Vehicle platoon control and signal optimization method and system in connected traffic environment

Abstract

The present invention provides a vehicle formation control and signal optimization method and system in a networked traffic environment, belonging to the field of intelligent traffic vehicle formation technology. The method obtains intersection geometry, road width, entrance road length, and roadside traffic equipment location parameters; predicts the traffic operation status of the intersection; makes formation decisions for vehicles entering the intersection; and optimizes the real-time signal timing of the intersection under multiple constraints, taking the environment, safety, and comfort as constraints, and using the lowest energy consumption and highest traffic efficiency as the optimization objective function to obtain the economic speed, green light display time for each phase, and phase sequence through the rolling time domain dynamic programming method. From both microscopic and macroscopic perspectives, the present invention greatly enhances the communication efficiency of intersection network nodes based on the vehicle formation control strategy and signal timing coupling optimization mechanism, while ensuring the safety and reliability of traffic operation, and has practical engineering application value.

Description

Technical Field

[0001] The present invention relates to the field of intelligent traffic vehicle platooning technology, and in particular to a vehicle platooning control and signal optimization method and system in a networked traffic environment. Background Art

[0002] From the perspective of traffic wave theory, queues and congestion at urban intersections occur because the dissipation wave velocity of vehicles leaving the intersection is slower than the assembly wave velocity of vehicles entering the intersection. The assembly wave is relatively constant for the same intersection. Generally speaking, since people's daily travel needs are relatively fixed, the number of vehicles entering the intersection at the same time of day is roughly the same, and the traffic flow density entering the intersection also varies roughly the same. Therefore, in addition to implementing traffic control measures to limit vehicle travel and reduce the assembly wave velocity, the problem of queues and congestion at urban intersections can also be solved by increasing the dissipation wave velocity at the intersection, that is, increasing the efficiency of vehicles leaving the intersection under the green light duration.

[0003] The main reason for the low dissipation wave speed at urban intersections is the varying reaction times of individual drivers and the varying acceleration capabilities of individual vehicles, which reduces the effective green light duration. Currently, the development of next-generation communication technologies, intelligent connected vehicles, and vehicle-infrastructure collaboration (VIC) technologies presents an opportunity to address the problem of queuing and congestion at urban intersections. Next-generation communication technologies, including 5G, DSRC, and LTE-V, offer higher-speed, more stable data transmission and lower latency, enabling faster information exchange between vehicles and between vehicles and infrastructure. The development of intelligent connected vehicles enables autonomous perception, decision-making, and control, improving road safety while also optimizing energy consumption and emissions based on road and vehicle conditions. VIC technology, through information exchange between vehicles and roadside equipment, enables vehicles to obtain real-time information on current traffic conditions, traffic light information, road conditions, and more. This assists drivers and vehicles in making proactive decisions regarding acceleration, deceleration, or lane changes, thereby improving traffic efficiency. As a key technology for multi-vehicle cooperative driving in connected traffic environments, platooning control treats each vehicle as an independent intelligent entity. They perceive and exchange information with each other through onboard sensors and communication equipment, and achieve coordinated driving through collaborative control algorithms. At urban intersections, platooning control helps improve the organization and orderliness of vehicle groups. By grouping incoming vehicles, the distance between them is minimized while ensuring safety, and they simultaneously start at the same speed to move through the intersection quickly, efficiently, and in an orderly manner.

[0004] However, existing research on the application of vehicle platooning control to urban intersections faces several challenges: most studies focus on optimizing vehicle platooning trajectories and fail to consider the relationship between vehicle platooning and intersection signal timing. The rationality of urban intersection signal timing largely determines the upper limit of vehicle platooning optimization capabilities. Even though some studies use roadside sensing devices to capture intersection traffic conditions and couple the number of vehicles on the current entrance lane with signal timing for optimization, despite the strong spatiotemporal dependence of intersection traffic conditions, the random nature of intersection traffic conditions makes this approach inherently hysteretic and unstable for discrete control systems such as signal control. Furthermore, existing research focuses on static platoon formation at intersections, lacking feedforward control to account for the random variations in urban intersection traffic conditions. This results in suboptimal platooning control. Summary of the Invention

[0005] The purpose of the present invention is to provide a method and system for vehicle formation control and signal optimization in a connected traffic environment, which realizes the autonomous operation of vehicle formations at urban intersections through a closed-loop process of prediction, formation, speed control, timing, departure, correction, and re-prediction, so as to solve at least one technical problem existing in the above-mentioned background technology.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] In a first aspect, the present invention provides a method for vehicle platoon control and signal optimization in a connected traffic environment, comprising:

[0008] Determine the static attribute parameters of urban intersections and divide them into functional areas to obtain intersection geometry, road width, entrance road length, and roadside traffic equipment location parameters;

[0009] Predict the traffic situation at the intersection and use the Conv-LSTM network based on the attention mechanism to predict the vehicle arrival situation and vehicle type ratio in each phase of the next signal cycle;

[0010] The system makes platooning decisions for vehicles entering an intersection, using a pilot vehicle-two-way preceding vehicle-follower vehicle communication topology to ensure platooning safety. Vehicle-road interaction is used to determine the attributes of incoming vehicles, and the optimal parking spacing and minimum spacing for platooning are determined based on the platooning strategy.

[0011] Real-time signal timing optimization is performed on intersections under multiple constraints, taking the environment, safety, and comfort as constraints. The economical vehicle speed, green light display time of each phase, and phase sequence are obtained by using the rolling horizon dynamic programming method with minimum energy consumption and maximum traffic efficiency as the optimization objective function.

[0012] Furthermore, the functional areas of the urban intersection are divided into three areas: the entrance of the urban intersection is divided into three areas, starting from the parking line, the parking area, the formation area and the free driving area, with lengths L and L respectively. i,p,0 , L i,p,1 , L i,p,2 , where i represents the number of the road section, which is consistent with the number of its downstream intersection, p represents the direction of the four entrance roads of the intersection, e represents the east entrance, s represents the south entrance, w represents the west entrance, and n represents the north entrance.

[0013] Furthermore, the future traffic situation of the intersection is predicted, including:

[0014] The location of the roadside vehicle information collection device on the intersection entrance road section is denoted as m, where the collected information includes traffic flow and vehicle attributes, and is stored in the form of a tuple in f t m In , the traffic state sequence from time tn to time t is recorded. Assuming there are k observation points, the historical traffic flow moment data of the intersection is represented by a matrix;

[0015] Use a one-dimensional convolution kernel to obtain the local perception domain, perform a one-dimensional convolution operation on the traffic state data at each time step t to obtain spatial features;

[0016] Based on the attention mechanism, the importance of traffic flow sequences in the spatiotemporal traffic flow matrix is ​​automatically mined in different time periods, and the hidden state of the output is calculated at each time step t. The weighted sum of is used to predict the traffic volume and vehicle attribute ratio of each entrance lane.

[0017] Furthermore, platooning decisions are made for vehicles entering the intersection, including:

[0018] The communication topology of the platooning system adopts a leader vehicle and a two-way leading vehicle following system. The first vehicle entering the platooning area at each stage is designated as the leader vehicle, and the remaining vehicles are designated as following vehicles. During the platooning process, the leader vehicle and following vehicle, as well as the following vehicles and following vehicles, perform vehicle pairing and authentication to establish a communication connection. The leader vehicle transmits real-time speed, direction, and braking information to the following vehicles. At the same time, the following vehicles receive and process the data sent by the leader vehicle in real time, responding to the actions of the leader vehicle, maintaining a safe following distance, and adjusting speed and direction in a timely manner.

[0019] Furthermore, when a vehicle enters a platooning area, it sends vehicle information to the roadside edge server. The roadside server then makes judgments and corrections based on the vehicle information, and sends the deceleration, vehicle spacing, starting acceleration, and economic speed to the vehicle via the vehicle-road communication device.

[0020] During platooning, only the longitudinal acceleration and deceleration of the vehicle are considered; the stopping distance needs to meet the optimal following distance after the vehicle starts and accelerates. The acceleration-displacement equation can be obtained to determine the optimal stopping distance.

[0021] Furthermore, real-time signal timing optimization is performed on intersections under multiple constraints, including:

[0022] Based on the predicted number of vehicles arriving at each entrance, as well as their types, attributes, and proportions, and with road conditions, weather conditions, comfort, and safety as constraints, and minimum fuel consumption and maximum traffic efficiency as the objective functions, a rolling-horizon dynamic programming method is used to determine the average energy-saving speed for each entrance, thereby minimizing overall fuel consumption for vehicles at intersection network nodes.

[0023] Adding a negative sign before the traffic efficiency objective function is equivalent to solving its minimum value, so the average energy-saving speed of each entrance lane is established to solve the optimization equation system.

[0024] Furthermore, vehicles entering the formation area send vehicle information to the roadside vehicle information collection device to update and iterate the road section database; the roadside edge server calculates the length of the vehicle queue in the next stage based on the predicted traffic volume and the proportion of historical vehicle attributes; with the goal of passing all queued vehicles at an energy-saving speed, the green light display time for each phase is obtained; while the vehicle formation leaves the intersection, the roadside edge server makes corrections based on the traffic flow and vehicle type ratio of the previous stage to prepare for the formation control in the next stage and complete a complete vehicle formation control.

[0025] In a second aspect, the present invention provides a vehicle platoon control and signal optimization system in a connected traffic environment, comprising:

[0026] The acquisition module is used to determine the static attribute parameters of urban intersections and divide them into functional areas, and obtain the intersection geometry, road width, entrance road length, and roadside traffic equipment location parameters;

[0027] The prediction module is used to predict the traffic situation at the intersection. It uses a Conv-LSTM network based on the attention mechanism to predict the vehicle arrival situation and vehicle type ratio in each phase of the next signal cycle.

[0028] The decision-making module is used to make platooning decisions for vehicles entering the intersection. It uses a pilot vehicle-two-way leading vehicle-follower vehicle communication topology to ensure platooning safety. It obtains the attributes of incoming vehicles through vehicle-road interaction and determines the optimal parking spacing and minimum spacing for platooning based on the platooning strategy.

[0029] The timing optimization module is used to optimize the real-time signal timing of intersections under multiple constraints. It takes the environment, safety, and comfort as constraints, and uses the lowest energy consumption and highest traffic efficiency as the optimization objective function to obtain the economic speed, green light display time of each phase, and phase sequence through the rolling horizon dynamic programming method.

[0030] In a third aspect, the present invention provides a non-transitory computer-readable storage medium, which is used to store computer instructions. When the computer instructions are executed by a processor, the vehicle formation control and signal optimization method in a connected traffic environment as described in the first aspect is implemented.

[0031] In a fourth aspect, the present invention provides a computer device comprising a memory and a processor, wherein the processor and the memory communicate with each other, the memory stores program instructions that can be executed by the processor, and the processor calls the program instructions to execute the vehicle formation control and signal optimization method in a connected traffic environment as described in the first aspect.

[0032] In a fifth aspect, the present invention provides an electronic device comprising: a processor, a memory, and a computer program; wherein the processor is connected to the memory, and the computer program is stored in the memory. When the electronic device is running, the processor executes the computer program stored in the memory, so that the electronic device executes instructions for implementing the vehicle formation control and signal optimization method in a connected traffic environment as described in the first aspect.

[0033] The beneficial effects of the present invention are as follows: the entrance road section of the urban intersection is divided into regions according to the position of the roadside vehicle information collection equipment, and the roadside vehicle information sensing equipment is moved from the vicinity of the intersection to the road section, thereby increasing the effective length of the vehicle formation; the future traffic situation of the urban intersection is predicted by the Conv-LSTM module based on the attention mechanism, and the traffic volume, vehicle type ratio and attributes of each entrance road are obtained, which improves the prediction accuracy and robustness, and at the same time adds the prediction of vehicle type ratio and attributes. The average energy-saving speed of each entrance lane is solved through the rolling time domain dynamic programming method. At the same time, the vehicle formation length of each entrance lane is calculated by combining the predicted traffic volume and vehicle type ratio and attributes, and then the green light display time of each phase is obtained. Through vehicle-road communication technology, dynamic formation is carried out according to the interacting vehicle attributes during vehicle driving, and the economic speed and real-time signal timing plan are obtained based on multiple constraints, which greatly increases the traffic efficiency of intersection road network nodes; a pilot vehicle-two-way leading vehicle and following vehicle communication topology is adopted to achieve high-efficiency communication of intelligent connected vehicles, and the parking queue spacing is defined according to different vehicle attributes. By determining the parking spacing of each incoming vehicle, it reaches the minimum platoon spacing at the same time when starting to accelerate to the economic speed, which greatly reduces the reduction in formation stability caused by different vehicle acceleration capabilities.

[0034] Additional advantages of the present invention will be more clearly given in the following description or learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0036] Figure 1 This is a structural diagram of the Conv-LSTM network module described in an embodiment of the present invention.

[0037] Figure 2 This is a diagram of the Conv-LSTM network structure based on the attention mechanism described in an embodiment of the present invention.

[0038] Figure 3 This is a topology diagram of the communication between the pilot vehicle and the leading vehicle in a two-way platoon according to an embodiment of the present invention.

[0039] Figure 4 This is a scene diagram of the vehicle platoon operation environment in a connected traffic environment according to an embodiment of the present invention.

[0040] Figure 5 This is a flow chart of the vehicle formation control and signal optimization method in a connected traffic environment according to an embodiment of the present invention.

[0041] Figure 6 Schematic diagram comparing the average delay of the method of the present invention and the traditional fixed timing intersection under different traffic flow inputs according to the embodiment of the present invention.

[0042] Figure 7 This is a comparison chart of the average delay optimization effects of the present invention and the traditional fixed timing intersection under different traffic flow inputs described in the embodiments of the present invention.

[0043] Figure 8 This is a comparison chart of the average speeds at intersections using the present invention and traditional fixed-timing intersections under different traffic flow inputs according to an embodiment of the present invention.

[0044] Figure 9 This is a diagram showing the average speed optimization effect of the present invention and the traditional fixed timing intersection under different traffic flow inputs described in an embodiment of the present invention.

[0045] Figure 10 This is a comparison chart of average queue lengths at intersections with the traditional fixed timing under different traffic flow inputs described in an embodiment of the present invention.

[0046] Figure 11This is a diagram showing the optimization effect of the average queue length at an intersection using the present invention and the traditional fixed timing method under different traffic flow inputs described in an embodiment of the present invention. DETAILED DESCRIPTION

[0047] The embodiments of the present invention are described in detail below. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention and are not to be construed as limiting the present invention.

[0048] Those skilled in the art will understand that unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by those skilled in the art to which this invention belongs.

[0049] It should also be understood that terms, such as those defined in commonly used dictionaries, should be understood to have a meaning consistent with their meaning in the context of the prior art and will not be interpreted in an idealized or overly formal sense unless as defined herein.

[0050] Those skilled in the art will appreciate that, unless otherwise stated, the singular forms "a," "an," "said," and "the" used herein may also include plural forms. It should be further understood that the term "comprising" used in the specification of the present invention refers to the presence of the stated features, integers, steps, operations, elements, and / or components, but does not preclude the presence or addition of one or more other features, integers, steps, operations, elements, and / or groups thereof.

[0051] In the description of this specification, reference to the terms "one embodiment," "some embodiments," "examples," "specific examples," or "some examples" means that the specific features, structures, materials, or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. Moreover, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. In addition, those skilled in the art may combine and integrate different embodiments or examples described in this specification, as well as features of different embodiments or examples, unless otherwise inconsistent.

[0052] To facilitate understanding of the present invention, the present invention is further explained below with reference to specific embodiments in conjunction with the accompanying drawings. However, the specific embodiments do not constitute a limitation on the embodiments of the present invention.

[0053] Those skilled in the art should understand that the drawings are merely schematic diagrams of embodiments, and the components in the drawings are not necessarily necessary for implementing the present invention.

[0054] Example 1

[0055] In this embodiment 1, a vehicle formation control and signal optimization system in a networked traffic environment is first provided, including: an acquisition module for determining static attribute parameters of urban intersections and performing functional area division, and obtaining intersection geometry, road width, entrance road length, and roadside traffic equipment location parameters; a prediction module for predicting the traffic operation situation at the intersection, and obtaining the vehicle arrival status and vehicle type ratio of each phase in the next signal cycle through a Conv-LSTM network based on an attention mechanism; a decision module for making formation decisions for vehicles entering the intersection, adopting a pilot vehicle-two-way preceding vehicle-following vehicle communication topology to ensure formation driving safety, obtaining the attributes of the entering vehicles through vehicle-road interaction, and obtaining the optimal formation parking spacing and the minimum formation driving spacing according to the formation strategy; a timing optimization module for performing real-time signal timing optimization under multiple constraints at the intersection, considering the environment, safety, and comfort as constraints, and taking the lowest energy consumption and the highest traffic efficiency as the optimization objective function to obtain the economic vehicle speed, the green light display time of each phase, and the phase sequence through the rolling time domain dynamic programming method.

[0056] In this embodiment, the above-mentioned system is used to implement a vehicle formation control and signal optimization method in a connected traffic environment. By predicting the future traffic operation situation at the intersection, a real-time optimization mechanism for signal timing is constructed. At the same time, multi-objective constraints for vehicle formation control at the intersection are established by considering fuel economy, traffic efficiency, comfort and safety.

[0057] The technical solution for vehicle platoon control and signal optimization in a connected traffic environment is as follows:

[0058] Step 1: Divide the intersection area mechanism under the connected traffic environment.

[0059] The entrance road of the urban intersection is divided into three areas, starting from the parking line, which are the parking area, the formation area and the free driving area, with lengths L i,p,0 , L i,p,1 , L i,p,2 , where i represents the number of the road section, which is consistent with the number of its downstream intersection, p represents the direction of the four entrances to the intersection, e represents the east entrance, s represents the south entrance, w represents the west entrance, and n represents the north entrance. i,p,1 The length of is determined on each road segment and is calculated as:

[0060] Step 2: Construct a method to predict future traffic conditions at intersections

[0061] Combining the spatiotemporal and periodic characteristics of urban intersection traffic status, this embodiment proposes a Conv-LSTM urban intersection traffic status prediction method based on the attention mechanism, so as to obtain the arrival status of each imported vehicle in the next stage. The module design is as follows: Figure 1 As shown, it is divided into the following steps:

[0062] Step 201: The location of the roadside vehicle information collection device on the intersection entrance road section is denoted as m, where the collected information includes traffic flow and vehicle attributes, and is stored in the form of a tuple in f t m In the example, the traffic state sequence from time tn to time t is Therefore, assuming there are k observation points, the historical traffic flow data of the intersection can be expressed in a matrix as follows:

[0063]

[0064] in It represents the traffic operation status of the observation area of ​​the intersection entrance road section at time t.

[0065] Step 202: Use a one-dimensional convolution kernel to obtain a local perception domain and perform traffic state data at each time step t. Perform a one-dimensional convolution operation to obtain spatial features, which can be expressed as follows:

[0066]

[0067] Among them, Y t S represents the output matrix of the convolutional layer, σ represents the ReLU activation function, W S represents the weight of the convolution kernel, represents the input spatiotemporal traffic flow matrix, b s After extracting the spatial features of the traffic flow data, this embodiment extracts the temporal features through a two-layer LSTM network.

[0068] Step 203: The custom attention mechanism enables the Conv-LSTM module to automatically mine the importance of traffic flow sequences in the spatiotemporal traffic flow matrix in different time periods, such as Figure 2 As shown, the hidden state of the Conv-LSTM network output is calculated at each time step t The weighted sum of is as follows:

[0069]

[0070] Where n-1 represents the length of the spatiotemporal traffic flow data sequence, t-(i-1) represents the time difference between the current time step t and the i-th historical time step, and β k represents the attention value when the time step is t-(i-1), β k The calculation can be expressed as follows:

[0071]

[0072] Where S=(s1,s2,...,s n+1 ) T It represents the importance score of each part of the spatiotemporal traffic flow data sequence. The calculation formula is as follows:

[0073]

[0074] in, W hs and W ls are the parameters that need to be learned, Represents the output matrix of the convolutional layer.

[0075] Based on the above prediction, the traffic flow Q of each entrance is obtained i,p,k and vehicle attribute ratio A i,p,type , where i represents the number of the road section, which is consistent with the number of its downstream intersection, p represents the direction of the four entrance roads of the intersection, k (k = l, s, r) represents the channelization scheme of the entrance road, where l represents left turn, s represents straight ahead, r represents right turn, and type represents the vehicle type.

[0076] Step 3: Determine the intelligent connected vehicle platooning strategy.

[0077] The Leader-Bidirectional-Predecessor Following (LBPF) is used as the formation communication topology. Figure 3 As shown in the figure, the first vehicle entering the platooning area at each stage is designated as the lead vehicle, and the remaining vehicles are designated as following vehicles. During the platooning process, the lead vehicle and following vehicles, and the following vehicles themselves, undergo vehicle pairing and authentication to establish a communication connection. The lead vehicle then transmits real-time speed, direction, braking status, and other information to the following vehicles. The following vehicles simultaneously receive and process this data in real time, responding to the lead vehicle's actions, maintaining a safe following distance and adjusting speed and direction accordingly.

[0078] Based on the unique advantages of intelligent connected vehicles, when a vehicle enters the formation area, on the one hand, it sends information such as the vehicle's own attributes, acceleration and deceleration capabilities, and fuel consumption-speed-acceleration function to the roadside edge server. On the other hand, the roadside end makes judgments and corrections based on this information, and sends the deceleration, vehicle spacing, starting acceleration and economic speed to the vehicle through the vehicle-road communication equipment.

[0079] During platooning, the present invention simplifies the vehicle model into a two-degree-of-freedom model, only considering the longitudinal acceleration and deceleration of the vehicle. After reaching the optimal following distance s min , the acceleration-displacement equation can be expressed as follows:

[0080]

[0081] Among them, a i+1 and a i The optimal parking distance s can be determined by specifying the acceleration of the i+1th vehicle and the ith vehicle to ensure the lowest energy consumption, t is the starting time, and the following relationship is satisfied. i It is expressed by the following formula:

[0082]

[0083] Ideally, when vehicles are platooning, the following distance can approach zero, thereby improving intersection efficiency. However, in real-world traffic, the tolerance for the ideal following distance is too low, posing a safety hazard. Therefore, the minimum following distance is defined as the braking distance after stopping at maximum braking acceleration, as expressed in the following formula:

[0084]

[0085] Among them, a -max Indicates the maximum braking acceleration. Considering the vehicle's braking capacity and the driver's comfort, the maximum braking acceleration should not be greater than 8.5m / s 2 .

[0086] Step 4: Construct a real-time signal timing method with autonomous optimization under multiple constraints

[0087] According to the above method, the number of vehicles arriving at each entrance and the type, attributes and proportion of vehicles are predicted. At the same time, the average energy-saving speed of each entrance is solved by the rolling horizon dynamic programming method with road conditions, weather environment, comfort and safety as constraints and minimum fuel consumption and maximum traffic efficiency as the objective function. From a macro perspective, the overall fuel consumption of vehicles at intersection network nodes is minimized.

[0088] In order to maximize traffic efficiency and facilitate the solution of the objective function, in this embodiment, adding a negative sign before the traffic efficiency objective function is equivalent to solving its minimum value. Therefore, the optimization equations for solving the average energy-saving speed of each entrance lane are established as shown in the following formula:

[0089]

[0090] Where n represents the number of intersection entrances, m represents the number of lanes in each entrance, and l represents the length of the platooning area in each lane. ij (t) represents the position of the vehicle in the jth lane of the i-th entrance lane at time t. vij (t) and ε aij (t) represents the random influence of the meteorological environment on the vehicle speed and acceleration at time t on the jth lane of the i-th entrance lane, and is described by normal distribution. max Indicates the maximum speed of the vehicle, v min The minimum speed is a max represents the maximum acceleration, a min is the minimum acceleration. f(v,a) represents the vehicle's fuel consumption function, which indicates the vehicle's fuel consumption rate when the vehicle's speed is v and the acceleration is a. Similarly, g(v,a) represents the traffic efficiency function of the intersection network node, h(v,a) represents the vehicle's comfort function, and s(v,a) represents the vehicle's safety function. Indicates the energy-saving speed at the intersection network node when the overall fuel consumption of vehicles is the lowest.

[0091] At the same time, vehicles entering the formation area send information such as vehicle attributes, acceleration capabilities, fuel consumption-speed-acceleration function, etc. to the vehicle information collection equipment on the roadside to update and iterate the road section database.

[0092] The roadside edge server obtains the traffic volume Q based on the prediction i,p,k and the proportion of historical vehicle attributes, calculate the next stage vehicle queue length d i,p,pha , the calculation formula is as follows:

[0093] d i,p,pha =∑s i +∑l i α i

[0094] Among them, l i Represents the length α of different types of vehicles i Indicates the proportion of different vehicle types in this stage, s i is the parking interval between the i-th car and the i+1-th car.

[0095] Finally, the goal is to pass all queued vehicles at an energy-saving speed and obtain the green light display time G of each phase.i,p,pha Where pha represents a certain phase, and the phase order is determined according to the traffic flow ratio.

[0096] As the vehicle formation leaves the intersection, the roadside edge server makes corrections based on the traffic volume and vehicle type ratio in the previous stage, and also prepares for the next stage of formation control, thus completing a complete vehicle formation control.

[0097] Example 2

[0098] This invention takes intelligent connected vehicles as the target, gives full play to the advantages of vehicle-road cooperative technology, and realizes the coordinated formation control and signal timing optimization of vehicles through interconnection with intersection signal control systems. Figure 4 As shown, the system mainly includes roadside vehicle information collection equipment, intelligent traffic signal control equipment, vehicle-road communication equipment, and roadside edge servers. Among them, the vehicle information collection equipment is used to interact with intelligent connected vehicles to collect vehicle attribute information. The intelligent traffic signal control equipment is used to receive signal control instructions from the roadside edge server and autonomously optimize signal timing in real time. The vehicle-road communication equipment is used to issue vehicle formation control instructions. The roadside edge server is used to provide traffic status prediction, formation length division, economic speed, and signal timing optimization algorithm support. In addition, in the specific implementation method, the sensitivity analysis of the method proposed in the present invention was carried out using the vissim2020 simulation software.

[0099] The specific implementation process is as follows Figure 5 As shown, each step is described in detail below.

[0100] (1) Intersection area division under the connected traffic environment

[0101] The entrance road of the urban intersection is divided into three areas, starting from the parking line, which are the parking area, the formation area and the free driving area, with lengths L i,p,0 , L i,p,1 , L i,p,2 , where i represents the road section number, which is consistent with the number of its downstream intersection, p represents the direction of the four entrance roads of the intersection, e represents the east entrance, s represents the south entrance, w represents the west entrance, and n represents the north entrance.

[0102] The parking area is used to park vehicles that have been arranged according to the longitudinal formation strategy. When the phase green light is on, all vehicles in the area start to the same speed at the same time and quickly leave the intersection. The length of the parking area of ​​each entrance lane is L i,p,0 The length of is affected by the predicted arrival vehicles and vehicle platooning strategy at the intersection entrance, which ensures the queue length corresponding to the maximum number of vehicles that can pass through the intersection under the green light duration of each phase.

[0103] The formation area is used to dynamically group vehicles entering the entrance lane and determine whether the vehicle formation service request can be met. If the formation can be performed, the optimal headway between adjacent vehicles is determined according to the longitudinal formation strategy before entering the parking area, and the acceleration and deceleration are determined according to different vehicle types. If it is not satisfied, the corresponding decision information is sent to the vehicle and it is included in the next formation cycle. The length of the formation area of ​​each entrance lane is L i,p,1 The length of L is affected by the vehicle-road communication delay and the computing performance and deployment location of the roadside edge server. It is the product of the time it takes for an incoming vehicle to send a platoon service request to receive an edge control instruction and the average driving speed of the road section in a connected traffic environment. According to the above principle, L i,p,1 The length of is determined on each road segment and is calculated as:

[0104] The free travel area is used by the driver to complete the turn and lane change before entering the intersection formation area. When the vehicle leaves the upstream intersection, according to different OD requirements, the vehicle performs the turn and lane change operation to prepare for entering the formation area, thereby improving the roadside edge server's focus on solving the vehicle formation strategy. The length of the free travel area of ​​each entrance lane is L i,p,2 It is the distance from the vehicle leaving the upstream intersection to entering the platooning area.

[0105] (2) Establishing a method for predicting future traffic conditions at intersections

[0106] Traffic state prediction is to allow intersections to predict vehicle arrivals in advance, so as to change the signal timing plan according to the real-time changing traffic flow state. Since most traffic travelers have certain regularity in their travel methods and habits, they have similar or repeated traffic states every day or every week. The traffic state of urban intersections has spatiotemporal and periodic characteristics within a certain period of time. Therefore, a custom Conv-LSTM urban intersection traffic state prediction method based on the attention mechanism is used to obtain the arrival status of each imported vehicle in the next stage. The module design is as follows: Figure 1 As shown, it is divided into the following steps:

[0107] Step 1: The input of traffic status data is regulated and unified to better extract the spatiotemporal and periodic features in traffic status data. The location of the roadside vehicle information collection device on the intersection entrance road section is denoted as m, where the collected information includes traffic flow and vehicle attributes, and is stored in the form of two tuples in f. t m In the example, the traffic state sequence from time tn to time t is Therefore, assuming there are k observation points, the historical traffic flow data of the intersection can be expressed in a matrix as follows:

[0108]

[0109] in It represents the traffic operation status of the observation area of ​​the intersection entrance road section at time t.

[0110] Step 2: Extract the spatial features of the traffic operation state of the urban intersection, use the one-dimensional convolution kernel to obtain the local perception domain, and perform traffic state data at each time step t. Perform a one-dimensional convolution operation to obtain spatial features, which can be expressed as follows:

[0111]

[0112] Among them, Y t S represents the output matrix of the convolutional layer, σ represents the ReLU activation function, W S represents the weight of the convolution kernel, represents the input spatiotemporal traffic flow matrix, b s After extracting the spatial features of traffic flow data, the temporal features are extracted through a two-layer LSTM network.

[0113] Step 3: Use an attention mechanism to automatically mine the importance of traffic flow sequences in the spatiotemporal traffic flow matrix in different time periods, as shown in the following formula:

[0114]

[0115] Where n-1 represents the length of the spatiotemporal traffic flow data sequence, t-(i-1) represents the time difference between the current time step t and the i-th historical time step, and β k represents the attention value when the time step is t-(i-1), β k The calculation can be expressed as follows:

[0116]

[0117] Where S=(s1,s2,...,s n+1 ) T It represents the importance score of each part of the spatiotemporal traffic flow data sequence. The calculation formula is as follows:

[0118]

[0119] in, W hs and W ls are the parameters that need to be learned, Represents the output matrix of the convolutional layer.

[0120] Based on the above prediction, the traffic flow Q of each entrance is obtained i,p,k and vehicle attribute ratio A i,p,type , where i represents the number of the road section, which is consistent with the number of its downstream intersection, p represents the direction of the four entrance roads of the intersection, k (k = l, s, r) represents the channelization scheme of the entrance road, where l represents left turn, s represents straight ahead, r represents right turn, and type represents the vehicle type.

[0121] (3) Determine the intelligent connected vehicle platooning strategy

[0122] Determine the ICV platooning strategy through the following four steps:

[0123] Step 1: Use Leader-Bidirectional-Predecessor Following (LBPF) as the formation communication topology, such as Figure 3 As shown in the figure, the first vehicle entering the platooning area at each stage is designated as the lead vehicle, and the remaining vehicles are designated as following vehicles. During the platooning process, the lead vehicle and following vehicles, and the following vehicles themselves, undergo vehicle pairing and authentication to establish a communication connection. The lead vehicle then transmits real-time speed, direction, braking status, and other information to the following vehicles. The following vehicles simultaneously receive and process this data in real time, responding to the lead vehicle's actions, maintaining a safe following distance and adjusting speed and direction accordingly.

[0124] Step 2: Based on the unique advantages of intelligent connected vehicles, when a vehicle enters the platoon area, it sends information such as the vehicle's own attributes, acceleration and deceleration capabilities, and fuel consumption-speed-acceleration function to the roadside edge server. The roadside end then makes judgments and corrections based on this information, and sends the deceleration, vehicle spacing, starting acceleration, and economic speed to the vehicle.

[0125] Step 3: During platooning, the stopping distance must be sufficient to allow the vehicle to start and accelerate to After reaching the optimal following distance s min , the acceleration-displacement equation can be expressed as follows:

[0126]

[0127] Among them, a i+1 and a i represents the acceleration of the i+1th vehicle and the ith vehicle to ensure the lowest energy consumption, t is the starting time, and the following relationship is satisfied, which is expressed by the following formula, from which the optimal parking distance s can be determined i .

[0128]

[0129] Step 4: Define the minimum vehicle following distance as the braking distance after stopping at maximum braking acceleration, as expressed in the following formula:

[0130]

[0131] Among them, a -max Indicates the maximum braking acceleration. Considering the vehicle's braking capacity and the driver's comfort, the maximum braking acceleration should not be greater than 8.5m / s 2 .

[0132] (IV) Constructing a real-time signal timing method with autonomous optimization under multiple constraints

[0133] The traffic flow at the intersection predicted in the previous step indicates that the number of vehicles arriving at each entrance to the intersection at a given moment changes in real time. This reduces the stability of the control effect of a single signal timing. Therefore, intelligent traffic signal control equipment is required to autonomously iteratively optimize the signal timing plan to adapt to the traffic flow entering the entrance at different time periods. The roadside edge server performs real-time optimization based on the future traffic flow at the intersection, which is divided into the following three steps:

[0134] Step 1: Based on the above method, the number of vehicles arriving at each entrance, as well as the vehicle type, attributes, and proportion are predicted. Taking road conditions, weather conditions, comfort, and safety as constraints, and minimum fuel consumption and maximum traffic efficiency as objective functions, the average energy-saving speed of each entrance is calculated using the rolling horizon dynamic programming method. From a macro perspective, the overall fuel consumption of vehicles on the road section is minimized.

[0135] In order to maximize traffic efficiency and facilitate the solution of the objective function, the present invention adds a negative sign before the traffic efficiency objective function, which is equivalent to solving its minimum value. Therefore, the optimal equation group for solving the average economic speed of each entrance lane is established as shown in the following formula:

[0136]

[0137] Where n represents the number of intersection entrances, m represents the number of lanes in each entrance, and l represents the length of the platooning area in each lane. ij (t) represents the position of the vehicle in the jth lane of the i-th entrance lane at time t. vij (t) and ε aij (t) represents the random influence of the meteorological environment on the vehicle speed and acceleration at time t on the jth lane of the i-th entrance lane, and is described by normal distribution. max Indicates the maximum speed of the vehicle, v min The minimum speed is a max represents the maximum acceleration, a minis the minimum acceleration. f(v,a) represents the vehicle's fuel consumption function, which indicates the vehicle's fuel consumption rate when the vehicle's speed is v and the acceleration is a. Similarly, g(v,a) represents the traffic efficiency function of the intersection network node, h(v,a) represents the vehicle's comfort function, and s(v,a) represents the vehicle's safety function. Indicates the energy-saving speed at the intersection network node when the overall fuel consumption of vehicles is the lowest.

[0138] At the same time, vehicles entering the formation area send information such as vehicle attributes, acceleration capabilities, fuel consumption-speed-acceleration function, etc. to the vehicle information collection equipment on the roadside to update and iterate the road section database.

[0139] Step 2: The roadside edge server calculates the traffic volume Q based on the prediction i,p,k and the proportion of historical vehicle attributes, calculate the next stage vehicle queue length d i,p,pha , the calculation formula is as follows:

[0140] d i,p,pha =∑s i +∑l i α i

[0141] Among them, l i Represents the length α of different types of vehicles i Indicates the proportion of different vehicle types in this stage, s i is the parking interval between the i-th car and the i+1-th car.

[0142] Step 3: Take the energy-saving speed to pass all queued vehicles and obtain the green light display time G of each phase i,p,pha . Where pha represents a certain stage.

[0143] As the vehicle formation leaves the intersection, the roadside edge server makes corrections based on the traffic volume and vehicle type ratio in the previous stage, and also prepares for the next stage of formation control, thus completing a complete vehicle formation control.

[0144] (V) Sensitivity analysis of vehicle platooning control methods

[0145] The present invention uses Vissim2020 to perform sensitivity analysis on vehicle formation control strategy and signal optimization method, which is divided into the following steps:

[0146] Step 1: Draw the Vissim intersection. First, use the road segment editor to draw each entrance road with a length of 1000 meters. Each entrance road has three lanes: right turn, straight ahead, and left turn, and a lane width of 3.5 meters. Then, use connectors to connect the entrance and exit road segments.

[0147] Step 2: Edit the driving behavior for the road segment. First, define the vehicle type and select a specific vehicle model as the intelligent connected vehicle simulation model. Then, specify the vehicle's driving behavior, including the following model, headway, desired speed, acceleration, braking acceleration, and minimum distance. Next, define the color of the intelligent connected vehicle input to distinguish between the lead vehicle and following vehicles. Finally, select the customized intelligent connected vehicle as the vehicle input for the road segment.

[0148] Step 3: Define vehicle input. First, define the intersection entrance vehicle input traffic volume and input vehicle type, and then set the vehicle turning ratio according to different ratios.

[0149] Step 4: Set up the traffic lights. First, each lane at the intersection, except the right-turn lane, is equipped with independent channeling control for phase protection. Then, a phase sequence diagram is created based on the phase switching pattern for east-west straight-through, east-west left-turn, north-south straight-through, and north-south left-turn. Finally, the resulting signal timing scheme is received and processed via the COM port for simulation control.

[0150] Step 5: Set up detectors to output evaluation results. Place nodes, data collectors, travel time detectors, queue counters, and delay detectors at the intersection. Nodes are used to automatically form a detection area for the entire intersection to obtain overall evaluation indicators. Data collectors are used to measure indicators such as vehicle flow, speed, density, occupancy, and emissions on a road section or lane. Travel time detectors are used to measure indicators such as the travel time, travel speed, and travel distance between two points. Queue counters can be used to measure indicators such as queue length, number of queued vehicles, and queue time on a road section or lane. Delay detectors can be used to measure indicators such as delay time, number of delayed vehicles, number of stops, and stop time on a road section or lane.

[0151] Step 6: Implement real-time control of signal timing by calling a Python script through the COM port. First, a trained Conv-LSTM model based on the attention mechanism is used to predict future traffic conditions. Experiments show that the mean absolute error is only 5.54 at a 5-minute interval. Then, based on the future traffic conditions, the length of each entrance to the queue is calculated, thereby determining the green light duration for each phase. Finally, the SignalControllers property of the vissim object is used to obtain a signal controller collection object and issue signal timing instructions to the signal controller in the simulation. The Net.Vehicles property is also used to obtain a vehicle collection object to correct the arrival volume and vehicle type ratio.

[0152] Step 7: Call the Python script through the COM port to control the platoon vehicles. Use the Net.VehiclePlatoons property to obtain a platoon collection object, which is used to access and control all platoons in the simulation. Use the methods and properties of the vehicle and platoon objects, such as GetAttValue, SetAttValue, AddVehicle, and RemoveVehicle, to create, delete, modify, and query vehicles and platoons.

[0153] Step 8: Design comparative experiments for sensitivity analysis. On the one hand, the traffic flow input of each entrance lane is changed to 1600pcu / h, 1400pcu / h, 1200pcu / h, 1000pcu / h, 800pcu / h, and 600pcu / h, respectively, to explore the stability of the vehicle formation control strategy and signal optimization method of the present invention under different traffic flows. On the other hand, compared with the traditional fixed timing, the average delay of the intersection, the average speed and the average queue length are used as direct indicators for evaluation, and the optimization potential of the method of this embodiment in actual intersection optimization control is explored. The experimental results show that it has good anti-interference ability when different traffic flow inputs are used, such as Figure 6 、 Figure 8 、 Figure 10 At the same time, the optimization effect of intersection control is significantly better than the traditional fixed timing scheme, such as Figure 7 、 Figure 9 、 Figure 11 shown.

[0154] Through sensitivity analysis, it is concluded that the vehicle platoon control and signal optimization method in this connected traffic environment has good optimization effect and practical value.

[0155] Example 3

[0156] This embodiment 3 provides a non-transitory computer-readable storage medium, which is used to store computer instructions. When the computer instructions are executed by a processor, the vehicle formation control and signal optimization method in the connected traffic environment as described above is implemented.

[0157] Example 4

[0158] This embodiment 4 provides a computer device, including a memory and a processor, wherein the processor and the memory communicate with each other, the memory stores program instructions that can be executed by the processor, and the processor calls the program instructions to execute the vehicle formation control and signal optimization method in the connected traffic environment as described above.

[0159] Example 5

[0160] This embodiment 5 provides an electronic device, including: a processor, a memory, and a computer program; wherein the processor is connected to the memory, and the computer program is stored in the memory. When the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to execute instructions for implementing the vehicle formation control and signal optimization method in a connected traffic environment as described above.

[0161] In summary, the vehicle formation control and signal optimization method and system in a connected traffic environment described in an embodiment of the present invention divides the entrance road section of an urban intersection into regions based on the location of the roadside vehicle information collection device. Unlike prior art research, this invention moves the roadside vehicle information sensing device from the vicinity of the intersection to the middle of the road section, thereby increasing the effective length of the vehicle formation. A Conv-LSTM module based on an attention mechanism is used to predict the future traffic situation of the urban intersection to obtain the traffic volume, vehicle type ratio, and attributes of each entrance lane. Unlike prior art research, this invention improves prediction accuracy and robustness while also adding predictions of vehicle type ratio and attributes. The average energy-saving speed of each entrance lane is solved using a rolling time domain dynamic programming method. The predicted traffic volume, vehicle type ratio, and attributes are combined to determine the vehicle formation length of each entrance lane, thereby determining the green light display time for each phase. Unlike prior art research, this invention uses vehicle-road communication technology to dynamically form vehicles based on interacting vehicle attributes during driving, and obtains an economical speed and real-time signal timing plan based on multiple constraints, greatly improving the traffic efficiency of intersection network nodes. This invention uses a pilot vehicle-two-way leading vehicle-following vehicle communication topology to achieve highly efficient communication between intelligent connected vehicles, and defines parking queue spacing based on different vehicle attributes. Unlike existing research, this invention determines the parking spacing for each incoming vehicle, ensuring that they simultaneously reach the minimum queue spacing when starting to accelerate to an economic speed, thereby reducing the reduction in queue stability caused by different vehicle acceleration capabilities. This shows that the present invention, based on the vehicle queue control strategy and signal timing coupling optimization mechanism from both micro and macro perspectives, greatly enhances the communication efficiency of intersection network nodes while ensuring the safety and reliability of traffic operation, and has practical engineering application value.

[0162] Although the above describes the specific embodiments of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art on the basis of the technical solutions disclosed in the present invention without the need for creative work should be included in the scope of protection of the present invention.

Claims

1. A vehicle formation control and signal optimization method in a connected traffic environment, characterized in that: include: Determine the static attribute parameters of urban intersections and divide them into functional areas, and obtain the intersection geometry, road width, entrance road length, and location parameters of roadside vehicle information collection equipment; The traffic situation at the intersection is predicted. The vehicle arrival situation and vehicle type ratio of each phase in the next signal cycle are predicted by the Conv-LSTM network based on the attention mechanism. The future traffic situation at the intersection is predicted, including: the position of the roadside vehicle information collection device on the intersection entrance road section is recorded as m, where the collected information includes traffic flow and vehicle attributes, and is stored in the form of two tuples in f t m In the algorithm, the traffic state sequence from time tn to time t is recorded. Assuming there are k observation points, the historical traffic flow data of the intersection is represented by a matrix. A one-dimensional convolution kernel is used to obtain the local perception domain. A one-dimensional convolution operation is performed on the traffic state data at each time step t to obtain spatial features. Based on the attention mechanism, the importance of the traffic flow sequence in the spatiotemporal traffic flow matrix is ​​automatically mined in different time periods. The hidden state of the output is calculated at each time step t. The weighted sum of is used to predict the traffic volume and vehicle type ratio of each entrance lane; The system makes platooning decisions for vehicles entering an intersection, using a pilot vehicle-two-way preceding vehicle-follower vehicle communication topology to ensure platooning safety. Vehicle-road interaction is used to determine the attributes of incoming vehicles, and the optimal parking spacing and minimum spacing for platooning are determined based on the platooning strategy. Real-time signal timing optimization is performed at intersections under multiple constraints, taking into account environmental, safety, and comfort constraints. Minimum energy consumption and maximum traffic efficiency are used as optimization objectives. The economic speed, green light display time for each phase, and phase sequence are obtained through the rolling horizon dynamic programming method. Optimize intersection signal timing in real time under multiple constraints, including: Based on the predicted number of vehicles arriving at each entrance, the proportion of vehicle types, and their attributes, with road conditions, meteorological environment, comfort, and safety as constraints, and minimum fuel consumption and maximum traffic efficiency as objective functions, the economic speed is solved through the rolling time domain dynamic programming method to achieve the lowest overall fuel consumption of vehicles at the intersection network nodes; vehicles entering the formation area send vehicle information to the roadside vehicle information collection device to update and iterate the road section database; the roadside edge server calculates the length of the vehicle queue in the next stage based on the predicted traffic volume and historical vehicle type ratio; with the goal of passing all queued vehicles at the economic speed, the green light display time for each phase is obtained; as the vehicle formation leaves the intersection, the roadside edge server makes corrections based on the traffic volume and vehicle type ratio of the previous stage to prepare for the formation control in the next stage, completing a complete vehicle formation control.

2. The vehicle formation control and signal optimization method in a connected traffic environment according to claim 1, characterized in that: The functional area division of the urban intersection includes: dividing the entrance of the urban intersection into three areas, starting from the parking line, which are the parking area, the formation area and the free driving area, with lengths L and L respectively. i,p,0 , L i,p,1 , L i,p,2 , where i represents the number of the road section, which is consistent with the number of its downstream intersection, p represents the direction of the four entrance roads of the intersection, e represents the east entrance, s represents the south entrance, w represents the west entrance, and n represents the north entrance.

3. The vehicle formation control and signal optimization method in a connected traffic environment according to claim 1, characterized in that: Make platooning decisions for vehicles entering the intersection, including: The communication topology of the platooning system adopts a leader vehicle and a two-way leading vehicle following system. The first vehicle entering the platooning area at each stage is designated as the leader vehicle, and the remaining vehicles are designated as following vehicles. During the platooning process, the leader vehicle and following vehicle, as well as the following vehicles and following vehicles, perform vehicle pairing and authentication to establish a communication connection. The leader vehicle transmits real-time speed, direction, and braking information to the following vehicles. At the same time, the following vehicles receive and process the data sent by the leader vehicle in real time, responding to the actions of the leader vehicle, maintaining a safe following distance, and adjusting speed and direction in a timely manner.

4. The vehicle formation control and signal optimization method in a connected traffic environment according to claim 3, characterized in that: When a vehicle enters the platooning area, it sends vehicle information to the roadside edge server. The roadside server then makes judgments and corrections based on the vehicle information, and sends the deceleration, inter-vehicle distance, starting acceleration, and economic speed to the vehicle via the vehicle-road communication device. During platooning, only the longitudinal acceleration and deceleration of the vehicle are considered; the stopping distance needs to meet the optimal following distance after the vehicle starts and accelerates to the economic speed. The acceleration-displacement equation is obtained to determine the optimal stopping distance.

5. A vehicle platoon control and signal optimization system in a connected traffic environment that implements the method according to any one of claims 1 to 4, characterized in that: include: The acquisition module is used to determine the static attribute parameters of urban intersections and divide them into functional areas, and obtain the intersection geometry, road width, entrance road length, and location parameters of roadside vehicle information collection equipment; The prediction module is used to predict the traffic situation at the intersection. It uses a Conv-LSTM network based on the attention mechanism to predict the vehicle arrival situation and vehicle type ratio in each phase of the next signal cycle. The decision-making module is used to make platooning decisions for vehicles entering the intersection. It uses a pilot vehicle-two-way leading vehicle-follower vehicle communication topology to ensure platooning safety. It obtains the attributes of incoming vehicles through vehicle-road interaction and determines the optimal parking spacing and minimum spacing for platooning based on the platooning strategy. The timing optimization module is used to optimize the real-time signal timing of intersections under multiple constraints. It takes the environment, safety, and comfort as constraints, and uses the lowest energy consumption and highest traffic efficiency as the optimization objective function to obtain the economic speed, green light display time of each phase, and phase sequence through the rolling horizon dynamic programming method.

6. A non-transitory computer-readable storage medium, characterized in that The non-transitory computer-readable storage medium is used to store computer instructions. When the computer instructions are executed by the processor, the vehicle formation control and signal optimization method in a connected traffic environment as described in any one of claims 1 to 4 is implemented.

7. A computer device, characterized in that: It includes a memory and a processor, the processor and the memory communicate with each other, the memory stores program instructions that can be executed by the processor, and the processor calls the program instructions to execute the vehicle formation control and signal optimization method in a connected traffic environment as described in any one of claims 1 to 4.

8. An electronic device, characterized in that: include: A processor, a memory, and a computer program; wherein the processor is connected to the memory, and the computer program is stored in the memory. When the electronic device is running, the processor executes the computer program stored in the memory to implement the vehicle formation control and signal optimization method in a connected traffic environment as described in any one of claims 1 to 4.

Images