Lane-changing control model construction method based on vehicle interaction potential

Through the lane change control model based on the vehicle interaction potential, the collision problem of vehicle lane change behavior in the intelligent networked environment is solved. By introducing the Lennard-Jones potential function with acceleration and steering angle correction, the potential field of the vehicle under different states is analyzed. , and achieve more realistic and safe lane change decisions, making up for the sensitivity and differences of existing models.

CN118810788BActive Publication Date: 2024-12-06SHANDONG TRANSPORT VOCATIONAL COLLEGE
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
CN202411305402.2
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-19
Publication Date
2024-12-06
Estimated Expiration
2044-09-19

AI Technical Summary

Technical Problem

In an intelligent networked environment, when the vehicle switches lanes, collisions are prone to occur due to the differences between the actual vehicle and the model. The existing molecular interaction potential model ignores the differences in the order of magnitude of the vehicle's mass and speed, resulting in model sensitivity High, it is difficult to implement a simple and applicable lane-changing decision algorithm.

Method used

Using a lane change control model based on the vehicle interaction potential, by constructing a plane coordinate system, the acceptable gap between the lane change vehicle and the surrounding vehicles is calculated, acceleration and steering angle correction are introduced, and the Lennard-Jones potential function is established to analyze the different vehicles. The potential field force in the motion state, formulate a lane-changing decision strategy.

Benefits of technology

By drawing the spatial distribution diagram of the potential field of the vehicle, we intuitively analyze the changes in the potential field with speed, acceleration, and steering angle, and clarify the relationship between the acceptable gap of lane change and the speed and acceleration of the interactive vehicle, which improves the safety of lane change behavior and the practical significance of decision-making , verifies the effectiveness in an intelligent connected environment.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118810788B_ABST
    Figure CN118810788B_ABST
Patent Text Reader

Abstract

The present invention discloses a method for constructing a lane-changing control model based on vehicle interaction potential, and relates to the technical field of unmanned driving path planning. The model construction method comprises the following steps: the first step is to calculate the acceptable gap between the vehicle M and the surrounding vehicles for lane changing; the second step is to construct a plane coordinate system, and judge whether there is a lane-changing trend according to the potential field force of the same lane received by the vehicle M; the third step is to judge whether to change lanes according to the potential field force of the target lane received by the vehicle M; the fourth step is that the vehicle M accelerates to change lanes under the force of the surrounding vehicles; the fifth step is to complete the lane-changing process. The present invention is based on the establishment of the vehicle interaction potential of the Lennard-Jones potential, and introduces the correction of the vehicle shape, speed, acceleration and steering angle. It can intuitively analyze the changes of the vehicle potential field with the vehicle speed, acceleration and steering angle, and can study the microscopic behavior of the vehicle based on molecular dynamics.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned driving path planning, and in particular to a method for constructing a lane change control model based on vehicle interaction potential. Background Art

[0002] Lane changing is one of the main behaviors of vehicle driving and one of the main causes of road traffic accidents. Compared with following behavior, lane changing is more complex. With the development of intelligent network technology, vehicles and vehicles, and vehicles and road environments can obtain driving information in real time, which puts forward higher requirements for the research of vehicle lane changing behavior. Vehicles can achieve the transmission and integration of traffic information through vehicle-to-vehicle communication and vehicle-to-road communication, and make real-time judgments and operations on driving behaviors to ensure driving efficiency and safety.

[0003] Vehicle lane changing behavior is a complex decision-making process. In the current research on lane changing behavior, lane changing behavior can be divided into three types according to the lane changing type: free lane changing, forced lane changing and cooperative lane changing; according to the order of lane changing, the lane changing process can be divided into three stages: lane changing preparation stage, lane changing execution stage and lane changing continuation stage. At present, in the research on lane changing safety conditions, a variety of lane changing models have been established. For example, the gap acceptance model is used to describe the execution of lane changing behavior, and a lane changing model for forced lane changing and judgment lane changing is established; the mutual cooperation mechanism is introduced into the lane changing process, and the gap acceptance model is established considering the differences in driver characteristics; according to the following safety of the lane changing vehicle after completing the lane changing, an acceptable gap model between the lane changing vehicle and the surrounding conflicting vehicles is established; an acceptable gap model based on different speeds and different driver types; based on the mutual interference of vehicles between multiple lanes, a neighboring vehicle influence model is established; from the perspective of molecular dynamics, the system considers the following demand safety characteristics to ensure that when the lane changing is completed, the lane changing vehicle and the vehicle behind the target lane can follow the vehicle at the required safety distance, and an expected safety spacing model simulating molecular dynamics is established.

[0004] At present, the earliest research on potential field proposed a mobile robot trajectory planning method based on the concept of artificial potential field. With the development of intelligent transportation, many scholars have extended the potential field theory to the field of traffic flow. For example, a real-time autonomous vehicle controller based on machine learning was proposed based on the gravitational field and repulsive field; corresponding potential field models were established for different objects based on the artificial potential field theory; the existing safety potential field model was improved by introducing acceleration parameters, and a safety potential field lane changing model was established; on the basis of objective quantification of lane changing intention, a molecular dynamics model based on molecular interaction potential was established to establish lane changing decision behavior.

[0005] Regarding the data sets used in the research, the data used can be roughly divided into three types, namely natural data sets, video observation data, and simulation test data. For example, NGSIM data was used to investigate the gap acceptance behavior in lane changing situations on highways; five general light vehicles equipped with SHRP2 NextGen data acquisition systems were used to collect real driving data, which were then used for empirical analysis; and some scholars, due to limitations in equipment conditions, used simulation experimental platforms, including Prescan software, to build models of traffic scenarios, vehicle control systems, etc.

[0006] Whether it is the study of safety distance or the study of safety potential field, they are all related to the safety of lane changing. Therefore, in the study of lane changing in the intelligent network environment, the influence of the motion parameters of the interactive vehicles in the surrounding environment of the lane changing vehicle on the lane changing safety should be fully considered, and the safety risks that the vehicle may cause to the lane changing vehicle under different motion states should be analyzed, so as to find a lane changing decision algorithm with a simple model and strong applicability to avoid the occurrence of safety accidents. Summary of the invention

[0007] When the above-mentioned existing molecular interaction potential is applied to the lane-changing behavior of vehicles in a connected traffic environment, there is a large difference between the lane-changing behavior of actual vehicles, which leads to the problem that vehicles are prone to collision when changing lanes.

[0008] In order to solve the above technical problems, the present invention adopts the following technical solutions: a method for constructing a lane-changing control model based on vehicle interaction potential comprises the following steps: S1: a lane-changing vehicle M follows a front vehicle L in the original lane in the original lane, and continuously calculates the lane-changing acceptable gaps between the lane-changing vehicle M and the front vehicle L in the original lane on the current lane, as well as the front vehicle TF in the target lane and the rear vehicle TL in the target lane; S2: a plane coordinate system is constructed to determine whether there is a lane-changing trend according to the potential field force of the front vehicle L in the original lane exerted on the lane-changing vehicle M, and if there is a lane-changing trend, the method enters S3, otherwise it enters S1; S3: in the case of a lane-changing trend, the method determines whether to change lanes according to the potential field force of the rear vehicle TL in the target lane and the front vehicle TF in the target lane exerted on the lane-changing vehicle M, and if there is a lane-changing trend, the method enters S4, otherwise it enters S2; S4: when changing lanes, the lane-changing vehicle M is subjected to the force of the rear vehicle TL in the target lane, the front vehicle TF in the target lane and the front vehicle L in the original lane, and the lane-changing vehicle M accelerates at an acceleration of |a x |Accelerate to change lanes to the target lane, acceleration|a x |Satisfy the relationship:

[0009]

[0010] In the formula, x is the vertical direction of the vehicle, defined as the x-axis, and a xis the acceleration of vehicle M in the x-axis direction, F x is the component of the force in the x-axis direction, m is the mass of vehicle M;

[0011] S5 completes the lane change process.

[0012] Furthermore, in S1, the acceptable gap between M and TL for lane change is expressed as:

[0013]

[0014] In the formula, is the acceptable gap between M and TL for lane change, v TL is the initial velocity of TL, v M is the initial lane-changing speed of M, a M is the lane change acceleration of M, a TL is the acceleration of TL, is the critical longitudinal distance between M and TL after the lane change, w is the width of M, θ is the angle between M and the horizontal direction of the road, t0, c v 、c a is the coefficient to be determined.

[0015] Furthermore, in S1, the acceptable gap expression for lane change between M and TF is:

[0016]

[0017] In the formula, is the acceptable gap between M and TF for lane change, v TF is the initial velocity of TF, α TF is the acceleration of TF, It is the critical longitudinal distance between M and TF after the lane change is completed.

[0018] Furthermore, in S1, the acceptable gap expression for lane change between M and L is:

[0019]

[0020] In the formula, is the acceptable gap between M and L for lane change, v L is the initial velocity of L, j is the angular collision state, t j The time it takes for M and L to show the illustrated state when they collide with each other, and the value is between 1.8s and 3s. It is the critical longitudinal distance between M and L after the lane change is completed.

[0021] Furthermore, in S2, a plane coordinate system is constructed to determine whether F y M-L ≥0, where y is the direction of the vehicle, defined as the y-axis, Fy M-L is the potential field force generated by L on M in the y-axis direction;

[0022] If the conditions are met, the system goes into S1 and M continues to follow L. If the conditions are not met, the system goes into S3.

[0023] Furthermore, in S3, when F y M-TL <0 or F y M-TF >0, indicating that M is subject to the longitudinal repulsion of TL or TF, and it is not recommended to change lanes and enter S2, where F y M-TL is the potential force generated by TL on M in the y-axis direction, F y M-TF is the potential field force generated by TF on M in the y-axis direction;

[0024] When F y M-TL ≥0 and F y M-TF When ≤0, it means that TL and TF exert gravitational force on M, and it is recommended to change lanes and enter S4;

[0025] In S4, the sum of the lateral gravitational forces on M is |F x M-TF |+|F x M-TL |, prompting M to accelerate into the target lane, where x is the vertical direction of the vehicle, defined as the x-axis, F x M-TF is the potential force generated by TF on M in the x-axis direction, F x M-TL is the potential field force generated by TL on M in the x-axis direction;

[0026] As M turns, L generates a lateral repulsive force on M, forcing M to accelerate laterally away from the original lane. At this time, the lateral force on M is equal to |F x |For:

[0027] |F x |=|F x M-L |+|F x M-TF |+|F x M-TL |;

[0028] In the formula, F x M-L is the potential field force generated by L on M in the x-axis direction.

[0029] Furthermore, in S2, the molecular interaction potential function between following vehicles is:

[0030]

[0031] Where U is the molecular interaction potential function, ε is the potential well depth, and X r The safety distance required for following is L, which is the distance between the front and rear vehicles;

[0032] The powers of the attractive and repulsive terms in the molecular interaction potential model are set as undetermined parameters, and the vehicle interaction potential function is constructed:

[0033]

[0034] In the formula, U is the vehicle interaction potential function, v and u are the powers of the attraction term and repulsion term, and v is greater than u;

[0035] Introducing distance, velocity, and acceleration for correction, we get the vehicle interaction potential function:

[0036]

[0037] Where a is the acceleration of the vehicle, d is * is the corrected distance, φ is the clockwise angle between the vehicle acceleration and the line connecting the centers of mass of the two vehicles, γ is an undetermined parameter, (x0, y0) is the coordinate position of the center of mass of the current vehicle, (x, y) is the center of mass position of any surrounding vehicle, and s0 is the acceptable parking gap;

[0038] Introducing the steering angle for correction, we get:

[0039]

[0040] In the formula, (x * ,y * ) is the corrected deflection angle coordinate;

[0041] The vehicle interaction potential model for lane changing behavior is:

[0042]

[0043] Furthermore, the potential of TL on M is:

[0044]

[0045] Where U M-TL is the potential generated by TL on M, L M-TL is the distance between TL and M, β is the angle between the center of mass of TL and M and the y-axis, d * is the corrected distance, γ is an undetermined parameter, (x TL ,y TL ) is the coordinate position of the center of mass of TL;

[0046] make The potential force generated by TL on M:

[0047]

[0048] make but:

[0049]

[0050] In the formula, η is a substitution variable, is the distance between TL and M raised to the power of u+1, It is the distance between TL and M raised to the power of v+1.

[0051] Furthermore, the potential field force generated by TF on M is:

[0052]

[0053] In the formula, is the distance between TF and M raised to the power of u+1, is the v+1th power of the distance between TF and M, and α is the angle between the line connecting the centroids of TF and M and the y-axis.

[0054] Furthermore, the potential field force generated by L on M is:

[0055]

[0056] In the formula, is the distance between L and M raised to the power of u+1, It is the distance between L and M raised to the power of v+1, and ξ is the angle between the line connecting the centers of mass of L and M and the y-axis.

[0057] The beneficial effects of the present invention are as follows: the present invention establishes the vehicle interaction potential based on the Lennard-Jones potential, introduces the correction of the vehicle shape, speed, acceleration and steering angle, and makes up for the fact that the previous research on the molecular lane-changing model directly uses the Lennard-Jones potential function as a prototype and ignores the huge difference in the magnitude of parameters such as mass and speed between particles and vehicles. In addition, the present invention can intuitively analyze the changes of the vehicle potential field with the vehicle speed, acceleration and steering angle by drawing the spatial distribution diagram of the vehicle potential field under different motion states, making the research on the microscopic behavior of the vehicle based on molecular dynamics more realistic.

[0058] In the previous research on molecular dynamics lane changing model, for the setting of longitudinal critical distance after lane changing, only the speed and acceleration of the current vehicle were considered, and the influence of the running state of the front vehicle in the original lane on the critical safety distance after lane changing was ignored. The present invention introduces the acceleration of the front vehicle in the original lane into the longitudinal critical distance, and through the analysis of the lane changing safety critical time, clarifies the relationship between the acceptable gap for lane changing and the speed difference and acceleration difference of the interacting vehicles, and constructs the expressions for the acceptable gaps for lane changing between the lane changing vehicle M and the front vehicle L in the original lane, the front vehicle TL in the target lane, and the rear vehicle TF in the target lane, respectively. Through the simulation analysis of different motion states of the interacting vehicles, reasonable suggestions are put forward for lane changing behavior in a networked environment.

[0059] The present invention establishes a lane-changing model based on vehicle interaction potential, analyzes the potential and potential field force generated by the surrounding interactive vehicles on the lane-changing vehicle, and gives the driving decision of vehicle M under different potential field forces. By comparing with the lane-changing model based on molecular dynamics and the SL2015 model, its effectiveness is verified, which has theoretical guiding significance for the vehicle lane-changing behavior decision in the future intelligent network environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 Shown is a schematic diagram of the potential field of the initial stationary vehicle.

[0061] Figure 2 Shown is a schematic diagram of the road environment.

[0062] Figure 3 The following is a schematic diagram of the lane change scenario setting.

[0063] Figure 4 The figure shows a schematic diagram of the lane changing process between vehicle M and vehicle TL.

[0064] Figure 5 The figure shows a schematic diagram of the lane changing process between vehicle M and vehicle TF.

[0065] Figure 6 The figure shows a schematic diagram of the lane changing process of vehicle M and vehicle L.

[0066] Figure 7 Figure 2 shows the force analysis of a lane-changing vehicle M.

[0067] Figure 8 Shown is the acceptable gap for lane change between vehicle M and vehicle TL.

[0068] Fig. 9 Shown is the acceptable gap for lane change between vehicle M and vehicle TF.

[0069] Fig.10 The acceptable gap between vehicle M and vehicle L for lane change is shown.

[0070] Fig.11 Shown is the lane-changing experiment simulation interface.

[0071] Fig.12 The figure shows the change of the average speed of traffic.

[0072] Fig.13 Shown is a schematic diagram of the changes in the average acceleration of traffic flow.

[0073] Fig.14 Shown is a schematic diagram of the average speed comparison of the vehicle potential lane changing model, molecular dynamics lane changing model and SL2015 lane changing model. DETAILED DESCRIPTION

[0074] The lane-changing control model construction method based on vehicle interaction potential includes the following steps:

[0075] S1: the lane-changing vehicle M follows the vehicle L in the original lane, and continuously calculates the acceptable gaps for lane change between the lane-changing vehicle M and the vehicle L in the original lane in the current lane, as well as the target lane vehicle TF and the target lane vehicle TL in the target lane;

[0076] S2 constructs a plane coordinate system and determines whether the lane-changing vehicle M has a lane-changing tendency according to the potential field force of the vehicle L in the original lane. If so, it proceeds to S3, otherwise it proceeds to S1;

[0077] S3: In the case of a lane-changing tendency, whether to change lanes is determined based on the potential field forces of the vehicle TL behind the target lane and the vehicle TF in front of the target lane on the lane-changing vehicle M. If the lane-changing is to be carried out, the process proceeds to S4; otherwise, the process proceeds to S2.

[0078] S4 When changing lanes, the lane-changing vehicle M is subjected to the forces of the vehicle behind the target lane TL, the vehicle in front of the target lane TF, and the vehicle in front of the original lane L. The lane-changing vehicle M accelerates at an acceleration of |a x |Accelerate to change lanes to the target lane, acceleration|a x |Satisfy the relationship:

[0079]

[0080] In the formula, x is the vertical direction of the vehicle, defined as the x-axis (the same below), a x is the acceleration of vehicle M in the x-axis direction, F x is the component of the force in the x-axis direction, m is the mass of vehicle M;

[0081] S5 completes the lane change process.

[0082] Furthermore, in S1, the acceptable gap between M and TL for lane change is expressed as:

[0083]

[0084] In the formula, is the acceptable gap between M and TL for lane change, v TL is the initial velocity of TL, v M is the initial lane-changing speed of M, a M is the lane change acceleration of M, a TL is the acceleration of TL, is the critical longitudinal distance between M and TL after the lane change, w is the width of M, θ is the angle between M and the horizontal direction of the road, t0, c v 、c a is the coefficient to be determined.

[0085] Furthermore, in S1, the acceptable gap expression for lane change between M and TF is:

[0086]

[0087] In the formula, is the acceptable gap between M and TF for lane change, v TF is the initial velocity of TF, α TF is the acceleration of TF, It is the critical longitudinal distance between M and TF after the lane change is completed.

[0088] Furthermore, in S1, the acceptable gap expression for lane change between M and L is:

[0089]

[0090] In the formula, is the acceptable gap between M and L for lane change, v L is the initial velocity of L, j is the collision state, t j The time it takes for M and L to show the illustrated state when they collide with each other, and the value is between 1.8s and 3s. It is the critical longitudinal distance between M and L after the lane change is completed.

[0091] Furthermore, in S2, a plane coordinate system is constructed with the direction perpendicular to the road as the x-axis and the direction in which the vehicle moves forward as the y-axis to determine whether F y M-L ≥0, where y is the direction of the vehicle, defined as the y-axis, F y M-L is the potential field force generated by L on M in the y-axis direction;

[0092] If the conditions are met, the system goes into S1 and M continues to follow L. If the conditions are not met, the system goes into S3.

[0093] Furthermore, in S3, when F yM-TL <0 or F y M-TF >0, indicating that M is subject to the longitudinal repulsion of TL or TF, and it is not recommended to change lanes and enter S2, where F y M-TL is the potential force generated by TL on M in the y-axis direction, F y M-TF is the potential field force generated by TF on M in the y-axis direction;

[0094] When F y M-TL ≥0 and F y M-TF When ≤0, it means that TL and TF exert gravitational force on M, and it is recommended to change lanes and enter S4;

[0095] In S4, the sum of the lateral gravitational forces on M is |F x M-TF |+|F x M-TL |, prompting M to accelerate into the target lane, where x is the direction perpendicular to the road, defined as the x-axis, F x M-TF is the potential force generated by TF on M in the x-axis direction, F x M-TL is the potential field force generated by TL on M in the x-axis direction;

[0096] As M turns, L generates a lateral repulsive force on M, forcing M to accelerate laterally away from the original lane. At this time, the lateral force on M is equal to |F x |For:

[0097] |F x |=|F x M-L |+|F x M,-TF |+|F x M-TL |;

[0098] In the formula, F x M-L is the potential field force generated by L on M in the x-axis direction.

[0099] Furthermore, in S2, the molecular interaction potential function between following vehicles is:

[0100]

[0101] Where U is the molecular interaction potential function, ε is the potential well depth, and X r The safety distance required for following is L, which is the distance between the front and rear vehicles;

[0102] The powers of the attractive and repulsive terms in the molecular interaction potential model are set as undetermined parameters, and the vehicle interaction potential function is constructed:

[0103]

[0104] In the formula, U is the vehicle interaction potential function, v and u are the powers of the attraction term and repulsion term, and v is greater than u;

[0105] Introducing distance, velocity, and acceleration for correction, we get the vehicle interaction potential function:

[0106]

[0107] Where a is the acceleration of the vehicle, d is * is the corrected distance, φ is the clockwise angle between the vehicle acceleration and the line connecting the centers of mass of the two vehicles, γ is an undetermined parameter, (x0, y0) is the coordinate position of the center of mass of the current vehicle, (x, y) is the center of mass position of any surrounding vehicle, and s0 is the acceptable parking gap;

[0108] Introducing the steering angle for correction, we get:

[0109]

[0110] In the formula, (x * ,y * ) is the corrected deflection angle coordinate;

[0111] The vehicle interaction potential model for lane changing behavior is:

[0112]

[0113] Furthermore, the potential of TL on M is:

[0114]

[0115] Where U M-TL is the potential generated by TL on M, L M-TL is the distance between TL and M, β is the angle between the center of mass of TL and M and the y-axis, d * is the corrected distance, γ is an undetermined parameter, (x TL ,y TL ) is the coordinate position of the center of mass of TL;

[0116] make The potential force generated by TL on M:

[0117]

[0118] make but:

[0119]

[0120] In the formula, η is a substitution variable, is the (u + 1)-th power of the distance between TL and M, is the (v + 1)-th power of the distance between TL and M.

[0121] Furthermore, the potential field force generated by TF on M:

[0122]

[0123] In the formula, is the (u + 1)-th power of the distance between TF and M, is the (v + 1)-th power of the distance between TF and M, and α is the angle between the line connecting the centers of mass of TF and M and the y-axis.

[0124] Furthermore, the potential field force generated by L on M:

[0125]

[0126] In the formula, is the (u + 1)-th power of the distance between L and M, is the (v + 1)-th power of the distance between L and M, and ξ is the angle between the line connecting the centers of mass of L and M and the y-axis.

[0127] The intermolecular force that exists between molecules refers to the attractive and repulsive interactions between molecules, which is the main reason for the different physical properties of molecules. Denote the zero distance of the resultant force of molecular attraction and repulsion as the equilibrium distance r0. When r > r0, the intermolecular force as a whole shows an attractive force; when r < r0, the intermolecular force as a whole shows a repulsive force.

[0128] Microscopically, the movement of vehicle flow is reduced to the particle level. It is like the current particle is pushed by the particle behind and pulled by the particle in front and moves forward. It is not easy for vehicles to get closer or farther away from each other. From the perspective of dynamics, it is moving under the combined action of attractive and repulsive forces. The interaction relationship between vehicles driving on the road has similar characteristics to the mechanical relationship between molecules. In this paper, the microscopic vehicles in traffic flow are compared to molecules, and a vehicle lane-changing behavior model based on molecular dynamics is constructed.

[0129] In the modeling research based on molecular dynamics, the Lennaed-Jones molecular interaction potential function is widely used. Based on the Lennrd-Jones molecular interaction potential between molecules, the molecular interaction potential function between following vehicles is established:

[0130]

[0131] In the formula, U is the molecular interaction potential function, ε is the depth of the potential well, L is the distance between the front and rear vehicles, X rA safe distance is required for following.

[0132] However, the current molecular interaction potential model is only a special form of Lennrd-Jones potential. Although it conforms to the interaction between microscopic particles in the fluid, it is directly applied to the real vehicle interaction behavior, ignoring the large differences in mass, speed and acceleration between the two scenes. The model causes the vehicle to be highly sensitive to various parameters. Therefore, it is necessary to improve it, set the power of the gravitational term and the repulsive term in the original model as undetermined parameters, and construct the vehicle interaction potential function:

[0133]

[0134] Where U is the vehicle interaction potential function, ε is the potential well depth, L is the distance between the front and rear vehicles, and X r To maintain a safe distance for following, v and u are the powers of the gravitational term and the repulsive term, and v is greater than u.

[0135] Research shows that when the actual distance between two vehicles is less than the required safety distance, they exhibit different degrees of short-range repulsion. The closer the distance, the greater the repulsion. The schematic diagram of the potential field of the initial stationary vehicle is as follows: Figure 1 The figure shows that the risk from approaching the vehicle from any direction is the same, but this is clearly unreasonable. Figure 1 The figure shows the contour projection of the vehicle potential field, and the potential field gradually decreases from the inside to the outside. In the actual road driving environment, the potential generated by the front and rear vehicles in the same lane on the target is significantly higher than the potential generated by the vehicles in the adjacent lane approaching from the side; dynamic vehicles should show different potentials from static vehicles.

[0136] The coordinate system is established with the position of the leftmost lane line of the road as the Y axis. Figure 2 As shown, Figure 2 A schematic diagram of a road environment is shown.

[0137] Taking distance, speed, and longitudinal acceleration into account in the model modification, the vehicle potential function is obtained as follows:

[0138]

[0139] Where a is the vehicle acceleration, d is * is the corrected distance, φ is the clockwise angle between the vehicle acceleration a and the line connecting the centers of mass of the two vehicles, s0 is the acceptable parking gap, γ is an undetermined parameter, the coordinate position of the center of mass of the current vehicle is (x0, y0), and the center of mass position of any surrounding vehicle is (x, y).

[0140] The model is applicable to car-following behavior, but in the process of lane changing, the vehicle has a certain steering angle, and the spatial distribution of the vehicle potential field should also be deflected as a whole in theory. The above formula cannot well represent this state, so the steering angle is introduced to correct the formula. (x, y) is the original coordinate of any vehicle, (x * ,y * ) is the corrected deflection angle coordinate, and we get:

[0141]

[0142] Where (x, y) is the original coordinate of any vehicle, (x * ,y * ) is the corrected deflection angle coordinate, and θ is the angle between the lane-changing vehicle's running direction and the road's horizontal direction.

[0143] In summary, the vehicle interaction potential model applicable to lane changing behavior is:

[0144]

[0145] When the distance between vehicles is less than the required safety distance, it manifests as short-range repulsion.

[0146] Taking the center of mass of the research vehicle as the origin, the contour lines of the vehicle potential field in five motion states are displayed, namely, stationary, uniform speed, acceleration, deceleration, and turning. The colors represent the different sizes of the potential field. White near the center of mass indicates that the potential field approaches infinity, which means that a vehicle will inevitably collide when it appears at this position, which is in line with practical significance. The spatial distribution of the vehicle potential field is analyzed according to the above five driving states:

[0147] (1) Stationary state: When the vehicle is stationary, it is equivalent to a stationary obstacle. Since the risk of approaching the vehicle from the front or rear is greater than that of approaching the vehicle from the side, this paper corrects the distance, so the potential field is elliptical as a whole.

[0148] (2) Uniform speed state: when the vehicle is moving at a constant speed v = 10 m·s -1 and c = 15 m·s -1 When driving at a constant speed, and the driving direction is consistent with the y-axis direction, because the speed has no component on the x-axis, the potential field will not change with the change of speed in the x-axis direction, so the potential fields displayed by the two on the x-axis are consistent. In the y-axis direction, the potential field increases with the increase of speed, which also well reflects that the risk range of vehicles to interactive vehicles is greater when driving at high speed than when driving at low speed. Therefore, the overall potential field presents an ellipse elongated along the direction of vehicle movement.

[0149] (3) Acceleration state, when the speed is v = 10 m·s -1The vehicles are respectively a=1.5m·s -2 and a=2m·s -2 When accelerating along the y-axis, it can be found that the vehicle potential field has a significant forward tilt phenomenon in the accelerated motion state, which means that the risk posed by the vehicle to the vehicle in front is significantly greater than that to the vehicle behind, so the required safety distance in front of the vehicle should be greater than that to the vehicle behind. As the acceleration increases, the forward tilt of the vehicle potential field increases, indicating that the greater the acceleration of the vehicle, the greater the risk to the front.

[0150] (4) Deceleration state, when the speed is v = 10 m·s -1 The vehicles are respectively a=-1.5m·s -2 and a=-2m·s -2 When driving at a reduced speed along the y-axis, it can be found that the vehicle potential field shows a significant backward tilt phenomenon in the deceleration state, which means that the risk posed by the vehicle to the vehicle behind is significantly greater than that to the vehicle in front, so the required safety distance behind the vehicle should be greater. The degree of backward tilt of the vehicle potential field increases with the increase of the deceleration value, that is, the risk level of the vehicle behind increases with the increase of the deceleration. Therefore, in the connected environment, when the rear vehicle obtains the deceleration information of the vehicle in front, it can perform deceleration operation in advance to improve driving safety.

[0151] (5) Turning state. When the vehicle turns left under acceleration and deceleration, it can be found that when the vehicle turns, the overall vehicle potential field shifts with the steering angle. Therefore, when studying lane changing behavior, it is necessary to consider the influence of the steering angle on the vehicle potential field, which is also in line with the correction of the vehicle potential field mentioned above.

[0152] In the following behavior, the vehicle only needs to consider the vehicle potential generated by the vehicles in front and behind in the same lane. The two vehicles only need to maintain a certain following safety distance X. r You can drive safely, but when the longitudinal following repulsion is not 0, it means that the vehicle in front of the original lane has exerted pressure on the lane-changing vehicle. The lane-changing vehicle needs to slow down to maintain a safe distance from the vehicle in front. If the vehicle travels at a speed lower than the expected speed, it will have an intention to change lanes.

[0153] When conducting lane-changing research, it is necessary to analyze whether the target vehicle and the interacting vehicle meet the lane-changing safety gap. The research shows that when the vehicle meets the acceptable gap for lane-changing, the vehicle potential is the smallest. At this time, the attraction and repulsion are both zero, reaching a balanced state in the lane-changing behavior. In other words, when the acceptable gap is met, the vehicle can be guaranteed not to slow down and change lanes. Therefore, it is necessary to study the acceptable gap in lane-changing behavior below.

[0154] When the driver is eager to achieve a higher speed or more comfortable driving conditions in the adjacent lane, or to reach the destination by diverging or merging, the vehicle lane change behavior will occur. Generally speaking, researchers divide vehicle lane change behaviors into two categories: forced lane change and active lane change. When the driver must leave the original lane to reach the planned destination, the forced lane change behavior is performed; when the driver wants to achieve a faster driving speed or have a shorter queue, the active lane change is performed.

[0155] The present invention simplifies the model by selecting automatic lane-changing behavior for analysis. The vehicle's autonomous lane-changing needs to meet two conditions: one is the speed space advantage, the vehicle cannot travel at the expected speed in the lane, and the driving conditions in the adjacent lane are good at this time; the other is the lane-changing safety condition, which means that the vehicle needs to meet a certain safety distance when changing lanes to avoid traffic accidents.

[0156] During the lane change process, the vehicle will constantly interact with the adjacent vehicles in the original lane or the target lane. Figure 3 As shown in the figure, M is a lane-changing vehicle, where vehicle L is the front vehicle in the original lane, F is the rear vehicle in the original lane, TL is the front vehicle in the target lane, and TF is the rear vehicle in the target lane. In this lane-changing process, the driving safety of M is affected by the surrounding traffic conditions. The study assumes that only the four neighboring vehicles adjacent to M are considered. Among the four neighboring vehicles, F and M are in a following state. The following safety relationship is mainly determined by F, so the influence of F is not considered here. In the lane-changing process, M needs to perceive and judge the motion states of the original three vehicles L, TL, and TF, and then make a decision and perform the operation in a very short time. In this study, the distance at which the interaction potential of the interactive vehicles is zero during the lane-changing process is called the acceptable gap for lane-changing. The following focuses on the conditions for safe lane-changing between M and L, TL, and TF. To simplify the model, the following assumptions are proposed: the overall running speed of the target lane is better than that of the original lane, which meets the speed space advantage in the lane-changing conditions; the lane-changing vehicle maintains the same state during the lane-changing process; and the influence of the steering angle on the speed is ignored.

[0157] The lane changing process between M and TL is as follows Figure 4 As shown in the figure, the interaction state between M and TL during the lane change process is shown. In order to avoid a collision between M and TL during the lane change process, the two vehicles need to meet the following conditions:

[0158]

[0159] In the formula, is the acceptable gap between vehicle M and vehicle TL for lane change, S M is the distance traveled by the lane-changing vehicle M after the lane-changing is completed, S TL is the distance traveled by the front vehicle TL in the target lane after the lane change is completed, is the critical longitudinal distance between the two vehicles at the end of the lane change, w is the width of the lane-changing vehicle, and θ is the angle between the lane-changing vehicle M and the horizontal direction of the road.

[0160] In order to change lanes in a connected environment, vehicles need to obtain the operating parameters of surrounding vehicles and calculate whether the acceptable gap for lane change is met. The determination of and ΔS is particularly critical. and ΔS were studied.

[0161] Longitudinal critical distance: The longitudinal critical distance indicates that after the vehicle completes the lane change, it can safely follow the vehicle in front while meeting this distance. Therefore, it can also be explained by the following requirement safety distance.

[0162] The required safety distance for following a car is related to the vehicle's speed and reaction time.

[0163] G min =ηv+γv 2 ;

[0164] In the formula, G min is the required safe distance for following, v is the speed of the vehicle, η is the vehicle reaction time, and γ is an undetermined parameter. Although this formula can be used to study the molecular dynamics lane-changing model, it lacks consideration of the parameters related to the preceding vehicle.

[0165] The required safety distance set in the VTH following strategy is related to the current vehicle speed and the speed difference between the front and rear vehicles. The expression is:

[0166] G min =(t0-c v Δv)v;

[0167] In the formula, t0, c v is the coefficient to be determined.

[0168] In the intelligent network environment, the acquisition, processing and transmission of information have been qualitatively improved, the vehicle reaction time can be basically ignored, and the driving state parameters between vehicles can be obtained in real time. Therefore, the setting of the longitudinal critical distance should also introduce the influence of the front and rear vehicle acceleration difference, and the expression is:

[0169] G min =(t0-c v Δv-c a Δa)v;

[0170] Where Δa is the acceleration difference between the front and rear vehicles.

[0171] According to the above analysis, the longitudinal critical distance between M and TL is:

[0172]

[0173] In the formula, v TL is the initial velocity of TL, v M v is the initial speed of lane change, a M is the acceleration of M, a TL is the acceleration of TL.

[0174] Lane changing safety critical time: Lane changing safety critical time represents the time it takes for the lane changing vehicle and the interacting vehicle to reach the same speed during the lane changing process. Because when the speeds of the two vehicles reach the same speed, if there is no collision risk, there will be no possibility of collision. Since in the actual lane changing process, the running state of the front vehicle in the target lane is little affected by the lane changing vehicle and can be ignored, the TL state generally does not change during the lane changing process. Assume that the initial lane changing speed and acceleration of M are v m 、a M , the initial velocity and acceleration of TL are v TL 、a TL The distance difference between M and TL is:

[0175]

[0176] And the lane-changing safety critical time satisfies the formula:

[0177]

[0178] have to:

[0179]

[0180] Where ΔS is the distance difference between M and TL, T is the lane change safety critical time, and v TL is the initial velocity of TL, v M is the initial speed of M’s lane change, a M is the acceleration of M, a TL is the acceleration of TL.

[0181] Therefore, the acceptable gap expression for lane change between M and TL is obtained as follows:

[0182]

[0183] In the formula, is the acceptable gap between M and TL for lane change, t0, c v 、c a is the unknown coefficient, Δa is the acceleration difference between the front and rear vehicles, v TL is the initial velocity of TL, v M is the initial speed of M changing lanes, a M is the acceleration of M, a TLis the acceleration of TL, w is the width of M, and θ is the angle between M and the horizontal direction of the road.

[0184] There is a risk of collision between M and TF during the lane changing process, so it is necessary to study the interaction between the two. Figure 5 As shown in the figure, in order to avoid a collision between M and TF during the lane change process, the two vehicles must meet the following conditions, namely:

[0185]

[0186] In the formula, is the acceptable gap between M and TF for lane change, S M is the distance traveled by lane change M after the lane change is completed, S TF is the distance traveled by the target lane following vehicle TF after the lane change is completed.

[0187] Based on the above lane-changing scenario assumption, TF's speed is greater than M's. In the initial stage of lane-changing, when the distance between the two vehicles is large enough, TF can move at a constant speed or even accelerate. When the distance between the two vehicles decreases, as the following vehicle after M changes lanes, it is a forced vehicle and will decelerate due to the repulsive force of following. Assume that the initial speed and acceleration of M are v M 、a M , the initial velocity and acceleration of TF are v TF 、a TF . Also assume that the two vehicles complete the lane change when their speeds are equal, and the distance difference does not change afterwards. Let v M <v TF It can be obtained that the safety critical time of M and TF in different motion states is:

[0188]

[0189] The distance difference between the two vehicles is:

[0190]

[0191] In summary, the acceptable gap expression for lane change between M and TF can be obtained as:

[0192]

[0193] In the formula, is the acceptable gap between M and TF for lane change, v TF is the initial velocity of TF, a TF is the acceleration of TF.

[0194] There is a risk of collision between M and L during the lane changing process, so it is necessary to study their interaction state. Figure 6As shown, in order to avoid a corner collision between M and L at the beginning of the lane change, the two vehicles need to meet the following conditions, namely:

[0195]

[0196] In the formula, is the acceptable gap between M and L for lane change, S L is the distance traveled by L during the lane change process.

[0197] Since the motion state of the front vehicle L in the original lane is basically not affected by M, L may move at a constant speed, accelerate, or decelerate. Suppose the initial speed and acceleration of M are v M 、a M , the initial velocity and acceleration of TF are v L 、a L According to the generation mechanism of lane-changing intention, M generates lane-changing intention only when it is subjected to the repulsive force of L. Therefore, this paper assumes that v M >v L .

[0198] In this lane-changing scenario, the most likely collision accident is a corner collision. The time when the two vehicles are in the state shown in the figure is defined as t j , according to experience, is usually between 1.8s and 3s. If the two vehicles do not collide at this time, there will be no collision risk. At this time, the distance difference can be obtained as:

[0199]

[0200] In summary, the acceptable gap expression for lane change between M and L can be obtained as:

[0201]

[0202] In the formula, is the acceptable gap between M and L for lane change, v L is the initial velocity of L, a L is the acceleration of L, j is the angular collision state, t j It is the time when the vehicle is in corner collision state.

[0203] When a vehicle is within the acceptable clearance distance for lane change, the potential field force it is subjected to is a short-range repulsive force, which aims to prevent other vehicles from approaching to ensure safe driving. Generally speaking, the greater the potential of the vehicle, the stronger the potential field force. During the lane change process, the vehicle is affected by the potential of multiple vehicles, so it is necessary to determine the potential field force generated by each vehicle on the lane change vehicle, and then determine whether the vehicle meets the safe lane change conditions. The force analysis diagram of M is shown in the figure below. Figure 7 As shown in the figure, the potentials of TL, TF and L on M are analyzed respectively.

[0204] The potential of TL on M:

[0205]

[0206] Where U M-TL is the potential generated by TL on M, L M-TL is the distance between TL and M, β is the angle between the center of mass of TL and M and the y-axis, d * is the corrected distance, γ is an undetermined parameter, (x M ,y M ) is the coordinate position of the center of mass of M, (x TL ,y TL ) is the coordinate position of the center of mass of TL, s0 is the acceptable parking space, (x * ,y * ) is the corrected deflection angle coordinate.

[0207] To simplify the formula, let The potential force generated by TL on M:

[0208]

[0209] make but:

[0210]

[0211] In the formula, η is a substitution variable, is the distance between TL and M raised to the power of u+1, It is the distance between TL and M raised to the power of v+1. β is the angle between the line connecting the centroids of TL and M and the y-axis.

[0212] The potential field force generated by TF on M is:

[0213]

[0214] In the formula, is the distance between TF and M raised to the power of u+1, is the v+1th power of the distance between TF and M, and α is the angle between the line connecting the centroids of TF and M and the y-axis.

[0215] The potential force generated by L on M:

[0216]

[0217] In the formula, is the distance between L and M raised to the power of u+1, It is the distance between L and M raised to the power of v+1, and ξ is the angle between the line connecting the centers of mass of L and M and the y-axis.

[0218] According to the vehicle interaction potential theory, when the distance between vehicles is less than the acceptable gap, the force generated by the potential is repulsive force; when the distance between vehicles is greater than the acceptable gap, the force generated is attractive force. However, combined with the actual driving of the vehicle, even if TF generates longitudinal gravitational force on M, it cannot represent that it will attract M to generate backward acceleration.

[0219] Therefore Figure 7 Based on the established coordinate system, the vehicle interaction potential combined with the actual lane change analysis can be obtained:

[0220] When F y M-L ≥0, M continues to follow the vehicle, where y is the direction of the vehicle, defined as the y axis, and F y M-L is the potential field force generated by L on M in the y-axis direction;

[0221] When F y M-- <0, M is repelled by the force and intends to change lanes. u M-TL <0 or F y M*TF >0, it means that M is subject to the longitudinal repulsion of TL or TF, and lane change is not recommended. y M-TL is the potential force generated by TL on M in the y-axis direction, F y M-TF is the potential field force generated by TF on M in the y-axis direction; when F y M-TK ≥0 and F y M-TF When ≤0, it means that TL and TF exert gravitational forces on M, and it is recommended to change lanes. At this time, the sum of the lateral gravitational forces on M is |F x M-TF |+|F x M-TL |, prompting M to accelerate into the target lane, where x is the vertical direction of the vehicle, defined as the x-axis, and F x M-TF is the potential force generated by TF on M in the x-axis direction, F x M-TL is the potential field force generated by TL on M in the x-axis direction; as M turns, L generates a lateral repulsive force on M, forcing M to accelerate laterally away from the original lane. At this time, the lateral force on M is equal to |F x |For:

[0222] |F x |=|F x M-L |+|F x M-TF |+|F xM-TL |.

[0223] Under the gravitational force of TL and TF and the repulsive force of L, M accelerates at an acceleration of |a x |Accelerate to change lanes to the target lane, acceleration|a x |Satisfy the relationship:

[0224]

[0225] Based on the above analysis of the acceptable gap for lane changing, numerical simulation is performed using Python software to analyze the changes in the acceptable gap for lane changing under different vehicle speeds and accelerations.

[0226] The simulation of the lane-changing acceptable gap of M and TL is to establish a spatial coordinate system with the speed of the two vehicles as the x-axis and the lane-changing acceptable gap as the z-axis, and output the relationship surface between the speed of the two vehicles and the lane-changing acceptable gap. M The value is between 2m·s -2 and 3m·s -2 Between and 0.2m·s -2 The parameter calibration determines t0 = 1.5, c v =0.05,c a =0.3, the vehicle length is w=3.5, the lane change angle is θ=3°, and the following parameters have the same values. Based on the previous assumption v M <v TL , plot TL at acceleration a TL =1m·s -2 The change of the acceptable gap between M and TL when changing lanes is as follows: Figure 8 shown.

[0227] The results show that the acceptable gap for lane change is determined by the speed and acceleration of M and TL. As the speed difference between TL and M increases, the acceptable gap for lane change decreases; as the acceleration of M increases, the acceptable gap for lane change increases; as the acceleration of the front vehicle TL in the target lane increases, the acceptable gap for lane change decreases. This can be inspired: in the process of lane change in a connected environment, in order to meet the speed expectations of the lane-changing vehicle, the lane change safety is ensured by accelerating the TL without decelerating.

[0228] Simulation of the acceptable gap between M and TF for lane change, with the same parameter settings as above. Based on the assumption v M <v TF , respectively draw the target lane following TF at acceleration a TF =1m·s -2 The change of the acceptable gap between M and TF when changing lanes is as follows: Fig. 9 shown.

[0229] The results show that the acceptable gap for lane change is determined by the speed and acceleration of M and TF. As the speed difference between TF and M increases, the acceptable gap for lane change increases; as the acceleration of M increases, the acceptable gap for lane change decreases; as the speed of the target lane following TF increases, the acceptable gap for lane change increases. This can be inspired: in the process of lane change in a connected environment, in order to meet the speed expectations of the lane-changing vehicle, M ensures lane change safety by slowing down TF when the speed is constant or accelerating.

[0230] The simulation of the acceptable gap between M and L lane change is as follows: the parameters are set as above. Based on the assumption v M >v L , respectively draw the original lane front car L at acceleration a L =1m·s -2 When , the change of the acceptable gap between M and L lane change is as follows Fig.10 shown.

[0231] The results show that the acceptable gap for lane change is determined by the speed and acceleration of M and L. As the speed difference between M and L increases, the acceptable gap for lane change increases; as the speed of M increases, the acceptable gap for lane change increases; as the speed of L increases, the acceptable gap for lane change decreases. This can be inspired: in the process of lane change in a connected environment, in order to meet the speed expectations of the lane-changing vehicle, M ensures lane change safety by accelerating L at a constant speed.

[0232] In summary, the acceptable gap for lane changing is affected by the different motion states of the lane changing interactive vehicles. The above simulation analysis shows that the lane changing acceptable gap model established in this study conforms to the actual lane changing state.

[0233] In order to further verify the effectiveness of the lane-changing model based on vehicle potential, a simulation experiment was conducted using SUMO software. The experimental scenario was set as a 2km long one-way three-lane road with a speed limit of 60km / h. The simulation interface of the lane-changing test is shown in the figure below. Fig.11 As shown in the figure, after the vehicle has the intention to change lanes, it will turn on the left turn signal to indicate it. The figure well reflects the environmental state of the lane-changing vehicle in the model. The reference literature calibrates the parameters of the vehicle potential model, taking u = 0.713, v = 1.648, γ = 0.032, s0 = 2.11, and combining the parameters determined above to determine the model.

[0234] The traffic flow operation parameters output by the experiment are analyzed by drawing, combined with Fig.12 and Fig.13As shown in the figure, it reflects the changes in the average speed and average acceleration of the traffic flow. In the intelligent network environment, vehicles can obtain the operating parameters of surrounding vehicles in real time, and then judge the environment in which they are driving, and achieve uninterrupted speed adjustment. Fig.12 This reflects the change in the average speed of the traffic. It can be seen from the figure that the traffic as a whole is traveling at a relatively ideal speed, and the average speed fluctuates within 0.6 m·s -1 , the fluctuation is relatively gentle. Fig.13 It reflects the average acceleration of the traffic flow. As can be seen from the figure, the average acceleration is distributed in the range of -0.6m·s -2 and 0.6 m·s -2 range, and the acceleration state is obviously more than the deceleration state, which is also the reason why the average speed of traffic flow is relatively ideal.

[0235] In order to better verify the effectiveness of the vehicle potential lane-changing model, a set of comparative experiments were designed between the vehicle potential lane-changing model, the molecular dynamics lane-changing model, and the SL2015 lane-changing model. Among them, SL2015 is a built-in model in the SUMO software, which can well simulate the lane-changing behavior of the vehicle. The molecular dynamics lane-changing model is based on the Lennard-Jones 6-12 The potential function is used as the prototype. The vehicle potential proposed in this study corrects various parameters based on the Lennard-Jones molecular interaction potential. The comparison between the three is of certain significance.

[0236] like Fig.14 The figure shows the average speed comparison of the vehicle potential lane changing model, the molecular dynamics lane changing model and the SL2015 lane changing model when the traffic flow is roughly stable. The blue line in the figure represents the vehicle potential lane changing model, the orange line represents the molecular dynamics lane changing model, and the gray line represents the SL2015 lane changing model. Fig.14 It can be seen that compared with the molecular dynamics lane changing model and the SL2015 lane changing model, the average speed of the traffic flow using the vehicle potential lane changing model has been greatly improved, the average speed fluctuation is relatively smooth, and the vehicle driving comfort is higher.

[0237] The present invention establishes the vehicle interaction potential based on the Lennard-Jones potential, introduces the correction of the vehicle shape, speed, acceleration, and steering angle, and makes up for the fact that the previous research on the molecular lane-changing model directly used the Lennard-Jones potential function as a prototype and ignored the huge difference in the magnitude of parameters such as mass and speed between particles and vehicles. By drawing the spatial distribution diagram of the vehicle potential field under different motion states, the changes of the vehicle potential field with vehicle speed, acceleration, and steering angle can be intuitively analyzed, making the study of vehicle microscopic behavior based on molecular dynamics more realistic.

[0238] In the previous research on molecular dynamics lane changing model, for the setting of longitudinal critical distance after lane changing, only the speed and acceleration of the current vehicle were considered, and the influence of the running state of the front vehicle in the original lane on the critical safety distance after lane changing was ignored. The present invention introduces the acceleration of the front vehicle in the original lane into the longitudinal critical distance, and through the analysis of the lane changing safety critical time, clarifies the relationship between the acceptable gap for lane changing and the speed difference and acceleration difference of the interacting vehicles, and constructs the expressions for the acceptable gap for lane changing between M and L, TL, and TF respectively. Through the simulation analysis of different motion states of interacting vehicles, reasonable suggestions are put forward for lane changing behavior in a networked environment.

[0239] A lane-changing model based on vehicle interaction potential was established, and the potential and potential field forces generated by the surrounding interactive vehicles on the lane-changing vehicle were analyzed. The driving decision of M under different potential field forces was given. Its effectiveness was verified by comparing it with the lane-changing model based on molecular dynamics and the SL2015 model. This has theoretical guiding significance for the lane-changing behavior decision of vehicles in the future intelligent network environment.

[0240] Of course, the above description is not a limitation of the present invention, and the present invention is not limited to the above examples. Changes, modifications, additions or substitutions made by technicians in this technical field within the essential scope of the present invention should also fall within the protection scope of the present invention.

Claims

1. A lane-changing control model construction method based on vehicle interaction potential, characterized in that: The following steps are involved: S1: the lane-changing vehicle M follows the vehicle L in the original lane, and continuously calculates the acceptable gaps for lane change between the lane-changing vehicle M and the vehicle L in the original lane in the current lane, as well as the target lane vehicle TF and the target lane vehicle TL in the target lane; S2 constructs a plane coordinate system and determines whether the lane-changing vehicle M has a lane-changing tendency according to the potential field force of the vehicle L in the original lane. If so, it proceeds to S3, otherwise it proceeds to S1; S3: In the case of a lane-changing tendency, whether to change lanes is determined based on the potential field forces of the vehicle TL behind the target lane and the vehicle TF in front of the target lane on the lane-changing vehicle M. If the lane-changing is to be carried out, the process proceeds to S4; otherwise, the process proceeds to S2. S4 When changing lanes, the lane-changing vehicle M is subjected to the forces of the vehicle behind the target lane TL, the vehicle in front of the target lane TF, and the vehicle in front of the original lane L. The lane-changing vehicle M accelerates at an acceleration of |a x |Accelerate to change lanes to the target lane, acceleration|a x |Satisfy the relationship: In the formula, x is the vertical direction of the vehicle, defined as the x-axis, and a x is the acceleration of vehicle M in the x-axis direction, F x is the component of the force in the x-axis direction, m is the mass of vehicle M; S5 completes the lane change process; In S1, the acceptable gap between M and TL for lane change is expressed as: In the formula, is the acceptable gap between M and TL for lane change, v TL is the initial velocity of TL, v M is the initial lane-changing speed of M, a M is the lane change acceleration of M, a TL is the acceleration of TL, is the critical longitudinal distance between M and TL after the lane change, w is the width of M, θ is the angle between M and the horizontal direction of the road, t0, c v 、c a is the coefficient to be determined.

2. The method for constructing a lane-changing control model based on vehicle interaction potential according to claim 1, characterized in that: In S1, the acceptable gap between M and TF for lane change is expressed as: In the formula, is the acceptable gap between M and TF for lane change, v TF is the initial velocity of TF, a TF is the acceleration of TF, It is the critical longitudinal distance between M and TF after the lane change is completed.

3. The method for constructing a lane-changing control model based on vehicle interaction potential according to claim 2, characterized in that: In S1, the acceptable gap expression for lane change between M and L is: In the formula, is the acceptable gap between M and L for lane change, v L is the initial velocity of L, a L is the acceleration of L, j is the angular collision state, t j The time it takes for M and L to show the illustrated state when they collide with each other, and the value is between 1.8s and 3s. It is the critical longitudinal distance between M and L after the lane change is completed.

4. The method for constructing a lane-changing control model based on vehicle interaction potential according to claim 3, characterized in that: In S2, construct a plane coordinate system to determine whether F y M-L ≥0, where y is the direction of the vehicle, defined as the y-axis, F y M-L is the potential field force generated by L on M in the y-axis direction; If the conditions are met, the system goes into S1 and M continues to follow L. If the conditions are not met, the system goes into S3.

5. The method for constructing a lane-changing control model based on vehicle interaction potential according to claim 4, characterized in that: In S3, when F y M-TL <0 or F y M-TF When F > 0, it means that M is subject to the longitudinal repulsion of TL or TF, and it is not recommended to change lanes and enter S2. y M-TL is the potential force generated by TL on M in the y-axis direction, F y M-TF is the potential field force generated by TF on M in the y-axis direction; When F y M-TL ≥0 and F y M-TF When ≤0, it means that TL and TF exert gravitational force on M, and it is recommended to change lanes and enter S4; In S4, the sum of the lateral gravitational forces on M is |F x M-TF |+|F x M-TL |, prompting M to accelerate into the target lane, where x is the direction perpendicular to the road, defined as the x-axis, F x M-TF is the potential force generated by TF on M in the x-axis direction, F x M-TL is the potential field force generated by TL on M in the x-axis direction; As M turns, L generates a lateral repulsive force on M, forcing M to accelerate laterally away from the original lane. At this time, the lateral force on M is equal to |F x |For: |F x |=|F x M-L |+|F x M-TF |+|F x M-TL |; In the formula, F x M-L is the potential field force generated by L on M in the x-axis direction.

6. The method for constructing a lane-changing control model based on vehicle interaction potential according to claim 5, characterized in that: In S2, the molecular interaction potential function between following vehicles is: Where U is the molecular interaction potential function, ε is the potential well depth, and X r The safety distance required for following is L, which is the distance between the front and rear vehicles; The powers of the attractive and repulsive terms in the molecular interaction potential model are set as undetermined parameters, and the vehicle interaction potential function is constructed: In the formula, U is the vehicle interaction potential function, v and u are the powers of the attraction term and repulsion term, and v is greater than u; Introducing distance, velocity, and acceleration for correction, we get the vehicle interaction potential function: Where a is the acceleration of the vehicle, d* is the corrected distance, φ is the clockwise angle between the vehicle acceleration and the line connecting the centers of mass of the two vehicles, γ is an undetermined parameter, (x0, y0) is the coordinate position of the center of mass of the current vehicle, (x, y) is the center of mass position of any surrounding vehicle, and s0 is the acceptable parking gap; Introducing the steering angle for correction, we get: In the formula, (x * ,y * ) is the corrected deflection angle coordinate; The vehicle interaction potential model for lane changing behavior is:

7. The method for constructing a lane-changing control model based on vehicle interaction potential according to claim 6, characterized in that: The potential of TL on M: Where U M-TL is the potential generated by TL on M, L M-TL is the distance between TL and M, β is the angle between the center of mass of TL and M and the y-axis, d* is the correction distance, γ is an undetermined parameter, (x M ,y M ) is the coordinate position of the center of mass of M, (x TL ,y TL ) is the coordinate position of the center of mass of TL; make The potential force generated by TL on M: make but: In the formula, η is a substitution variable, is the distance between TL and M raised to the power of u+1, It is the distance between TL and M raised to the power of v+1.

8. The method for constructing a lane-changing control model based on vehicle interaction potential according to claim 7, characterized in that: The potential field force generated by TF on M is: In the formula, is the distance between TF and M raised to the power of u+1, is the v+1th power of the distance between TF and M, and α is the angle between the line connecting the centroids of TF and M and the y-axis.

9. The method for constructing a lane-changing control model based on vehicle interaction potential according to claim 8, characterized in that: The potential force generated by L on M: In the formula, is the distance between L and M raised to the power of u+1, It is the distance between L and M raised to the power of v+1, and ξ is the angle between the line connecting the centers of mass of L and M and the y-axis.

Citation Information

Patent Citations

  • Lane changing decision-making method for networked autonomous vehicle based on risk potential field

    CN117485366A