CN108932856A - Intersection weighs setting method under a kind of automatic Pilot - Google Patents

Abstract

The invention discloses a method for setting the right of way at an intersection under automatic driving, which belongs to the field of intelligent transportation and can provide the best time for an automatic driving vehicle to enter the intersection and select the best exit lane. In the autonomous driving environment, the physical parameters of the intersection are determined, and the internal space of the intersection is divided into several square grids. Collect the number of the entrance lane where the vehicle is located, the direction of turning, and the predicted time when it reaches the stop line. Establish the vehicle route selection model, the calculation model for the time when the vehicle actually reaches the stop line, and the calculation model for the time when the vehicle enters and exits the grid; ensure that the same grid can only be occupied by one vehicle at the same time, and establish a multi-vehicle occupancy model. The constraint condition formula of the same grid; with the minimum total delay of vehicles arriving at the intersection as the objective function, optimize the best exit lane for each vehicle and the best time to pass through the intersection. The method of the invention can ensure the safe and efficient passage of the self-driving vehicle at the intersection.

Description

A method for setting the right of way at an intersection under automatic driving

technical field

The invention belongs to the field of intelligent transportation, relates to the technical field of traffic management and control for automatic driving vehicles on urban roads, and more specifically relates to a method for setting the right of way at an intersection under automatic driving.

Background technique

The development prospects of autonomous driving technology are very broad. Since 2018, Beijing, Shanghai, Chongqing, and Shenzhen have successively issued detailed rules for autonomous driving road tests for autonomous vehicles. Hangzhou, Guangzhou, and Wuhan are planning to build experimental bases for autonomous driving. In the autonomous driving environment, by using technologies such as wireless communication and the Internet, communication between vehicles and between vehicles and roads can be realized. This makes vehicles no longer need signal light control at the intersection, and vehicles can cooperate and cooperate with each other to pass through the intersection, realizing the transformation from signal phase-based signal control to "single vehicle" right-of-way control. However, when the amount of arriving traffic is large, how to coordinate the right-of-way of self-driving vehicles in the internal space of the intersection, and determine the optimal order of vehicles passing through the intersection between vehicles with different turning directions at different entrances, has become a need for research. important question.

On the other hand, most of the existing intersections have strictly divided the left-turn, straight-going, and right-turning lanes of the entrance lanes. Vehicles need to change lanes to the corresponding lanes before entering the intersection. However, there is no regulation which lane must be driven into, and any exit lane can be selected to exit. Under automatic driving conditions, it is necessary to strictly set the lanes for each vehicle entering and exiting the intersection. On the one hand, the safety of autonomous vehicles can be improved. traffic efficiency.

Therefore, the present invention provides a method for setting the right of way at an intersection under automatic driving. Under the condition of automatic driving, the entrance road at the intersection does not need to divide the lane function, that is, all the entrance roads can drive "straight from left to right". , while optimizing the transit time and exit lanes of autonomous vehicles. With the goal of minimizing the total delay, the internal area of the intersection is divided into several grids, and the time occupied by each grid is calculated when the vehicle chooses different exit lanes. By restricting a grid to be occupied by at most one vehicle, the safety and efficiency of vehicles can be achieved through the intersection area.

According to the literature search of the existing technology, the existing literature on the right of way at the intersection of automatic driving is mainly based on the first-come-first-served traffic method, and lacks the overall consideration of the total delay of arriving vehicles at the entire intersection.

Contents of the invention

Technical problem: Aiming at the problem of how to optimally determine the traffic sequence and select the best exit lane under the condition that all lanes of the entrance road can pass "straight from left to right" for automatic driving vehicles, the present invention provides a crossover under automatic driving. The method of setting the right of way at the intersection ensures that autonomous vehicles pass through the intersection safely, efficiently and orderly.

Technical solution: In order to solve the above technical problems, a method for setting the right of way at an intersection under automatic driving of the present invention includes the following steps:

Step 1: Determine the number of entrance lanes and exit lanes at the intersection and number them respectively, divide the internal space of the intersection into several small square grids and number the grids, and determine the coordinate range of each grid; collect The number of the entrance lane where the vehicle is located, the direction of turning, the predicted time of arrival at the stop line, the grid where all the vehicle passages at the intersection are pressed and the position points where the vehicle enters and exits the grid;

Step 2: Establish a vehicle driving route selection model, establish a calculation model for the time when the vehicle actually reaches the stop line, and establish a calculation model for the time when the vehicle enters the grid and the time when it leaves the grid;

Step 3: Ensure that the same square can only be occupied by one vehicle at the same time, and establish a constraint formula for multiple vehicles occupying the same square;

Step 4: Taking the minimum total delay of arriving vehicles at the intersection as the objective function, optimize the best exit lane for each vehicle and the best time to pass the intersection.

In the present invention, step 1 includes the following steps:

Step 11: Use the parameter O to indicate the intersection entrance, the parameter D to indicate the intersection exit, and the parameters E, W, S, and N to represent the east, west, south, and north directions respectively, and O∈{E,W,S,N} , D∈{E,W,S,N}; g means the gth vehicle on the lane; Oi→Dj means that the vehicle is driving from the i-th lane of the O entrance to the j-th lane of the D exit, regardless of Turn around, therefore, in Oi→Dj, O≠D; each entrance road contains i _O lanes, where i _O ∈ {1,2,…,n _O }, n _O represents the maximum number of lanes of the entrance road in the O direction , each exit road contains j _D lanes, where j _D ∈ {1,2,…,m _D }, m _D represents the maximum number of lanes of the exit road in the D direction. The speed of the vehicle is represented by the parameter v, considering that the vehicle is traveling at a constant speed inside the intersection and cannot stop. Divide the intersection into several small square grids and number them. R _pq means that the numbers corresponding to the grid R on the x-axis and y-axis are p and q respectively; establish a Cartesian coordinate system and collect the grids that each vehicle passes through and points of entry and exit from the grid, Indicates that the vehicle on the path Oi→Dj enters the position point of the ath square R _pq , then where a ∈ {1,2,…,A}, A represents the total number of entry points; Indicates the position point where the path Oi→Dj leaves the bth square R _pq , where b∈{1,2,…,B}, B represents the total number of exit location points.

In the present invention, step 2 establishes the vehicle travel path selection model, establishes the calculation model of the moment when the vehicle actually arrives at the stop line, and sets up the calculation model of the moment when entering the grid and the time when the grid exits, including the following steps:

Step 21: Vehicles turn left and right at the entrance road, and the direction of exiting the intersection is known and determined. Therefore, select a lane for all exits in the exiting direction to exit, which is calculated by formula (1).

formula, is a binary variable, Indicates that the gth vehicle chooses the path Oi→Dj, that is, it travels from the i-th entry road in the O direction to the j-th exit road in the D direction; Indicates that the vehicle does not drive out from exit j.

When the vehicle is going straight and the entrance lane i _O is less than or equal to the number of exit lanes j _D , that is, i _O ≤ j _D , the path chosen by the straight vehicle is calculated by formula (2):

When the vehicle is going straight and the entrance lane i _O is greater than the exit lane j _D , that is, i _O > j _D , the vehicles on the lane with more entrance lanes than the exit lane drive to the nearest exit lane, and the route selection is calculated by formula (3) :

Step 22: Use Indicates that the actual moment of arrival at the stop line is represented by Indicates that when leaving the stop line and entering the intersection, use express.

The predicted moment when the gth car arrives at the stop line is greater than or equal to the moment when the previous car g-1 leaves the stop line, that is Then there is no need to wait in line; if , then the gth car cannot arrive at the stop line at the predicted time and needs to wait in line, then the actual time at which it arrives at the stop line Calculated by formula (4).

In the formula, d _v is the distance from the vehicle to the stop line at the intersection, and G _Oi represents the total number of vehicles on an entry road.

When g=1, that is, when the vehicle is the first vehicle in the lane, there is no waiting vehicle in front, and the actual time of reaching the stop line is equal to the predicted time of reaching the stop line, which is calculated by formula (5).

Step 23: The gth vehicle enters the location point time to use Indicates that driving out of the position point time to use Indicates that the moment when the vehicle enters the first position is equal to the moment when the vehicle leaves the stop line, which is calculated by formula (6):

The moment when the gth car enters other positions is calculated by formula (7):

The moment when the gth car leaves the first location point is calculated by formula (8):

The moment when the gth car leaves other locations is calculated by formula (9):

In the present invention, step 3 ensures that the same square can only be occupied by one vehicle at the same time, and establishes a constraint formula for multiple vehicles occupying the same square, which specifically includes the following steps:

Step 31: There will be multiple vehicles with different driving paths arriving at the same grid R _pq , and the time when the vehicles arrive at and leave the grid must be compared pair by pair, and the set of all vehicles arriving at the intersection is defined as G. For any vehicle g ₁ , g ₂ ∈G, suppose the path of vehicle g ₁ is (Oi→Dj) ₁ , the path of vehicle g ₂ is (Oi→Dj) ₂ , for any O∈{E,W,S,N}, D ∈ {E, W, S, N}, i ∈ {1, 2,..., n _o }, j ∈ {1, 2,..., m _D }. Using formula (10) and formula (11) can ensure that the same square can only be occupied by one vehicle at the same time.

formula is a binary variable, when the grid R _pq is on the path Oi→Dj, When the square R _ab is not on the path Oi→Dj, M is a large positive number, y is a binary variable, which can be 0 or 1;

In the present invention, step 4 takes the minimum total delay of vehicles arriving at the intersection as the objective function, optimizes and obtains the best exit lane of each vehicle and the best time by the intersection, specifically comprising the following steps:

Step 41: The delay of the gth vehicle is calculated by the formula (12), and the delay is equal to the time of leaving the stop line minus the predicted time of arriving at the stop line. The total delay is calculated by formula (13), according to the formula (14) with the minimum total delay, and publicity (1)-(13), the best path for each vehicle in the intersection area can be determined That is, the best exit lane, and the best time to leave the stop line

MIN(Delay) (14)

Beneficial effect: compared with the prior art, the present invention has the following advantages:

The method of the invention can determine the right of way and the exit lane of the vehicle before the self-driving vehicle enters the intersection, so as to ensure that the vehicles on each entrance lane pass through the intersection safely and orderly with the lowest total delay.

Description of drawings

Fig. 1 is the flowchart of the inventive method;

Figure 2 is a schematic diagram of the embodiment.

Detailed ways

In conjunction with the accompanying drawings and embodiments, the technical solution of the present invention is described in detail as follows:

Example: Select a typical intersection in the city as the research object. In the autonomous driving environment, the real-time position of the arriving vehicle, the predicted time of arriving at the stop line, the turning direction, and the position of all traffic paths entering the grid inside the intersection can be obtained. and out of square position. In the example, the number of entrance lanes in each direction is: n _E =n _S =n _N =3, n _W =4; the number of exit lanes in each direction is: m _W =m _S =m _N =2, m _E = 3. The width of each lane at the intersection is 3m, and the side length of the set grid is 3m. The speed v=10m/s through the intersection at a constant speed. Randomly generate 13 self-driving vehicles, optimize the driving path of these 13 vehicles and the best time to enter the intersection. The predicted time and turning direction of the self-driving vehicles at each entrance road to the stop line are shown in Figure 2 and Table 1.

Table 1 Predicted time and turning direction table of the self-driving vehicles arriving at the stop line at each entrance road

Vehicle 1 is a straight vehicle, the number of the grid it has driven, the coordinates of the position point of the entry grid and the coordinates of the grid position point of the exit, the moment when the front of the vehicle enters the grid, and the moment when the rear of the vehicle exits the grid, as shown in the table 2 shown.

Table 2: The moment when the head of vehicle 1 enters the grid and the moment when the rear leaves the grid

Vehicle 2 is a straight vehicle, the number of the grid it has driven, the coordinates of the position point of the entry grid and the coordinates of the grid position point of the exit, the moment when the front of the vehicle enters the grid, and the moment when the rear of the vehicle exits the grid, as shown in the table 3 shown.

Table 3: The moment when the head of vehicle 2 enters the grid and the moment when the rear of the vehicle leaves the grid

Vehicle 3 is a straight vehicle, the number of the grid it passes by, the time when the front of the vehicle enters the grid and the time when the rear of the vehicle exits the grid is shown in Table 4.

Table 4: The moment when the front of vehicle 3 enters the grid and the moment when the rear exits the grid

Vehicles 4, 5, 6, 7, and 8 are straight vehicles, and the grids they pass, the moment when the front of the vehicle enters the grid, and the time when the rear of the vehicle exits the grid are shown in Table 5.

Table 5: The squares passed by vehicles 4, 5, 6, 7, and 8, the moment when the front of the vehicle enters the square, and the moment when the rear of the vehicle leaves the square

Vehicle 9 is a left-turning vehicle, and there are S1 and S2 to choose from at the exit lane. The time of grid-out is shown in Table 6.

Table 6: The grid that vehicle 9 passes through on the path E2→S1, the position point coordinates of entering and leaving the grid, the moment when the front of the vehicle enters the grid, and the moment when the rear of the vehicle exits the grid

Table 7 shows the position point coordinates of the grid, the entry grid, and the exit grid of the vehicle 9 on the path E2→S2, and the time when the vehicle enters the grid and exits the grid.

Table 7: The grid that vehicle 9 passes through on the route E2→S2, the position coordinates of the entry grid and the exit grid, the moment when the front of the vehicle enters the grid, and the moment when the rear of the vehicle exits the grid

The vehicle 10 is a left-turning vehicle, and there are N1 and N2 exit lanes to choose from. When the driving path is W3→N1, the coordinates of the grid that the vehicle 10 passes, the location point coordinates of the grid that it enters and the grid that it exits, and the Table 8 shows the grid and the time of leaving the grid, and it is shown in Table 9 when the driving path is W3→N1.

Table 8: The grid that the vehicle 10 passes through on the path W3→N1, the position coordinates of the entry grid and the exit grid, the moment when the front of the vehicle enters the grid, and the moment when the rear of the vehicle exits the grid

Table 9: The grid the vehicle passes through on the path W3→N2, the position coordinates of the vehicle entering the grid and the grid exiting the grid, and the coordinates projected to the outer boundary, the moment when the front of the vehicle enters the grid and the time when the rear of the vehicle travels out of the box moment

Vehicle 11 is a right-turning vehicle, and there are three lanes to choose from at the exit, namely E1, E2, and E3. When the driving path is S1→E1, the positions of the passing grid, entering grid and leaving grid The point coordinates, the time of entering and leaving the grid are shown in Table 10; when the driving route is S1→E2, it is shown in Table 11; when the driving route is S1→E3, it is shown in Table 12.

Table 10: The grid that the route S1→E1 passes through, the position coordinates of the entry grid and the exit grid, the moment when the front of the vehicle enters the grid, and the moment when the rear of the vehicle exits the grid

Table 11: The grid that the vehicle passes through on the route S1→E2, the position coordinates of the vehicle entering the grid and the grid exiting the grid, the moment when the front of the vehicle enters the grid, and the moment when the rear of the vehicle exits the grid

Table 12: The moment when the head of the route S1→E3 enters the grid and the moment when the rear of the car leaves the grid

Vehicle 12 is a right-turning vehicle, and there are three exit lanes to choose from, which are E1, E2, and E3 respectively. Table 13 shows the time when the front of the vehicle enters the grid and the time when the rear of the vehicle exits the grid on different driving paths.

Table 13: The moment when the front of the vehicle enters the grid and the moment when the rear exits the grid on different paths of vehicle 12

Vehicle 13 is a right-turning vehicle, and there are two optional lanes corresponding to the exit road, W1 and W2 respectively. Table 14 shows the time when the front of the vehicle enters the grid and the time when the rear of the vehicle exits the grid on different paths.

Table 14: The moment when the front of the vehicle enters the grid and the moment when the rear exits the grid on different paths of vehicle 13

According to the formula (14) and formula (1)-(13) in step 4, the total delay minimum MIN(Delay)=15.9s for 13 vehicles passing through the intersection can be obtained. At this time, the optimal exit lane and The best time for vehicles to leave the stop line is shown in Table 15; the predicted time of vehicles arriving at the stop line, the actual time of arrival at the stop line and the vehicle delay are shown in Table 16. If there is no vehicle in line ahead, the actual arrival time is equal to the predicted arrival time.

Table 15: Selection of the best exit lane for vehicles and the best timetable for vehicles leaving the stop line

Table 16: Predicted time of arrival at the stop line, actual time of arrival at the stop line, and vehicle delay table for vehicles at each entrance road

Claims (5)

1. intersection weighs setting method under a kind of automatic Pilot, which is characterized in that this method comprises the following steps:

Step 1: determining each entrance driveway in intersection, exit ramp number of track-lines and it is numbered respectively, by the inner space of intersection point For several squares lattice and grid is numbered, determine the coordinate range of each grid；Entrance driveway where acquiring vehicle Number, turn direction reach the prediction time of stop line, grid that all vehicle pass-through paths in input intersection are pressed onto and Vehicle enters and is driven out to the location point of grid；

Step 2: establishing vehicle running path preference pattern, establish the computation model that vehicle is actually reached the stop line moment, establish Computation model at the time of at the time of driving into grid and being driven out to grid；

Step 3: ensuring that the same grid can only be occupied in synchronization by a vehicle, establish more trolleys and occupy the same grid Constraint condition formula；

Step 4: the minimum objective function of total delay of vehicle is reached with intersection, optimization obtains the best exit lane of each car And the best time for passing through intersection.

2. intersection weighs setting method under a kind of automatic Pilot according to claim 1, which is characterized in that the step Rapid 1 includes the following steps:

Step 11: indicating crossing inlet road with parameter O, parameter D indicates intersection exit road, parameter E, W, S, and N is respectively indicated East, West, South, North direction, O ∈ { E, W, S, N }, D ∈ { E, W, S, N }；G indicates the g vehicle on lane；Oi → Dj indicates vehicle The j-th strip lane that D exit ramp is driven towards from i-th lane of O entrance driveway does not consider to turn around to travel, therefore, in Oi → Dj, and O ≠ D；Each entrance driveway includes i_OLane, wherein i_O∈{1,2,…,n_O, n_OIndicate the direction O entrance driveway maximum number of track-lines, each Exit ramp includes j_DLane, wherein j_D∈{1,2,…,m_D, m_DIndicate the direction D exit ramp maximum number of track-lines；Speed parameter v It indicates, considers that vehicle drives at a constant speed inside intersection, cannot stop；The lattice that intersection is divided into several squares is gone forward side by side Row number, R_pqExpression grid R is p, q respectively in the corresponding number of x-axis and y-axis；Rectangular coordinate system is established, acquisition each car presses through Grid and entrance and the location point for being driven out to grid,Indicate that the vehicle on the Oi → Dj of path enters a-th of grid R_pq's Location point, thenWherein a ∈ { 1,2 ..., A }, A indicate entry into the sum of location point；Indicate that path Oi → Dj is driven out to b-th of grid R_pqLocation point,Wherein b ∈ 1, 2 ..., B }, B indicates the sum for being driven out to location point.

3. intersection weighs setting method under a kind of automatic Pilot according to claim 1, which is characterized in that the step In rapid 2, vehicle running path preference pattern is established, establishes the computation model that vehicle is actually reached the stop line moment, foundation is driven into Computation model at the time of at the time of grid and being driven out to grid, includes the following steps:

Step 21: entrance driveway vehicle turns left and turns right, and the direction for being driven out to intersection is known and determining, therefore in the institute for being driven out to direction There is exit ramp that a lane is selected to be driven out to, is calculated by formula (1)；

In formula,For binary variable,Indicate the g vehicle selection path Oi → Dj, i.e. i-th from the direction O Entrance driveway drives to the j-th strip exit ramp in the direction D；Indicate that vehicle is not driven out to from j-th strip exit ramp；

When vehicle is straight trip and entrance driveway i_OLess than or equal to exit ramp number of track-lines j_D, i.e. i_O≤j_DWhen, the path of through vehicles selection It is calculated by formula (2):

When vehicle is straight trip and entrance driveway i_OGreater than exit ramp j_DThat is i_O>j_DWhen, the vehicle on lane that entrance driveway has more than exit ramp It drives towards apart from nearest exit ramp, Path selection is calculated by formula (3):

The g vehicle prediction on i-th article of entrance driveway in the direction step 22:O is used at the time of reaching stop lineIt indicates, reaches parking The practical moment of line is usedIt indicates, sails out of stop line, used at the time of into intersectionIt indicates；

At the time of the prediction time that the g vehicle reaches stop line sails out of stop line more than or equal to previous vehicle g-1, i.e., Do not have to then wait in line；IfWhen, then the g vehicle cannot reach stop line by prediction time, need to wait in line, then Reach the practical moment of stop lineIt is calculated by formula (4)；

D in formula_vFor vehicle to the distance of intersection parking line, G_OiIndicate the vehicle fleet on an entrance driveway；

As g=1, i.e., when vehicle is lane first car, front is actually reached stop line without waiting for current vehicle at this time At the time of be equal to prediction reach the stop line moment, by formula (5) calculate；

Step 23: the g vehicle drives into location pointAt the time of useIt indicates, is driven out to location pointAt the time of useIt indicates, at the time of leaving stop line equal to vehicle at the time of driving into first location point, is calculated by formula (6):

The g vehicle is calculated at the time of driving into other positions point by formula (7):

The g vehicle is calculated at the time of being driven out to first location point by formula (8):

The g vehicle is calculated at the time of being driven out to other positions point by formula (9):

4. intersection weighs setting method under a kind of automatic Pilot according to claim 1, which is characterized in that the step In rapid 3, it is ensured that the same grid can only be occupied in synchronization by a vehicle, establish the pact that more trolleys occupy the same grid Beam condition formula, includes the following steps:

Step 31: the vehicle that can have a plurality of different driving paths reaches the same grid R_pq, must compare two-by-two vehicle reach and At the time of being driven out to grid, the collection that definition reaches all vehicles in intersection is combined into G, for any one vehicle g₁,g₂∈ G, it is assumed that vehicle g₁Path be (Oi → Dj)₁, vehicle g₂Path be (Oi → Dj)₂, to arbitrary O ∈ { E, W, S, N }, D ∈ { E, W, S, N }, i∈{1,2,…,n_O, j ∈ { 1,2 ..., m_D}；It may insure the same grid in same a period of time using formula (10) and formula (11) Carving can only be occupied by a vehicle；

In formulaFor binary variable, as grid R_pqWhen on the Oi → Dj of path,As grid R_abNot in path When on Oi → Dj,M be big positive number, y is binary variable, can value be 0 or 1.

5. intersection weighs setting method under a kind of automatic Pilot according to claim 1, which is characterized in that the step In rapid 4, with intersection reach vehicle the minimum objective function of total delay, optimization obtain each car best exit lane and By the best time of intersection, include the following steps:

Step 41: the delay of the g vehicle is calculated by formula (12), and delay, which is equal at the time of sailing out of stop line to subtract prediction and reach, to stop At the time of fare；Total delay is calculated by formula (13), can be true according to the smallest formula of total delay (14) and publicity (1)-(13) Determine optimal path of each car in the region of intersectionI.e. best exit lane, and sail out of stop line optimal time

MIN(Delay) (14)。