A vehicle cooperative forced lane-changing control method for highway traffic incident section under V2X environment

Abstract

The application discloses a kind of V2X environment under highway traffic event section vehicle cooperative forced lane-changing control method, and the method is also a kind of V2X environment under highway traffic event section vehicle cooperative forced lane-changing control system.The application first proposes the road effective traffic capacity detection method based on improved neural network, detects traffic event section and operating state;Second, in traffic event section, based on V2X technology, a vehicle longitudinal driving control method considering lane-changing detection factor and vehicle collision avoidance position is proposed;Third, in the case of needing to carry out transverse forced lane-changing, a game model considering following vehicle driving stability, driving time and comfort is constructed to calculate the target lane merging sequence, and finally whether the lane-changing condition is met is determined by calculating the feasible point of forced lane-changing, to control the vehicle to complete the transverse forced lane-changing process.By this method, vehicle can be controlled to complete vehicle transverse and longitudinal driving safely, efficiently and smoothly on traffic event section.

Description

[Technical Field]

[0001] This invention belongs to the field of intelligent transportation system technology, and involves technical fields such as vehicle networking technology, sensor technology, and automatic control technology. Specifically, it relates to a vehicle cooperative forced lane change control method for highway traffic incident sections in a V2X environment. This method is a vehicle cooperative forced lane change control method that controls the longitudinal movement of vehicles in the traffic incident section and enables the vehicles to complete the forced lane change process after detecting the traffic incident section. [Background Technology]

[0002] In recent years, V2X technology has made good progress in the application of vehicle cooperative control. Through V2X technology, vehicles can perceive more information over a wider range and in greater variety, providing a rich data foundation for vehicle cooperative control. This enables vehicles to drive more safely, efficiently, and economically on highways, and can also reduce the occurrence of traffic incidents and improve the driving conditions of vehicles on the road sections where traffic incidents occur, thereby enhancing the traffic capacity and operational efficiency of the affected road sections.

[0003] Because traffic incidents can occur at any time on highways, and the computing power of roadside units in smart streetlights is limited, it is necessary to detect key road sections in the road network so that computing resources can be concentrated on solving traffic problems on these key sections. Bottleneck sections where traffic incidents reduce effective traffic capacity are among these key sections, thus necessitating the detection of effective road traffic capacity. With the development of artificial intelligence algorithms, their applications are becoming increasingly widespread. Among them, the Back Propagation Neural Network (BPNN) algorithm is widely used in traffic incident detection algorithms and can detect traffic incidents to a certain extent. However, the weights and thresholds of the BPNN are randomly selected, making it prone to getting trapped in local optima.

[0004] On highways where traffic incidents occur, some lanes become impassable, forcing vehicles in those lanes to change lanes to reach a passable one. This often leads to more vehicles changing lanes, causing traffic flow disorder and reducing vehicle safety and efficiency. By utilizing V2X technology to obtain information such as headway, vehicle trajectory, and queue length, lane-changing models and decision-making methods can be constructed to guide or control lane changes in bottleneck scenarios like weaving zones, enabling vehicles to maintain safe distances and avoid collisions. However, this application is typically limited to normally operating highways, with limited research and application in scenarios involving highway traffic incidents. [Summary of the Invention]

[0005] This invention provides a method for controlling vehicle cooperative forced lane change in a highway traffic incident section under a V2X environment, and also a control system for vehicle cooperative forced lane change in a highway traffic incident section under a V2X environment.

[0006] The working principle of this invention includes a road effective capacity detection module, a vehicle information collection and processing module, a vehicle longitudinal motion control module, a vehicle forced lane change merging sequence optimization module, and a forced lane change feasible point calculation module, the logical relationship of which is as follows: Figure 1 As shown.

[0007] The objective of this invention is achieved through the following technical solution:

[0008] A method for controlling vehicle cooperative forced lane change in a highway traffic incident segment under a V2X environment includes a road effective capacity detection module, a vehicle information collection and processing module, a vehicle longitudinal motion control module, a vehicle forced lane change merging sequence optimization module, and a forced lane change feasible point calculation module.

[0009] To detect road segments involved in traffic incidents in the road network, the effective capacity is calculated using a BP neural network based on the eagle search algorithm (BES-BP neural network) in the road effective capacity detection module, thereby more accurately detecting the road segments involved in traffic incidents.

[0010] In order to detect the road segment where a traffic incident caused a lane closure, the road effective capacity detection module is used to identify the phenomenon of reduced road effective capacity in the road network, thereby finding the bottleneck segment in the road network caused by the traffic incident.

[0011] To avoid vehicle collisions in traffic incident areas, it is crucial to accurately detect the position of each vehicle. The extended Kalman filter in the vehicle information acquisition and processing module is used to process vehicle state data from different sources to obtain high-precision position information of each vehicle.

[0012] To address the problem of disorderly and forced lane changes by vehicles in traffic incident sections, which leads to traffic flow disorder, low vehicle efficiency, and significant safety hazards, it is necessary to regulate and control the behavior of forced lane changes. By using the vehicle longitudinal motion control module to consider lane change detection factors and the collision avoidance position in front of the vehicle to obtain longitudinal acceleration control, the longitudinal motion control of the vehicle can be completed to stabilize the traffic flow and provide lane change space for vehicles changing lanes.

[0013] During forced lane changing, an unreasonable vehicle merging sequence can lead to a deterioration in vehicle efficiency, comfort, and the stability of following vehicles. The vehicle forced lane changing merging sequence optimization module utilizes a game model that considers the stability of following vehicles, vehicle travel time, and comfort to optimize the vehicle merging sequence to achieve the optimal result.

[0014] To ensure vehicle safety and prevent secondary accidents at traffic incident sections, a forced lane change feasibility calculation module is used to determine whether a vehicle meets the lane change conditions. Only when the lane change conditions are met can a forced lane change be executed, ensuring that the vehicle can safely pass through the traffic incident section.

[0015] Furthermore, when detecting traffic incidents on highway network sections, the BP neural network is first improved using the Bald Eagle Search algorithm. Then, the effective capacity detection module uses traffic flow parameters such as traffic volume, speed, and vehicle occupancy as input to the BES-BP neural network, and outputs the effective capacity to train the BES-BP neural network. Then, the traffic volume, speed, and vehicle occupancy of adjacent locations are collected and input into the trained BES-BP neural network, thereby detecting the effective capacity in real time and detecting the occurrence of traffic incidents.

[0016] In the event of a traffic incident, the vehicle information acquisition and processing module obtains vehicle driving information from vehicles and roadside equipment, and uses extended Kalman filtering to improve the accuracy of vehicle position information. Within the traffic incident section, after a vehicle takes over, the vehicle longitudinal motion control module plans and controls the vehicle's longitudinal motion, while simultaneously determining whether there are lane-changing vehicles in adjacent lanes; if so, it provides lane-changing space for the adjacent vehicle. During a forced lane change, the vehicle forced lane-change merging sequence optimization module optimizes the order of lane-changing vehicles and their adjacent vehicles in the target lane, resulting in an optimized target lane vehicle sequence.

[0017] Once the vehicle sequence in the target lane is obtained, the lane-changing position of the vehicles is mapped into the target lane. The forced lane-changing feasibility point calculation module is used to determine whether the vehicle meets the lane-changing conditions. If not, the vehicle longitudinal motion control module is used to change the driving state of the vehicles in front and behind in the target lane to provide more lane-changing space. If the lane-changing conditions are met, the vehicle performs a forced lane-changing behavior. Finally, the vehicle passes through the traffic event section through the vehicle longitudinal motion control module.

[0018] Furthermore, to better detect the effective capacity of road segments, the initial weights and thresholds of the BP neural network are optimized using the BES algorithm. This allows for the discovery of some better, smaller-range solutions within the optimal solution space of the weights and thresholds. The BP neural network algorithm is then used to continue searching for the optimal solution within this solution space. Using the BES-BP neural network achieves better overall performance. Key traffic flow parameters such as traffic volume, speed, and vehicle occupancy are used as inputs to the trained BES-BP neural network, enabling real-time detection of effective capacity, identifying bottleneck segments in the road network, and implementing precise management and control of vehicles on those segments. When a traffic event causing lane closure is detected, the vehicle information acquisition and processing module obtains precise vehicle location information, allowing vehicles to change lanes more safely and minimizing traffic accidents. This provides data support for vehicle-coordinated forced lane-changing control.

[0019] Compared with the prior art, the present invention has the following advantages:

[0020] 1. This invention discloses a vehicle cooperative forced lane-changing control method for highway traffic incident sections under a V2X environment. Addressing the forced lane-changing problem in highway traffic incident sections, this method constructs a vehicle cooperative forced lane-changing method under a V2X environment. This method includes a road effective capacity detection module, a vehicle longitudinal motion control module, a vehicle forced lane-changing merging sequence optimization module, and a forced lane-changing feasibility point calculation module. This method enables intelligent connected vehicles to safely and efficiently complete vehicle driving and forced lane-changing behaviors on highway traffic incident sections. Due to the high speeds in highway areas, this invention analyzes the vehicle driving state in traffic incident sections from two perspectives: longitudinal driving motion and lateral lane-changing motion. This simplifies the vehicle motion model and improves solution efficiency. Only when the lane-changing conditions calculated by the lane-changing feasibility point calculation module are met can the vehicle perform a lateral lane-changing motion, ensuring the safety of intelligent connected vehicles during driving, improving vehicle traffic efficiency and comfort, and stabilizing traffic flow in the section.

[0021] 2. The present invention provides a method for vehicle cooperative forced lane-changing control in a V2X environment for highway traffic incidents. It utilizes an improved BES-BP neural network to detect the effective capacity of highways in real time, improving the detection effect and enabling the detection of traffic incidents and identifying bottleneck sections in the road network. The information acquisition and processing module of this invention uses vehicle driving information collected from onboard and roadside sensors to construct a state prediction equation based on vehicle position, azimuth, and speed. Using the relative distance and azimuth between the vehicle and the roadside sensor as observation values, it employs extended Kalman filtering for fusion to obtain more accurate vehicle positions, providing more precise information for lane-changing.

[0022] 3. The present invention provides a vehicle cooperative forced lane-changing control method for highway traffic incident sections under a V2X environment. The vehicle forced lane-changing merging sequence optimization module addresses the problems of travel time, unstable following vehicle travel status, and poor comfort caused by disordered lane changing during forced lane changing. It uses game theory to construct a payoff function that considers the stability of following vehicle travel, vehicle travel time, and comfort to analyze and solve for the optimal vehicle sequence in the target lane after forced lane changing. Combining the characteristics of forced lane changing in highway traffic incident sections, it constructs a vehicle longitudinal motion model to ensure vehicle traffic efficiency and a forced lane-changing feasibility point generation module to ensure vehicle safe driving, thereby regulating and controlling the vehicle forced lane-changing behavior. [Attached Image Description]

[0023] Figure 1 This is a logic diagram of the traffic event forced lane change control method for a highway traffic event segment in a V2X environment, as described in this invention.

[0024] Figure 2 This is a flowchart of the BP neural network algorithm based on BES for a vehicle cooperative forced lane change control method for highway traffic incidents in a V2X environment, as described in this invention. 【Detailed Implementation Methods】

[0025] The following examples further illustrate specific embodiments of the present invention. These examples are intended to illustrate the invention and not to further limit it. Unless otherwise specified in the experimental examples, conventional methods can be followed.

[0026] Example:

[0027] A method for controlling vehicle-cooperative forced lane changing in highway traffic incidents under a V2X environment includes a road effective capacity detection module, a vehicle information collection and processing module, a vehicle longitudinal motion control module, a forced lane changing merging sequence optimization module, and a forced lane changing feasibility point calculation module. Their logical relationships are as follows: Figure 1 As shown;

[0028] The road effective capacity detection module utilizes an improved BES-BP neural network to monitor changes in effective capacity by inputting traffic volume, speed, and vehicle occupancy rate at two adjacent points. This BES-BP neural network requires pre-training, and after training, it can detect effective capacity in real time. The vulture search algorithm has three processes: selecting the hunting area, circling and searching for prey, and diving to capture the prey. An adaptive weight factor is introduced during the diving to capture prey process to enhance the algorithm's search capability, and its expression is as follows:

[0029] Pi,new =rand*P best +ω*(x1(i)*(P i -c1*P mean )+y1(i)*(P i -c2*P best )) (1)

[0030]

[0031]

[0032]

[0033] xr(i)=r(i)*sinh[θ(i)] (5)

[0034] yr(i)=r(i)*cosh[θ(i)] (6)

[0035] θ(i)=b*π*rand (7)

[0036] r(i)=θ(i) (8)

[0037] Where: c1 and c2 are parameters that increase the intensity of the vulture's movement towards the optimal point and the center point, and their values ​​range from 1 to 2; t is the current iteration number; t_max is the maximum number of iterations; P i,new The location of the vulture after the update; P best This is the location of the best hunting area for bald eagles at present; P mean P is the average of the searched hunting locations. i is the position of the i-th bald eagle; b is the parameter in polar coordinates that determines the angle of the search point, and its value ranges from 5 to 10; rand is a random number between 0 and 1.

[0038] The BPNN algorithm is further improved through the improved vulture search algorithm. The process is as follows: Figure 2 (BES-based BP neural network algorithm flow):

[0039] In the BES-BP neural network, the input layer has 6 neurons, representing 6 traffic feature values ​​(upstream and downstream vehicle speed, traffic flow, and vehicle occupancy rate), and the output layer has 1 neuron, which is the effective traffic capacity value. This forms a method for detecting the effective traffic capacity of road segments in traffic events, enabling real-time monitoring of the effective traffic capacity of highways and detecting whether traffic events that cause lanes to become impassable have occurred.

[0040] In the information acquisition and processing module, under the V2X environment, on-board equipment acquires data such as vehicle position, azimuth angle, and speed, while roadside sensor equipment senses the relative distance and relative orientation between the vehicle and the smart streetlights. A state prediction equation is constructed using the vehicle position, azimuth angle, and speed, with the relative distance and relative orientation between the vehicle and the roadside sensors as the observed values. The error detected by the on-board and roadside sensors is determined by the specific sensor model. An extended Kalman filter is used to form a high-precision vehicle position monitoring method, thereby obtaining more accurate vehicle position information.

[0041] In the vehicle longitudinal motion control module, to enable vehicles to drive safely and efficiently and to perform coordinated lane changes, the calculated acceleration is used to control the longitudinal motion of the vehicles. The calculation of the longitudinal acceleration of vehicles in traffic event segments is an improvement on the IDM model. The expression for the vehicle control acceleration of the i-th vehicle at time t is as follows:

[0042]

[0043] X c (t)=min(X i+1 (t),X k (t)) (10)

[0044]

[0045]

[0046] Where C i (t) is the lane change monitoring factor, which monitors whether there are vehicles with mandatory lane change requirements within 150m of the adjacent lane of the vehicle. Its value is 0 or 1, where 0 means there are no vehicles with mandatory lane change requirements in the adjacent lane, and 1 means there are vehicles with mandatory lane change requirements in the adjacent lane; a r To ensure safe driving, and when there are vehicles changing lanes, the acceleration of the vehicle is set to 0; a a,i (t) represents the acceleration of the i-th vehicle at time t, calculated according to the IDM model; X i (t) represents the distance between the position of vehicle i at time t and the starting point of the traffic event control zone, where k refers to the vehicle number in the original lane of the lane-changing vehicle; X c (t) represents the position in front of the vehicle that needs to be avoided, and d represents the minimum position of the vehicle in front of the i-th vehicle or the position of the vehicle changing lanes mapped in the target lane; s,i (t) represents the safe distance of the i-th vehicle; a max v is the vehicle's maximum acceleration. f α is the free-flow velocity of this section; α is the velocity power coefficient; s i *(t) represents the expected headway of the i-th vehicle at time t; s0 represents the minimum safe stopping distance for the vehicle; t h For the headway of vehicles in this section of road; a com To provide comfortable vehicle acceleration on this section of road;

[0047] Equations (9) to (12) indicate that, to ensure vehicle safety while driving on a road segment affected by a traffic incident, the vehicle is controlled using the acceleration of the IDM model when following another vehicle or when the safe distance between the vehicle and the vehicle changing lanes is not met. When the vehicle senses a vehicle changing lanes in an adjacent lane, the acceleration is adjusted to 0 m / s² in order to enable the adjacent vehicle to complete the forced lane change smoothly. 2 The vehicle is driven at a constant speed. During this process, the vehicle uses the forced lane change feasible point calculation model to calculate the order of vehicles after the lane change. When the order of vehicles after the lane change is determined, the target vehicle will map the predicted position of the lane-changing vehicle. The vehicle at this position will decelerate through the IDM model.

[0048] Furthermore, the expressions for vehicle position and vehicle speed are as follows:

[0049]

[0050] v i (t+t ξ ) = v i (t)+a i (t)t ξ (14)

[0051] 0≤v i (t)≤V i,max (15)

[0052] a i,min ≤a i (t)≤a i,max (16)

[0053]

[0054] Where x i (t) represents the longitudinal displacement of the i-th vehicle from the starting point of the traffic incident control zone at time t; t ξ v is the time interval; i (t) represents the velocity of the i-th car at time t; a i (t) represents the acceleration of the vehicle at time t; V i,max These represent the maximum speeds of the i-th vehicle in this lane; a i,min and a i,max These represent the maximum and minimum accelerations of the i-th vehicle in this lane, respectively.

[0055] Equations (13) and (14) represent the vehicle planning state iteration of displacement and velocity during vehicle driving; Equations (15) to (16) represent the speed limit, acceleration limit and time value limit of the vehicle during driving, respectively, wherein the upper and lower limits of acceleration are determined by the driving performance and braking performance of the vehicle.

[0056] In the forced lane change merging sequence optimization module, when the target lane detects a vehicle in the traffic event lane, it performs forced lane change planning to calculate the vehicle sequence. The forced lane change merging sequence optimization method is used to determine the order of vehicles involved in the lane change process in the target lane. Game theory is used to analyze and solve the sequence of vehicles in the target lane, i.e., to determine whether a vehicle becomes the preceding or following vehicle.

[0057] The game payoff matrix for the vehicle forced lane-changing and merging sequence optimization method is as follows:

[0058]

[0059] In the table, b represents the collision risk value. The higher the collision risk value, the more dangerous the strategy adopted by the participant. q1, q2, p1, and p2 represent the corresponding reward values ​​of the strategies, indicating the reward values ​​obtained by the participants after adopting different strategies.

[0060] By solving the game theory lane-changing model, the payoff function is as follows:

[0061] If both Player 1 and Player 2 simultaneously choose to be the car in front or behind, a collision risk will occur. The expression for the collision risk value b is as follows:

[0062]

[0063] Where t0 represents the start time, and ||*|| represents a L2 norm used to solve for the two state functions θ. i The distance between (t) indicates that the closer the initial states of the two are, the greater the probability of a collision and the smaller the reward corresponding to the collision risk. Conversely, the probability of a collision is smaller and the reward corresponding to the collision risk will increase.

[0064] The operating state function θ of vehicle i i (t), used to describe the vehicle's operating state at time t, specifically:

[0065] θ i (t)={x i (t),v i (t),a i (t)} (17)

[0066] Where, x i (t) represents the position of vehicle i at time t; vi (t) represents the velocity of vehicle i at time t; a i (t) represents the acceleration of vehicle i at time t; the starting time of vehicle i in the traffic incident control zone is set to t. The time of leaving the monitoring and control area is Then we can obtain the running status of vehicle i at the starting point of the monitoring and control area, denoted as: The expression is as follows:

[0067]

[0068] Its corresponding components are:

[0069]

[0070]

[0071]

[0072] Since the strategy adopted by the vehicle is directly related to the initial states of the vehicles before and after it in the target lane, let F be the sequence of the vehicle's parallel mapping to the target lane. i Let G be the sequence obtained by the proposed game theory method. i And G i There are two types: becoming the preceding vehicle or becoming the following vehicle, denoted as [missing characters]. Furthermore, defining the vehicles as k1 and k2, the revenue values ​​p1, p2, q1, and q2 can be expressed as:

[0073]

[0074]

[0075]

[0076]

[0077]

[0078] In the first item, the stability of the following vehicles can be evaluated by summing the squares of the following vehicle accelerations of each lane-changing vehicle.

[0079] In the second item, the travel time is defined as T. i The time when vehicle i enters the traffic incident section and the time of leaving the traffic incident site The relationship between them is established as shown in equation (25):

[0080]

[0081] In the third term, the jerk j i (t) is the derivative of acceleration, one of the most important factors affecting ride comfort. The smaller the jerk, the higher the comfort of the vehicle occupants. The jerk is calculated as follows:

[0082]

[0083] In the lane change feasibility point calculation module, the lane change feasibility point generation model can be used to ensure that the lane change position of the vehicle is reasonable and safe, as detailed below:

[0084] P k s ={p y,k |X i (t f,k )+d s,i -Δx k ≤p y,k ≤X i+1 (t f,k )+d s,k -Δx k} (27)

[0085] p y,k =X k (t f,k (28)

[0086] d s,k (t)=t h v k (t f,k (39)

[0087]

[0088] X k (t f,k ) < P incident (31)

[0089] In the formula: X i (t) represents the distance between the position of vehicle i at time t and the starting point of the traffic event control zone, and k refers to the vehicle number in the adjacent lane that needs to change lanes; p y,k t is the longitudinal distance between the lane-changing point and the starting point of the traffic incident section. f,k d represents the time when the i-th vehicle reaches the target lane and completes the lane change; s,k (t) represents the safe distance of the k-th vehicle at time t; t h For this section of road, the headway of vehicles; Δx k v represents the longitudinal displacement of the k-th vehicle during the lane-changing process; y,k Let Δt be the velocity of the k-th car at the feasible lane-changing point y; kP represents the time spent by car k during the lane change process. incident This refers to the distance between the location of the traffic incident and the starting point of the traffic incident section.

[0090] The lane change feasibility point calculation model determines whether the vehicle meets the lane change conditions. If it does, the vehicle performs a lane change and uses the vehicle longitudinal motion control module to travel through the traffic event section. If it does not meet the conditions, the vehicle longitudinal motion control module will continue to update the vehicle driving status planning information and use the lane change feasibility point calculation module to calculate and judge until the vehicle meets the forced lane change conditions, thereby performing a forced lane change and passing through the traffic event section.

[0091] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, any improvements and changes made without departing from the inventive concept of the present invention are within the protection scope of the present invention.

Claims

1. A method for controlling vehicle cooperative forced lane changing on highways in a V2X environment during traffic incidents, characterized in that: It includes a road effective traffic capacity detection module, a vehicle information collection and processing module, a vehicle longitudinal motion control module, a vehicle forced lane change merging sequence optimization module, and a forced lane change feasible point calculation module; The effective capacity is calculated by using the BP neural network based on the eagle search algorithm, namely the BES-BP neural network, in the road effective capacity detection module, so as to more accurately detect the road segments where traffic events occur. The extended Kalman filter in the vehicle information acquisition and processing module is used to process vehicle status data from different sources to obtain high-precision location information for each vehicle. By utilizing the longitudinal motion control module of the vehicle to obtain longitudinal acceleration control of the vehicle, considering the lane change detection factor and the collision avoidance position in front of the vehicle, the longitudinal motion control of the vehicle is completed to smooth the traffic flow and provide lane change space for vehicles changing lanes. The vehicle longitudinal motion control module is used to determine the lane change monitoring factor C. i (t) Monitor whether there are vehicles with mandatory lane-changing needs within 150m of the adjacent lane of the vehicle. The value is 0 or 1. A value of 0 means that there are no vehicles with mandatory lane-changing needs in the adjacent lane, and a value of 1 means that there are vehicles with mandatory lane-changing needs in the adjacent lane. The vehicle merging sequence is optimized by using a game model that considers the stability of following vehicles, vehicle travel time, and comfort in the vehicle forced lane change merging sequence optimization module. The forced lane change feasibility point calculation module is used to determine whether a vehicle meets the lane change conditions. Only when the lane change conditions are met can the forced lane change be executed to ensure that the vehicle can safely pass through the traffic incident section.

2. The method for cooperative forced lane changing control of vehicles on highway traffic incident sections in a V2X environment according to claim 1, characterized in that: When detecting traffic incidents on highway network sections, the BP neural network is first improved using the Bald Eagle Search algorithm. Then, the effective capacity detection module uses traffic flow parameters such as traffic volume, speed, and vehicle occupancy as input to the BES-BP neural network and outputs the effective capacity to train the BES-BP neural network. Then, the traffic volume, speed, and vehicle occupancy of adjacent locations are collected and input into the trained BES-BP neural network to detect the effective capacity in real time and detect the occurrence of traffic incidents. If a traffic incident occurs, the vehicle information collection and processing module obtains vehicle driving information from vehicles and roadside equipment, and uses extended Kalman filtering to improve the accuracy of vehicle position information. In the traffic incident section, after the vehicle takes over, the vehicle longitudinal motion control module plans and controls the longitudinal motion of the vehicle, and at the same time determines whether there are vehicles changing lanes in the adjacent lanes. If so, it provides lane-changing space for the adjacent vehicles. During the forced lane change process, the vehicle forced lane change merging sequence optimization module is used to optimize the sorting of lane-changing vehicles and their adjacent vehicles in the target lane, resulting in an optimized target lane vehicle sequence. Once the vehicle sequence in the target lane is obtained, the lane-changing position of the vehicles is mapped into the target lane. The forced lane-changing feasibility point calculation module is used to determine whether the vehicle meets the lane-changing conditions. If not, the vehicle longitudinal motion control module is used to change the driving state of the vehicles in front and behind in the target lane to provide more lane-changing space. If the lane-changing conditions are met, the vehicle performs a forced lane-changing behavior. Finally, the vehicle passes through the traffic event section through the vehicle longitudinal motion control module.

3. The method for cooperative forced lane-changing control of vehicles on highway traffic incident sections in a V2X environment according to claim 1, characterized in that: The initial weights and thresholds of the BP neural network are optimized using the BES algorithm, finding some better, smaller-range solutions within the optimal solution space of the weights and thresholds. The BP neural network algorithm is then used to continue searching for the optimal solution within this solution space. Using the BES-BP neural network achieves better overall performance. Key traffic flow parameters such as traffic volume, speed, and vehicle occupancy are used as inputs to the trained BES-BP neural network, enabling real-time detection of effective traffic capacity, identifying bottleneck sections in the road network, and implementing precise management and control of vehicles on those sections. When a traffic event causing lane closure is detected, the vehicle information acquisition and processing module obtains precise vehicle location information, allowing vehicles to change lanes more safely, avoiding traffic accidents, and providing data support for vehicle-coordinated forced lane-changing control.

Images