description

Technical Field

[0001] The present invention relates to the field of intelligent connected vehicle cluster decision-making technology, and in particular to a macro-micro integrated alliance decision-making method based on intelligent connected vehicle clusters. Background Art

[0002] Intelligent driving decision-making is a core component of vehicle operation. It receives and analyzes inter-vehicle traffic information in real time, making precise judgments and responses to ensure safe and efficient interaction between vehicles. This process not only ensures driving safety, ensuring every step is secure and reliable, but is also crucial for improving driving efficiency. It optimizes traffic flow, reduces congestion, and ultimately increases overall driving speed.

[0003] In recent years, a series of theories for intelligent connected vehicle (ICV) swarm decision-making have been proposed both domestically and internationally. Research on ICV swarm decision-making can be categorized into two mechanisms: centralized and distributed. Centralized ICV swarm decision-making mechanisms utilize a global central controller to formulate traffic plans for all ICV swarms. Distributed ICV swarm decision-making mechanisms utilize a global and local traffic information collection model, whereby ICV swarms autonomously formulate traffic plans without the need for a central controller.

[0004] Swarm decision-making is a complex, non-deterministic problem whose computational difficulty increases exponentially with the number of vehicles. This makes traditional algorithms face significant time challenges when handling multi-vehicle trajectory planning. The Adaptive Coevolutionary Algorithm (ACA) effectively addresses complex, high-dimensional optimization problems by executing evolutionary algorithms in parallel across multiple subsystems and exchanging information to collaboratively optimize the population. ACA not only improves search efficiency and accuracy but also fully utilizes the computing resources of intelligent connected vehicles to optimize the trajectories of vehicle groups within a limited timeframe on a local road network.

[0005] Decision-making for intelligent connected vehicle swarms in urban road networks has become a hot and challenging issue in the field of intelligent transportation. Enabling vehicles to travel safely, quickly, energy-efficiently, and comfortably in urban road networks has become a key research direction in intelligent transportation systems.

[0006] For example, Chinese patent application publication number: CN117116043A discloses a macro-path planning method for vehicles in urban road networks with frequent congestion, which belongs to the field of traffic path planning. The method adopts a two-layer planning model, including macro-path planning and local path planning. In macro-path planning, the existing traffic sub-area division method and macro-traffic flow distribution strategy are used to achieve effective management of large-scale urban road networks. Local path planning uses the A* algorithm to plan a time-consuming shortest path for vehicles from the previous sub-area boundary to the next sub-area boundary within each traffic sub-area based on the traffic sub-area passage sequence given by the macro-planning. In order to cope with the frequent traffic congestion in urban road networks, the invention uses the flow balance theory to distribute the traffic flow of the road network from a macro level from a global perspective. The invention schedules the flow between the current traffic sub-areas based on the macro-path planning algorithm, so that the urban traffic flow reaches the global optimum.

[0007] However, the prior art still has the following problems:

[0008] 1. In traffic network scenarios, individual vehicles currently rely primarily on their own intelligence for autonomous decision-making, lacking collaboration with other vehicles and roadside infrastructure. This makes it difficult to fully capture information about complex traffic environments and achieve globally optimized decision-making. To overcome the limitations of individual vehicle decision-making, information exchange and sharing, as well as collaborative perception and cooperation mechanisms among intelligent connected vehicle clusters enable them to obtain more comprehensive traffic information and make better decisions.

[0009] 2. Centralized swarm collaborative decision-making approaches face extremely high computational complexity when applied to traffic network scenarios, and the inherent causal loops in these scenarios exacerbate the difficulty of solving the problem. In this context, a distributed swarm collaborative decision-making mechanism in these scenarios is expected to achieve a balance between coordination performance and computational efficiency.

[0010] 3. Current research on trajectory optimization on urban roads contains some idealized assumptions, and there is still room for improvement in solutions that balance comfort, traffic efficiency, energy saving, and safety.

[0011] 4. In view of the dynamic topology and high mobility of intelligent connected vehicles, traditional platooning methods mainly focus on vehicle following behavior. However, on urban roads, a large number of other driving behaviors such as lane changing and turning are involved, and traditional platooning methods are difficult to meet the needs. Therefore, a macro-micro integrated alliance decision-making method for intelligent connected vehicle clusters is proposed. Summary of the Invention

[0012] To this end, the present invention provides a macro-micro integrated alliance decision-making method based on intelligent connected vehicle clusters to overcome the problem that the existing technology for the dynamic topology and high mobility of intelligent connected vehicles is difficult to meet the needs of traditional platooning methods. Traditional platooning methods mainly focus on vehicle following behavior, but they involve a large number of other driving behaviors such as lane changing and turning on urban roads.

[0013] To achieve the above objectives, the present invention provides a macro-micro integrated alliance decision-making method based on intelligent connected vehicle clusters, comprising:

[0014] Building an initial path for intelligent connected vehicle swarms;

[0015] Construct an alliance set for intelligent connected vehicle clusters and construct cost functions for individual intelligent connected vehicles in the alliance set, including driving safety cost function, ride comfort cost function, driving efficiency cost function, and energy consumption cost function;

[0016] The driving safety cost function is the sum of the longitudinal safety cost and the lateral safety cost;

[0017] The ride comfort cost function is determined based on the lateral and longitudinal accelerations of the intelligent connected vehicle;

[0018] The driving efficiency cost function is determined based on the longitudinal speed;

[0019] The energy consumption cost function is determined based on air density, vehicle frontal area, air resistance coefficient, rolling resistance coefficient, vehicle mass, gravitational acceleration, vehicle comprehensive acceleration, fuel consumption rate and engine efficiency;

[0020] The longitudinal safety cost is related to the longitudinal gap and the relative speed to the preceding vehicle, and the lateral safety cost is related to the relative distance and relative speed between the intelligent connected vehicle and the adjacent vehicle.

[0021] Constructing an alliance game model for the alliance set, and setting the optimization goal to minimize the cost function;

[0022] Determine the constraints for game decision-making, including kinematic constraints and comfort constraints;

[0023] The kinematic constraints are determined based on the acceleration value, speed value, front wheel steering angle, yaw angle, maximum acceleration, minimum steering angle, maximum steering angle, minimum yaw angle, and maximum yaw angle of the intelligent connected vehicle;

[0024] The comfort constraint is constructed based on the front wheel steering angle, the front wheel steering angle change rate, the acceleration increment and the jerk;

[0025] During the coalition model solution process, each intelligent connected vehicle is identified as an independent population for evolution. After a fixed number of evolutions, each intelligent connected vehicle exchanges information via V2X communication technology to achieve global optimization.

[0026] Each intelligent connected vehicle is identified as an independent population for evolution, where

[0027] Genetically encode the intelligent connected vehicle to obtain chromosomes specific to the intelligent connected vehicle;

[0028] Assign a population to each intelligent connected vehicle and perform fitness evaluation on chromosomes that meet the constraints to determine the evaluation value for each population;

[0029] Adjusting the proportion of global exploration and local optimization of individuals in the population during the evolution process according to the evaluation value;

[0030] The evaluation value is determined based on the Euclidean distance between the chromosomes corresponding to the intelligent connected vehicles;

[0031] The solution process is repeated until the termination condition is met and the optimal trajectory solution is obtained.

[0032] Furthermore, in step S1, the process of constructing the initial path of the intelligent connected vehicle cluster includes:

[0033] Pre-build road network data, including node division and construction of directed graph for the road network;

[0034] Each intelligent connected vehicle loads the road network data to determine the starting point and end point and independently runs the A* algorithm to generate the initial path.

[0035] Furthermore, the process of constructing a coalition game model for a coalition set includes:

[0036] Initialize the alliance and divide the alliance into straight alliance, left turn alliance, and right turn alliance based on the vehicle's direction label;

[0037] Determine the preference order of the alliance game process;

[0038] Determine the behavioral conditions for alliance mergers, alliance splits, and alliance exchanges.

[0039] Further, it includes determining whether the alliance is in a stable state, wherein,

[0040] If all intelligent connected vehicles adopt a preference order and each of the intelligent connected vehicles cannot reduce its own cost by joining or exiting the alliance, the alliance is in a stable state.

[0041] Furthermore, for local optimization, the individual mutation rate of intelligent connected vehicles is determined based on traffic flow.

[0042] Furthermore, during the population evolution process, the number of chromosomes for crossover operations of intelligent connected vehicles is determined, and crossover operations are performed through chromosome single-point crossover.

[0043] Furthermore, in step S6, the termination condition is that the number of cycles of global optimization reaches a set maximum value or the total fitness of the intelligent connected vehicle does not improve under a predetermined number of global optimizations.

[0044] Compared with existing technologies, the present invention has the beneficial effect of constructing a cluster alliance decision cost function by constructing an initial path for a cluster of intelligent connected vehicles, taking into account safety, comfort, efficiency, and energy consumption objectives. Under the conditions of satisfying the kinematic constraints and comfort requirements of the intelligent connected vehicles, an adaptive co-evolutionary algorithm is used to solve the problem. The adaptive co-evolutionary algorithm uses a diversity-based search and allocation mechanism to mutate population diversity and traffic flow guidance. It optimizes the global exploration and local optimization modules separately, and performs periodic global optimization operations to generate the optimal intelligent connected vehicle cluster decision, thereby reducing energy consumption and improving comfort while maximizing safety and improving efficiency.

[0045] In particular, the present invention comprehensively considers factors such as driving safety, ride comfort, driving efficiency and energy consumption cost when constructing the cost function, and then subsequently obtains the cluster's decision, so that intelligent connected vehicles can improve driving safety, improve ride comfort, improve driving efficiency and reduce energy consumption.

[0046] In particular, the present invention constructs an alliance game model for alliance sets. By analyzing alliance mergers, alliance splits, and preference ordering during alliance games, the model aims to find the optimal solution for the combined state of intelligent connected vehicles, thereby achieving a stable alliance state. The minimization and cost problem based on game decisions is ultimately transformed into a closed-loop iterative optimization process with multiple constraints, and solved using an efficient adaptive co-evolutionary algorithm. This enables more efficient, accurate, and reliable decision-making.

[0047] In particular, the present invention identifies each intelligent connected vehicle as an independent population for evolution, genetically encodes the intelligent connected vehicles, and obtains chromosomes for the intelligent connected vehicles. Next, a population is assigned to each intelligent connected vehicle, and fitness evaluation is performed on the chromosomes that meet the constraints to determine an evaluation value for each population. The proportion of global exploration and local optimization of individuals in the population during the evolution process is adjusted according to the evaluation value. DSAM aims to solve the problem of rapid decline in population diversity by dynamically balancing global exploration and local optimization, thereby effectively preventing the algorithm from falling into a local optimal solution. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 This is a step diagram of the macro-micro integrated alliance decision-making method based on the intelligent connected vehicle cluster of the present invention;

[0049] Figure 2 This is a schematic diagram of a city local area road network node according to an embodiment of the present invention;

[0050] Figure 3 A schematic diagram of a coordinate system for a kinematic model of an intelligent connected vehicle according to an embodiment of the present invention;

[0051] Figure 4 A schematic diagram of the gene encoding of the collaborative evolution algorithm according to an embodiment of the present invention;

[0052] Figure 5 Schematic diagram of a diversity-based search allocation mechanism according to an embodiment of the present invention;

[0053] Figure 6 A schematic diagram of a variation of traffic flow guidance according to an embodiment of the present invention. DETAILED DESCRIPTION

[0054] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are merely used to explain the present invention and are not intended to limit the present invention.

[0055] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0056] See also Figure 1 As shown, it is a step diagram of the macro-micro integrated alliance decision-making method based on the intelligent networked vehicle cluster of the present invention. The macro-micro integrated alliance decision-making method based on the intelligent networked vehicle cluster of this embodiment includes:

[0057] Step S1, constructing an initial path for the intelligent connected vehicle cluster;

[0058] Step S2: constructing an alliance set for the intelligent connected vehicle cluster and constructing a cost function for each intelligent connected vehicle in the alliance set, including a driving safety cost function, a riding comfort cost function, a driving efficiency cost function, and an energy consumption cost function;

[0059] The driving safety cost function is the sum of the longitudinal safety cost and the lateral safety cost;

[0060] The ride comfort cost function is determined based on the lateral and longitudinal accelerations of the intelligent connected vehicle;

[0061] The driving efficiency cost function is determined based on the longitudinal speed;

[0062] The energy consumption cost function is determined based on air density, vehicle frontal area, air resistance coefficient, rolling resistance coefficient, vehicle mass, gravitational acceleration, vehicle comprehensive acceleration, fuel consumption rate and engine efficiency;

[0063] The longitudinal safety cost is related to the longitudinal gap and the relative speed to the preceding vehicle, and the lateral safety cost is related to the relative distance and relative speed between the intelligent connected vehicle and the adjacent vehicle.

[0064] Step S3, constructing an alliance game model for the alliance set, and setting the optimization goal to minimize the cost function;

[0065] Step S4, determining the constraints for the game decision, including kinematic constraints and comfort constraints;

[0066] The kinematic constraints are determined based on the acceleration value, speed value, front wheel steering angle, yaw angle, maximum acceleration, minimum steering angle, maximum steering angle, minimum yaw angle, and maximum yaw angle of the intelligent connected vehicle;

[0067] The comfort constraint is constructed based on the front wheel steering angle, the front wheel steering angle change rate, the acceleration increment and the jerk;

[0068] Step S5: During the coalition model solution process, each intelligent connected vehicle is determined as an independent population for evolution. After a fixed number of evolutions, each intelligent connected vehicle exchanges information via V2X communication technology to achieve global optimization.

[0069] Each intelligent connected vehicle is identified as an independent population for evolution, where

[0070] Genetically encode the intelligent connected vehicle to obtain chromosomes specific to the intelligent connected vehicle;

[0071] Assign a population to each intelligent connected vehicle and perform fitness evaluation on chromosomes that meet the constraints to determine the evaluation value for each population;

[0072] Adjusting the proportion of global exploration and local optimization of individuals in the population during the evolution process according to the evaluation value;

[0073] The evaluation value is determined based on the Euclidean distance between the chromosomes corresponding to the intelligent connected vehicles;

[0074] Step S6, repeat step S5 until the termination condition is met and the optimal trajectory solution is obtained.

[0075] Specifically, in step S1, the process of constructing the initial path of the intelligent connected vehicle cluster includes:

[0076] Pre-build road network data, including node division and construction of directed graph for the road network;

[0077] See also Figure 2 As shown, it is a schematic diagram of the urban local road network nodes of an embodiment of the present invention. In implementation, those skilled in the art can divide the starting point and end point of each road section as nodes based on the urban local road network. These nodes represent the connection points of the roads, including intersections, the starting and ending positions of the road sections, etc., which will not be repeated here.

[0078] Specifically, to achieve efficient and accurate path planning, a directed graph for the road network can be constructed. The entire road network is represented as a directed graph G = (V, E), where V is the set of nodes and E is the set of edges. For each road segment, the connection between its starting point and end point is represented as a directed edge. Edge attributes include information such as the segment's length and speed limit. Furthermore, the reachability relationship between nodes is determined based on the actual road connectivity.

[0079] Specifically, the road network is first gridded, and the length and width of each grid are l grid The center of each grid is called a node. Each intelligent connected vehicle loads the road network data to determine the starting and ending points and independently runs the A* algorithm to generate an initial path.

[0080] It is understandable that, based on the traditional A* algorithm, a heuristic function is designed for the characteristics of urban road networks. Combining Manhattan distance (applicable to gridded roads) and direction-corrected Euclidean distance (for curved roads), the heuristic function is automatically switched according to the topological characteristics of the road section. The cost estimation function f(n) of an intelligent connected vehicle from the starting point to the end point is shown in formula (1)

[0081] f(n)=g(n)+h(n),

[0082] in,

[0083] Where g(n) is the actual cost from the starting point to the current node. h(n) is the estimated cost from the current node to the end point, also known as the heuristic function. n 、y n They are the horizontal and vertical coordinates of the current node respectively. goal 、y goal are the horizontal and vertical coordinates of the endpoint, respectively. α is the direction angle correction coefficient, which can be dynamically calculated based on the historical path curvature to ensure that the heuristic function is close to the actual road conditions.

[0084] The initial path is a discrete path point According to the channelization sign of the lane where the waypoint is located, a direction label is assigned to each vehicle, with straight corresponding to "S", left turn corresponding to "L", and right turn corresponding to "R".

[0085] Specifically, in step S2, when constructing an alliance set for an intelligent connected vehicle cluster, the set of intelligent connected vehicles is represented as M = {1, 2, ..., M}, where M is the number of intelligent connected vehicles in the intelligent connected vehicle cluster. Intelligent connected vehicles can cooperate with each other to form G different alliances, and the alliance set is represented as G = {1, 2, ..., G}.

[0086] Specifically, in step S2, the driving safety cost function is the sum of the longitudinal safety cost and the lateral safety cost;

[0087] In some possible implementations, the driving safety cost function is expressed by formula (2),

[0088]

[0089] In formula (2), Represents the vertical security cost. Represents the lateral safety cost.

[0090] Vertical security costs It is related to the longitudinal gap and the relative speed to the preceding vehicle, which can be expressed by equations (3) and (4).

[0091]

[0092] In formula (3) and formula (4), It represents the longitudinal speed of the intelligent connected vehicle i-1, that is, the longitudinal speed of the preceding vehicle. represents the longitudinal speed of the intelligent connected vehicle i. i-1 represents the longitudinal position of the intelligent connected vehicle i-1, X i Represents the longitudinal position of the intelligent connected vehicle i. represents the first weighting coefficient, represents the second weighting coefficient. ε represents the design parameter, the purpose of which is to avoid the occurrence of zero denominator. L i is the vehicle length. Is the switching function. If the front car is faster than the rear car, that is The switch function Then the cost of longitudinal safety is only related to the relative distance. Otherwise, longitudinal safety is related to both the relative distance and the relative speed.

[0093] Horizontal security costs It is related to the relative distance and relative speed to the neighboring vehicle (NV), which can be expressed by equations (5) and (6):

[0094]

[0095] In formulas (5) and (6), The longitudinal speeds of the intelligent connected vehicle i and its adjacent vehicles respectively. (X NV , Y NV ) is the position of the adjacent vehicle. represents the third weighting coefficient, represents the fourth weighting coefficient. Right now Then the cost of lateral safety is only related to the relative distance. Otherwise, it is related to both the relative distance and the relative speed.

[0096] The ride comfort cost function is determined based on the lateral and longitudinal accelerations of the intelligent connected vehicle;

[0097] In some possible implementations, according to formula (7),

[0098]

[0099] in, and are the longitudinal acceleration and lateral acceleration of the intelligent connected vehicle i, represents the fifth weighting coefficient, represents the sixth weighting coefficient.

[0100] The driving efficiency cost function is determined based on the longitudinal speed;

[0101] In some possible implementations,

[0102] Driving efficiency cost is a function of the longitudinal velocity of the intelligent connected vehicle i, as shown in equations (8) and (9):

[0103]

[0104] In formulas (8) and (9), They represent the longitudinal velocity and longitudinal expected velocity of intelligent connected vehicle i respectively. Indicates the speed limit of the lane, represents the speed of the preceding vehicle of the intelligent connected vehicle i, and is the seventh weighting coefficient.

[0105] The energy consumption cost function is determined based on air density, vehicle frontal area, air resistance coefficient, rolling resistance coefficient, vehicle mass, gravitational acceleration, vehicle comprehensive acceleration, fuel consumption rate and engine efficiency;

[0106] In some possible implementations,

[0107] The energy consumption of fuel vehicles is mainly related to factors such as engine efficiency, vehicle mass, air resistance, rolling resistance, etc. The following is the cost expression of the energy consumption of intelligent connected vehicles:

[0108]

[0109] In formula (10), ρ is the air density, A is the vehicle's frontal area, and C d is the air resistance coefficient, C g is the rolling resistance coefficient, m is the vehicle mass, g is the acceleration due to gravity, a is the vehicle's comprehensive acceleration, b is the fuel consumption rate, and η is the engine efficiency.

[0110] The goal of each intelligent connected vehicle is to maximize the driving safety, comfort, and passing efficiency of the intelligent connected vehicles while minimizing the energy consumption of the vehicles, that is, to minimize the decision cost function. To achieve this optimization goal, the decision cost function of each intelligent connected vehicle i is expressed as Equation (11).

[0111]

[0112] In formula (11), represents the driving safety cost function, represents the ride comfort cost function, represents the driving efficiency cost function and represents the energy cost function, represents the first driving characteristic weighting coefficient, represents the second driving characteristic weighting coefficient, represents the third driving characteristic weighting coefficient, represents the fourth driving characteristic weighting coefficient.

[0113] The optimization goal of the intelligent connected vehicle cluster is to minimize the total cost of the alliance, as shown in formula (12).

[0114]

[0115] In formula (12), the key to minimizing the optimization objective lies in dealing with the cooperative relationship between intelligent connected vehicles from the perspective of alliance game. This is because the total cost in (12) is composed of driving safety, riding comfort, driving efficiency, and energy consumption costs, which depends on the formation result of the alliance of intelligent connected vehicles. From the perspective of alliance game, the optimization objective in formula (12) is also transformed into the cost function minimization problem in (13).

[0116] Specifically, in step S3, the process of constructing the alliance game model for the alliance set includes:

[0117] Initialize the alliance and divide the alliance into straight alliance, left turn alliance, and right turn alliance based on the vehicle's direction label;

[0118] Determine the preference order of the alliance game process;

[0119] Determine the behavioral conditions for alliance mergers, alliance splits, and alliance exchanges.

[0120] Specifically, the alliance formation game model considers the paths of ICVs, the costs of the ICV alliance, and the available strategies. During the alliance initialization phase, vehicles with a direction label of "S" are automatically assigned to the straight-moving alliance (S-Coalition), vehicles with a direction label of "L" form the left-turning alliance (L-Coalition), and vehicles with a direction label of "R" form the right-turning alliance (R-Coalition).

[0121] The game model of intelligent connected vehicles can be expressed as {β i ,G g ,S i ,J i ,r J}, where the decision of the i-th intelligent connected vehicle is β i , if β i =g, indicating that intelligent connected vehicle i joins alliance g; G g is the set of intelligent connected vehicles in the g-th alliance; S i is the strategy set of intelligent connected vehicle i. The optional decision space of intelligent connected vehicle i includes all coalitions G = {1, 2, …, G}; J i is the cost function of the i-th intelligent connected vehicle, which is defined as shown in formula (11). The cost function of the g-th alliance is represented by r J (G g ) indicates that it can be defined as:

[0122]

[0123] Formula (13), the cost function of the g-th alliance is the entire set of intelligent connected vehicles G in the alliance g In addition, alliances are defined to be non-overlapping, and an intelligent connected vehicle can only join one alliance at a time.

[0124] Specifically, for the i-th intelligent connected vehicle i∈M, for the i-th intelligent connected vehicle, when At that time, the G Alliance G g is superior to the kth alliance G k , which means that the i-th intelligent connected vehicle is more inclined to join the alliance G g Instead of Gk M represents the set of all intelligent connected vehicles. Represents the priority ranking of potential alliances by intelligent connected vehicle i.

[0125] It is understandable that there are many types of preference orders, and different preference orders lead to different alliance results. This embodiment adopts the cooperation order.

[0126]

[0127] In formula (14), r J (G g ∪i) indicates that intelligent connected vehicle i joins the alliance G g The price after. J (G g ) indicates Alliance G g The price. J (G k ) indicates Alliance G k The price. J (G k ∪i) indicates that intelligent connected vehicle i joins the alliance G k The total cost after the alliance is formed. M represents the set of all intelligent connected vehicles. G represents the set of all intelligent connected vehicle alliances. If the total cost decreases when the i-th intelligent connected vehicle joins the alliance, it will choose to remain in the new alliance and exit the old one. The cooperation order aims to minimize the cost of the alliance, not the cost of each individual intelligent connected vehicle. Alliance members can join or leave, which is how the alliance changes its state. At the alliance level, this manifests as a merger or split.

[0128] Determine the conditions for alliance mergers, alliance splits, and alliance exchanges, among which:

[0129] Regarding alliance mergers,

[0130] If the new alliance (excluding the i-th intelligent connected vehicle) takes a merging action, and the original alliance cost function does not increase after the i-th intelligent connected vehicle joins the new alliance, the merger is allowed:

[0131]

[0132] From (15), we can see that when the i-th intelligent connected vehicle will exit the previous alliance and join the new alliance, the overall cost of all intelligent connected vehicles in the alliance will not increase, and the performance of the alliance can be improved.

[0133] Regarding the alliance split,

[0134] If the alliance containing the i-th intelligent connected vehicle splits, the i-th intelligent connected vehicle exits the alliance, and the cost of its original alliance does not increase, then the split is allowed. As shown in formula (16):

[0135]

[0136] It can be seen that the basic movement of the alliance is caused by the entry and exit of intelligent connected vehicles. However, when an intelligent connected vehicle joins or exits, a temporary increase in the total cost of all intelligent connected vehicles may occur.

[0137] Regarding alliance exchange,

[0138] Exchange behavior will only occur between alliances when the cost of the new alliance that the i-th intelligent connected vehicle joins is less than the cost of the old alliance that the intelligent connected vehicle exits.

[0139]

[0140] Formula (17) indicates that the total cost after the exchange is no higher than that before the exchange, to ensure that the global performance of the system is not degraded.

[0141] Specifically, when all intelligent connected vehicles adopt a preference order and each intelligent connected vehicle cannot reduce its own cost by joining or exiting an alliance, the alliance is in a stable state, that is, it reaches a Nash equilibrium state, as shown in formula (18):

[0142] i∈MG g ,r J (G g )≤r J (G g ∪i)(18)

[0143] Since the number of ICVs that can form alliances is limited, and the number of combined states of ICVs is limited, the goal is to find the optimal solution for the combined states of ICVs, converge to a stable alliance partition, and ultimately achieve a stable state.

[0144] Specifically, in step S4, the kinematic constraints are determined based on the acceleration value, speed value, front wheel steering angle, yaw angle, maximum acceleration, minimum steering angle, maximum steering angle, minimum yaw angle, and maximum yaw angle of the intelligent connected vehicle;

[0145] The comfort constraint is constructed based on the front wheel steering angle, the front wheel steering angle change rate, the acceleration increment, and the jerk.

[0146] Specifically, see Figure 3As shown in FIG, which is a schematic diagram of the coordinate system of the kinematic model of the intelligent connected vehicle according to an embodiment of the present invention. The present invention selects a two-degree-of-freedom kinematic model with high accuracy and low computing resource usage as the kinematic model of the intelligent connected vehicle. The kinematic model of the intelligent connected vehicle with the rear axle center as the origin of the vehicle body coordinate system is shown in formula (19).

[0147] Regarding kinematic constraints, the kinematic constraint equation can be expressed by equation (20).

[0148]

[0149] In formula (20), a i is the acceleration value of the intelligent connected vehicle i; v i is the speed value of the intelligent connected vehicle; δ i is the front wheel turning angle of the intelligent connected vehicle; is the yaw angle of the intelligent connected vehicle; a min and a max are the minimum acceleration and maximum acceleration of intelligent connected vehicles respectively; δ min and δ min are the minimum and maximum steering angles of intelligent connected vehicles respectively; and are the minimum and maximum yaw angles of intelligent connected vehicles respectively.

[0150] In addition to the motion constraints and physical limitations of connected vehicles, comfort constraints must also be considered during the decision-making process. These constraints include limits on the front wheel steering angle and its rate of change, as well as limits on acceleration and jerk, to ensure vehicle comfort.

[0151] The constraints of the front wheel steering angle and its increment are shown in formula (21):

[0152]

[0153] In formula 21, is the front wheel turning angle of the intelligent connected vehicle i, is the maximum front wheel turning angle of the intelligent connected vehicle i. and They represent the front wheel turning angle and the maximum front wheel turning angle of the intelligent connected vehicle i within a unit step, respectively.

[0154] Acceleration increment The constraint definition of is shown in formula (22).

[0155]

[0156] In formula 22, and They represent the longitudinal acceleration and maximum acceleration of intelligent connected vehicle i within a unit step, respectively.

[0157] The jerk constraint condition is shown in formula (23):

[0158]

[0159] In formula 23, and They represent the longitudinal acceleration and maximum longitudinal jerk of the intelligent connected vehicle i, and represent the lateral acceleration and maximum lateral acceleration of the intelligent connected vehicle i, respectively.

[0160] Specifically, in step S5, during the coalition model solution process, each intelligent connected vehicle is determined as an independent population for evolution. After a fixed number of evolutions, each intelligent connected vehicle exchanges information via V2X communication technology to achieve global optimization.

[0161] Genetically encode the intelligent connected vehicle to obtain chromosomes specific to the intelligent connected vehicle;

[0162] Assign a population to each intelligent connected vehicle and perform fitness evaluation on chromosomes that meet the constraints to determine the evaluation value for each population;

[0163] Adjusting the proportion of global exploration and local optimization of individuals in the population during the evolution process according to the evaluation value;

[0164] The evaluation value is determined based on the Euclidean distance between the corresponding chromosomes of the intelligent connected vehicles.

[0165] For local optimization, the individual mutation rate of intelligent connected vehicles is determined based on traffic flow.

[0166] During the population evolution process, the number of chromosomes for crossover operations of intelligent connected vehicles is determined, and crossover operations are performed through chromosome single-point crossover.

[0167] Regarding genetic coding,

[0168] See also Figure 4 As shown in FIG, it is a schematic diagram of the gene coding of the collaborative evolution algorithm according to an embodiment of the present invention. According to the preliminary discrete intelligent connected vehicle cluster driving path points, the individual level gene coding can be constructed. In the gene coding of the intelligent connected vehicle i, Represents the path point coding value of the intelligent connected vehicle i, l represents the number of rows in the path point coding area, corresponding to the number of lanes; the chromosome consists of a total of N W The genetic makeup of a person.

[0169] The solution of cluster decision-making includes the node set of the driving path points of each intelligent connected vehicle obtained by the A* algorithm Acceleration decision set Front wheel angle decision set Alliance decision set Turning to the decision set S means going straight, L means turning left, R means turning right, and n means the number of the waypoint. represents the acceleration of intelligent connected vehicle i at path point n, represents the front wheel turning angle of the intelligent connected vehicle i at the path point n, represents the alliance to which the intelligent connected vehicle i belongs at the path point n, Represents the direction label of intelligent connected vehicle i at path point n.

[0170] Regarding population initialization fitness evaluation,

[0171] To initialize the adaptive co-evolutionary algorithm, a population of P0 individuals is assigned to each intelligent connected vehicle based on the known lane-level path, generating an initial solution that satisfies the constraints.

[0172] The present invention evaluates the fitness of chromosomes that meet the constraints. The fitness evaluation function includes comfort, energy consumption, and intersection travel time, and sets a penalty weight P for the result of violating the constraints. pun At the same time, in order to reasonably construct the fitness function of intelligent connected vehicles, formula (11) is converted to obtain formula (24).

[0173]

[0174] Therefore, the fitness S of the i-th intelligent connected vehicle in the distributed genetic algorithm optimization process can be defined as i As shown in formula (25).

[0175]

[0176] in the formula Indicates driving safety, Indicates ride comfort, Indicates driving efficiency, Indicates the energy cost, represents the eighth weighting coefficient, represents the ninth weighting coefficient, represents the tenth weighted coefficient and represents the eleventh weighting coefficient.

[0177] About dynamic balance global exploration and local optimization,

[0178] This paper adopts a diversity-based search allocation mechanism (DSAM). DSAM aims to solve the problem of rapid decline in population diversity by dynamically balancing global exploration and local optimization, thereby effectively preventing the algorithm from falling into local optimal solutions.

[0179] The key idea behind this mechanism is to divide individuals into two independent parts based on population diversity: global exploration and local optimization. Global exploration individuals are used to explore unknown areas, while local optimization focuses on finding better solutions within the currently optimal area. After measuring population diversity, individuals are assigned to either the global exploration or local optimization part. A scaling factor, r, adjusts the relative sizes of the two parts.

[0180] This paper uses Euclidean distance as a measure of population diversity. The Euclidean distance between chromosomes is defined as the square root of the sum of the squares of the differences in the corresponding genetic parameters. In this paper, the individual with the highest fitness in the population is selected as the benchmark for assessing diversity, and the Euclidean distance between each individual and the optimal individual is calculated.

[0181] Standard population diversity (SPD) describes the degree of diversity of a population. The standard population diversity is denoted as D, which is the D of P0 individuals in the population. i The average value of the value is shown in formula (26). i The chromosome is marked as C i The contribution of individual i to the standard population diversity. Specifically, D i The chromosome C with the highest fitness best Individual chromosome C i The Euclidean distance between them is calculated as shown in Equations (26) and (27), where L is the encoding length.

[0182]

[0183] From formula (27), we can see that if there are more optimal individuals C best If the individuals are similar, the diversity of the population will decrease.

[0184] See also Figure 5 As shown in FIG, which is a schematic diagram of a diversity-based search allocation mechanism according to an embodiment of the present invention, based on the diversity of the population, the present invention divides the parent population into global exploration and local optimization, and independently evolves them. The ratio of global exploration to local optimization is adjusted by the proportional factor r, as shown in equation (28). The first preset parameter h and the second preset parameter l limit the value range of r to [l, h]. The number of individuals in the global exploration part S1 and the local optimization part S2 are given by equations (28) and (29), where P0 represents the total number of individuals in the population.

[0185]

[0186] S1=P0r,S2=P0(1-r) (29)

[0187] To ensure the algorithm possesses both global search capabilities and local optimization capabilities, the scaling factor r is adaptively adjusted within the range of [0.5, 0.9]. This ensures that at least half of the individuals perform local search, while no fewer than 10% perform global exploration. As population diversity decreases, reducing r increases the proportion of global exploration, encouraging the population to explore uncharted areas. Conversely, as diversity increases and the population's distribution becomes wider, the algorithm will focus more on the local area surrounding the individual with the highest fitness within the current population, performing local optimization search within this area.

[0188] See also Figure 6 As shown, it is a schematic diagram of the variation of traffic flow guidance in an embodiment of the present invention. For local optimization, the individual variation rate of intelligent connected vehicles is determined based on traffic flow;

[0189] Regarding traffic flow guided mutation, in the diversity-based search allocation mechanism (DSAM), the local optimization part focuses on improving the quality of the solution through traffic flow guided mutation. The probability of mutation operation P mut It is affected by the combined effects of individual fitness, contribution to population diversity, and real-time traffic flow.

[0190] The specific calculation method of mutation probability is shown in formula (30), where f par represents the parent individual, f max represents the maximum fitness of the population, f min represents the minimum fitness of the population.

[0191]

[0192] Designing individual mutation rates from the perspective of population diversity As shown in formula (31). In the local optimization of the partial subpopulation, the main purpose is to find a better solution around the current optimal individual. When the standard population diversity D is equal to 0, all individuals have the same chromosome, When the maximum value is reached, the algorithm will encourage the population to mutate and produce more new individuals.

[0193]

[0194] In order to enable the cluster to dynamically adjust the individual mutation rate based on the real-time traffic flow T Ensure that the algorithm remains robust under different traffic conditions, The calculation is shown in formula (32). The variation diagram of traffic flow guidance is shown in (32).

[0195]

[0196] The real-time traffic flow T refers to the number of vehicles passing through a specific road section in a unit of time. Define the maximum traffic flow T max is the maximum possible traffic volume on the road section. An adjustment coefficient μ is introduced to control the impact of traffic flow on the mutation rate. As T increases, the mutation rate increases linearly, helping to introduce genetic diversity and avoid premature convergence. By dynamically adjusting the mutation rate with real-time traffic flow, the genetic algorithm can adapt to varying traffic conditions: increasing the mutation rate during high traffic to increase population diversity, and decreasing it during low traffic to maintain the stability of superior genes.

[0197] The probability of mutation operation P mut The calculation of is shown in Equation (33). Under normal traffic conditions, the baseline mutation rate p0 is determined. This is the standard rate at which genetic mutation occurs in the algorithm, ensuring a balance between exploration and exploitation in the search space. σ is the weight parameter between the individual fitness and the population diversity mutation rate, and θ is the traffic flow mutation rate weight parameter.

[0198]

[0199] As the above formula shows, when individual fitness is low and population diversity is insufficient, the mutation rate will increase significantly, promoting the algorithm to explore new optimal solutions. Considering traffic flow as an independent factor enhances the model's adaptability and flexibility to varying traffic conditions. This dynamic adjustment mechanism enables the algorithm to adjust its search strategy based on real-time traffic data, achieving more effective optimization.

[0200] Regarding chromosome crossing over and mutation,

[0201] During the population evolution process, the number of chromosomes for the crossover operation of intelligent connected vehicles is set to N cros , and perform crossover operation by chromosome single-point crossover. Through crossover operation, N cros chromosomes to the next generation. In the process of iterative evolution of the population, chromosome mutation can make the algorithm approach the global optimal solution. The number of chromosomes involved in the mutation operation is N p -N cros , and the probability of mutation is P mut In the mutation process, mutation operations mainly include three categories:

[0202] ① Changes in the front wheel angle of intelligent connected vehicles;

[0203] ② Acceleration changes of intelligent connected vehicles;

[0204] ③ Changes in the alliance to which intelligent connected vehicles belong.

[0205] Regarding population periodic global optimization,

[0206] In the adaptive co-evolutionary algorithm, N p The initial population is assigned P0 individuals, and the optimization step size is set to T choose Each intelligent connected vehicle completes the population evolution process independently. During the evolution process, every T choose Second evolution, exchange information with the current population of other intelligent connected vehicles, and output N cho Optimizing the step size setting can reduce the amount of data in real-time information communication, thereby effectively reducing the communication load of the V2X network and accelerating the convergence of the algorithm.

[0207] Specifically, in step S6, step S5 is repeated until a termination condition is met to obtain an optimal trajectory solution. The termination condition is that the number of global optimization cycles reaches a set maximum value or the overall fitness of the intelligent connected vehicle does not improve under a predetermined number of global optimizations.

[0208] It is understandable that when the number of cycles of global optimization in the above chromosome iteration process reaches the set maximum value N global Or the total fitness of the intelligent connected vehicle is N window If there is no improvement within the global optimization times, the loop can be terminated and the optimal decision is output.

[0209] Thus far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art may make equivalent changes or substitutions to the relevant technical features, and the technical solutions after such changes or substitutions will fall within the scope of protection of the present invention.

[0210] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that the present invention is susceptible to various modifications and variations. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention are intended to be within the scope of protection of the present invention.

Claims

1. A macro-micro integrated alliance decision-making method based on intelligent connected vehicle clusters, characterized by: include: Step S1, constructing an initial path for the intelligent connected vehicle cluster; Step S2: constructing an alliance set for the intelligent connected vehicle cluster and constructing a cost function for each intelligent connected vehicle in the alliance set, including a driving safety cost function, a riding comfort cost function, a driving efficiency cost function, and an energy consumption cost function; The driving safety cost function is the sum of the longitudinal safety cost and the lateral safety cost; The ride comfort cost function is determined based on the lateral and longitudinal accelerations of the intelligent connected vehicle; The driving efficiency cost function is determined based on the longitudinal speed; The energy consumption cost function is determined based on air density, vehicle frontal area, air resistance coefficient, rolling resistance coefficient, vehicle mass, gravitational acceleration, vehicle comprehensive acceleration, fuel consumption rate and engine efficiency; The longitudinal safety cost is related to the longitudinal gap and the relative speed to the preceding vehicle, and the lateral safety cost is related to the relative distance and relative speed between the intelligent connected vehicle and the adjacent vehicle. Step S3, constructing an alliance game model for the alliance set, and setting the optimization goal to minimize the cost function; Step S4, determining the constraints for the game decision, including kinematic constraints and comfort constraints; The kinematic constraints are determined based on the acceleration value, speed value, front wheel steering angle, yaw angle, maximum acceleration, minimum steering angle, maximum steering angle, minimum yaw angle, and maximum yaw angle of the intelligent connected vehicle; The comfort constraint is constructed based on the front wheel steering angle, the front wheel steering angle change rate, the acceleration increment and the jerk; Step S5: During the coalition model solution process, each intelligent connected vehicle is determined as an independent population for evolution. After a fixed number of evolutions, each intelligent connected vehicle exchanges information via V2X communication technology to achieve global optimization. Each intelligent connected vehicle is identified as an independent population for evolution, where Genetically encode the intelligent connected vehicle to obtain chromosomes specific to the intelligent connected vehicle; Assign a population to each intelligent connected vehicle and perform fitness evaluation on chromosomes that meet the constraints to determine the evaluation value for each population; Adjusting the proportion of global exploration and local optimization of individuals in the population during the evolution process according to the evaluation value; The evaluation value is determined based on the Euclidean distance between the chromosomes corresponding to the intelligent connected vehicles; Step S6, repeat step S5 until the termination condition is met and the optimal trajectory solution is obtained.

2. The macro-micro integrated alliance decision-making method based on intelligent connected vehicle clusters according to claim 1 is characterized in that: In step S1, the process of constructing the initial path of the intelligent connected vehicle cluster includes: Pre-build road network data, including node division and construction of directed graph for the road network; Each intelligent connected vehicle loads the road network data to determine the starting point and end point and independently runs the A* algorithm to generate the initial path.

3. The macro-micro integrated alliance decision-making method based on intelligent connected vehicle clusters according to claim 1 is characterized in that: In step S3, the process of constructing the alliance game model for the alliance set includes: Initialize the alliance and divide the alliance into straight alliance, left turn alliance, and right turn alliance based on the vehicle's direction label; Determine the preference order of the alliance game process; Determine the behavioral conditions for alliance mergers, alliance splits, and alliance exchanges.

4. The macro-micro integrated alliance decision-making method based on intelligent connected vehicle clusters according to claim 3 is characterized in that: It also includes determining whether the alliance is in a stable state, in which If all intelligent connected vehicles adopt a preference order and each of the intelligent connected vehicles cannot reduce its own cost by joining or exiting the alliance, the alliance is in a stable state.

5. The macro-micro integrated alliance decision-making method based on intelligent connected vehicle clusters according to claim 1 is characterized in that: For local optimization, the individual mutation rate of intelligent connected vehicles is determined based on traffic flow.

6. The macro-micro integrated alliance decision-making method based on intelligent connected vehicle clusters according to claim 1 is characterized in that: During the population evolution process, the number of chromosomes for crossover operations of intelligent connected vehicles is determined, and crossover operations are performed through chromosome single-point crossover.

7. The macro-micro integrated alliance decision-making method based on intelligent connected vehicle clusters according to claim 1 is characterized in that: In step S6, the termination condition is that the number of cycles of global optimization reaches a set maximum value or the total fitness of the intelligent connected vehicle does not improve under a predetermined number of global optimizations.

Images