CN112258864A

CN112258864A - Automatic driving vehicle intersection scheduling method and system based on sequence selection

Info

Publication number: CN112258864A
Application number: CN202011119676.4A
Authority: CN
Inventors: 吕品; 何岳滨; 许嘉
Original assignee: Guangxi University
Current assignee: Guangxi University
Priority date: 2020-10-19
Filing date: 2020-10-19
Publication date: 2021-01-22
Anticipated expiration: 2040-10-19
Also published as: CN112258864B

Abstract

The invention discloses a method and a system for automatically scheduling vehicle intersections based on sequence selection, wherein the method comprises the following steps: s1, when a vehicle to be dispatched enters a dispatching area of a target crossing, the vehicle to be dispatched establishes communication with a coordinator of the target crossing, and enters a dispatched stage; s2, estimating the earliest time point of the vehicle to be dispatched to reach the intersection according to the distance between the vehicle to be dispatched and the target intersection and the current state information of the vehicle to be dispatched; s3, acquiring current road condition information sent by the coordinator, and selecting the earliest time of a vehicle to be dispatched to reach a target intersection under safety constraint as a planned arrival time point according to the acquired road condition information and the estimated earliest time point; and S4, setting a driving plan for the vehicle to be dispatched to reach the target intersection according to the planned arrival time point. The invention can realize the intelligent dispatching of FIFS traffic-light-free intersection vehicles and improve the vehicle passing efficiency of the intersection on the premise of ensuring the safety.

Description

Automatic driving vehicle intersection scheduling method and system based on sequence selection

Technical Field

The invention relates to the technical field of automatic driving vehicle control, in particular to a method and a system for automatically scheduling a vehicle intersection based on sequential selection.

Background

With the development of artificial intelligence, edge computing and other technologies, the autonomous vehicle and its related systems have become research hotspots in the fields of computing science and road traffic. Autonomous vehicles generally have significant advantages over conventional vehicles in terms of interaction and communication capabilities and computing capabilities. These advantages enable autonomous vehicles to travel more safely and efficiently in more complex traffic environments, and thus are increasingly gaining attention and being used in a variety of traffic scenarios. In fact, application examples consisting of autonomous vehicles only have started to appear in the scenes of logistics transportation, industrial parks, etc.

The intersections are key nodes of urban traffic, particularly crossroads, and for the long time, passive maneuvering schemes which need to depend on traffic lights in fixed time are mainly adopted for coordinated dispatching of passing vehicles at the intersections, but the schemes have the defect that dynamic adjustment is difficult to carry out according to road conditions, and along with the increase of the number of urban vehicles, the passive maneuvering schemes using the traffic lights can aggravate the traffic jam of the intersections. Moreover, in the future of further popularity of autonomous vehicles, such scheduling schemes may have difficulty exploiting the communication and computational advantages of the vehicle. Therefore, designing and researching intersection scheduling schemes for automatic driving vehicles has important meaning.

At present, aiming at the dispatching of the crossroads, the classical scheme is a dispatching method based on traffic lights, namely, the traffic lights are reserved, and the road conditions near the crossroads are used as reference factors to dynamically adjust the duration time of the light phases. For example, guiding the timing of the traffic light switching stage on the basis of the preset traffic time, or making a decision of the traffic light switching stage according to the real-time traffic condition, and some researches are also made to introduce artificial intelligent methods such as DQN and fuzzy theory into the field of traffic light control. Compared with classical traffic light control, the method has higher precision and road condition adaptability, however, the scheduling control scheme for reserving the traffic light cannot fully utilize the driving state information of the vehicle, does not prevent the collision risk when passing through the intersection from the control logic, and cannot accurately schedule the automatically driven vehicle at the intersection.

In order to more accurately schedule the automatic driving vehicles at the intersection, a solution is to realize the scheduling based on no traffic light, if a practitioner proposes a traffic light-free intersection scheduling method based on reserved time, a coordinator is established in the method to collect requests sent by vehicles in a certain range and coordinate the vehicles to pass through the intersection, and most of subsequently proposed improvement schemes take the improvement of the overall throughput of the intersection or the reduction of fuel consumption as an optimization target and improve the original time reserved scheduling to a certain extent according to the target. However, the above solutions are based on FIFO (First In First Out) strategy, that is, the sequence of vehicles scheduled determines the sequence of reaching the intersection, thereby limiting further improvement of intersection traffic efficiency. Therefore, it is desirable to provide a non-FIFO and more efficient intersection scheduling method. However, to implement non-FIFO traffic-light-free scheduling, there are problems of how to establish a vehicle scheduling rule more efficient than FIFO, so that intersection traffic efficiency is higher, and how to set more reasonable constraints based on vehicle kinematics analysis, so as to avoid collision that may occur when a vehicle passes through an intersection.

Disclosure of Invention

The technical problem to be solved by the invention is as follows: aiming at the technical problems In the prior art, the invention provides the method and the system for automatically scheduling the vehicle intersection based on the sequential selection, which have the advantages of simple implementation method, high scheduling performance and high safety, can realize the intelligent scheduling of the traffic-light-free intersection of the FIFS (First In First Select), improve the intersection passing efficiency and simultaneously avoid the collision possibly generated when the vehicle passes through the intersection.

In order to solve the technical problems, the technical scheme provided by the invention is as follows:

an automatic driven vehicle intersection scheduling method based on sequence selection comprises the following steps:

s1, scheduling and starting: when a vehicle to be dispatched enters a dispatching area of a target crossing, the vehicle to be dispatched establishes communication with a coordinator of the target crossing, and the vehicle to be dispatched enters a dispatched stage;

s2, earliest time point estimation: estimating the earliest time point of the vehicle to be dispatched reaching the intersection according to the distance between the vehicle to be dispatched and the target intersection and the current state information of the vehicle to be dispatched;

s3, selecting a planned arrival time point: acquiring current road condition information sent by the coordinator, wherein the road condition information comprises information of all vehicles in a scheduling area, and selecting the earliest time of a vehicle to be scheduled to reach a target intersection under safety constraint as a planned arrival time point according to the acquired road condition information and the estimated earliest time point, so that the time of the target vehicle reaching the intersection is selected according to the sequence of the scheduled vehicles, and the safety constraint is that the vehicles can be prevented from colliding;

s4, planning driving: and setting a driving plan for reaching the target intersection for the vehicle to be dispatched according to the planned arrival time point.

Further, the specific step of selecting the planned arrival time point in step S3 is:

s301, the earliest arrival time point is used as an alternative arrival time to generate a temporary information set, the generated temporary information set is added into a road condition information total set which comprises all vehicle information at present, and the vehicle information is reordered according to the arrival time to form a temporary information total set containing the vehicle information to be dispatched;

s302, judging whether the vehicle to be dispatched is confronted with collision risk according to the current alternative arrival time in the current vehicle sequence, if so, turning to the step S303 to perform a new round of attempt, otherwise, successfully attempting, and determining the planned arrival time of the vehicle to be dispatched according to the current alternative arrival time;

s303, new alternative arrival time is selected backwards again, a temporary information total set of the vehicles to be scheduled is updated according to the current new alternative arrival time, the information of each vehicle in the temporary information total set is reordered, and the step S302 is executed in a returning mode.

Further, in the step S302, when it is the first attempt, that is, the current alternative arrival time is the earliest arrival time point T₀(x) If the front vehicle and the rear vehicle do not exist in the vehicle to be dispatched, the vehicle to be dispatched is the only vehicle in the dispatching area, wherein the front vehicle is the vehicle to be dispatched when reaching the target intersectionThe vehicle closest to the front vehicle and the rear vehicle are the vehicles closest to the rear vehicle after the vehicle to be dispatched, and the current alternative arrival time is directly used as the planned arrival time of the vehicle to be dispatched;

if the vehicle to be dispatched only has the front vehicle, obtaining the minimum safe time interval between the vehicle to be dispatched and the front vehicle, and selecting the planned arrival time of the vehicle to be dispatched according to the following formula so as to select the time which is as early as possible and safe as the planned arrival time;

T(x)＝max(T(pre(x))+deal(x,pre(x)),T₀(x))

wherein x is a vehicle to be dispatched, T (x) is the planned arrival time of the vehicle to be dispatched, pre (x) is a front vehicle, T (pre (x)) is the time of the front vehicle to arrive at the destination intersection, deal (x, pre (x)) is the minimum interval between the vehicle to be dispatched and the front vehicle pre (x) which meets the safety; wherein if interval is satisfied<deal (pre (x), where interval is the earliest point in time T₀(x) At an interval from the time T (pre (x)) when the preceding vehicle arrives at the destination intersection, a time point apart from deal (x, pre (x)) from T (pre (x)) is directly selected as the planned arrival time.

Further, in the step S302, when it is the first attempt, that is, the current alternative arrival time is the earliest arrival time point T₀(x) And the vehicle to be dispatched only exists in the rear vehicle, wherein if the interval (x, next (x)) is judged to be satisfied>deal (x, next (x)), where x is the vehicle to be dispatched, next (x) is the rear vehicle, and interval (x, next (x)) is the earliest arrival time point T₀(x) The distance between the vehicle to be dispatched and the rear vehicle next (x), deal (x, next (x)) is the minimum distance which meets the safety between the vehicle to be dispatched x and the rear vehicle next (x), and the current alternative arrival time point is directly selected as the planned arrival time point; if the interval (x, next (x)) is satisfied<deal (x, next (x)), determine that there is a collision risk for the vehicle to be dispatched according to the current alternative arrival time, and go to step S303 to perform a new round of attempt.

Further, in step S302, when it is the first attempt, that is, the current alternative arrival time is the earliest arrival timeReach time point T₀(x) And the vehicle to be dispatched has a front vehicle and a rear vehicle at the same time, if the following conditions are met:

interval(pre(x),next(x))<deal(pre(x),x)+deal(x,next(x))

wherein, x is a vehicle to be dispatched, pre (x) is a front vehicle, next (x) is a rear vehicle, interval (pre (x), next (x)) is an interval between the front vehicle and the rear vehicle, deal (x, next (x)) is a minimum interval which meets safety between the vehicle x to be dispatched and the rear vehicle next (x); judging that the vehicle to be dispatched has a collision risk according to the current alternative arrival time, and turning to the step S303 to perform a new round of attempt;

if the interval (pre (x), next (x)) is satisfied>deal (pre (x), x) + deal (x, next (x)), and determine the earliest arrival time T₀(x) A value of (1), wherein if T₀(x)∈[T(next(x))-deal(x,next(x)),T(next(x))]If the current alternative arrival time is determined to cause the collision risk between the vehicle to be scheduled and the rear vehicle, the step S303 is executed to perform a new round of trial; if T₀(x)∈[0,T(pre(x))+deal(x,next(x))]The planned arrival time is selected as a planned arrival time T (pre (x)) + deal (x, pre (x)), T being a time as early as possible and safe as T (pre (x)) + max (x, pre (x)))₀(x))。

Further, in step S303, the new candidate arrival time is reselected backwards, specifically, the new candidate arrival time is the current candidate arrival time T _ temp (x), and the new candidate arrival time is obtained by moving to a position behind the arrival time of the current rear vehicle and keeping the minimum safety interval with the rear vehicle.

Further, the step S4 of setting the driving plan includes:

if the vehicle advances along a straight line, setting the time T for which the acceleration/deceleration of the vehicle to be dispatched should last_accComprises the following steps:

wherein, t₁(x)＝T(x)-t₀(x) Is that the vehicle x to be dispatched is planned to go from entering the dispatching area to the destinationT (x) is the planned arrival time of the vehicle x to be dispatched to the destination intersection, t₀(x) For the time when the vehicle x to be dispatched enters the current dispatching area, v₀(x) For the speed at which the vehicle x to be dispatched enters the current dispatching zone, a_xThe method comprises the steps that the acceleration of a vehicle x to be dispatched when the vehicle x enters a current dispatching area is obtained, L is the straight-line distance between an entrance of the current dispatching area and a junction area, and the junction area is an area where lanes intersect at a target intersection; acceleration/deceleration T of vehicle to be dispatched_accMaintaining the speed after a time until exiting the confluence region;

if the vehicle needs to turn to a lane, setting the vehicle to be dispatched to drive to a convergence area according to a first stage acceleration/deceleration stage, a constant speed stage and a second stage acceleration/deceleration stage in sequence, wherein the time of the first stage acceleration/deceleration stage is as follows:

wherein the content of the first and second substances,

v_turn(x) For the maximum steering speed, v, of the vehicle x to be dispatched in the junction zone_t(x) The speed of a vehicle x to be dispatched at time t;

the duration of the uniform speed stage is as follows:

wherein

Further, when the earliest time point is estimated in step S2, if the vehicle to be dispatched is a straight-ahead vehicle, wherein if the vehicle to be dispatched can accelerate to a limit speed before reaching the intersection, the earliest time point at which the vehicle to be dispatched reaches the destination intersection is estimated as follows:

if the vehicle to be dispatched cannot accelerate to the limit speed before reaching the intersection, estimating the earliest time point when the vehicle to be dispatched reaches the destination intersection as follows:

wherein, a_xFor acceleration, v, of the vehicle x to be dispatched₀(x) Is the speed, v, of the vehicle x to be dispatched at the moment of entering the current dispatching area_max(x) Is the maximum speed of the vehicle x to be dispatched.

Further, when the earliest time point is estimated in step S2, if the vehicle to be dispatched is a vehicle that needs to turn, if the vehicle to be dispatched can accelerate to the maximum speed during the driving process and decelerate to the maximum steering speed before reaching the confluence region, that is, the speed v when the vehicle to be dispatched enters the current dispatching region₀(x) Maximum steering speed v of the vehicle to be dispatched in the junction region_turn(x) Maximum speed v of the vehicle to be dispatched_max(x) Satisfies the following conditions:

estimating the earliest time point of the vehicle to be dispatched to reach the destination crossing as follows:

wherein, a_xFor the acceleration of the vehicle x to be dispatched, mu is the ground friction coefficient, g is the acceleration of gravity, and R is an element of { R ∈ [ R ]₁,R₂Is the turning radius, R₁And R₂Respectively the radius of a right-turn bend and the radius of a left-turn bend;

otherwise, estimating the earliest time point of the vehicle to be dispatched to reach the destination crossing as follows:

an automatic driving vehicle crossing dispatching system based on sequence selection comprises a coordinator arranged at a destination crossing end and a vehicle-mounted control terminal arranged at a vehicle end, wherein when a vehicle to be dispatched enters a dispatching area of the destination crossing, the vehicle to be dispatched establishes communication with the coordinator of the destination crossing through the vehicle-mounted control terminal, and the vehicle to be dispatched enters a dispatched stage; the vehicle-mounted control terminal estimates the earliest time point of the vehicle to be dispatched reaching the intersection according to the distance between the vehicle to be dispatched and the target intersection and the current state information of the vehicle to be dispatched; acquiring current road condition information sent by the coordinator, wherein the road condition information comprises information of all vehicles in a dispatching area, and selecting the earliest time of a vehicle to be dispatched to reach a target intersection under safety constraint as a planned arrival time point according to the acquired road condition information and the estimated earliest time point, so that the time of the target vehicle arriving at the intersection is selected according to the sequence of the vehicles to be dispatched, and the safety constraint is that the vehicles can be prevented from colliding; and setting a driving plan for reaching the target intersection for the vehicle to be dispatched according to the planned arrival time point.

Compared with the prior art, the invention has the advantages that:

1. the invention relates to a method and a system for dispatching an automatic driving vehicle intersection based on sequential selection, which firstly calculate the earliest time point capable of reaching the intersection according to the distance from the intersection and the state information of the vehicle, select the earliest reaching time point which is feasible under the safety constraint as the planned reaching time according to the estimated earliest time point and the road condition information sent by a coordinator, finally set a feasible driving plan according to the planned reaching time point, can realize the vehicle dispatching based on the sequential selection FIFS first-in first-out strategy, and the vehicle selects the time for reaching the intersection according to the dispatched sequence. Therefore, the intersection can obtain higher passing efficiency, particularly, the passing efficiency of the intersection can be greatly improved when the traffic flow pressure is higher, the running safety of the vehicle can be ensured, and the problem of collision when the obstacle avoidance vehicle passes through the intersection can be solved.

2. The invention discloses an automatic driving vehicle intersection scheduling method and system based on sequential selection, which realize intelligent scheduling of vehicles without traffic lights on intersections based on a sequential selection FIFS (first selection) strategy, can select a planned arrival time as early as possible for each scheduled vehicle by considering the difference of physical properties among the vehicles, and then set a specific driving scheme on the premise of considering safety and feasibility, so that high-speed vehicles obtain earlier results at the stage of estimating the earliest arrival time, and meanwhile, the vehicles with low speed but earlier approaching the intersection are prevented from being always positioned at the end of a sequence, thereby improving the crossing traffic efficiency and reducing the vehicle pause times.

3. According to the method and the system for dispatching the automatic driving vehicle intersection based on the sequence selection, the earliest arrival time point is further taken as the initial alternative arrival time, new alternative arrival time is continuously and iteratively searched according to the sequence of the vehicles in the current road condition information until the earliest arrival time meeting the safety constraint is obtained, and the vehicles to be dispatched can rapidly obtain the planned arrival time as early as possible on the premise of ensuring safety.

Drawings

Fig. 1 is a schematic flow chart of an implementation of the intersection scheduling method for the automatic driving vehicle based on the sequential selection according to the embodiment.

Fig. 2 is a schematic diagram of the intersection and the automatic driving vehicle model constructed in the embodiment.

Fig. 3 is a schematic diagram illustrating the selection of a planned arrival time point in a specific application embodiment of the present invention.

Fig. 4 is a schematic flow chart of an implementation of selecting a scheduled arrival time point in the present embodiment.

Fig. 5 is a schematic diagram of the first vehicle dispatching situation (2 nd situation) in the present embodiment.

Fig. 6 is a schematic diagram of the first vehicle dispatching situation (situations 3) to (ii) in the present embodiment.

Fig. 7 is a schematic diagram of a second vehicle dispatching situation (situations 4) - (i) in the present embodiment.

Fig. 8 is a schematic diagram of a third vehicle dispatching situation (situations 4) - (ii) - (a) in the present embodiment.

Fig. 9 is a schematic diagram of a fourth vehicle dispatching situation (situations 4) - (ii) - (b) in the present embodiment.

Fig. 10 is a schematic diagram illustrating the principle of the update method in the loop process when the scheduled arrival time point is selected in the present embodiment.

Fig. 11 is a detailed implementation flow diagram of implementing vehicle intersection scheduling in an embodiment of the present invention.

Fig. 12 is a diagram illustrating the results of driving time in the average dispatching area as a function of traffic pressure when the method of the present invention and two conventional methods are respectively adopted in a specific application embodiment.

Fig. 13 is a graph showing the results of the average delay with respect to the variation of the pressure of the vehicle stream when the method of the present invention and the two conventional methods are separately applied in specific application embodiments.

Fig. 14 is a diagram illustrating the variation of the total pause times with the traffic pressure when the method of the present invention and the two conventional methods are respectively applied to a specific application embodiment.

Fig. 15 is a graph showing the results of the average delay varying with the degree of imbalance of traffic flow when the method of the present invention and the two conventional methods are separately employed in a specific application example.

Fig. 16 is a graph showing the results of the variation of the average travel time with the degree of flow imbalance when the method of the present invention and the two conventional methods are respectively employed in practical application examples.

Fig. 17 is a diagram illustrating the results of the total number of stops as a function of the degree of imbalance between traffic flows when the method of the present invention and the two conventional methods are applied separately in a specific embodiment.

Detailed Description

The invention is further described below with reference to the drawings of the specification and to specific preferred embodiments, without thereby limiting the scope of protection of the invention.

As shown in fig. 1, the steps of the automatic driven vehicle intersection scheduling method based on sequential selection of the embodiment include:

s2, earliest time point estimation: estimating the earliest time point of the vehicle to be dispatched reaching the intersection without considering other vehicles according to the distance between the vehicle to be dispatched and the target intersection and the current state information of the vehicle to be dispatched, wherein the state information comprises parameter information such as current speed, acceleration performance and the like;

s3, selecting a planned arrival time point: acquiring current road condition information sent by a coordinator, wherein the road condition information comprises information of all vehicles in a dispatching area, specifically comprises planned arrival time, a preset track and the like of other existing vehicles in the dispatching area, and selecting the earliest time point of a vehicle to be dispatched to reach a target intersection under safety constraint as a planned arrival time point according to the acquired road condition information and the estimated earliest time point, so that the time of the target vehicle arriving at the intersection is selected according to the dispatching sequence of the vehicles, and the safety constraint is that the vehicles can be prevented from colliding;

The embodiment aims at the automatic driving vehicle scheduling of the intersection, when each vehicle establishes communication with an intersection coordinator, the vehicle enters a scheduled stage, the earliest time point which can reach the intersection under an ideal condition (namely, other vehicles are not considered) is calculated according to physical properties such as state information of the vehicle and the like in the scheduling process, then the feasible planned arrival time point of the vehicle is searched according to the time point and road condition information sent by the coordinator, the feasible earliest arrival time point under safety constraint is selected as planned arrival time, finally feasible driving planning is set for the vehicle according to the planned arrival time point, the intelligent scheduling of the vehicle without the traffic light intersection based on the sequential selection FIFS (fast FIFS) first-in first-out strategy can be realized, the vehicle selects the time for planning the arrival of the intersection according to the scheduled sequence, compared with the traditional FIFO scheduling strategy, on one hand, the physical properties of different vehicles can be considered, vehicles which are scheduled later but have higher initial speeds, accelerations and the like are allowed to arrive at the intersection earlier, on the other hand, the priority of entering the vehicles first can be guaranteed to a certain extent, so that the intersection can obtain higher passing efficiency, especially, the passing efficiency of the intersection can be greatly improved when the traffic pressure is higher, meanwhile, the driving safety of the vehicles can be ensured, the problem of collision when obstacle-avoiding vehicles pass through the intersection can be solved, and the method is particularly suitable for improving the passing efficiency of the vehicles in vehicle scheduling of the intersection.

The method of the present invention will be further described below by taking an intersection as an example as shown in fig. 2 (a).

In this embodiment, the center of the intersection is O, the O is taken as a symmetric center, and the square with the side length D is the administration area of the intersection. Two mutually perpendicular, two-way single lanes with a center line passing through O intersect at the junction region (marked by a square dashed line, which is a square with a side length of S). A coordinator is responsible for communicating with the arriving vehicles and coordinating the dispatch of the vehicles, the scope of which is indicated by the circular dashed line. The distances between the entries of the scheduling area and the merging area are the same, and are recorded as

At the intersection of the road and the jurisdiction there are edge terminals provided with sensors that can communicate with each other and record the total number of vehicles in the current jurisdiction, denoted n (t). When a vehicle x to be scheduled enters the jurisdiction area, the edge terminal generates a sequence number i ═ N (t) +1 for x; after receiving the sequence number, the vehicle x to be dispatched defines its unique travel track d (x) ═ g (x) +3road (x), where road (x) e {0,1,2,3} is the number of the road where the vehicle passes through the intersection, and the distribution is shown in fig. 2(b), and g (x) e {1,2,3} represents the target direction of the vehicle respectively: straight, right turn or left turn. For example, suppose the departure lane of a vehicle x to be dispatchedWhen road is road (x) 1 and the target direction is right turn, g (x) 2, the track number d (x) is 2+3 × 1 — 5. Similarly, it can be calculated that the vehicle is traveling straight when D (x) e {1,4,7,10} is known, and other trajectories include steering.

In this embodiment, a model for dispatching vehicles at a crossing is first built, wherein the communication between the vehicle to be dispatched and the coordinator and the information processing and exchange include:

(1) the method comprises the steps that a vehicle to be dispatched uploads a sequence number of the vehicle and an expected running direction to a coordinator, the coordinator transmits the current time t and an information aggregate (a road condition aggregate) B (t) of all vehicles in a dispatching area to the vehicle to be dispatched, wherein the information aggregate of each vehicle is represented as b (x) ═ x, T (x), D (x) } e (t), and T (x) is the predicted time of the vehicle x to be dispatched to reach an intersection;

(2) the scheduled time (namely the planned arrival time) of arriving at the intersection is calculated by the vehicle to be dispatched, a driving plan is set, and finally an information set belonging to the vehicle to be dispatched is generated and uploaded to the coordinator;

(3) and the coordinator adds the information set of the vehicle to be dispatched into B (t), and rearranges B (t) from early to late according to the arrival time of all vehicles, and the vehicle to be dispatched starts to execute the driving plan. When the vehicle leaves the dispatch area, the coordinator will delete the vehicle-related information set from the aggregate.

As mentioned above, in the present embodiment, when determining the planned arrival time for the vehicle to be scheduled, first only the initial speed and acceleration performance of the vehicle entering the scheduling area are considered to obtain the earliest time at which the vehicle can arrive at the intersection, then the time is adjusted in combination with b (t) obtained from the coordinator, and finally the driving scheme that can arrive at the intersection as quickly as possible on the premise of avoiding the collision is obtained, so that when determining the planned arrival time, the key factors that can avoid the collision with other vehicles and the physical constraints required for obtaining the initial scheme can be fully considered.

On the basis of the scheduling model described above, the present embodiment further makes the following assumptions to simplify the problem:

assume that 1: the coordination time and calculation time of all vehicles and the coordinator can be ignored in the model.

Assume 2: the moving speed of all vehicles in the merging area is constant.

Assume that 3: the acceleration or deceleration process of all vehicles is uniform while, outside the junction area, the speed of the vehicle must satisfy the following constraints:

0<v(x,t)<v_max(x) (1)

where v (x, t) denotes the speed of the vehicle x at time t (outside the junction area), v_max(x) Is the limit speed (maximum speed of the vehicle) determined by the vehicle itself and the road.

The speed constraint of the vehicle in the junction region is:

0<v(x,t)≤v_turn(x) (2)

v_turn(x) Is defined by the following formula (3):

wherein μ is the ground friction coefficient, g is the acceleration of gravity, R ∈ { R ∈ [ ]₁,R₂Is the turning radius, R₁And R₂Respectively the radius of the right-hand turn and the left-hand turn, R₁Corresponding to the locus of the direction 2 in FIG. 2(b), R₂Corresponding to the radius of the trajectory as in direction 3 in fig. 2 (b).

The symbols and meanings of the parameters used in the present example are specifically shown in table 1.

TABLE 1 symbol List

This example is further defined as follows:

defined in FIG. 2(a)Under the model, trajectory D₁,D₂Relation E (D) between₁,D₂) Belong to and only belong to one of three subsets:

1) theta is along D respectively₁,D₂There is a possibility of a side collision of the running vehicles.

2) Selecting D₁,D₂There is a possibility of rear-end collision.

3)Υ:D₁,D₂There is no intersection, and there is no risk of collision between vehicles having them as the planned path.

Referring to the partial vehicle trajectory in fig. 2(b), for example, the conflict relationship between a vehicle advancing along trajectory 5 and a vehicle on trajectory 1 belongs to Θ, while a vehicle advancing along trajectory 5 is not at risk of colliding with a vehicle scheduled to travel on trajectory 2, so the relationship is one of γ; in addition, vehicles on two

tracks

1 and 3 from the same road belong to Λ. Similarly, the collision relationship between every two of all 12 trajectories in the model presented in FIG. 2 can be determined and consolidated into a symmetric matrix

Thus, the collision type of any two vehicles x, y is further determined as:

wherein x and y represent two different vehicles.

And defining: the minimum interval of the planned arrival times between two different vehicles a, b, without the risk of collision between them, is called deal (a, b)), and its value is determined by the type of collision between the two vehicles, as shown in equation (5):

where δ (a, b) is the safe distance between two vehicles to avoid rear-end collisions, v_aAnd v_bThe predicted speeds of vehicles a and b when they arrive at the intersection, respectively, and d is the length of the trajectory of a in the junction region, which is given by equation (6):

from equation (6), it can be seen that for two vehicles at risk of side impact (c (a, b) ═ 0), only after one of them has completely passed through the junction area will the other be allowed to enter the intersection; for vehicles at risk of rear-end collision (c (a, b) ═ 1), only a safe distance needs to be kept between them.

The embodiment records the current time t when the vehicle to be dispatched enters the dispatching area₀(x) And simultaneously referring to a total road condition set B (t) sent by the coordinator, generating an information set such as b (x) ═ x, T (x), D (x) } before the next vehicle enters a dispatching area, uploading the information set to the coordinator, and simultaneously making a reasonable driving scheme according to the planned arrival time, wherein x is the number of the vehicle to be dispatched, T (x) is the predicted arrival intersection time point determined by the vehicle x, namely the planned arrival intersection time point, and for the vehicle x to be dispatched, T (x) is not changed once determined.

The shortest time for the vehicle x to reach the intersection under the assumption that no other vehicles are considered is the theoretical arrival time, and the earliest time point T estimated in step S2 in this embodiment is the shortest arrival time₀(x) I.e. the theoretical arrival time, for T₀(x) The estimation of (b) is specifically divided into two different cases depending on the trajectory of the vehicle (straight/curved):

the first case: earliest time point estimation for straight-ahead vehicle to reach intersection

Namely, the vehicle to be dispatched is a straight-going vehicle, wherein:

1) if the vehicle to be dispatched can accelerate to the limit speed before reaching the intersection, namely the conditional expression is

a_xAcceleration, v, of vehicle x₀(x) Is the speed of the vehicle entering the dispatching area, in this case, the vehicle will accelerate continuously until reaching the limit speed, then keep the speed cruising to the intersection merging area, then estimate the earliest time point when the vehicle to be dispatched reaches the target intersection as:

2) if the vehicle to be dispatched cannot accelerate to the limit speed before reaching the intersection, i.e. if the vehicle to be dispatched cannot accelerate to the limit speed

The second case: earliest time point estimation for turning vehicle to reach intersection

If the vehicle advances along the turning track, the speed at the intersection needs to be considered to meet the turning constraint, and the embodiment is divided into two cases to be considered separately, wherein:

1) if the vehicle to be dispatched can be accelerated to the maximum speed in the running process and decelerated to the maximum steering speed before reaching the confluence area, namely the speed v of the vehicle to be dispatched when entering the current dispatching area₀(x) Maximum steering speed v of the vehicle to be dispatched in the junction region_turn(x) Maximum speed v of the vehicle to be dispatched_max(x) Satisfies the following conditions:

then, the earliest time point when the vehicle to be dispatched reaches the destination intersection is estimated as follows:

2) if the condition 1) is not met, estimating the earliest time point of the vehicle to be dispatched to reach the destination intersection as follows:

after the earliest time point of the vehicle to be dispatched to reach the destination intersection is obtained through the estimation of the steps, the vehicle x to be dispatched can generate a temporary information set according to the earliest time point, namely b _ temp (x) ═ x, T₀(x) D (x), then adding it into the road condition data aggregate B (t) from the coordinator at the moment, reordering the road condition aggregate from small to large according to the arrival time, and forming an ordered temporary set B _ temp (x, t) ═ B (x) containing the information of the vehicles x to be dispatched₁)，b(x₂)…b_temp(x)…b(x_s) In which x₁,x₂……x_sAre all vehicles in the current dispatch area.

In this embodiment, the specific step of selecting the planned arrival time point in step S3 is:

s301, an earliest arrival time point is used as an alternative arrival time to generate a temporary information set, the generated temporary information set is added into a current road condition information total set comprising all vehicle information, and the vehicle information is reordered according to the arrival time to form a temporary information total set containing the vehicle information to be scheduled;

s303, new alternative arrival time is selected backwards again, the temporary information total set of the vehicles to be scheduled is updated according to the current new alternative arrival time, the vehicle information in the temporary information total set is reordered, and the step S302 is executed in a returning mode.

In the embodiment, an earliest arrival time point is taken as an initial alternative arrival time, new alternative arrival time is continuously tried to be selected for the vehicle to be scheduled according to the vehicle sequence in the current road condition information, whether the vehicle to be scheduled has collision risk according to the new alternative arrival time in the current vehicle sequence is judged in the trying process until the trying is successful, and the arrival time of the vehicle to be scheduled is searched iteratively, so that the vehicle to be scheduled can quickly acquire the earliest planned arrival time as possible on the premise of ensuring safety.

As shown in fig. 3 and 4, when the planned arrival time point is selected for the vehicle x to be scheduled in the present embodiment, the theoretical earliest arrival time T is used first₀(x) And simultaneously generating a temporary set B _ temp (x, t) as alternative arrival time, and attempting to acquire the earliest possible planned arrival time T (x) for the vehicle x to be dispatched under the arrival sequence of the vehicles which is consistent with the sequence described by the temporary set, and if the vehicle x is subjected to the unavoidable collision risk under the current sequence, generating a new alternative arrival time backwards, updating the temporary information set B _ temp (x) of the vehicle x to be dispatched, and rearranging the temporary set B _ temp (x, t), and performing a new round of attempt according to the new B _ temp (x, t) after that, and repeating the steps until the planned arrival time of the vehicle is determined. In the embodiment of the present invention, the above-mentioned execution process and the sequence of elements in the temporary information set B _ temp (x, t) are changed as shown in fig. 3, where a, B, and c are all vehicles for which the arrival time has been determined.

Among the above-mentioned several executions for selecting the scheduled arrival time point, the first-round execution (i.e. shown in fig. 3) is complicated due to the limitation of insufficient information amount and physical constraint conditions, and the subsequent attempts performed on the premise that the first-round execution fails to determine the scheduled arrival time are a more concise loop process, and the following specifically describes the first-round execution:

in the first round of execution, the alternative arrival time T _ temp (x) T₀(x) It is noteworthy that, due to T₀(x) Already the earliest a vehicle can reach the junction area under physical constraints, the planned arrival time chosen by the vehicle needs to be in accordance with T (x)>T₀(x) To describe this step more briefly, the following definitions are first made:

in one implementation of selecting the planned arrival time point, assuming that the alternative arrival time point of the vehicle x is T _ temp (x), it is set to:

1) front vehicle pre (x): from B _ temp (x, T), assuming L is the set of vehicles with arrival times before T _ temp (x), the last element of L is referred to as pre _1 (x)'. Meanwhile, L' is the set of all vehicles with rear-end collision risk with the vehicle x in the dispatching area, namely the pair

All have c (x, y) ═ 1. The last element in L' is called pre _2 (x). pre (x) represents the front vehicle to be considered by the vehicle x to be dispatched, given by:

2) rear vehicle next (x): representing the rear vehicle to be considered by vehicle x, is defined as follows:

(I) pre (x) exists, and the first element of R is called next (x) assuming that R is the set of vehicles whose planned arrival time is after T (pre (x)) (does not contain the vehicle x to be dispatched).

(II) pre (x) does not exist, assuming R1 is the set of vehicles with a projected arrival time after T _ temp (x), the first element of R1 is referred to as next (x).

In step S302, when the first attempt is made, that is, the current alternative arrival time is the earliest arrival time T₀(x) The selection of the planned arrival time point specifically includes the following casesShape:

the 1 st: the method comprises the steps that a front vehicle and a rear vehicle do not exist in a vehicle to be dispatched, namely, the vehicle to be dispatched is the only vehicle in a dispatching area, wherein the front vehicle is the vehicle which is closest to the vehicle to be dispatched before the time of reaching a target intersection, the rear vehicle is the vehicle which is closest to the time of reaching the target intersection after the vehicle to be dispatched, and then the current alternative arrival time is directly used as the planned arrival time of the vehicle to be dispatched, namely T (x) T (T ═ T-₀(x)。

The 2) kind: the method comprises the following steps that a vehicle to be dispatched only exists in a front vehicle, the vehicle obtains a minimum safe time interval between the vehicle to be dispatched and the front vehicle after determining the front vehicle pre (x), and then on the basis of ensuring that the two vehicles keep a safe time interval, the vehicle tends to select a time point as early as possible in [ T (pre (x)), + ∞ ] as a planned arrival time, and specifically selects the planned arrival time of the vehicle to be dispatched according to the following formula to select a time as early as possible and safe as the planned arrival time;

T(x)＝max(T(pre(x))+deal(x,pre(x)),T₀(x)) (12)

wherein, x is the vehicle to be dispatched, T (x) is the planned arrival time of the vehicle to be dispatched, pre (x) is the front vehicle, T (pre (x)) is the time of the front vehicle arriving at the destination crossing, deal (x, pre (x)) is the minimum interval which meets the safety between the vehicle to be dispatched and the front vehicle pre (x); wherein if interval is satisfied<deal (pre (x), where interval is the earliest point in time T₀(x) At an interval from the time T (pre (x)) when the preceding vehicle arrives at the destination intersection, a time point apart from T (pre (x)) by deal (x, pre (x)) is directly selected as the planned arrival time.

Taking FIG. 5 as an example, T₀(x) At interval of interval from T (pre (x))<deal (pre (x), meaning if the vehicle is determined to be T₀(x) The time of arrival, which is at risk of a pre (x) collision with the leading vehicle, will be reached, in which case the time point away from T (pre (x)) by deal (x, pre (x)) will be directly selected as the planned arrival time to ensure safety.

The 3) type: if the vehicle to be dispatched only exists in the rear vehicle, the vehicle x to be dispatched firstly considers the conflict relationship with the rear vehicle and the current time interval (x, next (x)) of the rear vehicle after determining the next (x) of the rear vehicle, and further judges according to the interval:

(i) if interval (x, next (x)) is satisfied>deal (x, next (x)), where x is the vehicle to be dispatched, next (x) is the rear vehicle, interval (x, next (x)) is the earliest arrival time point T₀(x) The distance between the vehicle x to be dispatched and the rear vehicle next (x) is the minimum distance which meets the safety between the vehicle x to be dispatched and the rear vehicle next (x), and the vehicle x and the next (x) do not have collision risk at the moment, and the current alternative arrival time point is directly selected as the planned arrival time point;

(ii) if interval (x, next (x)) is satisfied<deal (x, next (x)), which can not determine the scheduled arrival time on the premise of keeping the sequence of the elements in the current B _ temp (x, t), and determines that the vehicle to be dispatched has a collision risk according to the current alternative arrival time, and the process goes to step S303 to perform a new round of attempt. As shown in fig. 6, where the shaded portion represents that the vehicle cannot be selected to be earlier than T₀(x) Will collide with next (x) that has already determined the planned arrival time, so only T _ temp (x) can be updated and the next round of algorithm execution entered.

4) the following: the method comprises the following steps that a vehicle to be dispatched simultaneously has a front vehicle and a rear vehicle, the vehicle x to be dispatched firstly obtains pre (x) and next (x), and then further judgment is carried out according to the planned arrival time intervals of the vehicle to be dispatched and the vehicle to be dispatched, wherein the predicted arrival time intervals of the vehicle to be dispatched are pre (x), and the next (x) is T (next (x)) to T (pre (x)):

(i) if the following conditions are met: interval (pre (x), next (x)) and (x) + end (x, next (x)), wherein x is a vehicle to be dispatched, pre (x) is a front vehicle, next (x) is a rear vehicle, interval (pre (x) and next (x)) are intervals between the front vehicle and the rear vehicle, and end (x, next (x)) is the minimum interval which meets safety between the vehicle to be dispatched x and the rear vehicle; judging that the vehicle to be dispatched has a collision risk according to the current alternative arrival time, and turning to the step S303 to perform a new round of attempt; at this time, the planned arrival time cannot be determined on the premise of keeping the sequence of the elements in the current B _ temp (x, t), the alternative arrival time needs to be updated, and the algorithm is executed circularly to make the judgment. In the case of fig. 7, if the current element order {. b (pre (x)), b _ temp. (x), and b (next (x)) }ismaintained, no matter how much the value of T _ temp. (x) is selected, at least one of pre (x) and next (x) is collided with.

(ii) If the interval (pre (x), next (x)) is satisfied>deal (pre (x), x) + deal (x, next (x)), and determine the earliest time of arrival T₀(x) A value of (a), wherein:

(a) if T₀(x)∈[T(next(x))-deal(x,next(x)),T(next(x))]I.e. T₀(x) Causing x and the rear vehicle to have a collision risk, similar to cases 3) - (ii), and failing to determine the planned arrival time in the current order, it is determined that the current alternative arrival time would cause a collision risk between the vehicle to be dispatched and the rear vehicle, and the process proceeds to step S303 to perform a new round of trial, as shown in fig. 8.

(b) If T₀(x)∈[0,T(pre(x))+deal(x,next(x))]In this case, the vehicle will select a safe time as early as possible as the planned arrival time, and select the planned arrival time as early as possible as the planned arrival time T (pre (x)) + deal (x, pre (x))), T₀(x) As in the situation shown in fig. 9.

The planned arrival time of the preceding vehicle of the vehicle x to be dispatched may also be after T _ temp (x) due to T₀(x) Possibly earlier than T (pre _2(x)), but this type of situation does not affect the classification of the cases, and the above-mentioned processing in case 2) and cases 4) - (ii) - (b) is equally applicable to this type of situation.

For the case where the planned arrival time of the vehicle cannot be determined in the above first-round execution (the above-mentioned 3) - (ii), 4) - (i),4) - (ii) - (a)), it is necessary to enter a loop phase to try further, and the loop phase is executed in a specific application embodiment as shown in fig. 10.

In this embodiment, in step S303, the new candidate arrival time is reselected backwards, specifically, the new candidate arrival time is obtained by moving the current candidate arrival time T _ temp (x) to a position behind the arrival time of the current rear vehicle and keeping the minimum safety interval with the rear vehicle. In step S3, when the iteration is a non-first iteration, the current candidate arrival time T _ temp (x) is moved to a position after the arrival time of the current rear vehicle and just keeps the minimum safety interval with the rear vehicle to obtain a new candidate arrival time, the new candidate arrival time is used to update the temporary traffic information total set B _ temp (x, T), and for the updated temporary traffic information total set B _ temp (x, T), the vehicle x to be scheduled re-identifies the front vehicle and the rear vehicle, and then performs collision determination again.

In a specific application embodiment, the algorithm of the iteration stage in the selection of the planned arrival time point is specifically implemented by using the following algorithm 1.

Algorithm 1 iterative phase algorithm

Input, vehicle x-related temporary information aggregate B _ temp (x, t)

Output planned arrival time of vehicle x T (x)

The iterative stage algorithm comprises two key steps of updating and re-judging, wherein the updating main process is divided into the following three steps:

(i) the alternative arrival time T _ temp (x) of the vehicle is moved to after the arrival time of the current rear vehicle and is kept at the minimum safety interval with the rear vehicle just as shown in process 1 in fig. 10.

(ii, after the current alternative arrival time is changed, the update of the vehicle temporary road condition information aggregate B _ temp (x, t) needs to be carried out because the actually considered vehicle arrival sequence has changed.

(iii) For the updated B _ temp (x, t), the vehicle x to be dispatched re-identifies the preceding vehicle and the following vehicle, as shown in process 2 in fig. 10.

The step of judging again specifically is: since the alternative arrival time of the current vehicle x has been kept at the minimum safe time interval from the preceding vehicle (the original rear vehicle) during the update process, it is only necessary for the vehicle x to be scheduled to take into account the situation of the new rear vehicle:

(a) the rear vehicle does not exist, and can be classified as case 2) of the above first-wheel in-execution division;

(b) the rear vehicle exists and has T _ temp (x) > T (next (x)) new, which is the divided cases 4-i in the first execution, the determination of the planned arrival time can not be completed in the current round, and the next round of iterative loop needs to be entered.

(c) The rear vehicle exists and has T _ temp (x) < T (next (x)) new, which is the case 4) - (ii) - (b) divided in the above first-round execution, and the scheduled arrival time point of the vehicle x is determined using the same processing method.

Through the steps, the planned arrival time T (x) as early as possible can be obtained under various road conditions, the vehicle x to be scheduled replaces the alternative arrival time T _ temp (x) in the temporary set b _ temp (x) with T (x) after the planned arrival time T (x) is obtained, a determined vehicle scheduling total set b (x) containing information of the vehicle x to be scheduled is generated and sent to the coordinator, and the vehicle x to be scheduled then makes a complete movement plan according to T (x).

The setting of the driving plan in step S4 in this embodiment specifically includes the following two situations:

1) if the vehicle advances along a straight line, setting the time T for which the acceleration/deceleration of the vehicle to be dispatched should last_accComprises the following steps:

wherein, t₁(x)＝T(x)-t₀(x) Is the time that the vehicle x to be dispatched passes from the entering dispatching area to the destination crossing, T (x) is the planned arrival time of the vehicle x to be dispatched to the destination crossing, t₀(x) For the time when the vehicle x to be dispatched enters the current dispatching area, v₀(x) For the speed at which the vehicle x to be dispatched enters the current dispatching zone, a_xThe method comprises the following steps that (1) the acceleration of a vehicle x to be dispatched when the vehicle x enters a current dispatching area is obtained, L is the straight-line distance between an entrance of the current dispatching area and a junction area, and the junction area is an area where lanes intersect at a target intersection; acceleration/deceleration T of vehicle to be dispatched_accMaintaining the speed after a time until exiting the confluence region;

2) if the vehicle needs to turn to a lane, the vehicle to be dispatched is set to drive to a convergence area according to a first stage acceleration/deceleration stage, a constant speed stage and a second stage acceleration/deceleration stage in sequence, wherein the time of the first stage acceleration/deceleration stage is as follows:

wherein the content of the first and second substances,

v_turn(x) For the maximum steering speed, v, of the vehicle x to be dispatched in the junction zone_max(x) For the maximum speed, v, of the vehicle x to be dispatched_t(x) The speed of a vehicle x to be dispatched at time t;

the duration of the uniform speed stage is as follows:

wherein

If the first stage is accelerating, the term ω (x) takes negative, otherwise it takes positive. After undergoing the uniform velocity process, the vehicle will uniformly decelerate to the confluence region.

Through the steps, a reasonable driving plan can be made according to the determined planned arrival time T (x) and the vehicle advancing direction, so that the vehicle can arrive at the intersection as early as possible on the premise of safety, and the passing efficiency of the intersection is improved.

As shown in fig. 11, the detailed steps of the method for dispatching the vehicles at the intersection are as follows: when a vehicle x to be dispatched reaches the boundary of the jurisdiction area, self track and steering information are sent to an edge terminal, the edge terminal generates a unique number for the vehicle x to be dispatched and transmits the unique number to the vehicle, and when the vehicle x to be dispatched reaches the boundary of the jurisdiction area, a coordinator transmits a current road condition aggregate B (t) to the vehicle; judging whether the track of the vehicle x contains a turn or not, if not, judging that the track is straight, estimating the earliest arriving time according to the maximum straight speed, otherwise, estimating the earliest arriving time according to the maximum steering speed; according to the earliest arrival time and the current road condition aggregate B (t), a vehicle x to be dispatched generates a temporary information set B _ temp (x) of the vehicle x and an integral temporary aggregate B _ temp (x, t) containing the B _ temp (x), enters a planned arrival time selection process, selects the planned arrival time T (x) of the vehicle, determines the information set B (x) of the x and the corresponding B (x, t), transmits the information set B (x) and the corresponding B (x, t) to a coordinator, and formulates a driving plan for the vehicle according to the T (x) and executes the driving plan.

In order to verify the effectiveness of the invention, the method is verified on a SUMO (simulation of Urban mobility) platform, and the platform provides a large amount of API (application program interface) which is easy to use for various vehicle behavior control and motion control. First, a road network structure as shown in fig. 2(a) is built on the platform, initial parameters of the model are shown in table 2, and the coordinator records information such as average speed and number of vehicles in the dispatching area at a period of 100 ms.

TABLE 2 model parameters

The road speed limit refers to the highest speed that a vehicle can reach in each lane, and the vehicle safety distance refers to the shortest distance that two vehicles on the same lane should at least keep in order to avoid rear-end collision.

The present embodiment specifically uses the following indicators to evaluate the performance of the scheduling:

average travel time in dispatch area T _ a: the time between the vehicle entering the dispatching area and the vehicle entering the confluence area is called the driving time of the dispatching area. This indicator is the average of the travel times of all the experimental vehicle dispatch zones, with lower values indicating that the vehicle is able to reach the intersection faster.

Total number of pauses wait: and recording the index of the total number of times of stopping of all vehicles, wherein the index is the sum of the number of times of stopping of all vehicles at the end of the experiment. The number of stops per vehicle is defined by equation (16):

the total number of pauses can reflect the smoothness of the vehicle travel and thus the degree to which the scheduling scheme is accurately executed. The fewer the total number of pauses, the smoother the vehicle travels within the dispatch area.

Average Delay ratio Delay: an index reflecting the speed performance utilization rate of the vehicle entering the dispatch area to itself is defined by equation (17):

wherein speed (x)_iT) represents the vehicle x at time t_iN represents the total number of vehicles in the dispatch area at time t. The lower the average delay ratio, the more the vehicle in the dispatch area can exert speed performance closer to its upper limit.

The running speed of the vehicle is reflected by the index of the average running time and the average delay ratio of the dispatching area, and the smoothness degree of the running of the vehicle is reflected by the total pause times of the index.

In this embodiment, a simulation experiment is performed on the model shown in fig. 2(a) by using a conventional RA (reservation algorithm) fixed value reservation scheduling method, a FIFO-RA scheduling method, and the scheduling method of the present invention, respectively, to compare the performance of each scheduling method. In the experiment, the same road parameters and vehicle parameters are respectively used by all the dispatching methods, and the total number of vehicles reaching the junction area of the crossroads along the road is subjected to Poisson distribution. In addition, in order to simulate vehicles having different acceleration performances and maximum speeds appearing on actual roads, five different types of vehicles were experimentally set, as shown in table 3.

TABLE 3 vehicle types

This example was specifically divided into the following two sets of data for comparative experiments:

experimental group 1: on the premise that the departure probabilities of all lanes are the same, the overall pressure of the traffic flow is changed, and therefore the situation that indexes of the model and the comparison method change along with the overall pressure of the traffic flow under the condition of traffic flow equalization is observed.

Experimental group 2: on the premise that the total arrival rate of the vehicle is not changed (equal to 0.72), the difference value of the departure rates of the lanes in the north-south direction and the east-west direction is changed to observe the change of the imbalance degree of the vehicle-dependent flow under the condition that the overall pressure of the vehicle is the same through the indexes of the model and the comparison method.

The specific experimental results of this example are shown in FIGS. 12 to 17. As is clear from fig. 12 and 13, the FIFS method of the present invention has significant advantages in the average travel time and average delay of the dispatch area compared to the conventional FIFO-RA and RA methods. For example, when the vehicle arrival rate is 1.15 vehicles/second, the two indexes of the FIFS method are respectively reduced by 6.90% and 21.55% compared with the traditional FIFO-RA method, and are respectively reduced by 16.02% and 41.95% compared with the traditional RA method. For the traditional FIFO-RA method, the essence of the disadvantages is that the vehicle with higher speed upper limit and acceleration capability is limited to the low-speed performance vehicle entering the dispatching area first, and it is difficult to exert its own speed advantage, while the RA dispatching method is not first-in first-out, but on one hand, the planned arrival time obtained by only speed estimation when the vehicle enters the dispatching area is not necessarily accurate, and on the other hand, the efficiency and safety cannot be simultaneously considered by adopting constant-value time interval evasion conflict, so that the two indexes are also weaker than the FIFS of the present invention.

As can be seen from fig. 14, the conventional FIFO-RA method and the RA method are closer to each other in terms of the total number of pauses, while the FIFS method of the present invention has a lower value and is more stable, i.e., the FIFS method of the present invention has fewer vehicle pauses and the vehicle runs more smoothly. As mentioned above, the delay of vehicles in the conventional FIFO-RA and RA methods is high, i.e. it means that vehicles are easy to gather near the junction area, which is also the main reason for the occurrence of the standstill, and besides, the strategy of making vehicles arrive at the intersection at a constant speed in RA makes it take longer for vehicles to pass through the junction area, thereby causing more standstill.

It should be noted that, as the traffic pressure increases, the indexes of the above three scheduling methods all increase to some extent, because the possibility of collision of vehicles also increases with the increase of the traffic pressure. However, the influence degrees of the methods by traffic flow pressure are different, for example, under the condition that the traffic flow pressure is increased to a certain degree, the delay of the comparison method in the dispatching area and the running time of the average dispatching area are both greatly increased, and the FIFS method is still stable. On the total pause times, the disadvantage that the traditional RA method is inaccurate in scheduling is more obvious when the traffic pressure is high, so that more pauses are caused in a confluence area. In the conventional FIFO method, as the traffic pressure rises, the straight vehicles which are not limited by the maximum turning speed are scheduled to arrive more frequently after the turning vehicles, which causes the traffic flow at the junction area to be slow and crowded, thereby causing more stops. According to the FIFS method, under the guidance of the conflict matrix, the conflict can be solved by adopting a relatively precise time interval, meanwhile, the straight vehicles can exert the physical performance of the straight vehicles to a certain extent, the pause risk in the traditional FIFO is avoided, and therefore the influence degree is obviously lower than that of the traditional method.

As can be seen from fig. 15 and 16, as the arrival rate deviation of the vehicles increases, the average delay of the dispatch area decreases somewhat with the FIFS method of the present invention, mainly because the higher the traffic flow is concentrated in a certain direction, the more the type of collision between the vehicles tends to be rear-end collision risk and no risk, and the minimum safety interval between the vehicles decreases as a whole, so that the vehicles can pass through the intersection more efficiently, and the collision profile analysis is shown in table 4 when the vehicles are distributed in all directions in a balanced manner and are completely concentrated in one direction. The conventional RA method has a decrease in average delay time for similar reasons, but the travel time of the dispatch area is increased because more stops are generated during the straight-ahead travel of the vehicle, not only in the junction area, as the traffic pressure is concentrated.

TABLE 4 analysis of collision Risk types in different situations

The effect of the stopping condition of each method along with the degree of the imbalance of the traffic pressure is shown in fig. 17, and it can be seen from the figure that the stopping times of the FIFS method of the present invention are still low and stable, while the conventional FIFO-RA has an upward trend, and the conventional RA has an unstable trend. Similar to the analysis in experimental group 1, there are more pauses than in the present FIFS, since the traditional FIFO method will gradually accumulate slow vehicles in the junction area due to more concentrated traffic flow. For the conventional RA method, however, its inaccurate estimated time of arrival will result in an unstable total dwell time.

The effectiveness of the method can be verified through a large number of simulation experiments, and the experimental results show that the method can effectively improve the vehicle passing efficiency of the intersection, is superior to the traditional method in indexes such as pause, average running time and the like of the automatically driven vehicle, and particularly has obvious advantages when the intersection faces high-pressure traffic and the average arrival rate of the vehicle is high.

The embodiment further comprises an automatic driving vehicle crossing dispatching system based on sequence selection, wherein the system comprises a coordinator arranged at a target crossing end and a vehicle-mounted control terminal arranged at a vehicle end, when a vehicle to be dispatched enters a dispatching area of the target crossing, the vehicle to be dispatched establishes communication with the coordinator of the target crossing through the vehicle-mounted control terminal, and when the vehicle to be dispatched enters a dispatched stage, the vehicle-mounted control terminal uploads information of the vehicle to be dispatched to the coordinator; the vehicle-mounted control terminal estimates the earliest time point of the vehicle to be dispatched to reach the intersection without considering other vehicles according to the distance between the vehicle to be dispatched and the target intersection and the current state information of the vehicle to be dispatched; acquiring current road condition information sent by the coordinator, wherein the road condition information comprises information of all vehicles in a dispatching area, and selecting the earliest time of a vehicle to be dispatched to reach a target intersection under safety constraint as a planned arrival time point according to the acquired road condition information and the estimated earliest time point, so that the time of the target vehicle arriving at the intersection is selected according to the dispatching sequence of the vehicles, and the safety constraint is that the vehicles can be prevented from colliding; and setting a driving plan for reaching the target intersection for the vehicle to be dispatched according to the planned arrival time point.

In this embodiment, the vehicle-mounted control terminal specifically includes:

the earliest time point estimation unit is used for estimating the earliest time point of the vehicle to be dispatched reaching the intersection without considering other vehicles according to the distance between the vehicle to be dispatched and the target intersection and the current state information of the vehicle to be dispatched;

a scheduled arrival time point selection unit, configured to acquire current traffic information sent by the coordinator, where the traffic information includes information of all vehicles in a scheduling area, and select, according to the acquired traffic information and the estimated earliest time point, the earliest time point at which a vehicle to be scheduled arrives at a target intersection under safety constraints as a scheduled arrival time point, so as to select, according to a scheduled sequence of the vehicles, a time at which the target vehicle arrives at the intersection, where the safety constraints are that collisions between the vehicles can be avoided;

and the driving planning unit is used for setting the driving planning of the vehicle to be dispatched to the destination intersection according to the planned arrival time point.

The vehicle-mounted control terminal can be arranged on a vehicle and can also be realized in a remote control center and other modes according to actual requirements.

In this embodiment, the earliest time point estimating unit corresponds to the earliest time point estimating step (step S2), the scheduled arrival time point selecting unit corresponds to the scheduled arrival time point selecting step (step S3), and the driving planning unit corresponds to the driving planning step (step S4), which are not described in detail herein.

The foregoing is considered as illustrative of the preferred embodiments of the invention and is not to be construed as limiting the invention in any way. Although the present invention has been described with reference to the preferred embodiments, it is not intended to be limited thereto. Therefore, any simple modification, equivalent change and modification made to the above embodiments according to the technical spirit of the present invention should fall within the protection scope of the technical scheme of the present invention, unless the technical spirit of the present invention departs from the content of the technical scheme of the present invention.

Claims

1. An automatic driven vehicle intersection scheduling method based on sequential selection is characterized by comprising the following steps:

s2, earliest time point estimation: estimating the earliest time point of the vehicle to be dispatched reaching the intersection without considering other vehicles according to the distance between the vehicle to be dispatched and the target intersection and the current state information of the vehicle to be dispatched;

s3, selecting a planned arrival time point: acquiring current road condition information sent by the coordinator, wherein the road condition information comprises information of all vehicles in a dispatching area, and selecting the earliest time of a vehicle to be dispatched to reach a target intersection under safety constraint as a planned arrival time point according to the acquired road condition information and the estimated earliest time point, so that the time of the target vehicle arriving at the intersection is selected according to the sequence of the vehicles to be dispatched, and the safety constraint is that the vehicles can be prevented from colliding;

2. The sequence selection-based automatic driven vehicle intersection scheduling method of claim 1, wherein the specific step of selecting the planned arrival time point in step S3 is:

s302, judging whether the vehicle to be dispatched is confronted with collision risks according to the current alternative arrival time in the current vehicle sequence, if so, turning to the step S303 to perform a new round of attempt, otherwise, successfully attempting, and determining the planned arrival time of the vehicle to be dispatched according to the current alternative arrival time;

3. The sequence selection-based automatic vehicle intersection scheduling method of claim 2, wherein in step S302, when the first attempt is made, that is, the current alternative arrival time is the earliest arrival time T₀(x) If the vehicle to be dispatched does not have a front vehicle and a rear vehicle, namely the vehicle to be dispatched is the only vehicle in the dispatching area, wherein the front vehicle is the vehicle which is closest to the vehicle to be dispatched before the time of reaching the target intersection, and the rear vehicle is the vehicle which is closest to the time of reaching the target intersection after the vehicle to be dispatched, the current alternative arrival time is directly used as the planned arrival time of the vehicle to be dispatched;

if the vehicle to be dispatched only has the front vehicle, obtaining the minimum safe time interval between the vehicle to be dispatched and the front vehicle, and selecting the planned arrival time of the vehicle to be dispatched according to the following formula so as to select the earliest possible safe time as the planned arrival time;

T(x)＝max(T(pre(x))+deal(x,pre(x)),T₀(x))

wherein x is a vehicle to be dispatched, T (x) is the planned arrival time of the vehicle to be dispatched, pre (x) is a front vehicle, T (pre (x)) is the time of the front vehicle to arrive at the destination intersection, deal (x, pre (x)) is the minimum interval between the vehicle to be dispatched and the front vehicle pre (x) which meets the safety; wherein if interval is satisfied<deal (pre (x), where interval is the earliest point in time T₀(x)The interval from the time T (pre (x)) when the preceding vehicle arrives at the destination intersection is directly selected as the planned arrival time, a time point apart from T (pre (x)) by deal (x, pre (x)).

4. The sequence selection-based automatic vehicle intersection scheduling method of claim 2, wherein in step S302, when the first attempt is made, that is, the current alternative arrival time is the earliest arrival time T₀(x) And the vehicle to be dispatched only exists in the rear vehicle, wherein if the interval (x, next (x)) is judged to be satisfied>deal (x, next (x)), where x is the vehicle to be dispatched, next (x) is the rear vehicle, and interval (x, next (x)) is the earliest arrival time point T₀(x) The distance between the vehicle to be dispatched and the rear vehicle next (x), deal (x, next (x)) is the minimum distance which meets the safety between the vehicle to be dispatched x and the rear vehicle next (x), and the current alternative arrival time point is directly selected as the planned arrival time point; if the interval (x, next (x)) is satisfied<deal (x, next (x)), determine that there is a collision risk for the vehicle to be dispatched according to the current alternative arrival time, and go to step S303 to perform a new round of attempt.

5. The sequence selection-based automatic vehicle intersection scheduling method of claim 2, wherein in step S302, when the first attempt is made, that is, the current alternative arrival time is the earliest arrival time T₀(x) And the vehicle to be dispatched has a front vehicle and a rear vehicle at the same time, if the following conditions are met:

interval(pre(x),next(x))<deal(pre(x),x)+deal(x,next(x))

if the interval (pre (x), next (x)) is satisfied>deal(pre(x), x) + deal (x, next (x)), and determining the earliest arrival time T₀(x) A value of (1), wherein if T₀(x)∈[T(next(x))-deal(x,next(x)),T(next(x))]If the current alternative arrival time is determined to cause the collision risk between the vehicle to be dispatched and the rear vehicle, the step S303 is executed to perform a new round of attempt; if T₀(x)∈[0,T(pre(x))+deal(x,next(x))]The planned arrival time is selected as a planned arrival time T (pre (x)) + deal (x, pre (x)), T being as early and safe as possible as T (pre (x)) + max (x, pre (x)))₀(x))。

6. The method for automatically scheduling the intersection of the driven vehicle based on the sequential selection as claimed in any one of claims 2 to 5, wherein the new candidate arrival time is re-selected backwards in step S303, specifically, the new candidate arrival time is obtained by moving to a position behind the arrival time of the current rear vehicle and keeping a minimum safety interval with the rear vehicle.

7. The method for automatically driving vehicle intersection scheduling based on sequence selection according to any one of claims 2-5, wherein the step S4 of setting the driving plan comprises:

wherein, t₁(x)＝T(x)-t₀(x) Is the time that the vehicle x to be dispatched passes from the entering dispatching area to the destination crossing, T (x) is the planned arrival time of the vehicle x to be dispatched to the destination crossing, t₀(x) For the time when the vehicle x to be dispatched enters the current dispatching area, v₀(x) For the speed at which the vehicle x to be dispatched enters the current dispatching zone, a_xFor the acceleration of the vehicle x to be dispatched when entering the current dispatching area, L is the distance from the entrance of the current dispatching area to the convergence areaThe convergence region is a region where lanes intersect at a destination intersection; acceleration/deceleration T of vehicle to be dispatched_accMaintaining the speed after a time until exiting the confluence region;

wherein the content of the first and second substances,

the duration of the uniform speed stage is as follows:

wherein

8. The method for automatically driving vehicle intersection dispatching based on sequence selection as claimed in any one of claims 1-5, wherein when estimating the earliest time point in step S2, if the vehicle to be dispatched is a straight-going vehicle, wherein if the vehicle to be dispatched can accelerate to a limit speed before reaching the intersection, the earliest time point when the vehicle to be dispatched reaches the destination intersection is estimated as:

9. The method for automatically driving vehicles to cross road scheduling based on sequence selection as claimed in any one of claims 1-5, wherein when the earliest time point is estimated in step S2, if the vehicle to be scheduled is a vehicle that needs to turn, if the vehicle to be scheduled can accelerate to the maximum speed during the driving process and decelerate to the maximum turning speed before reaching the junction area, i.e. the speed v when the vehicle to be scheduled enters the current scheduling area₀(x) Maximum steering speed v of the vehicle to be dispatched in the junction region_turn(x) Maximum speed v of the vehicle to be dispatched_max(x) Satisfies the following conditions:

10. an automatic driving vehicle crossing dispatching system based on sequence selection is characterized by comprising a coordinator arranged at a destination crossing end and a vehicle-mounted control terminal arranged at a vehicle end, wherein when a vehicle to be dispatched enters a dispatching area of the destination crossing, the vehicle to be dispatched establishes communication with the coordinator of the destination crossing through the vehicle-mounted control terminal, and the vehicle to be dispatched enters a dispatched stage; the vehicle-mounted control terminal estimates the earliest time point of the vehicle to be dispatched reaching the intersection without considering other vehicles according to the distance between the vehicle to be dispatched and the target intersection and the current state information of the vehicle to be dispatched; acquiring current road condition information sent by the coordinator, wherein the road condition information comprises information of all vehicles in a dispatching area, and selecting the earliest time of a vehicle to be dispatched to reach a target intersection under safety constraint as a planned arrival time point according to the acquired road condition information and the estimated earliest time point, so that the time of the target vehicle arriving at the intersection is selected according to the sequence of the vehicles to be dispatched, and the safety constraint is that the vehicles can be prevented from colliding; and setting a driving plan for reaching the target intersection for the vehicle to be dispatched according to the planned arrival time point.