An information-assisted driving method based on car-road cooperation technology

Abstract

The application discloses an information assisted driving method based on a vehicle-road cooperation technology, a car-following driving assistance module considers the problem that a traditional intelligent vehicle can only refer to driving information of a single vehicle in front in car-following driving, detects information of a plurality of vehicles in front through subscription of a publisher, and performs an assisted car-following behavior according to the information. A lane-changing decision information assistance system considers the problem that a traditional vehicle is prone to inducing traffic accidents due to a detection blind area, performs real-time detection of a roadbed traffic unit, combines vehicle collision time, a vehicle head time interval, road conditions and other information, and performs decision of a lane-changing behavior based on a fuzzy rule. When a vehicle is faulty or performs a lane-changing behavior, the roadbed traffic unit and other traffic participants can be informed through message queue telemetry transmission (MQTT). Through the road condition information assistance based on MQTT, the safety of the car-following and lane-changing behaviors of the intelligent vehicle is effectively improved, and the probability of road traffic accidents induced by the detection blind area of the intelligent vehicle is reduced.

Description

Technical Field

[0001] This invention belongs to the field of vehicle control and communication, and relates to an intelligent vehicle information assistance method based on vehicle-road cooperative technology. It improves the following and lane-changing behavior of vehicles through information interaction between intelligent vehicles and roadbed traffic units. Background Technology

[0002] China is the world's largest single automobile market. With the continuous increase in the number of cars in China, various traffic problems have arisen, making the establishment of a more efficient and convenient transportation system an important research topic. Improving the efficiency of the transportation system is closely related to the improvement of road infrastructure and the enhancement of vehicle intelligence systems.

[0003] Lane changing and car-following technologies are crucial components of intelligent driving. Effective control relies on accurate perception information. Simply using the vehicle's own sensors may result in insufficient perception and blind spots, and significantly increase the cost of individual vehicles. This invention proposes an information-assisted method based on roadbed traffic units. This method sends the detection information from the roadbed traffic units to the cloud, where vehicles can subscribe, improving the vehicle's perception accuracy of the driving environment and enhancing the control of lane changing and car-following behaviors.

[0004] Vehicle following technology is a crucial component of platooning, and numerous studies have focused on this technology in recent years. These studies have employed various approaches, including PID control, adaptive control, sliding mode control, model predictive control, and deep learning. However, due to limitations in perception technology, the practical application of platooning theory remains constrained. For instance, in platooning studies, while theoretically feasible to achieve smooth following by considering the acceleration information of the two vehicles ahead, practical application is hampered by limitations in perceiving the state information of the vehicles ahead.

[0005] MQTT, or Message Queuing Telemetry Transport Protocol, is a publish-subscribe receiving protocol under the ISO standard. It is designed to improve the performance of network equipment hardware and networks, and has the advantages of easy implementation, low power consumption, and easy data transmission. It can achieve decoupling of data in space and time.

[0006] With the continuous development of intelligent transportation technology, vehicle-road cooperative technology has received increasing attention from society. Vehicle-road cooperative technology uses advanced wireless communication and Internet technology to implement dynamic information interaction between vehicles and roads in all aspects, and conducts road cooperative management based on all-time and space dynamic traffic information collection and fusion technology. This not only reduces the occurrence of traffic accidents, but also forms an efficient and environmentally friendly road traffic system. Summary of the Invention

[0007] The purpose of this invention is to provide a cloud-based publishing-subscription method for road traffic information based on the MQTT communication protocol, which helps intelligent vehicles to achieve deep traffic situational awareness and good lane changing and car-following control. To this end, an intelligent vehicle information assistance method based on vehicle-road cooperative technology is provided.

[0008] The present invention achieves the above-mentioned technical objectives through the following technical means.

[0009] The intelligent vehicle information assistance method based on vehicle-road cooperative technology proposed in this invention includes the following steps:

[0010] An information-assisted driving method based on vehicle-road cooperative technology includes the following steps:

[0011] (1) First, the roadbed traffic unit performs real-time traffic information detection on intelligent vehicles within the road segment; the detection information X can be represented as:

[0012] X = [a1, v1, x1, a2, v2, x2, ... a n ,v n ,x n (6)

[0013] In the formula, a n v n x n Let n represent the acceleration, velocity, and displacement of the target vehicle n, respectively.

[0014] (2) The roadbed traffic unit will transmit the detected information through 4G communication in a publish-subscribe mode based on the MQTT communication protocol;

[0015] (3) Vehicles traveling on this section of road that need to choose to follow or change lanes can subscribe to this traffic information to improve their driving experience.

[0016] (4) Vehicles following another vehicle can subscribe to the acceleration a of the first vehicle in front. LV1 Speed ​​v LV1 Displacement x LV1 Information and the acceleration of the second car ahead (a) LV2 Speed ​​v LV2 Displacement x LV2 To comprehensively solve for its own acceleration a hv Improving car-following behavior can be expressed as

[0017] a hv =f(a LV1 ,v LV1 ,x LV1 ,a LV2 ,v LV2 ,xLV2 (7)

[0018] This formula indicates the target vehicle's acceleration a hv It is a control variable related to the state variables of the two vehicles ahead. The specific calculation expression is shown in Formula 8.

[0019] a hv =k1(x LV1 -x hv -D ades )+k2(v LV1 -v hv )+k3*a LV1 +k4*a LV2 (8)

[0020] Where x hv For the displacement of this vehicle, v hv For the vehicle's speed, a hv For the acceleration of this vehicle, D ades The ideal distance between following vehicles is given by the formula, where k1, k2, k3, and k4 are the control parameters of the controller.

[0021] (5) Vehicles that are changing lanes can subscribe to receive information on vehicles in front and behind the target lane and the status information of vehicles in front of the original lane to make a comprehensive judgment on whether to change lanes; a lane change decision model is established with the time to collision (TTC), headway (THW) and the vehicle condition in the target lane as inputs to make a comprehensive judgment on whether to change lanes.

[0022] ξ=f(T TTC ,T THW ,γ) (9)

[0023] st:γ∈[0,1]ξ∈[0,1] (10)

[0024] Where T TTC Indicates the time of vehicle collision, T THW The distance to the front of the vehicle is indicated by ξ; γ represents the traffic condition; Equation 9 shows that the lane-changing behavior ξ is related to T. TTC T THW ξ is a quantity related to γ, specifically in the form of the two-level fuzzy lane-changing decision system mentioned later. st: γ is a constraint condition, where γ is between 0.5 and 1, indicating normal traffic conditions and lane changing is allowed, and between 0 and 0.5, indicating poor traffic conditions in the target lane and lane changing is not recommended; ξ represents the lane-changing decision behavior, with a value range of 0 to 1, where 0 to 0.5 indicates no lane changing behavior, and 0.5 to 1 indicates lane changing behavior.

[0025] Furthermore, before making lane-changing or following actions, the vehicle subscribes to information transmitted by the connected traffic unit of that road segment and combines it with road condition information to assist in making lane-changing and following actions.

[0026] Furthermore, the vehicle is equipped with millimeter-wave radar, which can detect the distance between the vehicle and the vehicle in front and the speed of the vehicle in front in real time; the headway is calculated from the distance between the two vehicles and the vehicle's speed, and the collision time is calculated from the distance between the two vehicles and the relative speed between the two vehicles; to avoid blind spots in sensor detection, road traffic status information is provided by the road network; the intelligent vehicle will comprehensively judge whether to change lanes based on the headway, the collision time, and the road traffic status information; the road network is equipped with sensors that can detect the displacement, speed, and acceleration information of vehicles in the road segment in real time, and the road network will update and transmit the detected road traffic scene and traffic condition information to each vehicle in real time.

[0027] Furthermore, the roadbed traffic unit and the intelligent vehicle act as publishers and receivers to each other. That is, the roadbed traffic unit can send the detected traffic information to the cloud via 4G communication for the intelligent vehicle to receive, while the intelligent vehicle can also act as a publisher to send information such as lane change, following, and fault to the road network to remind other traffic participants.

[0028] Furthermore, the lane-changing decision model takes TTC, THW, and road traffic conditions as input information and makes judgments on lane-changing behavior by formulating reasonable fuzzy rules. If the target lane has good traffic and lane changing will not cause congestion or traffic accidents, the vehicle can change lanes. The fuzzy decision system uses TTC, THW, and the safety of lane changing in the target lane as conditions to make a comprehensive decision on lane-changing behavior.

[0029] Compared with the prior art, the beneficial effects of the present invention are:

[0030] 1. By adopting advanced wireless communication and Internet technologies, we can implement dynamic information interaction between vehicles and roads in all aspects, and carry out collaborative road management based on dynamic traffic information collection and fusion technology in all time and space. This not only reduces the occurrence of traffic accidents, but also forms an efficient and environmentally friendly road traffic system.

[0031] 2. Cloud-based publishing and subscription of road traffic information based on the MQTT communication protocol helps intelligent vehicles to achieve deep traffic situational awareness and better lane changing and car-following control;

[0032] 3. When following another vehicle, the speed and acceleration of the vehicle in front are often considered. If only the vehicle's built-in sensors are used for real-time detection, its following performance is greatly limited. If a vehicle in the queue brakes suddenly, the vehicle may not have enough time to react, potentially causing a chain-reaction collision. Using roadbed traffic units as an information aid can provide a better understanding of traffic conditions and enable early warnings. Attached Figure Description

[0033] Figure 1 This is a schematic diagram of the roadbed traffic unit operation.

[0034] Figure 2 This is a schematic diagram of intelligent vehicle assisted driving.

[0035] Figure 3 This is a schematic diagram of roadbed traffic unit—vehicle information transmission.

[0036] Figure 4 This is a target detection map of roadbed traffic units.

[0037] Figure 5 It is a fuzzy lane-changing decision-making flowchart.

[0038] Figure 6 This is a picture of a car in front braking to a stop in an emergency.

[0039] Figure 7 It is a two-level fuzzy lane-changing structure diagram.

[0040] Figure 8 This is a scene depicting a lane change.

[0041] Figure 9 It is the input-output membership curve.

[0042] Figure 10 This is a diagram of the auxiliary system structure. Detailed Implementation

[0043] The present invention will be further described below with reference to the accompanying drawings:

[0044] 1. This invention relates to an intelligent vehicle information assistance method based on vehicle-road cooperative technology, the method comprising an information prompting module, a following-driving assistance module, and a lane-changing decision assistance module. For example... Figure 1 The diagram shows the operation of a roadbed traffic unit. First, the roadbed traffic unit monitors the vehicle status information within the road segment in real time, such as... Figure 4 The diagram shows a detection schematic of a roadbed traffic unit, which transmits the detected status information in real time via the MQTT communication protocol in a publish-subscribe mode. Furthermore, vehicles requiring lane changing or following can subscribe to receive this traffic information, such as... Figure 2 As shown, intelligent vehicles receive detection information from roadside traffic units as needed and use it to assist driving behavior. Finally, the vehicle will use the subscribed status information and its own detection information to perform lane-changing or car-following control, such as... Figure 3 As shown, by using information detection from roadbed traffic units and intelligent vehicle-assisted driving, lane-changing and following behaviors can be improved.

[0045] 2. This invention relates to an intelligent vehicle information assistance method based on vehicle-road cooperative technology. The method includes an information prompting module, a following-driving assistance module, and a lane-change decision assistance module. The main function of the information prompting module is to use roadbed traffic units to detect vehicle information in real time within the target road segment and establish a data connection with other driving assistance modules via 4G using the MQTT communication protocol. The main function of the following-driving assistance module is to establish a data connection with the information prompting module through 4G communication and calculate the following-driving acceleration based on the received road condition information, performing early braking in emergency situations. The main function of the lane-change decision assistance module is to obtain road condition detection information from the information prompting module through 4G communication and determine whether a lane-change action is permissible based on the headway, vehicle collision time, and road traffic conditions.

[0046] Real-time information detection is performed using roadbed traffic units, and the detected road status information is sent to vehicles within the road segment in real time in a publisher mode. The following will describe the information transmission based on MQTT.

[0047] (1) The roadbed traffic unit will first detect the vehicle status information in the road section in real time. The detection information includes the acceleration, speed and displacement of each vehicle.

[0048] (2) The roadbed transportation unit uses MQTT to connect to the Internet of Things platform, authenticates itself as a roadbed device, calls the API to connect to the Internet of Things platform, and establishes a connection between the roadbed transportation unit and the cloud platform.

[0049] (3) After the MQTT connection is established, the roadbed traffic unit will send the detected traffic status information to the cloud platform in the mode of publisher.

[0050] (4) Vehicles that need to follow or change lanes can receive information published on the cloud platform by the roadbed traffic unit through the subscriber mode.

[0051] (5) Following vehicles or lane-changing vehicles maintain data transmission with roadbed traffic units via MQTT and use the received information to assist in lane changing and following, thereby improving driving safety performance.

[0052] (6) When the vehicle ends its following or completes its lane change, it can cancel its subscription to disconnect the information connection between the intelligent vehicle and the cloud platform and roadbed traffic unit.

[0053] 3. The vehicle does not need to receive this information while driving freely. When the intelligent vehicle chooses to follow or change lanes, it will receive the information in receiver mode and optimize and improve the following and lane-changing behavior by using the received road condition information and the vehicle's own detection information, thereby improving the vehicle's driving safety.

[0054] The following sections will describe the car-following and lane-changing models separately:

[0055] (1) When a vehicle selects to follow another vehicle, a normal vehicle can only detect the status information of the vehicle in front during cruise control. Due to the obstruction of the vehicle in front, the sensor cannot continue to detect information about more vehicles in front. For example Figure 6 The image shows a scenario where the vehicle in front brakes suddenly. The controlled vehicle is the last vehicle in the right-hand lane. Assuming there are two vehicles in front of this vehicle, when a pedestrian crosses the road or an obstacle suddenly appears, the vehicle in front of this vehicle may brake suddenly to avoid the pedestrian or obstacle and ensure driving safety. If the vehicle behind fails to observe the road conditions, it may trigger a chain-reaction rear-end collision.

[0056] When the strategy employed in this invention is used, the roadbed traffic unit performs real-time detection of all traffic participants in the road segment; such as... Figure 6 As shown, when a vehicle brakes suddenly to avoid a pedestrian or due to a malfunction, the roadbed traffic unit will send traffic information in publisher mode. Even if following vehicles have obstructed vision or blind spots, they can receive timely warnings, expanding the intelligent vehicle's reaction range and allowing it to slow down in advance to avoid chain-reaction collisions.

[0057] As shown in Formula 11, the acceleration a of this vehicle hv It is a control variable related to the state of the vehicle in front and the vehicle ahead of it. Where a LV1 v LV1 x LV1 These are the acceleration, velocity, and displacement of the vehicle in front, a. LV2 v LV2 x LV2 These are the acceleration, velocity, and displacement of the vehicle in front.

[0058] a hv =f(a LV1 ,v LV1 ,x LV1 ,a LV2 ,v LV2 ,x LV2 (11)

[0059] Based on Equation 11 and the actual driving environment, a simple car-following model is constructed, as shown in Equation 12, which comprehensively considers the state information of the vehicle in front and the vehicle ahead of it. Where x hv For the displacement of this vehicle, v hv For the vehicle's speed, a hv For the acceleration of this vehicle, D adesThe ideal following distance is given by the formula, where k1, k2, k3, and k4 are the control parameters of the controller. By comprehensively considering the state information of the two vehicles in front, the following behavior can not only improve the following stability, but also avoid chain-reaction rear-end collisions when the vehicles in front brake suddenly, thus improving driving safety.

[0060] a hv =k1(x LV1 -x hv -D ades )+k2(v LV1 -v hv )+k3*a LV1 +k4*a LV2 (12)

[0061] As shown in Formula 13, a min a max For the upper and lower limits of acceleration, v min v max These are the upper and lower speed limit constraints. By imposing constraints on acceleration and speed, the safety of the vehicle during following traffic is ensured.

[0062] st:a min ≤a hv ≤a max ,v min ≤v hv ≤v max (13)

[0063] The above-mentioned car-following design fully considers the status and behavior information of the two vehicles ahead. With the assistance of information from the roadbed traffic unit, it can be put into practice to improve the driving safety of car-following.

[0064] (2) When a vehicle needs to change lanes

[0065] The intelligent vehicle information assistance system based on vehicle-road cooperative technology involved in this invention will send road condition information to the intelligent vehicle in the form of a publisher when the vehicle wants to change lanes. The intelligent vehicle will then determine whether to change lanes based on the lane change intention and the road condition information of the target vehicle.

[0066] Based on this lane-changing objective, this paper establishes a two-level fuzzy lane-changing model using time-to-collision (TTC), headway (THW), and target road condition information as inputs. Figure 5 The diagram shows a flowchart of a fuzzy lane-changing decision. Figure 7 The diagram shown is a two-level fuzzy lane-changing structure for the two-level decision-making system designed in this paper. First, the lane-changing intention is generated by TTC and THW, and then the lane-changing intention and the target vehicle's road condition information are used to determine whether to change lanes.

[0067] This invention designs a lane-changing decision model based on fuzzy control. Figure 8 The diagram illustrates a lane-changing scenario. First, the vehicle calculates the headway and collision time based on the positions and speeds of vehicles ahead and behind, serving as fuzzy inputs for a first-level fuzzy decision, thus determining its lane-changing intention. Then, acting as a receiver, it receives information from the roadbed traffic unit and, based on its lane-changing intention and the target vehicle's road conditions, makes a comprehensive decision on whether to proceed with the lane change.

[0068] The following will describe the two-level fuzzy decision-making system designed in this invention. The first-level fuzzy decision-making system uses the vehicle collision time (TTC) and headway (THW) as fuzzy inputs, and the intelligent vehicle's lane-changing intention (Change_aspiration) as the fuzzy output. The second-level fuzzy decision-making system uses the intelligent vehicle's lane-changing intention (Change_aspiration) and road traffic conditions (Traffic_condition) as fuzzy inputs, and the lane-changing decision behavior (Change_lane) as the fuzzy output. The fuzzy universe of discourse is set as shown in Equations 14 and 15.

[0069] st:TTC∈[0,7], ΤHW∈[0,6], Change_arpiration∈[0,1](14)

[0070] st:traffic_condition∈[0,1],Change_lane∈[0,1](15)

[0071] like Figure 9 The figure shows the fuzzy membership curves of the input and output quantities of the two-level fuzzy decision system designed in this invention.

[0072] Fuzzy rules, which are the correspondences between fuzzy inputs and outputs, have a significant impact on fuzzy control systems. The rationality of the fuzzy rules, to a certain extent, affects the reliability and effectiveness of the fuzzy decision-making system. Based on this, this invention formulates fuzzy rules for the two-level fuzzy decision-making system designed in this paper. Tables 1 and 2 show the fuzzy rule correspondence tables for the first-level and second-level fuzzy decision-making systems, respectively. In the tables: S, M, and B represent the small, medium, and large membership degrees of Time to Collision (TTC) and Time to Headway (THW), respectively. In Change aspiration, C indicates the generation of lane-changing intention, and K indicates the absence of lane-changing intention. B and G represent the poor and good traffic conditions in Traffic condition, respectively. In the second-level fuzzy decision-making system, keep and change represent lane-keeping behavior and lane-changing behavior, respectively. The first column in the table represents the first variable element in parentheses, and the first row represents the second variable element in parentheses.

[0073] Table 1. Fuzzy Rules for the First-Level Fuzzy Decision System

[0074]

[0075] Table 2 Fuzzy Rules for the Second-Level Fuzzy Decision System

[0076]

[0077] The configuration of fuzzy rules fully considers the lane-changing needs of individual traffic participants and the road traffic conditions of the target lane. The strength of an individual traffic participant's lane-changing intention is measured by TTC and THW; that is, at the current lane, the smaller the TTC and THW, the smaller the distance between the current vehicle and the vehicle in front, the more dangerous the driving, and the stronger the lane-changing intention. The two-level fuzzy decision system considers the road traffic conditions at the target lane, i.e., whether the target lane has safe driving conditions for lane changing. Only when a vehicle intends to change lanes and the target lane conditions are good (G) can the vehicle normally perform a lane-changing behavior.

[0078] By configuring fuzzy inputs and fuzzy rules as described above, the information released by the roadbed traffic unit can be well integrated with its own vehicle detection information. By comprehensively considering vehicle collision time, headway, and target lane road traffic conditions, lane-changing behavior can be determined, thus better taking into account the lane-changing needs of individual road traffic participants and the potential impact of lane-changing behavior on overall traffic.

[0079] 4. When a smart vehicle malfunctions or brakes suddenly, it can send information to the roadside traffic unit via MQTT to inform other road users to be cautious. The specific steps are as follows:

[0080] (1) When a smart vehicle malfunctions or brakes suddenly, it can be authenticated via MQTT, establish a connection between the smart vehicle and the cloud platform as a traffic participant, and report information to the cloud platform.

[0081] (2) The roadbed traffic unit establishes an information connection with the cloud platform through MQTT with other intelligent vehicles, and receives information sent to the cloud platform by intelligent vehicles in a receiver mode. This enables other traffic participants to avoid the road in advance.

[0082] By fusing information from roadbed traffic units and intelligent vehicles, the probability of road traffic accidents caused by insufficient perception or blind spots of individual vehicles can be effectively reduced. With the assistance of information from roadbed traffic units, intelligent vehicles can use the status information of multiple vehicles ahead as reference information during following maneuvers. When a vehicle ahead suddenly stops, it can perform preemptive braking, reducing the probability of chain-reaction rear-end collisions and improving the smoothness of following maneuvers. By establishing a two-level fuzzy decision-making system, the lane-changing needs of individual vehicles and the traffic conditions of the target lane can be effectively considered, better ensuring the safety performance of intelligent vehicles during lane-changing operations and fully taking into account the impact of vehicle conditions in the target lane.

[0083] 5. For example Figure 10 The diagram shows the main structure of the system. The information prompting module mainly consists of roadbed traffic units, including a main control console, vehicle detection devices, and information transmission devices; this part is primarily hardware-based. Real-time detection of vehicle displacement, speed, and acceleration information within the target road segment is achieved through road monitoring equipment, and the data is sent to the CPU control console for processing. The processed data is then transmitted via 4G communication through a transceiver device. The intelligent vehicle includes a following-driving assistance and lane-changing decision assistance module, primarily built using a combination of hardware and software. The software part establishes an information connection with the roadbed traffic units via 4G communication and uses the transmitted road traffic data to calculate following-driving acceleration and lane-changing decision results. The hardware part consists of a 4G communication device capable of data transmission and reception; by sending the acceleration and lane-changing decision information calculated by the driving assistance device to the vehicle's power unit and steering unit, the safety of following-driving and lane-changing can be improved.

[0084] In summary, this method uses MQTT (Message Queuing Telemetry Transport) to send traffic information to vehicles on a road segment in a publish-subscribe format, allowing vehicles to decide whether to receive the information based on their own needs. The car-following driving assistance module addresses the issue that traditional intelligent vehicles can only refer to the driving information of a single vehicle ahead during car-following. By subscribing to the publisher's detection information, it obtains information on the behavior of multiple vehicles ahead and uses this information to assist in car-following behavior. The lane-change decision information assistance system addresses the problem that traditional vehicles are prone to traffic accidents due to blind spots. Through real-time detection by the roadbed traffic unit, combined with information such as TTC (Time to Collision), THW (Time Headway), and road conditions, it makes lane-change decisions based on fuzzy rules. When a vehicle malfunctions or needs to change lanes, the message queue can also notify the intelligent transportation unit and other traffic participants. By using MQTT-based traffic information assistance, the safety of intelligent vehicle car-following and lane-changing behaviors can be effectively improved, reducing the probability of road traffic accidents caused by intelligent vehicles' blind spots.

Claims

1. A method for information-assisted driving based on car-infrastructure integration technology, characterized in that: Comprise the following steps: (1) First, the intelligent vehicle in the road section is detected by the subunit of the roadbed traffic to obtain real-time traffic information. The detection information X is expressed as: (1); wherein a n , v n , x n denote the acceleration, velocity and displacement of the target vehicle n, respectively; (2) The roadbed traffic unit transmits the detected information in a publish-subscribe mode through 4G communication based on the MQTT communication protocol; (3) The vehicle driving in the section subscribes the traffic information if it needs to select car following or lane changing behavior to improve driving effect; (4) Vehicles following the car in front subscribe to the acceleration a of the first vehicle in front. LV1 Speed ​​v LV1 Displacement x LV1 Information and the acceleration of the second car ahead (a) LV2 Speed ​​v LV2 Displacement x LV2 To comprehensively solve for its own acceleration a hv Improved car-following behavior is represented as: (2)； The formula shows that the target vehicle acceleration a hv is a control quantity related to the state quantities of the two vehicles in front, and the specific calculation expression is shown in formula 3. (3)； wherein x hv is the displacement of the vehicle, v hv is the speed of the vehicle, a hv is the acceleration of the vehicle, D ades is the ideal vehicle distance of the following vehicle, wherein k1, k2, k3, k4 are control parameters of the controller; (5) The vehicle performing lane changing behavior subscribes and receives the information of the vehicles before and after the target lane and the state information of the front vehicle in the original lane to comprehensively judge whether to perform lane changing driving; The lane changing decision model is established with the vehicle collision time TTC, the headway THW and the target lane vehicle condition as inputs to comprehensively judge whether to perform lane changing behavior: (4)； ， (5)； where T TTC represents the vehicle collision time, T THW represents the headway time; γ represents the traffic condition, and formula 4 shows the lane changing behavior is a quantity related to T TTC , T THW , and γ, the lane changing decision model takes TTC, THW, and road traffic conditions as input information, and establishes a fuzzy lane changing model for the second-level decision, the first-level fuzzy decision takes the vehicle collision time TTC and the headway time THW as fuzzy input quantities, and takes the intelligent vehicle lane changing intention as a fuzzy output quantity; the second-level fuzzy decision takes the intelligent vehicle lane changing intention and the road traffic condition as fuzzy input quantities, and takes the lane changing decision behavior as a fuzzy output quantity; st: γ is a constraint condition, γ represents normal traffic condition, allowing lane changing, 0 to 0.5, and 0 to 0.5, indicating that the target lane traffic condition is not good, and lane changing is not recommended; is expressed as a lane changing decision behavior, the value range is 0 to 1, 0 to 0.5, indicating that no lane changing behavior is performed, and 0.5 to 1 indicates that lane changing behavior is performed; Before making lane changing behavior and car following behavior, the vehicle subscribes the information transmitted by the road section networked traffic unit and combines the road condition driving information to assist in making lane changing behavior and car following behavior; The vehicle is equipped with a millimeter wave radar to detect the distance between the vehicle and the front vehicle and the speed of the front vehicle in real time; the headway is calculated from the distance between the two vehicles and the speed of the vehicle, and the vehicle collision time is calculated from the distance between the two vehicles and the relative speed between the two vehicles; In order to avoid the detection blind area of the sensor, the road traffic state information is provided by the road network end; the intelligent vehicle will comprehensively judge whether to perform lane changing based on the headway, the vehicle collision time and the road traffic state information; the road network end is configured with a sensor to detect the displacement, speed and acceleration information of the vehicle in the section in real time, and the road network end transmits the detected road traffic scene and traffic condition information to each vehicle in real time; The roadbed traffic unit and the intelligent vehicle are each other's publisher and receiver, that is, the roadbed traffic unit can send the detected traffic information to the cloud in the form of 4G communication to wait for the intelligent vehicle to receive, and the intelligent vehicle can also act as a publisher to send lane changing, car following, fault and other information to the road network to remind other traffic participants.

Images