Active longitudinal and lateral cooperative control method for vehicles in merging area in mixed flow environment

Abstract

The application provides a kind of mixed flow environment under the confluence area vehicle initiative transverse and longitudinal coordination control method, based on wide-area time, continuous variable and discrete variable, carry out multi-system input and output coordination control.First, by intelligent network connection vehicle speed control and variable speed limit area control, adjust the speed and the distance between vehicles when traffic flow meets motion wave, effectively slow down traffic motion wave and improve the efficiency of traffic flow;Second, by the behavior control of intelligent network connection vehicle, limit the influence of the change of vehicle speed caused by motion wave and main line control strategy on ramp merging area vehicle flow and speed, so as to avoid traffic conflict in the process of confluence;In addition, transverse lane changing rules are also designed to reduce transverse disturbance in interlaced area, and the disturbance of transverse and longitudinal behavior in the area on the traffic system model is considered.The strategy of the application strives to reduce traffic conflict in the pursuit of the coordinated optimization of traffic safety and efficiency, so as to achieve the optimal goal of the system.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of intelligent transportation, and particularly relates to a mixed flow environment underflow area vehicle active lateral and longitudinal collaborative control method. BACKGROUND

[0002] The behavior control of intelligent and networked vehicles CAV (Connected Autonomous Vehicle) is a new emerging intelligent traffic management and control technology on expressway, and its essence is a real-time dynamic vehicle-level micro-management process, which provides a new way for quickly, accurately and efficiently solving complex road traffic problems.

[0003] In the future for a long time, the popularization of intelligent and networked vehicles needs to go through a continuous development and evolution process, and in this process, there will be mixed traffic flow of human driving and vehicles of different automatic driving levels. In order to fully exert the advantages of CAV vehicles in energy saving and emission reduction, efficient execution, networked cooperation and other aspects, and effectively solve the problems existing in current traffic operation by using the driving advantages of CAV, based on the model construction of CAV vehicles, the efficiency analysis of mixed traffic flow, the traffic management and control mode of single vehicle optimization and linkage control of CAV, etc. Research contents are gradually valued by the academic circle.

[0004] Although there have been many breakthroughs in theoretical research, most of the existing technologies are concentrated on behavior control in pure intelligent and networked environment or in local single typical scene. Due to the mutual influence of different types of vehicles and the reaction delay of human drivers, the stability of mixed traffic flow is poor, and the advantage strategies such as platoon following and cooperative lane changing in pure networked environment cannot be realized. Therefore, the existing technology lacks systematic research on the influence process of traffic motion wave on mixed traffic flow in complex scenes. SUMMARY

[0005] The technical problem to be solved by the application is to provide a mixed flow environment underflow area vehicle active lateral and longitudinal collaborative control method, which is used for systematic control of the influence process of traffic motion wave on mixed traffic flow in complex scenes.

[0006] To solve the above technical problems, the technical scheme adopted by the application is as follows:

[0007] A mixed flow environment underflow area vehicle active lateral and longitudinal collaborative control method, comprising the following steps:

[0008] Step S1, collecting data, including collecting vehicle trajectory data and intelligent and networked mixed traffic flow trajectory data; the vehicle trajectory data is collected from a trajectory database, including vehicle number, vehicle speed, acceleration, spacing and lane information;

[0009] Step S2, using a characteristic parameter statistical analysis method, statistically analyzing the traffic flow characteristic parameters affected by the motion wave in the vehicle trajectory data and the intelligent network connection mixed traffic flow trajectory data, and revealing the behavior characteristics of the mixed traffic flow under the influence of the motion wave;

[0010] Step S3, constructing a mixed traffic flow vehicle cooperative control strategy under multi-lane complex scene elements, the cooperative control strategy including a vehicle lateral lane changing cooperative control strategy and a vehicle longitudinal following cooperative control strategy;

[0011] Step S4, designing a cooperative strategy cascade controller under a multi-lane scene;

[0012] Step S5, designing an optimization function, and deriving a working analytical process of the cascade controller according to a dynamic process of the cascade controller;

[0013] Step S6, in a merging area of a multi-lane with a ramp merging, based on the optimization function, on the basis of satisfying the basic vehicle running speed and acceleration interval, combining the rule constraints of lane changing behavior in the ramp merging process, and the constraint conditions of the cooperative control strategy, a cooperative control strategy of multi-system input and output is constructed.

[0014] Specifically, in step S2, the characteristic parameter statistical analysis method includes the following sub-steps:

[0015] S2.1, organizing the traffic flow characteristic parameters affected by the motion wave, and finding the corresponding data of each traffic characteristic parameter;

[0016] S2.2, drawing a statistical distribution frequency graph for each traffic characteristic parameter, obtaining the kurtosis value, standard deviation and mode of the traffic characteristic parameter through the distribution of the traffic characteristic parameter, and further obtaining the distribution characteristics of the traffic characteristic parameter;

[0017] S2.3, synthesizing the distribution characteristics of each traffic characteristic parameter to obtain the behavior characteristics of the mixed traffic flow under the influence of the motion wave;

[0018] The traffic flow characteristic parameters affected by the motion wave include: vehicle headway characteristics, vehicle headway characteristics, vehicle acceleration reaction delay time characteristics, and vehicle deceleration reaction delay time characteristics;

[0019] The behavior characteristics of the mixed traffic flow under the influence of the motion wave include: the overall characteristic parameter distribution of the vehicle under the mixed traffic flow state, the behavior mode of the vehicle, the driving mode and characteristics, the speed change reaction agility degree to the motion wave, and the traffic flow state change.

[0020] Specifically, in step S3, the vehicle lateral lane changing cooperative control strategy includes a vehicle non-mandatory lane changing behavior, and the strategy is as follows according to the condition of the vehicle in front of the target lane:

[0021] S3.1.1, the target lane front vehicle is performing speed change:

[0022] The lane changing process can be performed: the current driving vehicle is a human-driven vehicle and meets the following condition constraints,

[0023]

[0024] where d i (t) is the headway maintained by the current i vehicle at time t, v i (t) is the driving speed of the current i vehicle at time t, t ch is the total time of the lane changing process in the experiment, a chfront (t) is the speed change value of the target lane vehicle, d chfront (t) is the distance between the lane changing vehicle and the front vehicle in the target lane, d chback (t) is the distance between the lane changing vehicle and the rear vehicle in the target lane, d safe represents the minimum safety gap of vehicle lane changing;

[0025] The lane changing process cannot be performed: the current driving vehicle is an intelligent connected vehicle or the current driving is a human-driven vehicle, and does not meet the constraint condition of formula (1);

[0026] S3.1.2, the target lane front vehicle is not currently performing speed change:

[0027] The lane changing process can be performed: the current driving vehicle is a human-driven vehicle and meets the following constraints:

[0028]

[0029] The lane changing process cannot be performed: the current driving vehicle is an intelligent connected vehicle or the current driving is a human-driven vehicle, and does not meet the constraint condition of formula (2);

[0030] S3.1.3, there is no vehicle in front and behind the target lane: the current driving vehicle changes lane based on the free lane changing probability of the longitudinal safety distance model based on the headway in the connected environment.

[0031] Specifically, the vehicle lateral lane changing cooperative control strategy in step S3 also includes a vehicle forced lane changing behavior applied to ramp merging, according to the ramp intelligent connected vehicle speed limit control condition, the strategy is as follows:

[0032] S3.2.1, the vehicle subjected to ramp intelligent connected vehicle speed limit control:

[0033] The lane changing process can be performed, and the lane changing behavior is preferentially performed according to the lane changing model: the current driving vehicle exits the ramp and returns to normal speed, located in the acceleration lane;

[0034] Non-executable lane-changing process: the current driving vehicle does not perform the lane-changing process in the deceleration control process;

[0035] S3.2.2, vehicles not subject to ramp intelligent connected vehicle speed limit control:

[0036] The lane-changing process can be performed, and the lane-changing behavior will be preferentially performed according to the lane-changing model: the current driving vehicle is a human-driven vehicle and meets the following conditional constraints:

[0037]

[0038] Wherein, d0 represents the distance between the current vehicle and the front end of the acceleration lane, d begin , d end represent the distances between the current vehicle and the front end and the tail end of the acceleration lane; according to the constraint, the lane-changing behavior will be more urgent when the vehicle approaches the tail end of the acceleration lane;

[0039] Non-executable lane-changing process: the current driving vehicle is an intelligent connected vehicle or the current driving is a human-driven vehicle and does not meet the above constraint conditions;

[0040] S3.2.3, in the time period not disturbed by the motion wave: the current driving vehicle changes lanes with the free lane-changing probability of the longitudinal safety distance model based on the headway in the connected environment.

[0041] Further, in step S3, the longitudinal safety distance model based on the headway in the connected environment is represented as:

[0042] MFD(m,t)=v m *t tar (4)

[0043] Wherein, v m is the current driving speed of the host vehicle, t tar is the minimum following time of the target lane; (m, t) ∈ M × T, M × T is the full time domain and parameter vehicle, and MFD(m, t) is the minimum safety following distance of the host vehicle m at time t;

[0044] The vehicle free lane-changing behavior includes the following sub-steps:

[0045] S3.3.1, calculate the preferred alternative lane;

[0046] S3.3.2, calculate the safety speed of the vehicle located in the current lane, and judge whether the target lane meets the space requirement and speed requirement at the current time through the minimum safety distance formula;

[0047] S3.3.3, calculate the demand and probability of lane-changing through the lane-changing model;

[0048] S3.3.4, the speed of performing lane-changing action or calculating the gear shifting demand, whether the gear shifting is needed depends on the lane-changing request emergency degree.

[0049] Further, the vehicle longitudinal following cooperative control strategy in step S3 is determined by the vehicle lateral lane-changing cooperative control strategy described above, the number of vehicles changing lanes in each lane and the time and space position of vehicle lane-changing are determined, and after the lane-changing vehicle successfully changes lanes, the lane-changing vehicle driving behavior is integrated into the target lane vehicle following system. The vehicle longitudinal following cooperative control strategy is specifically as follows:

[0050] The intelligent connected vehicle speed control JAD strategy is executed, according to the traffic state of the current multi-lane scene, through the control starting strategy, the downstream traffic motion wave is generated, and the optimal control position x e is obtained according to the control algorithm a ; at the optimal position x e , the intelligent connected control vehicles are searched from each lane to the upstream respectively, and the process is executed in the full time domain, the main line intelligent connected vehicle cancels the free lane-changing behavior, that is, the lane-changing probability is 0;

[0051] Based on the uncertainty of human driving vehicle behavior in a multi-lane scene, the robustness control tuning speed is adopted to reduce the deviation of the control end position from the optimal control position x e , and the formula is as follows:

[0052] v a ″=β v v a (5)

[0053] Wherein, β v is a tuning coefficient, the tuning coefficient is smaller than the optimal value relative to the design ratio, so as to ensure that the position of the control vehicle end point satisfies:

[0054]

[0055] Wherein, x end-sw1 is the overlapping position of the downstream motion wave acceleration section and the ideal JAD head vehicle, so as to ensure the JAD flow of the control vehicle at this time;

[0056] When the nearest intelligent connected vehicle is indexed to different lanes, it is judged whether the execution vehicle position of the adjacent lane satisfies the condition:

[0057] |x ei -x ej |≤l (7)

[0058] Wherein, x ej , x ej are the positions of the intelligent connected vehicles along the lane line in different lanes, and l is the length of the vehicle body.

[0059] A variable speed limit control (VSL) strategy is performed for the upstream region to incorporate the merging vehicle, which is relatively consistent in multi-lane scenarios and single-lane situations due to the VSL control mechanism. The control input for the vehicle longitudinal car following behavior in a wide time domain is:

[0060]

[0061] wherein, is the target speed limit value at the current time step; is the recommended speed limit change rate at the current time step; K', K are gains for adjusting the amplitude of the control signal; ∈ is a very small positive number to avoid division by zero; ∈ α , ∈ β is the speed change response model parameter; is the speed limit state variable at the current time step; represents a function of the control gain, which determines the gain size of the control signal based on the change range of the speed limit state variable;

[0062] When the main line triggers the JAD strategy execution process, the control speed v conramp is adjusted to:

[0063]

[0064] wherein, ∈ ρ is the intelligent connected vehicle penetration rate model coefficient, ∈ j is the main line speed change model coefficient, ∈ ramp is the ramp model coefficient, β r is the ramp speed control rate, v ramp is the free flow speed limit under the ramp control-free state.

[0065] Further, in the vehicle longitudinal car following cooperative control strategy of step S3, the vehicle following is determined by the following formula:

[0066] v i (t+1) = v i (t) + a i (t)T m (10)

[0067]

[0068] s(t+1) = s(t) + Δt(v i (t+1) - v i (t)) (12)

[0069]

[0070] wherein, formula (10)-(12) calculate vehicle longitudinal position, speed and headway, x i (t) / x i (t+1), v i (t) / v i (t+1), s(t) / s(t+1) respectively represent vehicle position, speed and headway at t and t+1 time, a i (t) represents vehicle longitudinal acceleration at t time, Δt represents time interval between t and t+1 time, T m is simulation time step;

[0071] Formula (13)-(14) determine each lane vehicle following state system through wave model matrix, function F(X,U,V) is trigger function, U represents external disturbance factor, V represents traffic demand factor, X=(X1,X2…X i …X n ), n represents total number of vehicles in motion wave area, i represents the ith vehicle, s i represents headway between the ith vehicle and its front vehicle in motion wave area, u i represents acceleration of the ith vehicle in motion wave area, Δv i represents vehicle speed difference between the i-1th vehicle and the ith vehicle in motion wave area.

[0072] Specifically, in step S4, a cooperative strategy cascade controller under multi-lane scenario is designed, and the cascade controller system comprises: initial input flow state F wave1 under influence of motion wave, F wave2 , ramp merging flow F ramp , intermediate output flow following process r c under JAD control, flow lane changing process r change , following response process E v of upstream area flow, following response process E ramp of upstream area flow of ramp merging flow, control process r j executed by the JAD strategy, control process r v executed by the VSL strategy, and following response control E j in JAD process, flow state F cont1 and F cont2 under output control strategy.

[0073] Specifically, the solving process of the optimization function in step S5 is: using the solving library in python to solve the system multivariable extreme value search algorithm problem and the feedback controller design, the feedback controller is used to solve the output of the discrete variable, that is, the speed limit; the system multivariable extreme value search algorithm selects the target function mainly containing the total travel time TTT and the collision time TTC based on the optimization effect of the traffic efficiency and the traffic safety of the control strategy for optimizing the multi-lane traffic scene area, and the target function needs to be counted by lane in the multi-lane scene, and the specific form is:

[0074]

[0075] Wherein, J h is the target function value of each lane, H is the number of lanes, w1, w2 is the index evaluation gradient coefficient, x t，i is the current position of the ith vehicle at time t, v t，i is the driving speed of the ith vehicle at time t, l is the vehicle length, TTC t，i and TTT t，i are the corresponding collision time, that is, the relative position and speed ratio and the total travel time, and are the target values of TTC and TTT under the optimization target, and are the normalized bases of the corresponding index parameters.

[0076] Formulas (15) and (16) are used to calculate the optimal value of the target function of the whole road, and the optimal control strategy is selected; formulas (17) and (18) are used to evaluate the optimization effect of the control strategy on traffic efficiency and traffic safety.

[0077] Specifically, in step S6, a coordinated control system based on wide-area time, continuous variables and multi-system input and output is constructed in the scene of multi-lane and ramp merging, and the optimization function needs to meet the basic vehicle running speed and acceleration interval, and also needs to meet the rule constraints of lane changing behavior in the ramp merging process, that is, to meet the vehicle lateral lane changing coordinated control strategy and the vehicle longitudinal following coordinated control strategy, and also needs to meet the following constraint conditions of the coordinated control strategy, including:

[0078] S6.1, the control efficiency of the coordinated control strategy is constrained to ensure that the JAD control can well absorb the downstream transmitted motion wave during the execution process in each lane,

[0079]

[0080] t+k1=t start，1 =t jam，1 (21)

[0081] t+k'1=t end，1 =t l1 (22)

[0082] t+k2=t start，2 =t jam，2 (23)

[0083] t+k'2=t end，2 =t l2 (24)

[0084]

[0085] wherein, formula (19), (20) are used to constrain the JAD executed intelligent connected vehicle speed control strategy to completely eliminate the motion wave in each lane, and no secondary wave is generated, wherein, v a，1m / v a，2n , v e，1m / v e,2n , v f,1m / v f，2n , h a,1m / h a，2n , l1 / l2, v1 / v2 respectively represent the JAD execution head vehicle deceleration speed, compression wave speed, dissipation wave speed, headway when JAD vehicle decelerates, headway when JAD vehicle accelerates, lane free flow speed in the control process of lane 1 and lane 2;

[0086] Formula (21) to (24) respectively represent the start time (t start，1 , t start,2 ) and the end time (t end,1 , t end,2 ) of the JAD execution head vehicle deceleration control in lane 1 and lane 2, wherein, t jam,1 / t jam,2 respectively represent the motion wave formation time in lane 1 and lane 2, that is, the time when the first vehicle accelerates to recover to the free flow speed before deceleration when it is trapped in the motion wave, t l1 / t l2 respectively represent the time when the headway between the JAD execution head vehicle and its front vehicle in lane 1 / lane 2 is l1 / l2, x 1(m-1) (t+k'1) and x 2(n-1) (t+k'2) represent the longitudinal positions of the (m-1)th and (n-1)th vehicles in lane 1 and lane 2 at (t+k'1) and (t+k'2) time;

[0087] Formula (25), (26) gives the constraint that the control speed in the JAD strategy needs to satisfy the constraint of not generating secondary wave;

[0088] S6.2, for the ramp in speed control of the radiation constraint, ensure that in the process of controlling the merging ramp, the secondary ramp congestion BOR phenomenon caused by the ramp flow back overflow does not appear, the following formula is used as the constraint condition of the intelligent connected vehicle behavior control in the control coordination algorithm:

[0089]

[0090] Wherein, l ramp is the length of the ramp, t w is the control time;

[0091] S6.3, safety constraint, using the following formula to constrain the vehicle deceleration to meet the vehicle maximum deceleration limit:

[0092] v(t+k-1)-v(t+k)≤Δt·d max (28)

[0093] Wherein, d max Indicates the maximum deceleration of the vehicle, and Δt indicates the time interval between t+k-1 and t+k time;

[0094] S6.4, variable speed limit control constraint: the difference between the speed limits displayed on the same variable speed limit sign in two consecutive time steps shall not exceed K, and the speed difference between the speed limit of variable speed limit control and JAD speed limit control in the same time period shall not exceed K:

[0095]

[0096] Wherein, Is the speed of variable speed limit control at time t; Is the speed of JAD speed limit control at time t.

[0097] The above technical scheme is adopted in the present application compared with the prior art, and the following technical effects are obtained:

[0098] 1. By intelligent connected vehicle speed control and variable speed limit area control, the present application adjusts the speed and distance between vehicles when traffic flow meets motion wave, effectively slows down traffic motion wave and improves traffic flow efficiency.

[0099] 2. By intelligent connected vehicle behavior control, the present application limits the influence of vehicle speed change caused by motion wave and main line control strategy on ramp merging area vehicle flow and speed, avoids traffic conflict in the merging process, designs lateral lane changing rules to reduce lateral disturbance in interlaced area, and considers the interference of lateral and longitudinal behavior in the area to the traffic system model.

[0100] 3, The mixed flow environment under the merging area vehicle active lateral and longitudinal collaborative control method provided by the application reduces traffic conflicts in the pursuit of collaborative optimization of traffic safety and efficiency, so as to achieve the optimal goal of the system. The reliability of the strategy in the complex environment is verified through traffic simulation experiment, and a collaborative control system based on wide-area time, continuous variables and discrete variables and multi-system input and output is constructed. BRIEF DESCRIPTION OF DRAWINGS

[0101] Figure 1 The mixed flow environment under the merging area vehicle active lateral and longitudinal collaborative control method provided by the application is a hierarchical system structure diagram under the collaborative control framework in the multi-lane scene.

[0102] Figure 2 The mixed flow environment under the merging area vehicle active lateral and longitudinal collaborative control method provided by the application is a hierarchical system structure diagram under the collaborative control framework in the multi-lane scene. DETAILED DESCRIPTION

[0103] In order to make the purpose, technical scheme and advantages of the application more clear, the technical scheme of the application will be further described in detail below in combination with the drawings. The described embodiments are only a part of the embodiments involved in the application. All non-innovative embodiments of other researchers in the field on the basis of the embodiments belong to the protection scope of the application.

[0104] The application provides a mixed flow environment under the merging area vehicle active lateral and longitudinal collaborative control method, which comprises the following steps:

[0105] Step S1: collecting data, including collecting vehicle trajectory data and intelligent network connection mixed traffic flow trajectory data; the vehicle trajectory data is collected from a trajectory database, including vehicle number, vehicle speed, acceleration, spacing and lane information;

[0106] Step S2: using a feature parameter statistical analysis method, statistically analyzing the traffic flow feature parameters affected by the motion wave in the vehicle trajectory data and the intelligent network connection mixed traffic flow trajectory data, and revealing the behavior characteristics of the mixed traffic flow under the influence of the motion wave;

[0107] Step S3: constructing a mixed traffic flow vehicle collaborative control strategy under a multi-lane complex scene element, wherein the collaborative control strategy comprises a vehicle lateral lane changing collaborative control strategy and a vehicle longitudinal following collaborative control strategy;

[0108] Step S4: designing a collaborative strategy cascade controller in a multi-lane scene;

[0109] Step S5: designing an optimization function, and deriving a working analytical process of the cascade controller according to the dynamic process of the cascade controller;

[0110] Step S6: In the merging area of multiple lanes and ramp merging, based on the above optimization function, on the basis of meeting the basic vehicle running speed and acceleration interval, combined with the rule constraints of lane changing behavior in the ramp merging process, and the constraint conditions of the cooperative control strategy, a cooperative control strategy of multiple system input and output is constructed.

[0111] Specifically, in one embodiment of the present application,

[0112] Step S1: Trajectory data is collected from the UTE (The Ubiquitous Traffic Eyes) trajectory database in Nanjing, China. The length of the road section is about 386m. The number of lanes of the ramp bottleneck is reduced from five lanes at a distance of 150m from the starting position to four lanes, and finally to three lanes at a distance of 270m from the starting position. The video is taken during the morning peak period, with a time accuracy of 0.1s and a position accuracy of 0.01m. A total of 608 cars and 13 large vehicles are detected during the collection period. The data set is recorded in the form of a list, with each row representing the vehicle driving information of a detected vehicle at that time frame, and each column representing the detected data type, including the speed, acceleration, spacing, lane, etc. of each vehicle. The data types are shown in the following table.

[0113]

[0114] Step S2: The main traffic flow characteristic parameters affected by the motion wave in the UTE driving vehicle trajectory data and the intelligent network connected hybrid traffic flow trajectory data are statistically analyzed by using the characteristic parameter statistical analysis method. These include the vehicle headway feature, the vehicle headway time feature, and the acceleration / deceleration reaction delay time feature. The behavior characteristics of the hybrid flow under the influence of the motion wave are revealed.

[0115] Step S3: A hybrid traffic flow system cooperative control strategy under the complex scene elements of multiple lanes is constructed.

[0116] As shown in Figure 1 , a vehicle cooperative control strategy architecture for jointly improving traffic safety and efficiency in a multi-lane complex scenario is shown. The strategy includes a vehicle lateral lane changing cooperative control strategy and a vehicle longitudinal following cooperative control strategy.

[0117] On the one hand, the vehicle lateral lane changing cooperative control strategy and the system execution rules consider the collision risk intensity of the target vehicle and the vehicles before and after it in the current lane, the collision risk intensity of the vehicles before and after it in the adjacent lane, and the available speed improvement level of the adjacent lane, and set the following vehicle lateral lane changing rules:

[0118] S3.1, Non-mandatory lane changing behavior:

[0119] S3.1.1, the target lane vehicle is executing speed change:

[0120] (i) can perform the lane-changing process: the current vehicle is human-driven and meets the constraint condition.

[0121]

[0122] where d i (t) is the headway distance maintained by the current i vehicle at time t, v i (t) is the driving speed of the current i vehicle at time t, t ch is the total duration of the lane-changing process in the experiment, a chfront (t) is the speed change value of the target lane vehicle, d chfront (t) is the distance between the lane-changing vehicle and the front vehicle in the target lane, d chback (t) is the distance between the lane-changing vehicle and the rear vehicle in the target lane, d safe represents the minimum safety gap of vehicle lane changing.

[0123] (ii) cannot perform the lane-changing process: the current vehicle is an intelligent connected vehicle or the current vehicle is a human-driven vehicle and does not meet the above constraint condition.

[0124] S3.1.2, the target lane front vehicle does not currently perform speed change:

[0125] (i) can perform the lane-changing process: the human-driven vehicle and meets the constraint condition.

[0126]

[0127] (ii) cannot perform the lane-changing process: the current vehicle is an intelligent connected vehicle or the current vehicle is a human-driven vehicle and does not meet the above constraint condition.

[0128] S3.1.3, there is no vehicle in front of the target lane: the current vehicle changes lane with the free lane-changing probability of the longitudinal safety distance model based on the headway in the connected environment.

[0129] S3.2 Forced lane-changing behavior (ramp merging):

[0130] S3.2.1, vehicles subject to ramp intelligent connected vehicle speed control:

[0131] (i) can perform the lane-changing process and will preferentially perform high-probability lane-changing behavior according to the lane-changing model: exit the ramp and return to normal speed, located in the acceleration lane.

[0132] (ii) cannot perform the lane-changing process: does not perform the lane-changing process during the speed reduction control process.

[0133] S3.2.2, vehicles not subject to ramp intelligent connected vehicle speed control:

[0134] (i) can perform the lane-changing process and will prefer to perform high-probability lane-changing behavior according to the lane-changing model: the current vehicle is a human-driven vehicle and meets the condition constraints:

[0135]

[0136] where d0 represents the distance between the current vehicle and the front end of the acceleration lane, d end represents the distance between the current vehicle and the tail end of the acceleration lane. According to the constraint, the lane-changing behavior is more urgent when the vehicle approaches the tail end of the acceleration lane.

[0137] (ii) cannot perform the lane-changing process: the current vehicle is an intelligent connected vehicle or the current vehicle is a human-driven vehicle and does not meet the above constraint conditions.

[0138] S3.2.3, in the time period not disturbed by the motion wave: the current vehicle changes lanes with the free lane-changing probability of the longitudinal safety distance model based on the headway in the connected environment.

[0139] On the other hand, the vehicle longitudinal following cooperative control strategy and system execution rules are as follows:

[0140] The number of lane-changing vehicles and the space-time position of vehicle lane-changing can be determined by the vehicle lateral lane-changing cooperative control strategy, and the driving behavior of the lane-changing vehicle can be integrated into the target lane vehicle following system after the lane-changing vehicle successfully changes lanes. The vehicle following characteristics are determined by the following formula,

[0141] v i (t+1)=v i (t)+a i (t)T m

[0142]

[0143] s(t+1)=s(t)+Δt(v i (t+1)-v i (t))

[0144] where x i (t) / x i (t+1), v i (t) / v i (t+1), s(t) / s(t+1) represent the position, speed and headway of the vehicle at time t and t+1, a i (t) represents the longitudinal acceleration of the vehicle at time t, Δt represents the time interval between time t and t+1, T m is the simulation time step; the above formula is used to calculate the longitudinal position, speed and headway of the vehicle;

[0145]

[0146] The car-following state system of each lane is determined by the wave model matrix, the function F(X, U, V) is a trigger function, U represents external disturbance factors, V represents traffic demand factors, X=(X1, X2...X i ...X n ), n represents the total number of vehicles in the moving wave area, i represents the ith vehicle, s i represents the headway between the ith vehicle and the vehicle in front of it in the moving wave area, u i represents the acceleration of the ith vehicle in the moving wave area, Δv i represents the vehicle speed difference between the i-1th vehicle and the ith vehicle in the moving wave area.

[0147] The control logic of the JAD strategy in a multi-lane scenario is as follows: according to the current traffic state of the multi-lane scenario, the generation of the downstream traffic motion wave is perceived by controlling the start-up strategy. At this time, the optimal control position x e and the control speed v a are obtained according to the control algorithm, and at the optimal position x e , intelligent connected vehicles are searched from each lane upstream, and this process is executed in the full time domain. The main line intelligent connected vehicle cancels the free lane changing behavior (lane changing probability is 0). If there is a human-driven vehicle, the robust control tuning speed v a ″=β v v a .

[0148] In this embodiment, in the ramp merging scenario of a two-lane, the tuning parameter value

[0149] When performing specific JAD control, considering the uncertainty of human-driven vehicle behavior in a multi-lane scenario, this coefficient is relatively smaller than the optimal value, to ensure that the position of the control endpoint satisfies:

[0150]

[0151] where x end-sw1 is the overlapping position of the downstream motion wave acceleration section and the ideal JAD leading vehicle, to ensure that the JAD flow of the control vehicle at this time eventually stabilizes and dissipates the speed change caused by the motion wave. At the same time, when the nearest intelligent connected vehicle is indexed to a different lane, the following judgment process is performed: whether the position of the executing vehicle in the adjacent lane satisfies the condition:

[0152] |x ei -x ej |≤l

[0153] where xei x ej Let l represent the lane position of the connected vehicle implementing the control strategy in different lanes, and l be the vehicle length. If the constraints are met, the vehicle is controlled directly using a tuned speed v. a "Perform JAD control; if the condition constraints are not met, then relative to x..." e The vehicle at the closest distance adopts the tuned speed v in advance a "Perform JAD control; subsequent vehicles maintain normal speed until the constraints are met, then adjust the speed and enter the JAD control process."

[0154] When implementing a Variable Speed ​​Limit (VSL) strategy for merging vehicles in the upstream area, since the VSL control mechanism is relatively consistent in multi-lane and single-lane scenarios, the control input for the longitudinal following behavior of the vehicle in the wide time domain during implementation is as follows:

[0155]

[0156] Specifically, during the time domain when the main line triggers the JAD strategy execution process, the control speed v of the intelligent connected vehicle traveling on the ramp is... conramp Adjusted to:

[0157]

[0158] Where ∈ α For the velocity change response model parameters, ∈ ρ For the penetration rate model coefficients of intelligent connected vehicles, ∈ j The coefficients of the main linear velocity variation model, ∈ ramp For the ramp model coefficients, β r For ramp speed control rate, v ramp Speed ​​limits are set for free-flowing ramps in uncontrolled conditions.

[0159] Thus, the strategy employs a multi-lane JAD control method to adjust the driving behavior of following vehicles by controlling the driving behavior of the lead vehicle in parallel, thereby eliminating the motion wave process. The upstream variable speed limit control mitigates the secondary wave interference caused by JAD speed changes on the main line. At the same time, the speed control of connected vehicles on the ramps mitigates the lateral interference in the merging zone within the motion wave propagation time domain.

[0160] Step S4: Design a cascade controller for multi-lane scenarios.

[0161] like Figure 2 As shown, this cascaded controller system is influenced by the traffic flow status (F) of each lane under the influence of the initial input motion wave. wave1 F wave2 ), the traffic flow merging from the ramp (F ramp Traffic flow state (F) under the action of output control strategycont The intermediate output JAD controls the vehicle flow following process (r) c ), traffic lane changing process (r change The following response process of traffic merging from the upstream area (E) v The response process of merging traffic on ramps to mainline vehicles, and the car-following response process of merging traffic in the upstream area (E). ramp JAD executes control process (r) j VSL executes control procedures (r) v ) and follow-car response control (E) during the execution of JAD j It consists of structures such as )

[0162] Step S5: Design the optimization function and derive its analytical working process based on the dynamic process of the cascaded controller. Consider the main core control parameters affected by cooperative control in the hybrid flow system, and construct the traffic system throughout the entire time domain during the control process.

[0163] In the time-domain recursive process, this embodiment takes a 3-lane road as an example. The derivation process of the main traffic flow parameter variables is shown in the following formula:

[0164] S1(t+1)=S1(t)+BV1(t)

[0165] S2(t+1)=S2(t)+BV2(t)

[0166] S3(t+1)=S3(t)+BV3(t)

[0167] In the formula,

[0168]

[0169]

[0170] Where x(t) / x(t+k) and v x (t) / v x (t+k) and s(t) / s(t+k) represent the longitudinal position, velocity, and headway of the vehicle at times t and t+k, respectively. x (t) represents the longitudinal acceleration of the vehicle at time t, and Δt represents the time interval between t and t+k. d Let S1(t) / S1(t+1), S2(t) / S2(t+1), and S3(t) / S3(t+1) represent the headway matrices of all vehicles in lanes 1, 2, and 3 (acceleration lanes) at times t and t+k, respectively. Let V1(t), V2(t), and V3(t) represent the velocity matrices of all vehicles in lanes 1, 2, and 3 at time t. Let B represent the coefficient matrix. and This indicates that lane 1 is at position K1+k. 21 -k 12 Vehicle, lane 2, number (K2-k) 21 +k 12 +k 32 Vehicles and lanes, K3-k 32 The distance between the front ends of the vehicles. and This indicates that lane 1 is at position K1+k. 21 -k 12 Vehicle, lane 2, number (K2-k) 21 +k 12 +k 32 Vehicles and lanes, K3-k 32 The longitudinal speed of the vehicle. K1, K2, and K3 are the original total number of vehicles in lanes 1, 2, and 3, respectively. ij The number of vehicles that changed lanes from lane i to lane j to enter or exit the lane.

[0171] The expansion is to show the vehicle status of all vehicles in lane 1, lane 2, and acceleration lane 3 at all times, as shown in the following formula.

[0172] v i (t+k)=v i (t)+a i (t)T m

[0173]

[0174] S1 (t)= A1 S1(t)+ BV1 (t)

[0175] S2 (t)= A2 S2(t)+ BV2 (t)

[0176] S3 (t)= A3 S3(t)+ BV3 (t)

[0177] Furthermore, it can be derived that in the formula,

[0178]

[0179] In the formula, T p H represents the prediction time domain length; i This is the system identification matrix, used to index specific vehicles in the entire system; Q p，ij Indicates at t+T p The number of vehicles changing lanes from lane j to lane i at any given time. Rp，ij represents the number of vehicles that change lanes from lane j to lane i at time t+T p

[0180] For the target function optimization solving process of this control strategy, the solving library in python is used to analyze the system multivariable extreme value search algorithm problem and feedback controller design. The feedback controller solving provides the output of discrete variables (i.e. speed limit). Similarly, for the traffic scene area optimized by the algorithm, the optimization effect of the management and control strategy on traffic efficiency and traffic safety is still considered, and the target function selected mainly includes the total travel time (TTT) and the collision time (TTC).

[0181] In the multi-lane scene, the target function needs to be counted by lane. In this embodiment, three lanes are preferred, and the specific form is:

[0182] Min J = J1 + J2 + J3

[0183]

[0184] Where J h is the target function value of each lane, H is the number of lanes, w1 and w2 are index evaluation gradient coefficients, x t，i is the current position of the ith vehicle at time t, v t，i is the driving speed of the ith vehicle at time t, l is the vehicle length, TTC t，i and TTT t,i are the corresponding collision time, i.e. the relative position and speed ratio and the total travel time, and are the target values of TTC and TTT under the optimization target, and are the normalized bases of the corresponding index parameters.

[0185] Step S6: In the multi-lane and ramp merging scene, the traffic efficiency and traffic safety in the control area under the regional strategy need more strict constraints. The above target function needs to meet the basic vehicle running speed and acceleration interval, and also needs to meet the lane changing behavior rule constraints in the ramp merging process. At the same time, the following four constraints based on the cooperative control strategy need to be met:

[0186] S6.1, execute the cooperative control strategy control efficiency constraint to ensure that the JAD control can well absorb the downstream transmitted motion wave in the execution process of each lane. Specifically:

[0187] The intelligent connected vehicle speed control strategy (JAD) executed using the following formula can completely eliminate the motion wave in each lane and not produce a secondary wave, ​

[0188]

[0189] where v a,1m / v a,2n , v e，1m / v e，2n , v f,1m / v f，2n , h a，1m / h a，2n , l1 / l2, v1 / v2 represent the deceleration speed of the leading vehicle, the compression wave speed, the dissipation wave speed, the headway of the leading vehicle when decelerating, the headway of the leading vehicle when accelerating, and the free flow speed of the lane, respectively.

[0190] The following equations represent the start time (t start，1 , t start，2 ) and the end time (t end，1 , t end，2 ) of the deceleration control of the leading vehicle in lane 1 and lane 2, respectively.

[0191] t+k1=t start，1 =t jam，1

[0192] t+k'1=t end，1 =t l1

[0193] t+k2=t start，2 =t jam，2

[0194] t+k'2=t end，2 =t l2

[0195] where t jam，1 / t jam，2 represent the time when the motion wave is formed in lane 1 and lane 2, i.e., the time when the first vehicle accelerates to the free flow speed before deceleration, t l1 / t l2 represent the time when the headway between the leading vehicle and the preceding vehicle in lane 1 / lane 2 is l1 / l2, x 1(m-1) (t+k'1) and x 2(n-1) (t+k'2) represent the longitudinal positions of the (m-1)th and (n-1)th vehicles in lane 1 and lane 2 at times (t+k'1) and (t+k'2).

[0196] The following equations give the constraints that the control speed in the JAD strategy in lane 1 and lane 2 needs to satisfy to not generate secondary waves:

[0197] where v a,1m / v a,2n , v e，1m / v e，2n , v f,1m / v f，2n , h a，1m / h a，2n , l1 / l2, v1 / v2 represent the deceleration speed of the leading vehicle, the compression wave speed, the dissipation wave speed, the headway of the leading vehicle when decelerating, the headway of the leading vehicle when accelerating, and the free flow speed of the lane, respectively.

[0190] The following equations represent the start time (t start，1 , t start，2 ) and the end time (t end，1 , t end，2 ) of the deceleration control of the leading vehicle in lane 1 and lane 2, respectively.

[0191] t+k1=t start，1 =t jam，1

[0192] t+k'1=t end，1 =t l1

[0193] t+k2=t start，2 =t jam，2

[0194] t+k'2=t end，2 =t l2

[0195] where t jam，1 / t jam，2 represent the time when the motion wave is formed in lane 1 and lane 2, i.e., the time when the first vehicle accelerates to the free flow speed before deceleration, t l1 / t l2 represent the time when the headway between the leading vehicle and the preceding vehicle in lane 1 / lane 2 is l1 / l2, x 1(m-1) (t+k'1) and x 2(n-1) (t+k'2) represent the longitudinal positions of the (m-1)th and (n-1)th vehicles in lane 1 and lane 2 at times (t+k'1) and (t+k'2).

[0196] The following equations give the constraints that the control speed in the JAD strategy in lane 1 and lane 2 needs to satisfy to not generate secondary waves:

[0197]

[0198] S6.2, for the radiation constraint of the on-ramp speed control, ensure that there is no on-ramp flow back overflow caused by BOR phenomenon in the process of controlling the merging on-ramp, aiming at the constraint condition of the on-ramp intelligent connected vehicle behavior control in the cooperative algorithm, wherein l ramp is the on-ramp length, and the formula is as follows:

[0199]

[0200] S6.3, safety constraint: aiming at restricting the vehicle deceleration to meet the vehicle maximum deceleration limit, d max represents the maximum deceleration of the vehicle, and the formula is as follows:

[0201] v(t+k-1)-v(t+k)≤Δt·d max

[0202] S6.4, variable speed limit control constraint:

[0203] In this embodiment, preferably, the difference between the speed limits displayed on the same variable information sign in two consecutive time steps should not exceed 10 km / h. In the same time period, the speed difference between the variable information control speed limit and the JAD speed limit control should not exceed 10 km / h:

[0204]

[0205] wherein, is the speed of the variable speed limit control at time t; is the speed of the JAD speed limit control at time t.

[0206] Thus, a cooperative control system based on wide-area time, continuous variables and discrete variables and multi-system input and output is constructed.

[0207] The above merely describes specific embodiments of the present application and is not used to limit the present application, and any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the protection scope of the present application.

Claims

1. A method for active longitudinal and lateral cooperative control of vehicles in a merging area in a mixed traffic environment, characterized in that, The method comprises the following steps: Step S1, collecting data, including collecting vehicle trajectory data and intelligent networked mixed traffic flow trajectory data; the vehicle trajectory data is collected from a trajectory database, including vehicle number, vehicle speed, acceleration, spacing and lane information; Step S2, using a feature parameter statistical analysis method, statistically analyzing the traffic flow feature parameters affected by the motion wave in the vehicle trajectory data and the intelligent networked mixed traffic flow trajectory data, and revealing the behavior characteristics of the mixed traffic flow under the influence of the motion wave; Step S3, constructing a mixed traffic flow vehicle cooperative control strategy under multi-lane complex scene elements, the cooperative control strategy including a vehicle lateral lane changing cooperative control strategy and a vehicle longitudinal following cooperative control strategy; Step S4, designing a cooperative strategy cascade controller in a multi-lane scene; Step S5, designing an optimization function, and deriving a working analysis process of the cascade controller according to a dynamic process of the cascade controller; The solving process of the optimization function is: using a solving library in python to analyze a system multivariable extreme value search algorithm problem and a feedback controller design, the feedback controller is used to solve the output of discrete variables, i.e. speed limit; the system multivariable extreme value search algorithm is used for optimizing the traffic scene area of multiple lanes, and based on the optimization effect of the control strategy on traffic efficiency and traffic safety, the target function mainly includes the efficiency index total travel time TTT and the collision time TTC, and in the multi-lane scene, the target function needs to be counted by lane, and the specific form is: wherein J h is the target function value of each lane, H is the number of lanes, w1, w2 are index evaluation gradient coefficients, x t，i is the current position of the ith vehicle at time t, v t，i is the driving speed of the ith vehicle at time t, l is the vehicle body length, TTC t，i and TTT t，i are the corresponding collision times, i.e., the relative position and speed ratio and the total travel time, and are the target values of TTC and TTT under the optimization target, and are the normalization bases of the corresponding index parameters; Formulas (15) and (16) are used to calculate the optimal value Min J of the target function of the whole road, and the optimal control strategy is selected; formulas (17) and (18) are used to evaluate the optimization effect of the control strategy on traffic efficiency and traffic safety; Step S6, in a merging area of multiple lanes and ramp merging, based on the above optimization function, on the basis of meeting the basic vehicle running speed and acceleration interval, combining the rule constraints of lane changing behavior in the ramp merging process and the constraint conditions of the cooperative control strategy, a cooperative control strategy of multiple system input and output is constructed; In the scene of multiple lanes and ramp merging, a cooperative control system of multiple system input and output based on wide-area time, continuous variables and discrete variables is constructed, the optimization function meets the basic vehicle running speed and acceleration interval, also meets the rule constraints of lane changing behavior in the ramp merging process, i.e. meets the vehicle lateral lane changing cooperative control strategy and the vehicle longitudinal following cooperative control strategy, and also meets the following constraint conditions of the cooperative control strategy, including: S6.1, performing constraints on the control efficiency of the cooperative control strategy to ensure that the JAD control can well absorb the motion wave transmitted downstream in the execution process of each lane, t + k1 = t start，1 = t jam，1 (21) t + k'1 = t end，1 = t l1 (22) t + k2= t start,2 = t jam,2 (23) t + k'2= t end,2 = t l2 (24) wherein, formula (19) and (20) are used to constrain the executed intelligent networked vehicle speed control strategy JAD to completely eliminate the motion wave in each lane, and no secondary wave is generated, wherein, v a，1m / v a，2n , v e，1m / v e，2n , v f，1m / v f，2n , h a，1m / h a,2n , l1 / l2, v1 / v2 respectively represent the JAD execution head vehicle deceleration speed, compression wave speed, dissipation wave speed, headway when JAD vehicle decelerates, headway when JAD vehicle accelerates, lane free flow speed in the control process of lane 1 and lane 2. Equations (21) to (24) represent the start time (t start,1 , t start，2 ) and the end time (t end,1 , t end，2 ) of the head vehicle deceleration control by JAD in lane 1 and lane 2, respectively, where t jam,1 / t jam，2 represent the motion wave formation time in lane 1 and lane 2, i.e., the time when the first vehicle accelerates to the free flow speed before deceleration when trapped in the motion wave, t l1 / t l2 represent the time when the head vehicle and the preceding vehicle in lane 1 / lane 2 have a headway of l1 / l2, x 1(m-1) (t+k'1) and x 2(n-1) (t+k'2) represent the longitudinal positions of the (m-1)th and (n-1)th vehicles in lane 1 and lane 2 at times (t+k'1) and (t+k'2), respectively. Formulas (25) and (26) give the constraint that the control speed in the JAD strategy of lane 1 and lane 2 needs to meet the constraint of not producing a secondary wave; S6.2, for the radiation constraint of the speed control in the ramp, to ensure that there is no secondary ramp congestion BOR phenomenon caused by the ramp flow back overflow in the process of controlling the merging ramp, the following formula is used as the constraint condition of the intelligent connected vehicle behavior control in the control coordination algorithm: wherein, l ramp is the ramp length; ∈ α is the speed change response model parameter, ∈ ρ is the intelligent connected vehicle penetration rate model coefficient, ∈ j is the main line speed change model coefficient, ∈ ramp is the ramp model coefficient, β r is the ramp speed control rate, v ramp is the free flow speed limit under the ramp uncontrolled state, t w is the control time; S6.3, safety constraint, using the following formula to constrain the vehicle deceleration to meet the vehicle maximum deceleration limit: v(t+k-1) - v(t+k) < At - d max (28) where d max denotes the maximum deceleration of the vehicle, and Δt denotes the time interval between the time instants t+k-1 and t+k. S6.4, variable speed limit control constraint: the difference between the speed limits displayed on the same variable speed limit sign in two consecutive time steps shall not exceed K, and the speed difference between the variable speed limit control and the JAD speed limit control in the same time period shall not exceed K: wherein, V(t) is the speed of the variable speed limit control at time t; VJAD(t) is the speed of the JAD speed limit control at time t.

2. The method of claim 1, wherein In step S2, the characteristic parameter statistical analysis method comprises the following sub-steps: S2.1, sorting the traffic flow characteristic parameters affected by the motion wave, and finding the corresponding data of each traffic characteristic parameter; S2.2, drawing a statistical distribution frequency diagram for each traffic characteristic parameter, obtaining the kurtosis value, standard deviation and mode of the traffic characteristic parameter through the distribution of the traffic characteristic parameter, and further obtaining the distribution characteristics of the traffic characteristic parameter; S2.3, synthesizing the distribution characteristics of each traffic characteristic parameter to obtain the behavior characteristics of the mixed traffic flow under the influence of the motion wave; The traffic flow characteristic parameters affected by the motion wave include: vehicle headway characteristics, vehicle headway characteristics, vehicle acceleration reaction delay time characteristics and vehicle deceleration reaction delay time characteristics; The behavior characteristics of the mixed traffic flow under the influence of the motion wave include: the overall characteristic parameter distribution of the vehicle in the mixed traffic flow state, the behavior mode of the vehicle, the driving mode and characteristics, the reaction sensitivity to the speed change caused by the motion wave, and the traffic flow state change.

3. The method of claim 2, wherein The vehicle lateral lane changing coordination control strategy in step S3 comprises a vehicle non-forced lane changing behavior, and the strategy is as follows according to the condition of the vehicle in front of the target lane: S3.1.1, the vehicle in front of the target lane is performing speed change: The lane changing process can be performed: the current driving vehicle is a human-driven vehicle and meets the following condition constraints, wherein d i (t) is the headway distance maintained by the current i vehicle at time t from the front vehicle, v i (t) is the driving speed of the current i vehicle at time t, t ch is the total time length of the lane changing process in the experiment, a chfront (t) is the speed change value of the target lane vehicle, d chfront (t) is the distance between the lane changing vehicle and the front vehicle in the target lane, d chback (t) is the distance between the lane changing vehicle and the rear vehicle in the target lane, d safe represents the minimum safety gap of vehicle lane changing; The lane changing process cannot be performed: the current driving vehicle is an intelligent connected vehicle or the current driving is a human-driven vehicle, and does not meet the constraint condition of formula (1); S3.1.2, the vehicle in front of the target lane is not currently performing speed change: The lane changing process can be performed: the current driving vehicle is a human-driven vehicle and meets the following constraint conditions: The lane changing process cannot be performed: the current driving vehicle is an intelligent connected vehicle or the current driving is a human-driven vehicle, and does not meet the constraint condition of formula (2); S3.1.3, there is no vehicle in front of and behind the target lane: the current driving vehicle changes lane according to the free lane changing probability of the longitudinal safety distance model based on the headway in the connected environment.

4. The method of claim 3, wherein The vehicle lateral lane changing coordination control strategy in step S3 also comprises a vehicle forced lane changing behavior, which is applied to the ramp merging, and the strategy is as follows according to the intelligent connected vehicle speed limit control condition of the ramp: S3.2.1, the vehicle controlled by the intelligent connected vehicle speed limit of the ramp: The lane changing process can be performed, and the lane changing behavior is preferentially performed according to the lane changing model: the current driving vehicle exits the ramp and returns to normal speed, and is located in the acceleration lane; Non-executable lane-changing process: the current driving vehicle does not execute the lane-changing process in the deceleration control process; S3.2.2, vehicles not subject to ramp intelligent connected vehicle speed limit control: Can perform lane-changing process, and will preferentially perform lane-changing behavior according to the lane-changing model: the current driving vehicle is a human-driven vehicle and meets the following conditional constraints: where d0represents the distance between the current vehicle and the front end of the acceleration lane, d begin , d end represent the distances between the current vehicle and the front end and the rear end of the acceleration lane; according to the constraint, the lane-changing behavior is forced to be more urgent when the vehicle approaches the rear end of the acceleration lane; Non-executable lane-changing process: the current driving vehicle is an intelligent connected vehicle or the current driving is a human-driven vehicle and does not meet the above constraint conditions; S3.2.3, in the time period not disturbed by the motion wave: the current driving vehicle changes lanes with the free lane-changing probability of the longitudinal safety distance model based on the vehicle headway in the connected environment.

5. The method of claim 4, wherein In the step S3, the longitudinal safety distance model based on the vehicle headway in the connected environment is represented as: MFD(m, t) = v m * t tar (4) where v m is the current speed of the host vehicle, t tar is the minimum following time distance of the target lane; (m, t) ∈ M × T, M × T is the full time domain and parameter vehicle, and MFD(m, t) is the minimum safe following distance of the host vehicle m at time t. The vehicle free lane-changing behavior includes the following sub-steps: S3.3.1, calculate the preferred alternative lane; S3.3.2, calculate the safety speed of the vehicle assuming it is located in the current lane, and determine whether the target lane meets the space requirements and speed requirements at the current time through the minimum safety distance formula; S3.3.3, calculate the demand and probability of lane-changing through the lane-changing model; S3.3.4, perform the lane-changing action or calculate the speed of the speed change demand, whether the speed change is needed depends on the urgency of the lane-changing request.

6. The method of claim 5, wherein The vehicle longitudinal following cooperative control strategy in step S3 is determined by the vehicle lateral lane-changing cooperative control strategy, the number of lane-changing vehicles in each lane and the time and space point position of the vehicle lane-changing are determined, and after the vehicle lane-changing is successful, the lane-changing vehicle driving behavior is integrated into the target lane vehicle following system. The vehicle longitudinal following cooperative control strategy is as follows: The intelligent network connected vehicle speed control JAD strategy is executed, according to the traffic state of the current multi-lane scene, through control starting strategy, the downstream traffic motion wave is perceived to be generated, and the optimal control position x is obtained according to the control algorithm e With the control speed v a ; at the optimal position x e Lane respectively from each lane to the upstream, the process is executed in the full time domain, the main line intelligent network connected vehicle cancels the free lane changing behavior, that is, the lane changing probability is 0; Based on the uncertainty of human driving vehicle behavior in multi-lane scenarios, robust control is used to tune the speed to reduce the deviation of the control end position from the optimal control position x e , which is given by the following formula: v a "= β v v a (5) where β v is a tuning coefficient, which sets the position to be ensured to control the end of the vehicle satisfies: where x end-sw1 is the position of the downstream moving wave acceleration section coinciding with the ideal JAD head vehicle to ensure that the JAD flow controls the vehicle at this time; When the nearest intelligent connected vehicle is determined for different lane indexes, it is judged whether the execution vehicle position of the adjacent lane meets the conditions: x ei -x ej |≤l(7) wherein x ei , x ej is the position of the vehicle along the lane line, and l is the length of the vehicle. The variable speed limit control VSL strategy for the upstream area is executed for the merging vehicle, since the VSL control mechanism is relatively consistent in multi-lane and single-lane scenarios, the control input of the VSL control strategy for the vehicle longitudinal following behavior in the wide time domain is: wherein, is the target speed limit value at the current time step; is the proposed speed limit change rate at the current time step; K, K' are gains used to adjust the magnitude of the control signal; ∈ is a very small positive number used to avoid division by zero; ∈ α , ∈ β is the speed change response model parameter; is the speed limit state variable at the current time step; is a function of the control gain that determines the magnitude of the control signal based on the range of variation of the speed limit state variable; When the main line triggers the JAD strategy execution process, the control speed v of the intelligent network-connected vehicle driving in the loop conramp Adjustment: wherein, ∈ ρ is the intelligent network connected vehicle penetration rate model coefficient, ∈ j is the main line speed change model coefficient, ∈ ramp is the ramp model coefficient, β r is the ramp speed control rate, v ramp is the free flow speed limit under the ramp uncontrolled state.

7. The method of claim 6, wherein In the vehicle longitudinal following cooperative control strategy in step S3, the vehicle following is determined by the following formula: v i (t+1) = v i (t) + a i (t) T m (10) s(t + 1) = s(t) + At(v i (t + 1) - v i (t)) (12) wherein, formula (10)-(12) calculates vehicle longitudinal point, speed and vehicle head distance, x i (t) / x i (t+1), v i (t) / v i (t+1), s(t) / s(t+1) respectively represent vehicle position, speed and vehicle head distance at t and t+1 time, a i (t) represents vehicle longitudinal acceleration at t time, Δt represents time interval between t and t+1 time, T m is simulation time step; Equations (13)-(14) determine the car-following state system of each lane by the wave model matrix, the function F(X, U, V) is a trigger function, U represents external disturbance factors, V represents traffic demand factors, X=(X1, X2…X i ...X n ), n represents the total number of vehicles in the moving wave area, i represents the ith vehicle, s i represents the headway between the ith vehicle and the vehicle in front of it in the moving wave area, u i represents the acceleration of the ith vehicle in the moving wave area, Δv i represents the vehicle speed difference between the ith-1 vehicle and the ith vehicle in the moving wave area.

8. The method of claim 7, wherein, The step S4 designs the cooperative strategy cascade controller in the multi-lane scene, the cascade controller system includes: the vehicle flow state F of each lane under the initial input motion wave influence wave1 , F wave2 , the vehicle flow F of ramp merging ramp With, the following process r of vehicle flow under the intermediate output JAD control c , the following process r of vehicle flow change , the following response process E of upstream area vehicle flow merging v , the following response process E of upstream area vehicle flow merging to the response process of ramp merging vehicle flow to main line vehicle ramp , the following process r of vehicle flow through the JAD strategy execution control j , the following process r of vehicle flow through the VSL strategy execution control v And the following response control E in the execution JAD process j , the vehicle flow state F under the output control strategy action cont1 And F cont2 .

Images