CACC fleet cooperative lane changing control method in mixed traffic flow environment

Abstract

The invention discloses a CACC fleet cooperative lane changing control method in a mixed traffic flow environment. The method comprises the following steps: determining an initial lane changing gap of a target lane and a target network connection automatic driving vehicle cooperatively controlled by a CACC fleet; obtaining vehicle motion state information of the CACC fleet, the target networked autonomous vehicle and a vehicle in front of the target networked autonomous vehicle; constructing a longitudinal control method, generating lane changing gaps one by one through cooperative deceleration of the vehicles, and controlling the vehicles to complete lane changing by using a transverse control algorithm; and after the lane change of the CACC motorcade is finished, the longitudinal control method is used again to control the vehicle to cruise at the expected distance. According to the method, when the vehicle gap of the target lane does not meet the lane changing condition, lane changing of the CACC motorcade is effectively achieved through cooperative control over the vehicles.

Description

Technical Field

[0001] The present invention relates to the field of intelligent networked autonomous driving vehicles and traffic control, and specifically to a CACC fleet collaborative lane changing control method in a mixed traffic flow environment. Background Art

[0002] In recent years, with the rapid development of Internet of Vehicles, V2X and autonomous driving technologies, connected autonomous driving vehicle technology has attracted more and more attention from researchers. As an important control technology, Cooperative Adaptive Cruise Control (CACC) is widely used in the field of connected autonomous driving vehicles. CACC vehicles have higher sensitivity, shorter reaction time, and smaller time intervals between vehicles in a convoy. By coordinating the speed and spacing between vehicles, it can maximize the use of road space and improve traffic efficiency.

[0003] Existing research on CACC vehicles focuses on longitudinal control. In actual environments, road traffic conditions are complex and changeable. When CACC vehicles face problems such as passing through merging areas, exiting the main road, or encountering an accident in front of the lane, they will inevitably have to change lanes. However, existing CACC vehicle control methods lack research on lane changing issues for the fleet as a whole. When controlling the CACC fleet to change lanes, the initial composition structure of the fleet is often destroyed; how to control when there is no suitable gap on the target lane to meet the lane change of the fleet or any vehicle, how to coordinate with the connected autonomous vehicle (CAV) on the target lane, and how to adjust the vehicle's motion state according to the real-time working conditions during the control process. Existing control methods rarely consider the above issues; there is a lack of research on the coordination of lateral control and longitudinal control based on control theory at the fleet level.

[0004] Therefore, in a mixed traffic flow environment, how to help the CACC fleet complete lane change when the target lane gap is insufficient is an urgent problem to be solved. Summary of the invention

[0005] The technical problem to be solved by the present invention is as follows: In view of the above-mentioned problems in the prior art, a CACC fleet collaborative lane changing control method is provided which can create lane changing gaps in a mixed traffic flow environment to solve the problem of insufficient gaps in the target lane when the CACC fleet changes lanes.

[0006] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0007] A CACC convoy coordinated lane-changing control method in a mixed traffic flow environment comprises the following steps:

[0008] S1) selecting any connected autonomous driving vehicle on the target lane as the target connected autonomous driving vehicle, and taking the gap between the target connected driving vehicle and the vehicle in front of it as the initial lane change gap of the CACC fleet;

[0009] S2) Real-time acquisition of vehicle motion status information of the CACC fleet, the target connected autonomous driving vehicle and its preceding vehicle;

[0010] S3) creating lane change gaps for the vehicles starting from the rear vehicle of the CACC convoy in sequence, specifically using a preset longitudinal control method, based on the expected spacing before the lane change is completed and the control target, and according to the vehicle motion state information of the CACC convoy, the target connected autonomous driving vehicle and its front vehicle, controlling the target connected autonomous driving vehicle and the CACC convoy to decelerate in coordination, so that the size of the lane change gap corresponding to the vehicle meets the requirement;

[0011] S4) changing lanes for the vehicles starting from the rear vehicle of the CACC fleet in sequence. Specifically, when the size of the lane change gap corresponding to the vehicle meets the requirement, planning the lane change trajectory for the vehicle, and then using the preset vehicle lateral dynamics model and lateral control algorithm to control the vehicle to change lanes along the planned trajectory. For the rear vehicle after the lane change, the rear vehicle after the lane change is controlled to continue to travel at a constant speed in the longitudinal direction;

[0012] S5) Using a preset longitudinal control method, based on the desired spacing after the lane change is completed and the control target, the vehicles of the CACC fleet are controlled to cruise at the desired spacing after the lane change is completed.

[0013] Furthermore, the longitudinal control method specifically includes:

[0014] Before the CACC convoy completes the lane change, the rear vehicle of the CACC convoy is used as the guide vehicle of the target networked autonomous driving vehicle, and the rear vehicle of the CACC convoy uses the front vehicle in the corresponding lane change gap as the guide vehicle, and the remaining vehicles of the CACC convoy use the adjacent rear vehicle as the guide. If the distance between the vehicle of the CACC convoy and the front vehicle in the corresponding lane change gap is less than the lane change safety distance, the target networked autonomous driving vehicle and all vehicles in the CACC convoy are controlled to follow the movement state of the corresponding guide vehicle and decelerate until the distance between the vehicle of the CACC convoy and the front and rear vehicles in the corresponding lane change gap is greater than or equal to the lane change safety distance;

[0015] After the CACC convoy completes the lane change, each vehicle is controlled to use its adjacent preceding vehicle as a guide vehicle and cruise at the desired spacing and speed.

[0016] Furthermore, the longitudinal control method is constructed with the vehicle driving safety, following performance and comfort as the objectives, and the construction process of the longitudinal control method is as follows:

[0017] The system state space equation is obtained according to the vehicle motion state information of the controlled vehicle and the corresponding guide vehicle;

[0018] Consider safety, following and comfort to design the longitudinal performance indicators of the system, and set control targets before and after the lane change is completed;

[0019] Set up system longitudinal constraints and introduce relaxation factors to expand the scope of constraints;

[0020] A system longitudinal prediction model is established according to the system state space equation and the system longitudinal performance index, and a comprehensive performance function is constructed with the goal of optimizing the following performance and comfort while ensuring the driving safety of the vehicle;

[0021] Through the fuzzy control algorithm, the weights of the controller's safety, following and comfort are adaptively optimized and adjusted. Furthermore, the expression of the system state space equation is as follows:

[0022] x(t+1)=Ax(t)+Bu(t)+Cw(t)

[0023]

[0024] Where A, t, C are coefficient matrices; u(t), w(t) represent the control amount and the disturbance amount respectively; x(t) = [Δx(t), v(t), v rel (t),a(t),j(t)] T , is the system state quantity; Δx(t) is the following distance between the controlled vehicle and the leading vehicle at time t; v rel (t) is the relative speed between the controlled vehicle and the guide vehicle at time t; v(t) is the speed of the controlled vehicle at time t; a(t) is the acceleration of the controlled vehicle at time t; j(t) is the acceleration change rate of the controlled vehicle at time t; T is the system sampling time;

[0025] The expression of the longitudinal performance index of the system is as follows:

[0026] y(t)=[ξ(t)v rel (t)a(t)j(t)] T

[0027] Where ξ(t) is the following distance error representing safety, specifically the error between the actual following distance and the expected following distance before the lane change is completed or the expected following distance after the lane change is completed; v rel (t) is the relative speed between the controlled vehicle and the leading vehicle at time t, representing the following performance; a(t) and j(t) are the acceleration and acceleration change rate of the controlled vehicle at time t, representing the comfort performance;

[0028] The constraints of the system longitudinal constraints are as follows:

[0029]

[0030] Where Δx min , v min , a min , j min ,u min are the following distance between the controlled vehicle and the leading vehicle, the speed of the controlled vehicle, the acceleration of the controlled vehicle, the acceleration change rate of the controlled vehicle, and the hard constraint lower limit of the control amount; Δx max , v max , a max , j max ,u max is the following distance between the controlled vehicle and the leading vehicle, the speed of the controlled vehicle, the acceleration of the controlled vehicle, the acceleration change rate of the controlled vehicle, and the hard constraint upper limit of the control amount; λ1, λ2, λ3, λ4, λ5 are relaxation factors; Δx min_a , v min_a , a min_a , j min_a ,u min_a is the following distance between the controlled vehicle and the leading vehicle, the speed of the controlled vehicle, the acceleration of the controlled vehicle, the acceleration change rate of the controlled vehicle, and the soft constraint lower limit of the control amount; Δx max_a , v max_a , a max_a , j max_a ,u max_a The following distance between the controlled vehicle and the leading vehicle, the speed of the controlled vehicle, the acceleration of the controlled vehicle, the acceleration change rate of the controlled vehicle, and the soft constraint upper limit of the control amount;

[0031] The system longitudinal prediction model is a system longitudinal prediction model that considers constraints in the quadratic optimization problem, and the expression is as follows:

[0032]

[0033] Among them, A c and b c They are the coefficient matrices generated in the process of quadratic derivation, U(t+N c ) represents the control quantity matrix; M Xmin 、M Xmax are the hard constraint matrices of the lower and upper limits of the system state quantity, M Umin 、M Umax are the hard constraint matrices of the upper and lower limits of the control quantity, M Xmin_s 、M Xmax_s are the soft constraint matrices of the lower and upper limits of the system state quantity, MUmin_s 、M Umin_s are the soft constraint matrices of the upper and lower limits of the control quantity, λ is the relaxation factor matrix, L a , L b , L c , L h is the coefficient matrix;

[0034] The comprehensive performance function expression is as follows:

[0035]

[0036] Among them, O, P and are the weight matrices of performance error, control amount and soft constraint coefficient, y(t+i|t) represents the system longitudinal performance index, u(t+i) represents the control amount, y ref (t+i|t) represents ideal performance, represents a quadratic function, λ i is the soft constraint coefficient;

[0037] Through the fuzzy control algorithm, the weights of the controller's safety, following and comfort are adaptively optimized and adjusted.

[0038] Furthermore, the control target before the lane change is completed is to control the speed difference between the front and rear vehicles, the acceleration of the vehicles and the acceleration change rate to approach zero under the premise that the head distance between the vehicles is not less than the lane change safety gap.

[0039] Furthermore, the expected distance before the lane change is completed is greater than or equal to the lane change safety gap, and the expression of the lane change safety gap is as follows:

[0040]

[0041] in, and Respectively represent the front and rear safety clearances required for lane changing; t r represents the vehicle reaction time; t d Represents the braking delay time; a min Represents the minimum acceleration of the vehicle; v r represents the speed of the vehicle behind during the lane change interval; v f represents the speed of the vehicle in front during the lane changing gap; v represents the speed of the vehicle changing lanes.

[0042] Furthermore, when planning the lane-changing trajectory of the vehicle, a quintic polynomial is used to plan the lane-changing trajectory of the vehicle, and the expression is as follows:

[0043]

[0044] Among them, x r (t) and yr (t) is the planned longitudinal position and lateral position at time t, a i 、b i , i=0,1,…,5 are the coefficients of the polynomial.

[0045] Furthermore, the vehicle lateral dynamics model uses a linear two-degree-of-freedom model, as shown below:

[0046]

[0047] in, is the yaw angle of the controlled vehicle, m is the mass of the controlled vehicle, and are the cornering stiffness of the front and rear wheels of the controlled vehicle, v x is the longitudinal velocity of the controlled vehicle, a and b are the distances from the front and rear axles of the controlled vehicle to the center of mass, I is the moment of inertia of the controlled vehicle around the vertical ground axis, and δ is the front wheel turning angle.

[0048] Furthermore, the cost function of the lateral control algorithm is constructed using the Lagrange multiplier method with error elimination as the core, as shown below:

[0049]

[0050] Among them, Q rr and R rr They are the error state matrices E rr and the control quantity u rr The penalty matrix, and is the coefficient matrix obtained by discretization and simplification by Euler method, λ is the Lagrange multiplier, E k and u k is the error state matrix and control quantity after discretization, E n is the error state matrix E after discretization k One of the items.

[0051] Furthermore, the control goal after the lane change is to control the speed difference between the front and rear vehicles, the acceleration of the vehicles, and the acceleration change rate to approach zero, while making the distance between the vehicles approach the expected distance for cruising. The expression of the expected distance after the lane change is as follows:

[0052] Δx des (t) = t h v+d0-cv rel_head

[0053] Among them, t h is the time distance coefficient between vehicles; d0 is the minimum distance between vehicles; v rel_headis the speed difference between the controlled vehicle and the head vehicle of the convoy; c is a constant coefficient; v is the speed of the controlled vehicle.

[0054] Compared with the prior art, the advantages of the present invention are:

[0055] The lateral control that takes vehicle dynamics into consideration is combined with the longitudinal control to achieve integrated control of the CACC fleet's lane change preparation, lane change execution, and cruising after the lane change is completed, and the CACC fleet is controlled to complete the lane change in a stable and orderly manner.

[0056] By performing coordinated deceleration control on the CACC fleet and the target lane CAV, the distance to the vehicle in front of the target gap is expanded as quickly as possible while ensuring the rear safety gap, so as to generate a front safety gap that meets the CACC vehicle lane change requirements.

[0057] By dividing the vehicle's safety status, the performance indicator weights in the multi-objective system longitudinal prediction model are adjusted in real time to output the optimal results suitable for the current working conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 This is a flow chart of CACC fleet collaborative lane changing control according to an embodiment of the present invention.

[0059] Figure 2 The figure is a schematic diagram of a scene of an embodiment of the present invention.

[0060] Figure 3 Schematic diagram of a control method according to an embodiment of the present invention.

[0061] Figure 4 It is a schematic diagram of the application of fuzzy control in the longitudinal control algorithm in an embodiment of the present invention.

[0062] Figure 5 Schematic diagram of vehicle motion trajectory according to an embodiment of the present invention. DETAILED DESCRIPTION

[0063] The present invention is further described below in conjunction with the accompanying drawings and specific preferred embodiments, but the protection scope of the present invention is not limited thereby.

[0064] This example provides a cooperative adaptive cruise control (CACC) fleet cooperative lane change control method in a mixed traffic flow scenario on a highway. When the gap between the front and rear vehicles in the target lane is insufficient, the CACC fleet and the target connected autonomous vehicle (CAV) are cooperatively controlled to generate a gap that meets the lane change requirements, and the CACC fleet is controlled to complete the lane change in a stable and orderly manner. Figure 1 As shown, the following steps are included:

[0065] S1) Select any CAV on the target lane as the target CAV. In this embodiment, the lane that the CACC team travels after changing lanes is called the target lane, and the lane that the CACC team travels before changing lanes is called the current lane. The gap between the target CAV and the vehicle in front of it is selected as the initial lane-changing gap of the CACC team.

[0066] S2) Based on the Internet of Vehicles, vehicle-road collaboration, V2X and other technologies, real-time acquisition of vehicle motion status information of the CACC fleet, target CAV and its preceding vehicle;

[0067] S3) creating lane-changing gaps for the vehicles starting from the rear vehicle of the CACC convoy in sequence, specifically, using a preset longitudinal control method, based on the expected spacing before the lane change is completed and the control target, and according to the vehicle motion state information of the CACC convoy, the target CAV and its leading vehicle, controlling the target CAV and the CACC vehicle to decelerate in coordination, so that the size of the lane-changing gap corresponding to the vehicle meets the requirement;

[0068] S4) changing lanes for the vehicles starting from the rear vehicle of the CACC fleet in sequence. Specifically, when the size of the lane change gap corresponding to the vehicle meets the requirement, planning the lane change trajectory for the vehicle, and then using the preset vehicle lateral dynamics model and lateral control algorithm to control the vehicle to change lanes along the planned trajectory, and after the rear vehicle completes the lane change, controlling the rear vehicle to continue to travel at a constant speed in the longitudinal direction;

[0069] S5) Using the same longitudinal control method as in step S3, based on the desired spacing after the lane change is completed and the control target, the vehicles of the CACC fleet are controlled to cruise at the desired spacing after the lane change is completed.

[0070] It should be noted that the method of this embodiment is applied to mixed traffic flow in an intelligent network environment, with the following assumptions: there is a CAV that can be used for collaborative control in the traffic flow on the target lane; the vehicles in the CAV and CACC fleets are the same type of autonomous driving intelligent network vehicles, and they can be collaboratively controlled according to different goals. Figure 2 When the lane change gap is insufficient, the control method proposed in this embodiment starts to be implemented after the corresponding position of the rear vehicle of the CACC fleet on the current lane is located in the gap between the target CAV and its front vehicle on the target lane. Each step is described in detail below.

[0071] In step S1 of this embodiment, the preceding vehicle of the target connected autonomous driving vehicle CAV may be any vehicle, such as Figure 3 As shown, when the position of the rear vehicle of the CACC fleet in the current lane is at a position corresponding to the gap between the target connected autonomous driving vehicle CAV and the vehicle in front of it in the target lane, collaborative control begins.

[0072] In step S2 of this embodiment, the vehicle motion state information includes the speed, acceleration and position information of the vehicle, which is used for the coordinated control of the CACC fleet and the target CAV in step 3.

[0073] Step S3 of this embodiment is based on the model predictive control algorithm of vehicle dynamics and adaptive weights, and constructs a longitudinal control method with the goals of vehicle driving safety, following and comfort, and generates lane change gaps for each vehicle starting from the rear vehicle of the convoy through coordinated deceleration of the vehicles.

[0074] Specifically, in step S3 of this embodiment, the objects to be collaboratively controlled are Figure 3 The CACC fleet and target CAV shown in the figure are based on the model predictive control algorithm with adaptive weights to perform coordinated deceleration control on the control object. The longitudinal control method specifically includes:

[0075] Before the CACC convoy completes the lane change, the rear vehicle of the CACC convoy is used as the guide vehicle of the target CAV. The rear vehicle of the CACC convoy uses the preceding vehicle in the corresponding lane change gap (i.e. the preceding vehicle of the target CAV) as the guide vehicle, and the remaining vehicles of the CACC convoy use the adjacent following vehicles as the guide vehicles.

[0076] If the distance between a vehicle in the CACC fleet and the vehicle in front of the corresponding lane change gap is less than the lane change safety distance, the control target CAV and all vehicles in the CACC fleet will follow the movement state of the corresponding guide vehicle and decelerate until the distance between the vehicle and the vehicles in front and behind the corresponding lane change gap is greater than or equal to the lane change safety distance.

[0077] After the CACC convoy completes the lane change, each vehicle is controlled to use its adjacent preceding vehicle as a guide vehicle and cruise at the desired spacing and speed.

[0078] Before the lane change is completed, step S3 performs longitudinal control to sequentially create the safety gap required for the CACC vehicle to change lanes. Figure 3 As shown, the target CAV in the target lane moves with the rear vehicle of the CACC fleet as the guide vehicle, and maintains a sufficient lane change safety gap with the rear vehicle. The rear vehicle takes the front vehicle of the initial lane change gap, that is, the front vehicle of the target CAV, as the guide vehicle, and adjusts itself according to the longitudinal distance between the guide vehicle and the movement state of the guide vehicle. When the distance between the guide vehicle and the guide vehicle is less than the lane change safety gap, the rear vehicle will slow down to create the required gap. The remaining vehicles in the CACC fleet all move with their adjacent rear vehicles as guide vehicles, and slow down following the rear vehicle. To ensure that before changing lanes, sufficient front safety gaps are created between the front vehicles corresponding to the lane change gaps of the target lanes. Once the front and rear distances of the CACC vehicles meet the requirements, the lane change is performed.

[0079] The construction process of the longitudinal control method in step S3 of this embodiment is as follows:

[0080] S3.1) Constructing the system vehicle state space expression: This embodiment simplifies the vehicle into a moving particle in the longitudinal direction, ignoring the influencing factors such as air resistance, rolling resistance, communication delay and failure. The operation of the CACC fleet system in the longitudinal direction is affected by each vehicle, so the following distance, speed, relative speed with the leading vehicle, acceleration and acceleration change rate during the vehicle driving process are selected as the system state quantity, and the system state space equation is obtained according to the vehicle motion state information of the controlled vehicle and the corresponding guide vehicle. The expression is as follows:

[0081] x(t+1)=Ax(t)+Bu(t)+Cw(t)

[0082]

[0083] Where A, B, and C are coefficient matrices; u(t) and w(t) represent the control variable and the disturbance variable, respectively; x(t) = [Δx(t), v(t), v rel (t),a(t),j(t)] T , is the system state quantity; Δx(t) is the following distance between the controlled vehicle and the leading vehicle at time t; v rel (t) is the relative speed between the controlled vehicle and the guide vehicle at time t; v(t) is the speed of the controlled vehicle at time t; a(t) is the acceleration of the controlled vehicle at time t; j(t) is the acceleration change rate of the controlled vehicle at time t; T is the system sampling time;

[0084] S3.2) Construct system longitudinal performance indicators: Consider safety, following and comfort to design system longitudinal performance indicators:

[0085] y(t)=[ξ(t)v rel (t)a(t)j(t)] T

[0086] Where ξ(t) is the following distance error representing safety, specifically the error between the actual following distance and the expected following distance before the lane change is completed or the expected following distance after the lane change is completed; v rel (t) is the relative speed of the front and rear vehicles, which is used to represent the following performance; a(t) and j(t) are the acceleration and acceleration change rate, respectively, which are used to represent the comfort.

[0087] In order to ensure that the vehicles in the convoy can comfortably and stably follow the leading vehicle at a safe following distance in the longitudinal direction, this embodiment selects the actual following distance and the expected following distance error, vehicle speed, relative speed, acceleration and acceleration change rate as performance indicators of model predictive control. Because the longitudinal control of the vehicle is divided into two stages, before and after the lane change is completed, the expected distance and control target in different stages are also different.

[0088] Specifically, the control goal before the lane change is to control the speed difference between the front and rear vehicles, the acceleration of the vehicles, and the acceleration change rate to approach zero, while making the headway between the vehicles (i.e., the expected distance before the lane change is completed) not less than the lane change safety gap. The lane change safety gap expression is:

[0089]

[0090]

[0091] in, and Respectively represent the front and rear safety clearances required for lane changing; t r represents the vehicle reaction time; t d Represents the braking delay time; a min Represents the minimum acceleration of the vehicle; v r represents the speed of the vehicle behind during the lane change interval; v f represents the speed of the vehicle in front during the lane changing gap; v represents the speed of the vehicle changing lanes.

[0092] It can be seen that due to the existence of an actual following distance and an expected following distance, when the actual following distance is smaller than the expected following distance, the vehicle will be decelerated by the longitudinal control method of this embodiment, and the acceleration will be adjusted in real time according to the change of the distance.

[0093] S3.3) Construct system longitudinal constraints: Set system longitudinal constraints and introduce relaxation factors to expand the scope of constraints. Considering that strict physical constraints may cause vehicle instability, relaxation factors are introduced to expand the feasible domain. Under the premise of not affecting the normal operation of the system, the scope of constraints is moderately expanded. The constraints are as follows:

[0094]

[0095] Where Δx min , v min , a min , j min ,u min are the following distance between the controlled vehicle and the leading vehicle, the speed of the controlled vehicle, the acceleration of the controlled vehicle, the acceleration change rate of the controlled vehicle, and the hard constraint lower limit of the control amount; Δx max , vmax , a max , j max ,u max is the following distance between the controlled vehicle and the leading vehicle, the speed of the controlled vehicle, the acceleration of the controlled vehicle, the acceleration change rate of the controlled vehicle, and the hard constraint upper limit of the control amount; λ1, λ2, λ3, λ4, λ5 are relaxation factors; Δx min_a , v min_a , a min_a , j min_a ,u min_a is the following distance between the controlled vehicle and the leading vehicle, the speed of the controlled vehicle, the acceleration of the controlled vehicle, the acceleration change rate of the controlled vehicle, and the soft constraint lower limit of the control amount; Δx max_a , v max_a , a max_a , j max_a ,u max_a The following distance between the controlled vehicle and the leading vehicle, the speed of the controlled vehicle, the acceleration of the controlled vehicle, the acceleration change rate of the controlled vehicle, and the soft constraint upper limit of the control amount;

[0096] S3.4) Construction of system longitudinal prediction model: Based on the system state space equation and system longitudinal performance index constructed previously, a prediction model in the form of constraint matrix in the quadratic optimization problem is established and used as the prediction model of the MPC controller. The prediction model expression is as follows:

[0097]

[0098] Among them, A c and b c They are the coefficient matrices generated in the process of quadratic derivation, U(t+N c ) represents the control quantity matrix; M Xmin 、M Xmax are the hard constraint matrices of the lower and upper limits of the system state quantity, M Umin 、M Umax are the hard constraint matrices of the upper and lower limits of the control quantity, M Xmin_s 、M Xmax_s are the soft constraint matrices of the lower and upper limits of the system state quantity, M Umin_s 、M Umin_s are the soft constraint matrices of the upper and lower limits of the control quantity, λ is the relaxation factor matrix, L a , L b , L c , L h is the coefficient matrix;

[0099] Through the analysis of the longitudinal performance indicators of the system, under the condition of ensuring the driving safety of CACC vehicles, the comprehensive performance function constructed with the goal of optimizing the following performance and comfort can be expressed as:

[0100]

[0101] Among them, O, P and are the weight matrices of performance error, control amount and soft constraint coefficient respectively, y(t+i|t) represents the system longitudinal performance index, specifically the system longitudinal performance index at time t+i, u(t+i) represents the control amount, specifically the control amount at time t+i, y ref (t+i|t) represents ideal performance, represents a quadratic function, λ i is the soft constraint coefficient;

[0102] S3.5) Construct an adaptive weighting method: As a supplement to step S3.2, this embodiment uses a fuzzy control algorithm to adaptively optimize and adjust the weights of the controller's safety, following, and comfort. Vehicles have different requirements for safety, following, and comfort under different driving conditions. When there is a risk of collision, the safety index weight needs to be increased, and vice versa. The fuzzy control theory is used to adaptively optimize and adjust the safety, following, and comfort weights of the MPC controller. Figure 4 As shown, this embodiment selects the inverse of the headway between vehicles and the time between collisions as the basis for collision risk judgment and inputs them into the fuzzy controller to obtain the safety, following performance, and comfort weights after adaptive optimization adjustment.

[0103] Step S4 of this embodiment performs lateral control on the CACC vehicle that meets the lane change gap requirement. First, a lane change trajectory is planned using a quintic polynomial. Then, based on a two-degree-of-freedom vehicle model and an LQR algorithm, the vehicle is controlled to change lanes laterally. Figure 3 As shown, whenever the coordinated control is performed through step S3 so that the lane change gap corresponding to the vehicle meets the lane change requirement, the vehicle is laterally controlled to achieve lane change one by one starting from the rear vehicle of the CACC team, specifically including:

[0104] S4.1) Lane-changing trajectory planning: Use a quintic polynomial to plan the vehicle's lane-changing trajectory, as expressed below:

[0105]

[0106] Among them, x r (t) and y r (t) is the planned longitudinal position and lateral position at time t, a i 、b i, i = 0, 1, ..., 5 are the coefficients of the polynomial. The lane-changing trajectory of the vehicle will be repeatedly calculated as the vehicle moves until the vehicle completes the lane change;

[0107] S4.2) Select the vehicle lateral dynamics model: Select the linear two-degree-of-freedom model as the lateral dynamics control model. The vehicle state space equation under two degrees of freedom is as follows:

[0108]

[0109] in, is the yaw angle of the controlled vehicle, m is the mass of the controlled vehicle, and are the cornering stiffness of the front and rear wheels of the controlled vehicle, v x is the longitudinal speed of the controlled vehicle, a and b are the distances from the front and rear axles of the controlled vehicle to the center of mass, I is the moment of inertia of the controlled vehicle around the vertical ground axis, and δ is the front wheel steering angle;

[0110] S4.3) Calculate the lateral control cost function: This embodiment focuses on eliminating errors and uses the Lagrange multiplier method to construct the cost function of the lateral control algorithm, as shown below

[0111]

[0112] Among them, Q rr and R rr They are the error state matrices E rr and the control quantity u rr The penalty matrix, and is the coefficient matrix obtained by discretization and simplification by Euler method, λ is the Lagrange multiplier, E k is the error state matrix after discretization, u k is the control quantity after discretization, E n is the error state matrix E after discretization k One of the items.

[0113] like Figure 3 As shown, in step S4 of this embodiment, after the rear vehicle of the CACC fleet completes the lane change, the gap between the rear vehicle and the original front vehicle of the target CAV becomes the lane change gap of the penultimate vehicle of the CACC fleet. Since the rear vehicle continues to travel at a constant speed after completing the lane change, and the tail vehicle has been decelerated before the lane change is completed, the lane change gap of the penultimate vehicle is continuously increasing. In this process, the longitudinal control method of step S3 can also be executed to control the coordinated deceleration of the vehicles to ensure that the penultimate vehicle and the front vehicle (i.e. Figure 3 The original front vehicle HDV of the target CAV) and the rear vehicle (i.e. Figure 3 The distance between the last vehicle in the CCAC team is greater than or equal to the lane change safety distance. When the lane change clearance of the second-to-last vehicle meets the requirement, the second-to-last vehicle is controlled to change lanes again by executing step S4. By analogy, by repeating steps S3 and S4, the remaining vehicles in the team can complete lane changes in sequence on the basis of meeting the lane change clearance.

[0114] In step 5 of this embodiment, when the CACC convoy completes the lane change, the vehicle is no longer subject to lateral control, and the CACC vehicle is cruised at the desired distance through longitudinal control. Specifically, when the CACC convoy completes the lane change, each vehicle in the CACC convoy uses the adjacent front vehicle as a guide vehicle. When the distance between the vehicle in the CACC convoy and the guide vehicle is not equal to the desired distance after the lane change is completed, the longitudinal control method is used to adjust itself based on the desired distance after the lane change is completed and the control target after the lane change is completed, so that the CACC convoy cruises at the desired distance and speed.

[0115] Correspondingly, the control target after the lane change is completed is to control the speed difference between the front and rear vehicles, the acceleration of the vehicles and the rate of change of acceleration to approach zero, while making the distance between the vehicles approach the expected distance for cruising (i.e. the expected distance after the lane change is completed). Therefore, after the lane change is completed, the CAAC fleet is controlled to cruise at the expected distance.

[0116] In this embodiment, the expected spacing after the lane change is completed is expressed as follows:

[0117] Δx des (t) = t h v+d0-cv rel_head

[0118] Among them, t h is the time distance coefficient between vehicles; d0 is the minimum distance between vehicles; v rel_head is the speed difference between the controlled vehicle and the head vehicle of the convoy; c is a constant coefficient; v is the speed of the controlled vehicle.

[0119] As described above, the longitudinal control method used in step 5 of this embodiment is basically the same as the longitudinal control method used in step 3, except that the expected spacing and control target of the system longitudinal performance indicator of the longitudinal control method adopt the expected spacing and control target after the lane change is completed.

[0120] The effect of the method of this embodiment is verified by specific experiments below.

[0121] Use CarSim / Matlab / Simulink to build a joint simulation platform to verify the feasibility and effectiveness of the control method. Input the vehicle model in CarSim into Matlab / Simulink, and use relevant modules, S functions, and script files in Matlab / Simulink to control the vehicle.

[0122] Specifically, a speed of 72 km / h is selected. The CACC fleet consists of three vehicles, numbered vehicle 2, vehicle 3 and vehicle 4, with vehicle 2 being the last vehicle in the fleet. The rear vehicle in the target lane is vehicle 5, and the front vehicle is vehicle 1. Vehicle 5 is a controllable connected autonomous driving vehicle, and vehicle 1 is an uncontrolled human-driven vehicle. The initial lateral position of the CACC fleet is -1.65 m. When all CACC vehicles reach the lateral position of 1.65 m, it means that the CACC fleet has completed the lane change. The movement trajectory of each vehicle is shown in the figure below. Figure 5 As shown, at the beginning, the lateral position of the vehicles was -1.65m. Since the longitudinal lane-changing gap did not meet the conditions, no vehicle changed lanes. Later, with the implementation of the control method, the longitudinal gap met the conditions and the convoy began to change lanes one by one.

[0123] The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. A CACC convoy coordinated lane-changing control method in a mixed traffic flow environment, characterized in that: The following steps are involved: S1) selecting any connected autonomous driving vehicle on the target lane as the target connected autonomous driving vehicle, and taking the gap between the target connected driving vehicle and the vehicle in front of it as the initial lane change gap of the CACC fleet; S2) Real-time acquisition of vehicle motion status information of the CACC fleet, the target connected autonomous driving vehicle and its preceding vehicle; S3) creating lane change gaps for the vehicles starting from the rear vehicle of the CACC convoy in sequence, specifically using a preset longitudinal control method, based on the expected spacing before the lane change is completed and the control target, and according to the vehicle motion state information of the CACC convoy, the target connected autonomous driving vehicle and its leading vehicle, controlling the target connected autonomous driving vehicle and the CACC vehicle to coordinately decelerate so that the size of the lane change gap corresponding to the vehicle meets the requirements; S4) changing lanes for the vehicles starting from the rear vehicle of the CACC fleet in sequence. Specifically, when the size of the lane change gap corresponding to the vehicle meets the requirement, planning the lane change trajectory for the vehicle, and then using the preset vehicle lateral dynamics model and lateral control algorithm to control the vehicle to change lanes along the planned trajectory. For the rear vehicle after the lane change, the rear vehicle after the lane change is also controlled to continue to travel at a constant speed in the longitudinal direction; S5) Using a preset longitudinal control method, based on the desired spacing after the lane change is completed and the control target, the vehicles of the CACC fleet are controlled to cruise at the desired spacing after the lane change is completed.

2. The CACC convoy coordinated lane-changing control method in a mixed traffic flow environment according to claim 1, characterized in that: The longitudinal control method specifically includes: Before the CACC convoy completes the lane change, the rear vehicle of the CACC convoy is used as the guide vehicle of the target networked autonomous driving vehicle, and the rear vehicle of the CACC convoy uses the front vehicle in the corresponding lane change gap as the guide vehicle, and the remaining vehicles in the CACC convoy use the adjacent rear vehicle as the guide vehicle. If the distance between the vehicle of the CACC convoy and the front vehicle in the corresponding lane change gap is less than the lane change safety distance, the target networked autonomous driving vehicle and all vehicles in the CACC convoy are controlled to follow the movement state of the corresponding guide vehicle and decelerate until the distance between the vehicle of the CACC convoy and the front and rear vehicles in the corresponding lane change gap is greater than or equal to the lane change safety distance. After the CACC convoy completes the lane change, each vehicle is controlled to use its adjacent preceding vehicle as a guide vehicle and cruise at the desired spacing and speed.

3. The CACC convoy coordinated lane change control method in a mixed traffic flow environment according to claim 2, characterized in that: The longitudinal control method is constructed with the goal of vehicle driving safety, following and comfort. The construction process of the longitudinal control method is as follows: The system state space equation is obtained according to the vehicle motion state information of the controlled vehicle and the corresponding guide vehicle; Consider safety, following and comfort to design the longitudinal performance indicators of the system, and set control targets before and after the lane change is completed; Set up system longitudinal constraints and introduce relaxation factors to expand the scope of constraints; A system longitudinal prediction model is established according to the system state space equation and the system longitudinal performance index, and a comprehensive performance function is constructed with the goal of optimizing the following performance and comfort while ensuring the driving safety of the vehicle; Through the fuzzy control algorithm, the weights of the controller's safety, following and comfort are adaptively optimized and adjusted.

4. The CACC convoy coordinated lane-changing control method in a mixed traffic flow environment according to claim 3 is characterized in that: The expression of the system state space equation is as follows: x(t+1)=Ax(t)+Bu(t)+Cw(t) Where A, B, and C are coefficient matrices; u(t) and w(t) represent the control variable and the disturbance variable, respectively; x(t) = [Δx(t), v(t), v rel (t),a(t),j(t)] T , is the system state quantity; Δx(t) is the following distance between the controlled vehicle and the leading vehicle at time t; v rel (t) is the relative speed between the controlled vehicle and the guide vehicle at time t; v(t) is the speed of the controlled vehicle at time t; a(t) is the acceleration of the controlled vehicle at time t; j(t) is the acceleration change rate of the controlled vehicle at time t; T is the system sampling time; The expression of the longitudinal performance index of the system is as follows: y(t)=[ξ(t)v rel (t)a(t)j(t)] T Where ξ(t) is the following distance error representing safety, specifically the error between the actual following distance and the expected following distance before the lane change is completed or the expected following distance after the lane change is completed; v rel (t) is the relative speed between the controlled vehicle and the leading vehicle at time t, representing the following performance; a(t) and j(t) are the acceleration and acceleration change rate of the controlled vehicle at time t, representing the comfort performance; The constraints of the system longitudinal constraints are as follows: Where Δx min , v min , a min , j min ,u min are the following distance between the controlled vehicle and the leading vehicle, the speed of the controlled vehicle, the acceleration of the controlled vehicle, the acceleration change rate of the controlled vehicle, and the hard constraint lower limit of the control amount; Δx max , v max , a max , j max ,u max is the following distance between the controlled vehicle and the leading vehicle, the speed of the controlled vehicle, the acceleration of the controlled vehicle, the acceleration change rate of the controlled vehicle, and the hard constraint upper limit of the control amount; λ1, λ2, λ3, λ4, λ5 are relaxation factors; Δx min_a , v min_a , a min_a , j min_a ,u min_a is the following distance between the controlled vehicle and the leading vehicle, the speed of the controlled vehicle, the acceleration of the controlled vehicle, the acceleration change rate of the controlled vehicle, and the soft constraint lower limit of the control amount; Δx max_a , v max_a , a max_a , j max_a ,u max_a The following distance between the controlled vehicle and the leading vehicle, the speed of the controlled vehicle, the acceleration of the controlled vehicle, the acceleration change rate of the controlled vehicle, and the soft constraint upper limit of the control amount; The system longitudinal prediction model is a system longitudinal prediction model that considers constraints in the quadratic optimization problem, and the expression is as follows: Among them, A c and b c They are the coefficient matrices generated in the process of quadratic derivation, U(t+N c ) represents the control quantity matrix; M Xmin 、M Xmax are the hard constraint matrices of the lower and upper limits of the system state quantity, M Umin ,< Umax are the hard constraint matrices of the upper and lower limits of the control quantity, M Xmin_s 、M Xmax_s are the soft constraint matrices of the lower and upper limits of the system state quantity, M Umin_s 、M Umin_s are the soft constraint matrices of the upper and lower limits of the control quantity, λ is the relaxation factor matrix, L a , L b , L c , L h is the coefficient matrix; The comprehensive performance function expression is as follows: Among them, O, P and are the weight matrices of performance error, control amount and soft constraint coefficient, y(t+i|t) represents the system longitudinal performance index, u(t+i) represents the control amount, y ref (t+i|t) represents ideal performance, represents a quadratic function, λ i is the soft constraint coefficient; Through the fuzzy control algorithm, the weights of the controller's safety, following and comfort are adaptively optimized and adjusted.

5. The CACC convoy coordinated lane-changing control method in a mixed traffic flow environment according to claim 3, characterized in that: The control goal before lane change is to control the speed difference between the front and rear vehicles, the acceleration of the vehicles, and the rate of change of acceleration to approach zero, while ensuring that the head-to-head distance between the vehicles is not less than the safe gap for lane change. The control goal after the lane change is completed is to control the speed difference between the front and rear vehicles, the acceleration of the vehicles, and the rate of change of acceleration to approach zero while making the distance between the vehicles approach the expected distance for cruising.

6. The CACC convoy coordinated lane-changing control method in a mixed traffic flow environment according to claim 4, characterized in that: The expected distance before the lane change is completed is greater than or equal to the lane change safety gap. The lane change safety gap expression is: in, and Respectively represent the front and rear safety clearances required for lane changing; t r represents the vehicle reaction time; t d Represents the braking delay time; a min Represents the minimum acceleration of the vehicle; v r represents the speed of the vehicle behind during the lane change interval; v f represents the speed of the vehicle in front during the lane changing gap; v represents the speed of the vehicle changing lanes.

7. The CACC convoy coordinated lane change control method in a mixed traffic flow environment according to claim 4, characterized in that: The expected spacing after the lane change is completed is expressed as follows: Δx des (t)=t h v+d0-cv rel_head Among them, t h is the time distance coefficient between vehicles; d0 is the minimum distance between vehicles; v rel_head is the speed difference between the controlled vehicle and the head vehicle of the convoy; c is a constant coefficient; v is the speed of the controlled vehicle.

8. The CACC convoy coordinated lane change control method in a mixed traffic flow environment according to claim 1, characterized in that: When planning the lane-changing trajectory of a vehicle, a quintic polynomial is used to plan the lane-changing trajectory of the vehicle. The expression is as follows: Among them, x r (t) and y r (t) is the planned longitudinal position and lateral position at time t, a i , b i , i=0,1,…,5 are the coefficients of the polynomial.

9. The CACC convoy coordinated lane change control method in a mixed traffic flow environment according to claim 1, characterized in that: The vehicle lateral dynamics model uses a linear two-degree-of-freedom model, as shown below: in, is the yaw angle of the controlled vehicle, m is the mass of the controlled vehicle, and are the cornering stiffness of the front and rear wheels of the controlled vehicle, v x is the longitudinal velocity of the controlled vehicle, a and b are the distances from the front and rear axles of the controlled vehicle to the center of mass, I is the moment of inertia of the controlled vehicle around the vertical ground axis, and δ is the front wheel turning angle.

10. The CACC convoy coordinated lane change control method in a mixed traffic flow environment according to claim 1, characterized in that: The cost function of the lateral control algorithm is centered on eliminating errors and is constructed using the Lagrange multiplier method, as shown below: Among them, Q rr and R rr They are the error state matrices E rr and the control quantity u rr The penalty matrix, and is the coefficient matrix obtained by discretization and simplification by Euler method, λ is the Lagrange multiplier, E k is the error state matrix after discretization, u k is the control quantity after discretization, E n is the error state matrix E after discretization k One of the items.

Images