CN106910350A - A kind of method for finding signalized crossing group's critical path - Google Patents

Abstract

The invention discloses a method for finding a key path of a signal control intersection group, which comprises the following steps: (1) obtaining various data information of all road sections and intersections of the intersection group; The discrete correlation index I ₁ of the starting and ending points of the path; (3) calculating the blocking correlation index I ₂ of each section of the intersection group; (4) order As the total relevance index of a path, I ₂ is the average value of all sections of the path, use the population algorithm to search the path of the intersection group, and get the path with the highest I as the critical path. The present invention fine-tunes the path correlation model, searches for the path with the highest correlation degree, does not need to traverse all paths, greatly reduces the calculation amount, thereby reduces the calculation time of the entire signal control intersection group space-time resource coordination and optimization system, and enhances the efficiency of coordination control Real-time and effectiveness, increasing the scale of the intersection group it can handle.

Description

A Method of Finding the Critical Path of Signal Controlled Intersection Group

technical field

The invention relates to the field of traffic signal control, in particular to a method for finding a key path of a signal-controlled intersection group.

Background technique

Most of the intersections on the arterial roads of the urban road network in my country are signal-controlled intersections, and single-point signal control is carried out respectively. Vehicles frequently stop at intersections during single-point signal control, which leads to traffic problems such as inefficient road network operation and increased travel delays. In order to reduce the parking time of vehicles at each intersection, the time-space resource optimization system of intersection group monitors the traffic volume in real time, and coordinates and controls multiple intersections as a whole to reduce the congestion of intersection groups.

The existing coordination control system of intersection group is divided into five parts, including five steps of target intersection identification, intersection group range delimitation, critical path retrieval, spatio-temporal resource optimization and online real-time adjustment. Among them, in the model adopted by the critical path search module, the calculation of the path correlation degree has a dimensionless step, so that the calculation of the correlation degree of each path depends on the extreme value of all path blocking correlation degrees and discrete correlation degrees. All paths at the intersection are unavoidable, which makes the calculation of this module take a lot of time, often reaching several minutes. The traffic flow has changed during this period, which weakens the real-time performance of the entire coordinated control system and affects the adjustment effect.

Contents of the invention

The technical problem to be solved by the present invention is to provide a method for finding the critical path of signal-controlled intersection groups, which can reduce the amount of computation and increase the scale of the intersection groups that can be processed.

In order to solve the above-mentioned technical problems, the present invention provides a method for finding the key path of signal control intersection group, comprising the following steps:

(1) Obtain various data of the intersection group, including the range X of the intersection group, the number of intersections N, the key intersection K, the connection mode between the intersection and the road section, and the number of lanes of each entrance lane at each intersection, channel optimization scheme, queuing length, vehicle speed, and traffic volume in each direction;

(2) Traversing the data of all edge intersections, calculating the discrete correlation degree I ₁ of all paths in the intersection group, the calculation formula is as follows:

I ₁ =n _d /n ₀

n _d ——the number of vehicles passing by the convoy within the green light time at the starting point of the route (veh)

n _o ——the number of vehicles passing by the convoy within the green light time at the end of the route (veh);

(3) Traversing all road sections in the intersection group, calculate the blocking correlation degree I ₂ of the intersection at both ends, and the blocking correlation degree of the one-way N-lane road section m The calculation formula is:

L——length of road section (m)

——The functional area length of the nth lane of the road section (m)

——The vehicle queuing length of the nth lane of road section m (m)

——Deceleration distance (m)

——perception-response distance (m);

(4) Then obtain the path with the largest overall correlation degree I in the road network by genetic algorithm, there are M road sections, and its overall correlation degree is as follows:

Preferably, in step (4), the specific steps of genetic algorithm are:

(41) Number each intersection, that is, X={0,1,...,N}, represent a path with a number sequence, that is, R={x _n }, R is a path, and x _n represents the nth path of the path Intersection: Randomly search outward from the key intersection, generate a path R with both ends at the boundary of the intersection group, and repeatedly generate multiple R to form the initial population

(42) in Randomly choose two intersecting paths in :

rand(set) - randomly selects an element in the set

Let X ₁₂ ＝R ₁ ∩R ₂ , if |X ₁₂ |≥2, randomly select two intersection points of these two paths:

x ₁ =rand(X ₁₂ ), x ₂ =rand(X ₁₂ )

Exchange the parts of R ₁ and R ₂ between x ₁ and x ₂ to generate child paths R _F1 and R _F2 ; if |X ₁₂ |=1, exchange the parts after the intersection point x∈X ₁₂ to generate R _F1 , R _F2 ; repeat this process many times, and form all the generated offspring paths into the offspring population Simulate the process of mating and reproduction in nature;

(43) calculation The associativity I of all paths, and in Randomly select multiple paths R:

Taking adaptive probability, the selected probability formula of each path R is as follows:

Both p _min and p _max are adjustable parameters

f _max —— The maximum value of f _Rn for all paths in

f _min —— The minimum value of f _Rn in

i _R ——the correlation degree I _R of the path R in Descending rank in all paths

Randomly select an intersection x _i in the selected R, if i≠0∧i≠|R|, then select another adjacent intersection x′ _i of x _i-1 :

x _i = rand(R)

x′ _i ＝rand({x|x∈X∧join(x, _xi-1 )})

Otherwise choose another edge intersection x′ _i ;

If x′ _i is directly connected to the subsequent intersection x _{i+i ,} replace x _i in R with x′ _i ; otherwise find the shortest path r′={x′ _i ,x′ _{i+ 1} ,...,x′ _i+i′ }, replace _xi in R with r′; this step simulates the biological gene mutation process.

(44) Recalculate The association degree I of all paths is screened, and only a part of the paths with the same number of parental populations are left, forming The probability of path R being removed is:

(45) If the correlation degree of the path of the population does not meet the convergence condition and the predetermined maximum number of iterations, make And return to step (42); otherwise, stop the iteration and use the current path R _max with the highest correlation degree as the critical path.

The beneficial effects of the present invention are: the present invention fine-tunes the path correlation model, and the correlation degree of each path is decoupled from other paths of the intersection group, and can be calculated separately, so that an approximate solution can be obtained by using a heuristic algorithm, and the correlation degree can be searched The highest path does not need to traverse all paths, which greatly reduces the amount of computation, thereby reducing the computation time of the entire signal control intersection group space-time resource coordination and optimization system, enhancing the real-time performance and effectiveness of coordination control, and increasing the number of intersections it can handle population size.

Description of drawings

Fig. 1 is a schematic flow chart of the method of the present invention.

detailed description

As shown in Figure 1, a method for finding the critical path of a signal-controlled intersection group includes the following steps:

(1) Obtain various data of the intersection group, including the range X of the intersection group, the number of intersections N, the key intersection K, the connection mode between the intersection and the road section, the number of lanes M of each entrance road, channelization Plans, etc., and obtain real-time traffic flow information such as queue length, vehicle speed, traffic volume and vehicle speed in each flow direction through electronic equipment.

(2) Traversing the data of all edge intersections, calculating the discrete correlation degree I ₁ of all paths in the intersection group, the calculation formula is as follows:

I ₁ =n _d /n ₀

n _d ——the number of vehicles passing by the convoy within the green light time at the starting point of the route (veh)

n _o ——the number of vehicles passing by the fleet within the green light time at the end of the route (veh)

(3) Traverse all road sections in the intersection group, and calculate the blocking correlation degree I ₂ of the intersections at both ends. Relevance degree of retardation of one-way N-lane road section m The calculation formula is:

L——length of road section (m)

——The functional area length of the nth lane of the road section (m)

——The vehicle queuing length of the nth lane of road section m (m)

——Deceleration distance (m)

——perception-reaction distance (m)

(4) Then use the genetic algorithm to obtain the path with the largest overall correlation degree I in the road network. For a path with M road segments, its overall relevance is as follows:

The specific steps of genetic algorithm are:

(40) Set the condition for the end of the iteration:

And set the maximum number of iterations T;

(41) Let all intersections be a set X, and number each intersection, that is, X={0,1,...,N}, and use a numbering sequence to represent a path. That is, R={x _n }, R is a path, and x _n represents the nth intersection of the path. Randomly search outward from the critical intersection to generate a path from the critical intersection to the boundary of the intersection group:

R ₁ ＝{x _n |x _n ＝rand(X)∧(n＝0∨join(x _n ,x _n-1 )}

R ₂ ＝{x _n |x _n ＝rand(X)∧join(x _n-1 ,x _n )}

rand(set) - randomly selects an element in the set

Generate multiple R repeatedly to form the initial population

(42) in Randomly choose two intersecting paths in :

Let X ₁₂ ＝R ₁ ∩R ₂ , if |X ₁₂ |≥2, randomly select two intersection points of these two paths:

x ₁ =rand(X ₁₂ ), x ₂ =rand(X ₁₂ )

Exchange parts of R ₁ and R ₂ between x ₁ and x ₂ to generate descendant paths R _F1 and R _F2 . If |X ₁₂ |=1, exchange the part after the intersection point x∈X ₁₂ to generate R _F1 and R _F2 . Repeat this process multiple times to form all the generated offspring paths into offspring populations

(43) calculation The associativity I of all paths, and in Randomly select multiple paths R:

Taking adaptive probability, the formula of the selected probability of path R is as follows:

Both p _min and p _max are adjustable parameters

f _max —— The maximum value of f _Rn for all paths in

f _min —— The minimum value of f _Rn in

i _R ——the correlation degree I _R of the path R in Descending rank among all path associations in

Randomly select an intersection x _i in R, if i≠0∧i≠|R|, then select another adjacent intersection x′ _i of x _i-1 :

x _i = rand(R)

x′ _i ＝rand({x|x∈X∧join(x, _xi-1 )})

Otherwise choose another edge intersection x′ _i .

If x′ _i is directly connected to the subsequent intersection x _{i+i ,} replace x _i in R with x′ _i ; otherwise find the shortest path r′={x′ _i ,x′ _{i+ 1} ,…,x′ _i+i′ }, replace _xi in R with r′.

(44) Recalculate The association degree I of all paths is screened, and only a part of the paths with the same number of parental populations are left, forming The probability of path R being removed is:

(45) If the correlation degree of the path of the population does not meet the convergence condition P, and the number of iterations does not reach the predetermined maximum number of iterations T, then let And return to step 42); otherwise, stop the iteration and use the current path R _max with the highest correlation degree as the critical path.

This method aims to improve the critical path retrieval module of the urban road signal control intersection group spatio-temporal resource coordination system and improve the solution speed.

Although the present invention has been illustrated and described in terms of preferred embodiments, those skilled in the art should understand that various changes and modifications can be made to the present invention without departing from the scope defined by the claims of the present invention.

Claims (2)

1. it is a kind of find signalized crossing group critical path method, it is characterised in that comprise the following steps：

(1) obtain each item data of intersection group, including intersection group scope X, intersection quantity N, key intersection K, friendship It is the connected mode in prong and section and the number of track-lines of each each entrance driveway of intersection, canalization scheme, queue length, speed, each The volume of traffic of flow direction；

(2) data of all edge intersections are traveled through, the discreteness degree of association I in all paths in intersection group is calculated₁, meter Calculate formula as follows：

I₁=n_d/n₀

n_d--- the vehicle number (veh) that fleet passes through in the starting point green time of path

n_o--- the vehicle number (veh) that fleet passes through in path termination green time；

(3) all sections in traversal intersection group, calculate the retardancy degree of association I of its two ends intersection₂, unidirectional N tracks road The retardancy degree of association of section mComputing formula is：

$I_2^m=\frac{max(D_1^m,D_2^m,...,D_n^m,...,D_N^m)}{L}$

$D_n^m=d_{1n}^m+d_{2n}^m+d_{3n}^m$

L --- road section length (m)

--- function section length (m) in n-th of section track

--- the vehicle queue length (m) in section m nth bars track

--- deceleration distance (m)

--- perception-reaction distance (m)；

(4) path of population interconnection degree I maximums in road network and then by genetic algorithm is drawn, has the path in M bars section, its totality The degree of association is as follows：

$I=I_1+\stackrel{\‾}{I_2^m}$

$\stackrel{\‾}{I_2^m}={\mathrm{\Σ}}_{m=1}^M\frac{I_2^m}{M}.$

2. the method for finding signalized crossing group's critical path as claimed in claim 1, it is characterised in that step (4) In, genetic algorithm is concretely comprised the following steps：

(41) for each intersection is numbered, i.e. X={ 0,1 ..., N } represents a paths, i.e. R={ x with numbered sequence_n, R is one Paths, x_nRepresent n-th of path intersection；The outside random search since crucial intersection, one two ends of generation are in The path R on intersection group border, repeatedly generates multiple R, constitutes initial population

(42) existTwo intersecting paths of middle random selection：

Rand (set) --- an element is randomly choosed in set

Make X₁₂=R₁∩R₂If, | X₁₂| >=2, two intersection points of this two paths are taken at random：

x₁=rand (X₁₂),x₂=rand (X₁₂)

By R₁、R₂In x₁、x₂Between part exchange, generation filial generation path R_F1、R_F2；If | X₁₂|=1, then by intersection point x ∈ X₁₂Afterwards Part exchange, generate R_F1、R_F2；The process is repeated several times, all filial generation paths that will be generated constitute progeny populationSimulate mating, the reproductive process of nature；

(43) calculateThe degree of association I in all paths, andRandomly select multiple path R：

Adaptive probability is taken, the new probability formula being selected of each path R is as follows：

$p\left(R\right)=\left\{\begin{array}{cc}{p}_{\min },& I_x\&GreaterEqual;\stackrel{\&OverBar;}{I}\\ p_{\max }-\frac{(p_{max}-p_{\min })(f_R-f_{min})}{f_{max}-f_{min}},& I_x<\stackrel{\&OverBar;}{I}\end{array}\right.$

p_min、p_maxIt is customized parameter

f_max——In all paths f_RnMaximum

f_min——Middle f_RnMinimum value

i_R--- the degree of association I of path R_R Descending ranking in all paths

An intersection x is randomly choosed in selected R_iIf i ≠ 0 ∧ i ≠ | R | chooses x_i-1Another adjacent intersection Mouth x '_i：

x_i=rand (R)

x′_i=rand (x | x ∈ X ∧ join (x, x_i-1)})

Otherwise choose other edge intersection x '_i；

If x '_iWith follow-up intersection x_i+iIt is joined directly together, then by the x in R_iReplace with x '_i；Otherwise search x_i+iMost short path R '={ x '_i,x′_i+1,…,x′_i+i′, replace the x in R with r '_i；The step simulates biological gene mutation process.

(44) recalculateThe degree of association I in all paths, screens to it, only leaves and parental generation population quantity identical one Part path, is constitutedR removed probability in path is：

(45) if the degree of association in the path of population is not up to the condition of convergence, also not up to predetermined maximum iteration, makesAnd return to step (42)；Otherwise stop iteration and by current degree of association highest path R_maxAs critical path.