CAV information physical cooperative passing method in multi-intersection mixed traffic environment

Abstract

The invention belongs to the technical field of intelligent traffic, and discloses a CAV information physical cooperative passing method in a multi-intersection mixed traffic environment. Aiming at the problem that the traditional control method is difficult to cope with the mixed traffic flow space-time coordination dilemma, the invention constructs a three-layer coordination control architecture of vehicle and road cloud hierarchy-vehicle and road hierarchy-vehicle and vehicle hierarchy based on an intelligent vehicle information physical system (IVCPS), wherein the vehicle and road cloud hierarchy optimizes the trunk signal period and phase difference through model predictive control (MPC); the vehicle-road level adopts a dual-scale MPC to realize coordination of signal timing and CAV target arrival time; and the vehicle-vehicle hierarchy optimizes vehicle behaviors through dynamic functional area division and an ML-SLSG lane changing strategy. The multi-intersection traffic efficiency and the fuel economy are improved, and a theoretical support and a technical path are provided for an intelligent network connection traffic system.

Description

Technical Field

[0001] This invention belongs to the field of intelligent transportation technology, specifically relating to a CAV (Car-Road-Cloud) cyber-physical cooperative passage method in a multi-intersection mixed traffic environment. It is applicable to the vehicle-road-cloud integrated cyber-physical system (IVCPS) architecture and can be implemented in multi-intersection scenarios such as urban trunk lines and transportation hubs, realizing collaborative control of global coordination between vehicles, roads and cloud, local optimization of vehicles and roads, and individual execution of vehicles. Background Technology

[0002] With the continuous expansion of urban road network and the rapid growth of motor vehicle ownership, the spatiotemporal uncertainty and dynamic complexity of traffic flow have significantly increased. This has led to intersections gradually becoming bottlenecks restricting the efficiency of the road network, resulting in severe traffic congestion and pollution problems. Coordinated control of multiple intersections on urban arterial roads is a key link in improving the efficiency of the road network. Traditional traffic control methods (such as adaptive traffic signal control and sensor signal control) are limited by perception capabilities and computational efficiency, making it difficult to cope with the spatiotemporal coordination dilemmas caused by the dynamic evolution of heterogeneous traffic flows. They also have inherent defects such as "short-sighted" control strategies and insufficient adaptability.

[0003] Thanks to the rapid development of next-generation information and communication technologies and intelligent connected vehicles (CAVs), deep integration of vehicle-road-cloud has provided a new path to solving the aforementioned challenges. Through V2X technology and the dynamic multi-agent data interaction of vehicle-road-cloud systems, intelligent connected vehicles (CAVs) and connected vehicles (CVs) can share status information in real time and actively participate in collaborative decision-making, optimizing the spatiotemporal distribution of traffic flow at the physical-information spatial coupling level, thus creating conditions for building a closed-loop "perception-decision-control" system. This has driven a shift in traffic control paradigms from "passive response" to "active collaboration," becoming a research hotspot in the field of intelligent transportation. Against this backdrop, current research faces three key challenges:

[0004] (1) The driving behavior of human drivers is an unpredictable game process. The heterogeneity of traffic flow exhibited by different driving behaviors increases the uncertainty of traffic flow and vehicle arrival time between different intersections, and the difficulty of multi-intersection collaborative control is also increased accordingly. The introduction of connected vehicles (CVs) and intelligent connected vehicles (CAVs) will gradually evolve traffic flow from ordinary vehicles to a mixed flow of RVs, CVs and CAVs coexisting, and this will exist for a long time. The randomness of RV driving behavior, the incomplete controllability of CV driving behavior, the coupling of CV / CAV lateral and longitudinal trajectories, and the spatiotemporal sparsity of CV / CAVs pose serious challenges to the research on optimization control for improving future road capacity.

[0005] (2) Closed-loop control of signal timing and vehicle speed coordination. Based on the segmented coordinated control of traffic flow, a reasonable signalized intersection control strategy is crucial to further improve the overall traffic efficiency of roads. At present, intersection control optimization mainly focuses on two core methods: control methods based on intelligent traffic signals (roadside) and control methods based on intelligent vehicles (vehicleside). Existing research mostly adopts an open-loop control architecture, which does not fully consider the dynamic feedback mechanism between CAV trajectory optimization and signal timing. However, in actual traffic systems, the two have a close bidirectional coupling relationship. Signal timing directly affects the motion state of individual vehicles, thus significantly affecting their fuel economy, while real-time vehicle trajectory data is a key input to the traffic signal control model.

[0006] (3) Global optimization of multiple intersections

[0007] In real-world urban traffic systems, adjacent intersections are interconnected to form continuous traffic corridors. Adjustments to the control strategies of any intersection will affect traffic flow at adjacent intersections, and vehicle passage will also be influenced by the control measures of upstream and downstream intersections. Thanks to the development of connected technologies such as Vehicle-to-Everything (V2X), intersections can exchange information through network communication, laying the technological foundation for implementing coordinated control of arterial roads.

[0008] In summary, this invention proposes a multi-level control architecture based on Intelligent Vehicle Cyber-Physical Systems (IVCPS) to characterize the cross-scale interaction mechanism of traffic objects in multi-intersection mixed traffic environments, addressing the Sustainable Vehicle Collaborative Traffic (SVCC) problem. Based on this architecture, a multi-level SVCC model is established, and a hierarchical cloud control algorithm is proposed for solving it. This invention explores the integrated collaborative mechanism of "vehicle-road-cloud" from a cyber-physical fusion perspective, aiming to improve intersection traffic efficiency and system resilience, and providing theoretical support and technical pathways for intelligent connected transportation systems. Summary of the Invention

[0009] In view of this, this invention addresses the problem of cooperative passage of intelligent connected vehicles (CAVs) in mixed traffic environments at multiple intersections, providing a CAV cyber-physical cooperative passage method. The aim is to improve intersection traffic efficiency and system resilience, providing theoretical support and technical pathways for intelligent connected transportation systems.

[0010] To achieve the above objectives, the present invention adopts the following technical solution:

[0011] A CAV (Cyber-Physical Access Control) method for navigating mixed traffic at multiple intersections includes the following steps:

[0012] S1. Construct a three-layer cooperative control architecture based on IVCPS to decouple the multi-intersection cooperative passage problem into vehicle-road-cloud level, vehicle-road level and vehicle-to-vehicle level control problems;

[0013] S2. Vehicle-Road-Cloud Hierarchical Modeling: Based on road network traffic status data, construct a green wave coordinated control MPC model to optimize the trunk line public signal cycle, phase difference, and green wave guidance speed;

[0014] S3. Vehicle-Road Hierarchical Modeling: Based on vehicle group information obtained from V2I communication, a signal timing and vehicle speed coordination MPC model is constructed, and the green light ratio and CAV target arrival time are output.

[0015] S4. Vehicle-to-vehicle hierarchical modeling: Dynamically divide road segment functional areas and achieve coordinated control of mixed traffic flow through ML-SLSG lane-changing strategy and NMPC trajectory optimization.

[0016] Furthermore, step S2 includes the following sub-steps:

[0017] S2.1 Defines the green wave bandwidth reduction rate P B With the reduction rate P of traffic speed V As an evaluation indicator;

[0018] The formula for calculating the green wave bandwidth reduction rate is:

[0019]

[0020]

[0021]

[0022] In the formula, B u B is the uplink green wave bandwidth; d C represents the downlink green wave bandwidth; C represents the common signal period. This represents the theoretical maximum total green wave bandwidth; This represents the actual total green wave bandwidth; λ is the green signal ratio in each direction. This indicates the green light ratio at the first intersection or road segment in the uphill direction; This indicates the green light ratio at the first intersection or road segment in the downhill direction; This represents the ideal green wave bandwidth in the upward direction; This represents the ideal green wave bandwidth in the downlink direction;

[0023] Traffic speed reduction rate P V It is the percentage of the maximum difference between the green wave design speeds and the total speed reduction rate of the area. The calculation formula is:

[0024]

[0025] Section Traffic speed reduction rate The calculation formula is:

[0026]

[0027] In the formula, This represents the set of all road segments within the area to be evaluated; Indicates road segment The maximum design speed; Indicates road segment The minimum design speed; Indicates road segment The actual traffic speed;

[0028] S2.2 Constructing the vehicle-road-cloud hierarchical MPC objective function and constraints;

[0029] The objective function for the vehicle-road-cloud layer is obtained by weighted summation of the green wave bandwidth reduction rate and the traffic speed reduction rate:

[0030]

[0031] Its constraints are:

[0032]

[0033] In the formula, K represents the objective function of vehicle-road-cloud hierarchical MPC; B K represents the green wave bandwidth weighting coefficient, with a value range of (0,1), used to balance the optimization priority of green wave bandwidth and traffic speed; when K B When K → 0, prioritize optimizing traffic speed; when K B When →1, priority is given to optimizing green wave bandwidth; The predicted time domain length indicates that the time range for optimal control includes the period from the current time. arrive of total At that moment; This represents the time index within the prediction time domain; Indicates the time of prediction The green wave bandwidth reduction rate; Indicates the time of prediction Traffic speed reduction rate; This indicates the current optimization time, and its value range is... ; Indicates road segment The signal phase difference, that is, the difference between the downstream intersection j and the upstream intersection i at time... The time difference between the start of the green light; Indicates road segment Length; Indicates road segment At any moment The target traffic speed; Indicates road segment At any moment The period number correction coefficient, which is an integer, is used to adjust the signal period matching. Indicates the duration of the current signal cycle; , This represents the minimum and maximum duration of the signal period; Indicates road segment The phase difference compensation is used to correct system deviations in signal timing; This represents the cycle number correction coefficient for the ring network. It is an integer and is used for signal coordination optimization of the ring network.

[0034] Furthermore, step S3 includes the following sub-steps:

[0035] S3.1 Core Constraints;

[0036] I. Feasibility constraints regarding vehicle arrival time:

[0037]

[0038]

[0039]

[0040] In the formula, The core output of the model represents the target time when the lead car of the CAV team q reaches the stop line; Let q represent the theoretical minimum feasible time when the lead car of the CAV team reaches the stop line. This represents the theoretical maximum feasible time when the lead car of CAV team q reaches the stop line; This indicates the initial position of CAV team q at the entrance of the controlled area; This represents the initial velocity of the CAV vehicle q at the entrance to the controlled area; Indicates the maximum comfortable acceleration; Indicates the maximum comfortable deceleration; This represents the set of all CAV (Continuous Access Vehicle) vehicles entering the intersection control area within the prediction time domain;

[0041] II. Signal Logic and Coordination Constraints;

[0042] Signal logic:

[0043]

[0044] For any pair of conflicting phases Linearization needs to be achieved by introducing auxiliary binary variables to ensure that their green light periods do not overlap; Represents the set of conflicting phase pairs;

[0045] In the formula, Indicates signal phase The duration of the green light; Yϕ represents the signal phase. The duration of the yellow light; ARϕ represents the signal phase. The duration of the all-red sequence; The set of all phases contained in phase group r; Represents the set of all signal phase groups. This indicates that a periodic constraint is applied to all phase groups; It indicates the duration of the current signal cycle, which is equal to the sum of the "green + yellow + all red" times of each phase in the phase group;

[0046] III. Signal-vehicle coupling;

[0047] To prevent traffic flow conflicts, only one phase can be activated at any given time; therefore, there are constraints between phases:

[0048] Phase uniqueness:

[0049]

[0050] Time window constraints:

[0051]

[0052]

[0053] In the formula, It is a binary variable. This indicates that CAV team q has been assigned to the phase. Passage; Φq represents the set of passable phases for CAV vehicle q; Indicates phase The green light start time; a large M constant, used to linearize logical constraints; This indicates the number of vehicles included in the CAV fleet q; Indicates the safe headway of the CAV team;

[0054] IV. Mixed Traffic Management:

[0055]

[0056]

[0057] In the formula, Indicates phase The queue clearing time for the HDV / CV traffic flow prediction served at time k is a dynamically changing parameter that depends on real-time traffic demand. Represents the set of all signal phases at the intersection; Indicates phase HDV / CV traffic flow prediction queue length for services; Indicates phase The saturation flow rate of the inlet channel;

[0058] S3.2 Dual-Scale MPC Objective Function:

[0059] The vehicle-road control layer uses individual intersections as basic control units, responsible for tactical-level decision-making and planning. Its core objective is to coordinate the optimization of signal timing and controlled vehicle (CAV) arrival times for mixed traffic flows entering its control area within a finite prediction time domain. This optimization aims to minimize the total delay at local intersections while adhering to macro-coordination parameters (as soft constraints) issued from the cloud, and indirectly creates conditions for downstream energy consumption optimization.

[0060] The mathematical expression for the optimization objective is:

[0061]

[0062]

[0063] In the formula, This represents the objective function of the vehicle-road control layer. This represents the set of all entrance lanes; Indicates import channel The CAV team gathered inside; This represents the total number of vehicles in fleet q; To delay the weighting; Indicates the vehicle delay time; This refers to the actual arrival time. For the arrival time of the free flow; To coordinate the penalty coefficient; This represents the actual phase difference at intersection k at the current time. This is the reference phase difference issued at the vehicle-road-cloud level.

[0064] Furthermore, step S4 includes the following sub-steps:

[0065] S4.1 Dynamic Functional Area Division:

[0066] Based on the real-time lane-changing demand of traffic flow and the queuing prediction information fed back by downstream ICL, the road segment is dynamically divided into three mutually exclusive functional areas: lane-changing area, speed adjustment area, and queuing area.

[0067] lane change area Located upstream of the road section, it is used to handle lateral lane changing needs;

[0068]

[0069] Speed ​​adjustment zone Located in the middle of the road section, it is used for fine-tuning of vehicle longitudinal speed and trajectory optimization;

[0070]

[0071] Follow the entire team area Located downstream of the road section, it is used to stabilize the convoy, suppress disturbances, and prepare for entering the intersection;

[0072]

[0073] In the formula, Indicates the total length of the road segment This indicates the moment when vehicle i completes the lane change in lane group b and road segment l; Indicates vehicle correspondence Location; This indicates the number of vehicles queuing as predicted downstream. To ensure safe distance between vehicles; This refers to the length of the vehicle body. This refers to the speed at which traffic passes through the intersection. Indicates the maximum comfortable deceleration;

[0074] S4.2 ML-SLSG lane-changing strategy;

[0075] Within the lane change zone, the ML-SLSG lane change strategy is implemented to efficiently and safely resolve lane change conflicts in mixed traffic flows.

[0076] I. Vehicles entering the lane-changing area are divided into queue A (which requires lane changing) and queue B (which does not require lane changing) based on the relationship between the vehicle's current lane and the target lane.

[0077] Lane changing queue:

[0078] Non-lane-changing queue:

[0079] In the formula, i is the vehicle number; This represents the current lane of the i-th vehicle; Indicates the target lane for the i-th vehicle; This represents a queue of all vehicles whose current lane differs from the target lane and require lane changing. This represents a queue of all vehicles whose current lane is the same as the target lane and do not require lane changing.

[0080] II. The objective of the ML-SLSG lane-changing strategy is to assign differentiated induced speeds to lane-changing and non-lane-changing queues, enabling vehicles in queue A to form a predetermined longitudinal offset distance with vehicles in queue B, thereby safely merging into the target lane; the induced speeds are as follows:

[0081]

[0082] In the formula, and This is a lane-changing queue. Non-lane-changing queues The set induction speed;

[0083] Safety distance constraints:

[0084]

[0085]

[0086] In the formula, Indicates the longitudinal position of the vehicle ahead; Indicates the current longitudinal position of the vehicle; Indicates the longitudinal position of the vehicle behind; Indicates the length of the vehicle ahead; Indicates the length of the vehicle behind; It is a dynamic safety distance related to speed;

[0087] Safe lane changing time:

[0088]

[0089] In the formula, This indicates the initial longitudinal position of the lead car (B) in the queue; This indicates the longitudinal position of the last car in queue A at the initial moment;

[0090] Lane change zone length:

[0091]

[0092] In the formula, This indicates the longitudinal position of the lead car A in the queue at the initial moment; Indicates the lane change operation time; Indicates an additional safety margin distance;

[0093] S4.3 Vehicle-Road Hierarchical Trajectory Optimization;

[0094] Once the vehicle enters the speed adjustment zone and the platooning zone, the vehicle's main task is to perform refined longitudinal trajectory coordination, which is achieved through a nonlinear model predictive control framework (NMPC) running on the CAV onboard platform.

[0095] Objective function:

[0096]

[0097] In the formula, The decision variable is the acceleration sequence of CAV in the prediction time domain; Indicates the length of the prediction time domain; This represents the fuel consumption weighting coefficient; Indicates the future number Fuel consumption rate at any given time; This represents the comfort weighting coefficient; Indicates the future number Acceleration at any moment; This represents the arrival time weighting coefficient; Indicates the predicted time of arrival at the target point; Indicates the expected arrival time;

[0098] This represents a multi-objective optimization function that minimizes the acceleration sequence as the decision variable. This objective function integrates three optimization objectives—fuel consumption, comfort, and arrival time deviation—into a single index through weighted summation. By adjusting the weighting coefficients to balance the priorities of the different objectives, the final solution yields the acceleration sequence that minimizes the overall objective. This enables multi-objective collaborative control of intelligent connected vehicles in longitudinal trajectory optimization.

[0099] Furthermore, in step S4.2, The calculation expression is:

[0100]

[0101] In the formula, Indicates the vehicle's speed; Indicates the safe time interval; This indicates the minimum safe distance.

[0102] Beneficial effects:

[0103] 1. This invention significantly improves the traffic efficiency and fuel economy of intersections under mixed traffic flow through cross-level coordination of vehicle-road-cloud control, vehicle-road control, and vehicle-to-vehicle control.

[0104] 2. Through cyber-physical fusion, integrated collaboration of "vehicle-road-cloud" has been achieved, providing an innovative control framework for intelligent connected transportation systems.

[0105] 3. By employing dual-scale model predictive control, coordinated optimization of signal timing and vehicle trajectory is achieved, avoiding the limitations of traditional open-loop control.

[0106] 4. The ML-SLSG strategy and dynamic functional area division method proposed in this invention effectively solve the problems of lane changing conflicts and fleet organization in mixed traffic flow. Attached Figure Description

[0107] Figure 1 This is the overall framework of the CAV (Cyber-Physical Access Control) method for a mixed traffic environment at multiple intersections, as described in this invention.

[0108] Figure 2 A schematic diagram of dynamic modeling for vehicle-road-cloud collaborative control of vehicle groups. Detailed Implementation

[0109] To make the technical solutions, advantages, and objectives of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the described embodiments of the present invention without creative effort are within the protection scope of this application.

[0110] like Figure 1 and Figure 2 As shown, this invention provides a CAV (Cyber-Physical Access Control) method for navigating mixed traffic environments at multiple intersections, comprising the following steps:

[0111] S1. Construct a three-layer cooperative control architecture based on IVCPS to decouple the multi-intersection cooperative passage problem into vehicle-road-cloud level, vehicle-road level and vehicle-to-vehicle level control problems;

[0112] To address the collaborative control problem of mixed traffic flow on urban arterial roads, this invention proposes a hierarchical and segmented collaborative control framework that integrates cycle coordination, signal-vehicle speed coordination, and road segment traffic flow organization. Based on the Cyber-Physical Systems (CPS) concept, this framework decouples the complex network-level collaborative traffic problem in the spatiotemporal dimensions into three independent yet tightly coupled control levels: vehicle-road-cloud level, vehicle-road level, and vehicle-to-vehicle level.

[0113] (1) Vehicle-Road-Cloud Control Level. Traditional trunk line coordinated control often uses fixed cycle durations and phase differences, which are difficult to adapt to the dynamic fluctuations of mixed traffic flows. To this end, this invention introduces a global coordination mechanism at the vehicle-road-cloud control level of the hierarchical control framework. This level ignores micro-vehicle dynamics and treats trunk line traffic flow as a macro-entity. Its core task is to achieve optimal allocation of spatiotemporal resources at the strategic level of trunk lines from a global perspective.

[0114] Specifically, the vehicle-road-cloud control layer operates on a relatively long time step (e.g., minutes). Based on aggregated information such as regional traffic flow and saturation from the lower layer (vehicle-road control layer), it dynamically optimizes the common cycle time (C) of the entire arterial coordination zone and the phase difference (Offset) of each intersection according to the principle of comprehensive "bandwidth-speed" optimization. Furthermore, this layer is also responsible for generating a basic speed guidance profile based on road segment physical characteristics and historical data. Ultimately, these vehicle-road-cloud coordination parameters will be sent as guiding instructions to the lower-level controllers.

[0115] (2) Vehicle-Road Control Level: This level uses intersections as the basic control units and employs distributed model predictive control (MPC). After receiving the coordination parameters from the vehicle-road cloud control level, each intersection controller performs local signal-vehicle speed coordination optimization based on a spatiotemporal dual-scale approach. At the slow scale (second-level), signal timing is optimized to ensure local traffic efficiency while aligning with the phase difference of the vehicle-road cloud control level as much as possible. A key output of this optimization level is generating a series of key location sequences for connected vehicles (CAVs) within the control range. At the fast scale (sub-second-level), the CAV performs optimal trajectory planning based on the key location sequences planned at the slow scale, with the goal of minimizing fuel consumption and improving comfort.

[0116] (3) Vehicle-to-Vehicle (V2V) Control Level: This level is responsible for organizing traffic flow between two intersections. First, based on the real-time lane-changing demand of the traffic flow, the road segment is dynamically divided into a lane-changing zone, a speed adjustment zone, and a platooning zone. In the lane-changing zone, lane-based speed-induced staggered lane changing (ML-SLSG) strategy is used to efficiently organize platooning lane changes. In the speed adjustment zone, the vehicle-to-infrastructure (CAV) precisely tracks the trajectory issued by the V2V control level and indirectly guides the following human-driven vehicles (HDVs) through restraint control based on "1+N" logical subgroups using an intelligent driver model (IDM). Finally, a regular and stable traffic flow queue is formed in the platooning zone, and its highly deterministic state information (such as estimated time of arrival (ETA)) is fed back to the downstream V2V control level controller.

[0117] S2. Vehicle-Road-Cloud Hierarchical Modeling: Based on road network traffic status data, construct a green wave coordinated control MPC model to optimize the trunk line public signal cycle, phase difference, and green wave guidance speed;

[0118] The vehicle-road-cloud hierarchy is the top layer of the three-layer collaborative control architecture proposed in this invention. Its core objective is to break the limitation of fixed vehicle speed in traditional green wave coordination. By integrating and coordinating the optimization of signal coordination parameters and green wave guidance speed, it generates a globally optimal and dynamically adaptive traffic plan for trunk traffic flow.

[0119] S2.1 Defines the green wave bandwidth reduction rate PB With the reduction rate P of traffic speed V As an evaluation indicator;

[0120] Considering that the goal of arterial road green wave coordinated control is to ensure that as many controlled vehicles as possible pass through the arterial road intersection continuously at the most stable speed possible, green wave bandwidth and green wave speed are selected as evaluation elements of this model scheme. To ensure the comparability of regional green wave coordinated control evaluation indicators, the total regional green wave bandwidth reduction rate and traffic speed reduction rate are defined as evaluation indicators of the model scheme.

[0121] The formula for calculating the green wave bandwidth reduction rate is:

[0122]

[0123]

[0124]

[0125] In the formula, B u B is the uplink green wave bandwidth; d C represents the downlink green wave bandwidth; C represents the common signal period. This represents the theoretical maximum total green wave bandwidth; This represents the actual total green wave bandwidth; λ is the green signal ratio in each direction. This indicates the green light ratio at the first intersection or road segment in the uphill direction; This indicates the green light ratio at the first intersection or road segment in the downhill direction; This represents the ideal green wave bandwidth in the upward direction; This represents the ideal green wave bandwidth in the downlink direction;

[0126] Traffic speed reduction rate P V It is the percentage of the maximum difference between the green wave design speeds and the total speed reduction rate of the area. The calculation formula is:

[0127]

[0128] Section Traffic speed reduction rate The calculation formula is:

[0129]

[0130] In the formula, This represents the set of all road segments within the area to be evaluated; Indicates road segment The maximum design speed; Indicates road segment The minimum design speed; Indicates road segment The actual traffic speed;

[0131] S2.2 Constructing the vehicle-road-cloud hierarchical MPC objective function and constraints;

[0132] The objective function for the vehicle-road-cloud layer is obtained by weighted summation of the green wave bandwidth reduction rate and the traffic speed reduction rate:

[0133]

[0134] Its constraints are:

[0135]

[0136] In the formula, K represents the objective function of vehicle-road-cloud hierarchical MPC; B K represents the green wave bandwidth weighting coefficient, with a value range of (0,1), used to balance the optimization priority of green wave bandwidth and traffic speed; when K B When K approaches 0, the model will prioritize the coordinated control scheme that maximizes the green wave vehicle speed while meeting the green wave design requirements; when K... B When K approaches 1, the model will prioritize the coordinated control scheme with the largest possible green wave bandwidth within the green wave speed optimization range; when K... B As the range (0,1) is gradually increased, the model will assign greater weight to the green wave bandwidth of the optimization scheme. The predicted time domain length indicates that the time range for optimal control includes the period from the current time. arrive of total At that moment; This represents the time index within the prediction time domain; Indicates the time of prediction The green wave bandwidth reduction rate; Indicates the time of prediction Traffic speed reduction rate; This indicates the current optimization time, and its value range is... ; Indicates road segment The signal phase difference, that is, the difference between the downstream intersection j and the upstream intersection i at time... The time difference between the start of the green light; Indicates road segment Length; Indicates road segment At any moment The target traffic speed; Indicates road segment At any moment The period number correction coefficient, which is an integer, is used to adjust the signal period matching. Indicates the duration of the current signal cycle; , This represents the minimum and maximum duration of the signal period; Indicates road segment The phase difference compensation is used to correct system deviations in signal timing; This represents the cycle number correction coefficient for the ring network. It is an integer and is used for signal coordination optimization of the ring network.

[0137] S3. Vehicle-Road Hierarchical Modeling: Based on vehicle group information obtained from V2I communication, a signal timing and vehicle speed coordination MPC model is constructed, and the green light ratio and CAV target arrival time are output.

[0138] S3.1 Core Constraints;

[0139] I. Vehicle arrival time feasibility constraint: In order to ensure the executability of instructions in the hierarchical control framework, the intersection model should follow the vehicle arrival time feasibility constraint.

[0140]

[0141]

[0142]

[0143] In the formula, The core output of the model represents the target time when the lead car of the CAV team q reaches the stop line; The value represents the theoretical minimum feasible time when the lead car of the CAV team q reaches the stop line. This value is derived from the vehicle's kinematic model, taking into account its initial state at the entrance of the control area. This indicates the latest time that the lead car of CAV team q can reach the stop line without stopping and while maintaining minimum operating efficiency, taking into account its minimum permissible acceleration or maximum comfortable deceleration. and minimum operating speed ; This indicates the initial position of CAV team q at the entrance of the controlled area; This represents the initial velocity of the CAV vehicle q at the entrance to the controlled area; Indicates the maximum comfortable acceleration; Indicates the maximum comfortable deceleration; This represents the set of all CAV fleets (or individual CAVs) entering the intersection control area within the prediction time domain;

[0144] II. Signal Logic and Coordination Constraints;

[0145] As an important part of the road network, signalized intersections are where vehicles from different directions merge, cross, and diverge, inevitably resulting in overlapping travel routes and traffic conflicts due to competition for road resources.

[0146] For any two consecutive phases within the signal loop r:

[0147]

[0148]

[0149] Signal logic:

[0150]

[0151] For any pair of conflicting phases Linearization needs to be achieved by introducing auxiliary binary variables to ensure that their green light periods do not overlap; Represents the set of conflicting phase pairs;

[0152] In the formula, Indicates signal phase The duration of the green light; Yϕ represents the signal phase. The duration of the yellow light; ARϕ represents the signal phase. The duration of the all-red sequence; The set of all phases contained in phase group r; Represents the set of all signal phase groups. This indicates that a periodic constraint is applied to all phase groups; It represents the duration of the signal period at the current moment, which is equal to the sum of the "green + yellow + all red" times of each phase in the phase group;

[0153] The above formula forces the total duration of all signal loops to equal the macro-coordination cycle, which is the foundation for achieving regional green waves. By defining the timing relationship between phases and the minimum safe interval (yellow light + all red), the safety of signal switching is fundamentally guaranteed, preventing conflicting traffic flows.

[0154] III. Signal-vehicle coupling;

[0155] To prevent traffic flow conflicts, only one phase can be activated at any given time; therefore, there are constraints between phases:

[0156] Phase uniqueness:

[0157]

[0158] Time window constraints:

[0159]

[0160]

[0161] In the formula, It is a binary variable. This indicates that CAV team q has been assigned to the phase. Passage; Φq represents the set of passable phases for CAV vehicle q; Indicates phase The green light start time; a large M constant, used to linearize logical constraints; This indicates the number of vehicles included in the CAV fleet q; Indicates the safe headway of the CAV team;

[0162] when When = 1, the term of the larger M term is zero, and this set of inequalities forces the arrival time of the lead car in the convoy q. The passage of the last vehicle in the convoy must be after the green light begins, and the passage time of the last vehicle in the convoy (estimated by the time of the first vehicle, convoy size, and headway) must be before the green light ends. When M = 0, the large M term relaxes the inequality and does not produce any constraint.

[0163] IV. Mixed traffic management;

[0164] To maintain the overall stability of the intersection, the forced optimizer, while seeking the optimal solution for CAV, must allocate sufficient green light duration to serve existing or soon-to-be-formed HDV queues, and should adhere to the following constraints:

[0165]

[0166]

[0167] In the formula, Indicates phase The queue clearing time for the HDV / CV traffic flow prediction served at time k is a dynamically changing parameter that depends on real-time traffic demand. Represents the set of all signal phases at the intersection; Indicates phase HDV / CV traffic flow prediction queue length for services; Indicates phase The saturation flow rate of the inlet channel;

[0168] S3.2 Dual-Scale MPC Objective Function:

[0169] The vehicle-road control layer uses individual intersections as basic control units, responsible for tactical-level decision-making and planning. Its core objective is to coordinate the optimization of signal timing and controlled vehicle (CAV) arrival times for mixed traffic flows entering its control area within a finite prediction time domain. This optimization aims to minimize the total delay at local intersections while adhering to macro-coordination parameters (as soft constraints) issued from the cloud, and indirectly creates conditions for downstream energy consumption optimization.

[0170] The mathematical expression for the optimization objective is:

[0171]

[0172]

[0173] In the formula, This represents the objective function of the vehicle-road control layer. This represents the set of all entrance lanes; Indicates import channel The CAV team gathered inside; This represents the total number of vehicles in fleet q; To delay the weighting; Indicates the vehicle delay time; This refers to the actual arrival time. For the arrival time of the free flow; To coordinate the penalty coefficient; This represents the actual phase difference at intersection k at the current time. This is the reference phase difference issued at the vehicle-road-cloud level.

[0174] Its constraints are:

[0175]

[0176] When the local optimization result of ICL causes a large deviation from the phase difference set by MCL, try to maintain coordination with the macro green wave band as much as possible. Within the current control cycle k, the actual green light start time (s) of the coordinated phase at the local intersection; Within the current control cycle k, the target phase difference (Offset) at intersection i is issued by the macro coordination layer (MCL). The actual arrival time of the vehicle represents the target time when the CAV reaches the stop line after optimization, as well as the predicted arrival time of the vehicle under the current signal scheme, as predicted by the CV / RV through car-following models such as IDM and queue dissipation models. For the arrival time of the free flow.

[0177] S4. Vehicle-to-vehicle hierarchical modeling: Dynamically divide road segment functional areas and achieve coordinated control of mixed traffic flow through ML-SLSG lane-changing strategy and NMPC trajectory optimization.

[0178] S4.1 Dynamic Functional Area Division:

[0179] Based on the real-time lane-changing demand of traffic flow and the queuing prediction information fed back by downstream ICL, the road segment is dynamically divided into three mutually exclusive functional areas: lane-changing area, speed adjustment area, and queuing area.

[0180] lane change area Located upstream of the road section, it is used to handle lateral lane changing needs;

[0181]

[0182] Speed ​​adjustment zone Located in the middle of the road section, it is used for fine-tuning of vehicle longitudinal speed and trajectory optimization;

[0183]

[0184] Follow the entire team area Located downstream of the road section, it is used to stabilize the convoy, suppress disturbances, and prepare for entering the intersection;

[0185]

[0186] In the formula, Indicates the total length of the road segment This indicates the moment when vehicle i completes the lane change in lane group b and road segment l; Indicates vehicle correspondence Location; This indicates the number of vehicles queuing as predicted downstream. To ensure safe distance between vehicles; This refers to the length of the vehicle body. This refers to the speed at which traffic passes through the intersection. Indicates the maximum comfortable deceleration;

[0187] S4.2 ML-SLSG lane-changing strategy;

[0188] Within the lane change zone, the ML-SLSG lane change strategy is implemented to efficiently and safely resolve lane change conflicts in mixed traffic flows.

[0189] I. Vehicles entering the lane-changing area are divided into queue A (which requires lane changing) and queue B (which does not require lane changing) based on the relationship between the vehicle's current lane and the target lane.

[0190] Lane changing queue:

[0191] Non-lane-changing queue:

[0192] In the formula, i is the vehicle number; This represents the current lane of the i-th vehicle; Indicates the target lane for the i-th vehicle; This represents a queue of all vehicles whose current lane differs from the target lane and require lane changing. This represents a queue of all vehicles whose current lane is the same as the target lane and do not require lane changing.

[0193] II. The goal of the ML-SLSG lane-changing strategy is to allocate differentiated induced speeds to lane-changing queues and non-lane-changing queues so that vehicles in queue A can form a predetermined misalignment distance with vehicles in queue B in the longitudinal position, thereby safely merging into the target lane.

[0194] For the lead vehicle (CAV) in the team, its target speed in the lane change zone is as follows:

[0195]

[0196] In the formula, and This is a lane-changing queue. Non-lane-changing queues The set induction speed. To achieve efficient misalignment, it is usually set to... These two speed values ​​can be determined by the macro-coordination layer (MCL) based on global optimization results (such as the optimal green wave speed). Issued, or set according to the following principles:

[0197]

[0198]

[0199] in, and It is the gain / loss of speed, used to create a speed difference.

[0200] Safety distance constraints:

[0201]

[0202]

[0203] In the formula, Indicates the longitudinal position of the vehicle ahead; Indicates the current longitudinal position of the vehicle; Indicates the longitudinal position of the vehicle behind; Indicates the length of the vehicle ahead; Indicates the length of the vehicle behind; It is a speed-related dynamic safety distance, given by the safe time-to-distance (THW) model; The calculation expression is:

[0204]

[0205] In the formula, Indicates the vehicle's speed; Indicates the safe time interval; This indicates the minimum safe distance.

[0206] We are focusing on the two most critical vehicles: the vehicle at the very back of queue A (the last vehicle in queue A). ) and the last vehicle in queue B (the last car of queue B) ), and the vehicle at the very front of queue A (the lead vehicle of queue A) ).

[0207] Safe lane changing time:

[0208]

[0209] In the formula, This indicates the initial longitudinal position of the lead car (B) in the queue; This indicates the longitudinal position of the last car in queue A at the initial moment;

[0210] The length of the lane-changing zone must ensure that all vehicles needing to change lanes (queue A) can complete the lane change within this area. Therefore, its length is determined by the position of the first vehicle in queue A at the latest point in time when it completes its lane change. In our model, the last vehicle to complete its lane change is the last vehicle in queue A, and its completion time is... Lane change zone length:

[0211]

[0212] In the formula, This indicates the longitudinal position of the lead car A in the queue at the initial moment; Indicates the lane change operation time; This indicates an additional safety margin distance.

[0213] S4.3 Vehicle-Road Hierarchical Trajectory Optimization;

[0214] Once the vehicle enters the speed adjustment zone and the platooning zone, the vehicle's main task is to perform refined longitudinal trajectory coordination, which is achieved through a nonlinear model predictive control framework (NMPC) running on the CAV onboard platform.

[0215] Objective function:

[0216]

[0217] In the formula, The decision variable is the acceleration sequence of CAV in the prediction time domain; Indicates the length of the prediction time domain; This represents the fuel consumption weighting coefficient; Indicates the future number Fuel consumption rate at any given time; This represents the comfort weighting coefficient; Indicates the future number Acceleration at any moment; This represents the arrival time weighting coefficient; Indicates the predicted time of arrival at the target point; Indicates the expected arrival time;

[0218] This represents a multi-objective optimization function that minimizes the acceleration sequence as the decision variable. This objective function integrates three optimization objectives—fuel consumption, comfort, and arrival time deviation—into a single index through weighted summation. By adjusting the weighting coefficients to balance the priorities of the different objectives, the final solution yields the acceleration sequence that minimizes the overall objective. This enables multi-objective collaborative control of intelligent connected vehicles in longitudinal trajectory optimization.

[0219] for The instantaneous fuel consumption rate at any given time is calculated using a standard energy consumption model (such as VT-CPFM); The square of acceleration is used to penalize abrupt acceleration and deceleration, serving as a proxy for comfort. The target arrival time (s) of the CAV issued by the Intermediate Crossing Layer (ICL) is an input parameter; Based on the current acceleration sequence The predicted time to reach the stop line. It is a function of the decision variables and acts as a soft-terminal constraint.

[0220] For each time step in the prediction time domain The following constraints must be met:

[0221] Vehicle dynamics model:

[0222]

[0223]

[0224] Physical boundary constraints:

[0225]

[0226]

[0227] Safety following constraint:

[0228] CAVs must maintain a safe distance from the vehicle in front (which could be another CAV or an HDV).

[0229]

[0230] In the formula, For the car ahead at future moments The predicted location, The length of the vehicle in front. For the minimum safe time interval, This is the minimum safe distance when stationary.

[0231] To obtain the predicted position of the preceding vehicle within the aforementioned safety constraints We need an HDV behavior prediction model only if the preceding vehicle is an HDV. Based on the concept of "1+N" logical subgroups, we use an Intelligent Driver Model (IDM) to perform forward simulation prediction of HDV behavior.

[0232]

[0233]

[0234] In the formula, This represents the acceleration of the HDV at time t, calculated by the IDM model, reflecting the car-following behavior characteristics of the HDV. Indicates the maximum acceleration of HDV; This represents the HDV's speed at time t, its current actual speed; This indicates the desired speed of the HDV, the target speed of the vehicle in a free-moving state; is the acceleration exponent (dimensionless), usually taken as 2, used to describe the acceleration decay characteristics when the velocity is close to the desired velocity; The expected safety clearance for HDV is a function of the current speed and the relative speed. The minimum distance is T; the safe time interval is b; and the deceleration is b. This indicates the actual distance between the HDV and the vehicle in front (unit: meters), representing the current real distance between the two vehicles.

[0235] HDV is recursively calculated using kinematic equations throughout the prediction time domain. Trajectory within .

[0236] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the present invention, and all such modifications or substitutions should be covered within the protection scope of the present invention.

Claims

1. A CAV (Cyber-Physical Access Control) method for cooperative passage in a mixed traffic environment at multiple intersections, characterized in that, Includes the following steps: S1. Construct a three-layer cooperative control architecture based on IVCPS to decouple the multi-intersection cooperative passage problem into vehicle-road-cloud level, vehicle-road level and vehicle-to-vehicle level control problems; S2. Vehicle-Road-Cloud Hierarchical Modeling: Based on road network traffic status data, construct a green wave coordinated control MPC model to optimize the trunk line public signal cycle, phase difference, and green wave guidance speed; S3. Vehicle-Road Hierarchical Modeling: Based on vehicle group information obtained from V2I communication, a signal timing and vehicle speed coordination MPC model is constructed, and the green light ratio and CAV target arrival time are output. S4. Vehicle-to-vehicle hierarchical modeling: Dynamically divide road segment functional areas and achieve coordinated control of mixed traffic flow through ML-SLSG lane-changing strategy and NMPC trajectory optimization.

2. The CAV (Cyber-Physical Access Control) information-physical cooperative passage method in a multi-intersection mixed traffic environment according to claim 1, characterized in that, Step S2 includes the following sub-steps: S2.1 Defines the green wave bandwidth reduction rate P B With the reduction rate P of traffic speed V As an evaluation indicator, the formula for calculating the green wave bandwidth reduction rate is: In the formula, B u B is the uplink green wave bandwidth; d C represents the downlink green wave bandwidth; C represents the common signal period. This represents the theoretical maximum total green wave bandwidth; This represents the actual total green wave bandwidth; λ is the green signal ratio in each direction. This indicates the green light ratio at the first intersection or road segment in the uphill direction; This indicates the green light ratio at the first intersection or road segment in the downhill direction; This represents the ideal green wave bandwidth in the upward direction; The ideal green wave bandwidth in the downlink direction; the speed reduction rate P. V It is the percentage of the maximum difference between the green wave design speeds and the total speed reduction rate of the area. The calculation formula is: Section Traffic speed reduction rate The calculation formula is: In the formula, This represents the set of all road segments within the area to be evaluated; Indicates road segment The maximum design speed; Indicates road segment The minimum design speed; Indicates road segment The actual traffic speed; S2.2 Constructing the vehicle-road-cloud hierarchical MPC objective function and constraints; The objective function for the vehicle-road-cloud layer is obtained by weighted summation of the green wave bandwidth reduction rate and the traffic speed reduction rate: Its constraints are: In the formula, K represents the objective function of vehicle-road-cloud hierarchical MPC; B K represents the green wave bandwidth weighting coefficient, with a value range of (0,1), used to balance the optimization priority of green wave bandwidth and traffic speed; when K B When K → 0, prioritize optimizing traffic speed; when K B When →1, priority is given to optimizing green wave bandwidth; The predicted time domain length indicates that the time range for optimal control includes the period from the current time. arrive of total At that moment; This represents the time index within the prediction time domain; Indicates the time of prediction The green wave bandwidth reduction rate; Indicates the time of prediction Traffic speed reduction rate; This indicates the current optimization time, and its value range is... ; Indicates road segment The signal phase difference, that is, the difference between the downstream intersection j and the upstream intersection i at time... The time difference between the start of the green light; Indicates road segment Length; Indicates road segment At any moment The target traffic speed; Indicates road segment At any moment The period number correction coefficient, which is an integer, is used to adjust the signal period matching. Indicates the duration of the current signal cycle; , This represents the minimum and maximum duration of the signal period; Indicates road segment The phase difference compensation amount is used to correct system deviations in signal timing; This represents the cycle number correction coefficient for the ring network. It is an integer and is used for signal coordination optimization of the ring network.

3. The CAV (Cyber-Physical Access Control) information-physical cooperative passage method in a multi-intersection mixed traffic environment according to claim 2, characterized in that, Step S3 includes the following sub-steps: S3.1 Core constraints; I. Feasibility constraints regarding vehicle arrival time: In the formula, The core output of the model represents the target time when the lead car of the CAV team q reaches the stop line; Let q represent the theoretical minimum feasible time when the lead car of CAV team q reaches the stop line; This represents the theoretical maximum feasible time when the lead car of CAV team q reaches the stop line; This indicates the initial position of CAV team q at the entrance of the controlled area; This represents the initial velocity of the CAV vehicle q at the entrance to the controlled area; Indicates the maximum comfortable acceleration; Indicates the maximum comfortable deceleration; This represents the set of all CAV (Continuous Access Vehicle) vehicles entering the intersection control area within the prediction time domain; II. Signal Logic and Coordination Constraints; Signal Logic: For any pair of conflicting phases Linearization needs to be achieved by introducing auxiliary binary variables to ensure that their green light periods do not overlap; Represents the set of conflicting phase pairs; where, Indicates signal phase The duration of the green light; Yϕ represents the signal phase. The duration of the yellow light; ARϕ represents the signal phase. The duration of the all-red sequence; The set of all phases contained in phase group r; Represents the set of all signal phase groups. This indicates that a periodic constraint is applied to all phase groups; The current signal cycle duration is equal to the sum of the "green + yellow + all red" times for each phase within the phase group; III. Signal-vehicle coupling; To prevent traffic flow conflicts, only one phase can be activated at any given time, therefore, there is a constraint between phases: Phase uniqueness: Time window constraints: In the formula, It is a binary variable. This indicates that CAV team q has been assigned to the phase. Passage; Φq represents the set of passable phases for CAV vehicle q; Indicates phase The green light start time; a large M constant, used to linearize logical constraints; This indicates the number of vehicles included in the CAV fleet q; Indicates the safe headway of the CAV team; IV. Mixed Traffic Management: In the formula, Indicates phase The queue clearing time for the HDV / CV traffic flow prediction served at time k is a dynamically changing parameter that depends on real-time traffic demand. Represents the set of all signal phases at the intersection; Indicates phase HDV / CV traffic flow prediction queue length for services; Indicates phase The saturation flow rate of the inlet channel; S3.2 Dual-Scale MPC Objective Function: The vehicle-road control layer uses a single intersection as the basic control unit, responsible for tactical-level decision-making and planning. The core objective of the vehicle-road control layer is to perform coordinated optimization of signal timing and the arrival time of controlled vehicles for mixed traffic flows entering its control area within a finite prediction time domain. The mathematical expression for the optimization objective is: In the formula, This represents the objective function of the vehicle-road control layer. This represents the set of all entrance channels; Indicates import channel The CAV team gathered inside; This represents the total number of vehicles in the fleet q. To delay the weighting; Indicates the vehicle delay time; This refers to the actual arrival time. For the arrival time of the free flow; To coordinate the penalty coefficient; This represents the actual phase difference at intersection k at the current time. This is the reference phase difference issued at the vehicle-road-cloud level.

4. The CAV (Cyber-Physical Access Control) information-physical cooperative passage method in a multi-intersection mixed traffic environment according to claim 3, characterized in that, Step S4 includes the following sub-steps: S4.1 Dynamic functional zone division: Based on the real-time lane-changing demand of traffic flow and the queuing prediction information fed back by downstream ICLs, the road segment is dynamically divided into three mutually exclusive functional zones: lane-changing zone, speed adjustment zone, and following-the-car queueing zone; lane-changing zone Located upstream of the road section, it is used to handle lateral lane changing needs; Speed ​​adjustment zone Located in the middle of the road section, it is used for fine-tuning of vehicle longitudinal speed and trajectory optimization; Follow the whole team area Located downstream of the road section, it is used to stabilize the convoy, suppress disturbances, and prepare for entering the intersection; In the formula, Indicates the total length of the road segment This indicates the moment when vehicle i completes the lane change in lane group b and road segment l; Indicates vehicle correspondence Location This indicates the number of vehicles queuing as predicted downstream. To ensure safe distance between vehicles; This refers to the length of the vehicle body. This refers to the speed at which traffic passes through the intersection. Indicates the maximum comfortable deceleration; S4.2 ML-SLSG lane change strategy; In the lane change zone, the ML-SLSG lane change strategy is implemented to efficiently and safely resolve lane change conflicts in mixed traffic flows; I. Vehicles entering the lane-changing zone are divided into two queues based on the relationship between their current lane and the target lane: queue A (requiring a lane change) and queue B (not requiring a lane change); Lane-changing queues: Non-lane-changing queue: In the formula, i is the vehicle number; This represents the current lane of the i-th vehicle; Indicates the target lane for the i-th vehicle; This represents a queue of all vehicles whose current lane differs from the target lane and require lane changing. I. Represents all vehicles whose current lane is the same as the target lane and do not require lane changing; II. The goal of the ML-SLSG lane-changing strategy is to assign differentiated induced speeds to lane-changing and non-lane-changing queues, so that vehicles in queue A can form a predetermined longitudinal offset distance with vehicles in queue B, thereby safely merging into the target lane; Induced speed: In the formula, and This is a lane-changing queue. Non-lane-changing queues Set induction speed; safety distance constraints: In the formula, Indicates the longitudinal position of the vehicle ahead; Indicates the current longitudinal position of the vehicle; Indicates the longitudinal position of the vehicle behind; Indicates the length of the vehicle ahead; Indicates the length of the vehicle behind; It is a speed-related dynamic safety distance; safe lane-changing timing: In the formula, This indicates the initial longitudinal position of the lead car (B) in the queue; Indicates the initial longitudinal position of the last car in queue A; lane change zone length: In the formula, This indicates the longitudinal position of the lead car A in the queue at the initial moment; Indicates the lane change operation time; This represents the additional safety margin distance; S4.3 Vehicle-Road Hierarchical Trajectory Optimization; After the vehicle enters the speed adjustment zone and the platooning zone, the main task of the vehicle is to perform refined longitudinal trajectory coordination. This process is achieved through a nonlinear model predictive control framework (NMPC) running on the CAV onboard platform; Objective function: In the formula The decision variable is the acceleration sequence of CAV in the prediction time domain; Indicates the length of the prediction time domain; This represents the fuel consumption weighting coefficient; Indicates the future number Fuel consumption rate at any given time; This represents the comfort weighting coefficient; Indicates the future number Acceleration at any moment; This represents the arrival time weighting coefficient; Indicates the predicted time of arrival at the target point; Indicates the expected arrival time; This represents a multi-objective optimization function that minimizes the acceleration sequence as the decision variable. This objective function integrates three optimization objectives—fuel consumption, comfort, and arrival time deviation—into a single index through weighted summation. By adjusting the weighting coefficients to balance the priorities of the different objectives, the final solution yields the acceleration sequence that minimizes the overall objective. This enables multi-objective collaborative control of intelligent connected vehicles in longitudinal trajectory optimization.

5. The CAV (Cyber-Physical Access Control) information-physical cooperative passage method in a multi-intersection mixed traffic environment according to claim 4, characterized in that, In step S4.2, The calculation expression is: In the formula, Indicates the vehicle's speed; Indicates the safe time interval; This indicates the minimum safe distance.

Images