CN105678421A - Genetic algorithm-based guide sign guidance reachability optimization method - Google Patents

Abstract

The invention aims to solve the problem in the prior art that the guide is non-reachable based on the arrangement of existing guide signs and provides a genetic algorithm-based guide sign guide reachability optimization method. The method comprises the steps of designing a fitness function in consideration of the guide length and the number of additionally arranged guide signs; generating corresponding initial populations for multiple non-reachable guide paths; coding the chromosomes of the above populations in the path coding mode; calculating the fitness values of all chromosomes in the above populations and an optimal fitness value for the combination of chromosomes among the populations, and conducting the cross-correction genetic manipulation on a guide path of better fitness in each population; and conducting the variant genetic manipulation on guide paths of poor fitness based on probability selection. According to the technical scheme of the invention, the above process is repeated and the genetic operation is stopped only when the specified evolving algebra is completed. Through additionally arranging the guide information, the road-guiding sign arrangement scheme is optimized. Therefore, a global optimization method for multiple paths to reach the same destination is realized.

Description

A kind of fingerpost based on genetic algorithm guides accessibility optimization method

Technical field

The present invention relates to traffic planninng and design field, guide accessibility optimization method more particularly, to a kind of fingerpost based on genetic algorithm.

Background technology

Fingerpost is a kind of traffic guidance handling facility, the relevant information (direction, place, distance etc.) of road is transmitted principally for traffic participant, it is also the core of static inducible system simultaneously, smooth and easy to realizing traffic organization, and balanced road network traffic flow plays an important role. The present situation of urban road fingerpost still allows of no optimist, Guide Sign System lacks globality, and there is information instruction deficiency or disappearance, cause that road traveler can not successfully arrive at problems such as (namely guiding unreachable), it is impossible to give full play to the function of fingerpost. At present, most domestic city has had the road Guide Sign System of shaping, and fingerpost is laid research and also increasingly paid close attention to by research institution and government department and pay attention to. Particularly arrive at, significant for lifting road overall service levels, the trip delay promoting the efficiency of operation of traffic system, minimizing road user etc.

Nowadays, research about fingerpost is concentrated mainly on design, arranges, four aspects such as analysis and optimization, wherein the design of fingerpost is mainly manifested in the size of sign board, font, format etc., and arranges and be mainly manifested in the aspects such as the position of setting, quantity of information. Overall laying and information to fingerpost are chosen aspect and are concentrated mainly on some aspects such as message level division, level guide. Assay is concentrated mainly on the evaluation etc. to overall fingerpost inducible system and guide property continuously. And for the optimization of fingerpost, major embodiment is for the optimization to configuration information of the rules such as GB, but less accessibility is guided to be optimized research fingerpost inducible system. Fingerpost inducible system guides accessibility to refer to, and user passes through the guide of fingerpost inducible system and can arrive at smoothly. And, original research is concentrated mainly on single source guided path, and rarely has the optimization to multi-source guided path. Single source refers to only one of which starting point, and multi-source refers to the situation that there is multiple starting point.

Summary of the invention

The present invention is directed to fingerpost to guide and unreachable be optimized research, under the premise considering original fingerpost layout scheme, for guiding inaccessible layout scheme to find the multi-source optimum guided path of new bidding will, it is provided that a kind of multi-source fingerpost based on genetic algorithm guides accessibility optimization method.

For solving above-mentioned technical problem, technical scheme is as follows:

A kind of multi-source fingerpost based on genetic algorithm guides accessibility optimization method, comprises the following steps:

S1. design considers the length of guided path and needs the fitness function of the fingerpost director information number set up;

S2. to guiding inaccessible multi-source coordinates measurement initial population;

S3. population chromosome is encoded by the mode of path code;

S4. calculating in each population chromosomal adaptive optimal control angle value between chromosomal fitness value and population, selective staining body carries out intersecting-correct operation;

S5. the genetic manipulation made a variation is carried out by probability selection chromosome;

S6. repeat to select-intersect-genetic manipulation of variation until stopping evolving after arriving the evolutionary generation specified, obtain population and can optimize inaccessible infeasible paths, it is achieved multi-source infeasible paths is to the global optimization of destination.

Preferably, in described step S1, the fitness function of the fingerpost director information number that the length of design consideration guided path and needs are set up, specifically: by the director information number (N) of the fingerpost that the length (L) of optimizing index consideration guided path is set up with needs; Wherein the length of guided path formula (1) represents:

$F_1=\underset{i=1}{\overset{n}{\Σ}}f\left(d_i\right)---\left(1\right)$

Wherein F₁Represent the length function of guided path; F (d_i) representing the i-th paths length, n represents the quantity in path.

Owing to consider that existing fingerpost lays situation, therefore when calculating needs to set up fingerpost director information, existing fingerpost information in guided path need to be considered; Represent with formula (2):

F₂=∑ P (2)

Wherein F₂What represent director information sets up quantity function; ∑ P represents the sum referring to set up guide label entry;

There is the difference of dimension in two big index L and N, index need to carry out dimension unified, all convert two big indexs to spent Financial cost;

The totle drilling cost of loss is needed to be about C if travelling 1km₁Ten thousand, laying a sign board needs to spend about C₂Ten thousand; Adopting Y-factor method Y to define fitness function, its function expression is with shown in formula (3):

$F=\frac{1}{F_1\×C_1+F_2\×C_2}---\left(3\right).$

Based on multi-source infeasible paths, when all of infeasible paths globality is optimized, a plurality of optimization guided path can be produced, thus there will be the situation that guided path repeats and fingerpost is public; For this calculate total when setting up fingerpost Information Number, it is necessary to remove the part repeated, with reference to formula (3) combinations of definitions fitness function F_com, its function expression formula (4) represents:

$F_{com}=\frac{1}{F_1\×C_1+(F_2-F_2^{*})\×C_2}---\left(4\right)$

Wherein:Represent public fingerpost number.

Preferably, in described step S2, to guiding inaccessible multi-source coordinates measurement initial population; Concrete mode is: obtain the fingerpost information on all indicative purposes ground, guide and default rule searching route according to label entry, when destination unreachable, then being called infeasible paths from pathfinding beginning to the path of fingerpost loss of learning, distal point is the crossing node of fingerpost loss of learning; In road network, a plurality of infeasible paths be would be likely to occur for same destination.

Preferably, in described step S3, encoding population chromosome by the mode of path code, concrete mode is: the population of genetic algorithm is the chromosome congression of each multi-source infeasible paths; The chromosome of genetic algorithm is the ordered queue that multi-source infeasible paths forms from the node of origin-to-destination; Wherein starting point refers to the distal point of road network entrance or infeasible paths, and terminal refers to that destination, node refer to the gene in chromosome;The starting point of chromosomal first gene some infeasible paths i.e.;

For chromosome, it is stipulated that the gene in every chromosome is not allow for the gene of repeated encoding, chromosomal adjustable length, but its length needs to meet the node sum less than road network.

Preferably, in described step S4, calculating in each population chromosomal adaptive optimal control angle value between chromosomal fitness value and population, selective staining body carries out intersecting-correct operation, specifically includes:

Operation is selected to determine that the individuality of intersection; First calculate all chromosomal fitness function value F, be ranked up according to the height of adaptedness; Secondly by selective staining body, carrying out next step intersection operation, wherein ideal adaptation degree its selected probability more big is also more high; The chromosomal quantity produced after selecting and the chromosome that the size of population is constant and in population, nothing repeats;

The operation that intersects is two individual part-structure phase interchangeable selection obtained, and generates new individual; Specifically include: select all chromosomal common node in population, intermediate node identical with head and the tail node different chromosome be homologous chromosome; In homologous chromosome, randomly choose a common node as cross point, exchange the node after cross point;

Correct operation is to produce loop in the new individuality in order to avoid producing after intersecting; If producing loop, its chromosome is corrected operation, cancellation loop.

Preferably, in step S5, carried out the genetic manipulation made a variation by the guided path that probability selection fitness is poor. Specifically include: in a chromosome chosen, randomly choose a gene as mutant gene (except a node and destination node), from node (distal point of entrance or infeasible paths) constant to the path of mutant gene, gene after mutant gene then starts in the way of being similar to chromosome coding from mutant gene, and random didactic selection is until terminal (destination).

Compared with prior art, technical solution of the present invention provides the benefit that:

(1) present invention is based on genetic algorithm, and algorithm is practical, it is possible to meet engineering demand.

(2) the present invention is directed to the multi-source optimization problem of node, optimizing index had both considered guided path length, have also contemplated that the impact of original fingerpost layout scheme, it is possible to the scheme of systematic acquisition global optimization.

Accompanying drawing explanation

Fig. 1 is the flow chart that a kind of fingerpost based on genetic algorithm of the present invention guides the specific embodiment up to optimization method.

Fig. 2 is the schematic diagram of chromosome coding in the specific embodiment of the invention.

Fig. 3 a is the schematic diagram that in the specific embodiment of the invention, chromosomal chiasma operates front path.

Fig. 3 b is the schematic diagram of chromosomal chiasma operation rear path in the specific embodiment of the invention.

Fig. 4 a is the schematic diagram in the front path of chromosome correct operation in the specific embodiment of the invention.

Fig. 4 b is the schematic diagram of chromosome correct operation rear path in the specific embodiment of the invention.

Fig. 5 a is the schematic diagram in the front path of chromosome disorder in the specific embodiment of the invention.

Fig. 5 b is the schematic diagram of chromosome disorder rear path in the specific embodiment of the invention.

Fig. 6 is that the present invention is embodied as case schematic diagram.

Detailed description of the invention

Accompanying drawing being merely cited for property explanation, it is impossible to be interpreted as the restriction to this patent;

In order to the present embodiment is better described, some parts of accompanying drawing have omission, zoom in or out, and do not represent the size of actual product;

To those skilled in the art, in accompanying drawing, some known features and explanation thereof are likely to omission and will be understood by.

Below in conjunction with drawings and Examples, technical scheme is described further.

Embodiment 1

As it is shown in figure 1, be the flow chart of a kind of fingerpost guide accessibility optimization method specific embodiment based on genetic algorithm of the present invention. Referring to Fig. 1, a kind of fingerpost based on genetic algorithm of this specific embodiment guides the concrete steps of accessibility optimization method to include:

Step one: consider guide length F₁With the fingerpost quantity F set up₂, fitness function F closes in design team_com. Optimizing index considers the length (L) of guided path and needs the director information number (N) of the fingerpost set up. Wherein the length of guided path can represent by equation below, wherein F₁Represent the length function of guided path; F (d_i) represent the i-th paths length.

$F_1=\underset{i=1}{\overset{n}{\Σ}}f\left(d_i\right)---\left(1\right)$

Because considering that existing fingerpost lays situation, therefore when calculating needs to set up fingerpost director information, original fingerpost information in guided path need to be considered. Can represent by equation below. Wherein F₂What represent director information sets up quantity function; ∑ P represents the sum referring to set up guide label entry.

F₂=∑ P (2)

There is the difference of dimension in two big index L and N, index need to carry out dimension unified, all convert two big indexs to spent Financial cost. According to related data, consider environment, oil take, under the expense such as fee of material, construction cost, it is proposed to travelling 1km needs the totle drilling cost of loss to be about C1 ten thousand, lays sign board needs and spends about C2 ten thousand. Adopting Y-factor method Y to define fitness function, its function expression can be used shown in equation below:

$F=\frac{1}{F_1\×C_1+F_2\×C_2}---\left(3\right)$

Have multiple for infeasible paths distal point, when all of infeasible paths globality is optimized, a plurality of optimization guided path can be produced, thus there will be the situation that guided path repeats and fingerpost is public. For this calculate total when setting up fingerpost Information Number, it is necessary to remove the part repeated, with reference to formula (3) combinations of definitions fitness function F_com, its function expression can represent by equation below, wherein:Represent public fingerpost number.

$F_{com}=\frac{1}{F_1\×C_1+(F_2-F_2^{*})\×C_2}---\left(4\right)$

Step 2: infeasible paths is generated set OptV. Obtain the fingerpost information on all indicative purposes ground, according to guide and the default rule searching route of label entry, find and set up all infeasible paths set OptV of road network;

Step 3: produce several populations and be encoded population and chromosome processing. In fingerpost optimization method, encode population chromosome by the mode of path code. Specifically include: the population of genetic algorithm is the chromosome congression of each multi-source infeasible paths; The chromosome of genetic algorithm is the ordered queue that multi-source infeasible paths forms from starting point (distal point of road network entrance or infeasible paths) to the node (gene) of terminal (destination), and item chromosome represents a feasible solution. Chromosomal first gene and starting point, second gene is to randomly choose or heuristically select from the node of connection adjacent with first gene, so on up to destination's node. Instantiation is as shown in Figure 2. Assume that multi-source starting point is S₁And S₂, chromosome coding comprises from starting point S₁And S₂Ordered queue to destination. The queue "/" between homology starting point does not separate.

Chromosome:S₁-4-9-10-11-16-D/S₂-14-15-16-D

Step 4: judge whether population iterations i reaches the evolutionary generation Gen specified, if i<Gen, then proceeds genetic manipulation, if i>=Gen terminates genetic manipulation and obtains optimal solution;

Step 5: calculate in each population chromosomal adaptive optimal control angle value between chromosomal fitness value and population, selective staining body carries out intersecting-correct operation. The individuality of intersection is mainly determined in the operation that selects in genetic algorithm. In instantiation, adopting the mode of sequencing selection to carry out genetic algorithm selection, this operation is divided into two and walks greatly: (1) calculates all chromosomal fitness function value F_com, it is ranked up according to the height of adaptedness. (2) individuality that selection fitness is high is as parent, carries out next step intersection operation, and wherein ideal adaptation degree its selected probability more big is also more high. The chromosomal quantity produced after selecting and the chromosome that the size of population is constant and in population, nothing repeats.

The operation that intersects is two individual part-structure phase interchangeable selection obtained, and generates new individual. Specifically include: select all chromosomal common node in population, with the first chromosome segment that node is identical, intermediate node is different for homologous fragment; In homologous chromosome, randomly choose a common node as cross point, exchange the node after cross point. Instantiation as shown in Figure 3 a, by selecting the parent chromosome that obtains:

Chromosome1:S₁-4-9-10-11-16-D/S₂-14-15-16-D

Chromosome2:S₁-4-5-10-15-16-D/S₂-14-15-16-D

Its public crossbar contact is S₁、4、10、16、D、S₂, 14,15,16, D. Choose homologous fragment S₁To D, intersect with common node 10, then the chromosome after intersection is as shown in Figure 3 b:

Chromosome*1:S₁-4-9-10-15-16-D/S₂-14-15-16-D

Chromosome*2:S₁-4-5-10-11-16-D/S₂-14-15-16-D

Correct operation is to produce loop in the new individuality produced after avoiding intersection, and namely same node can only select once, if producing loop, its chromosome is corrected operation. Instantiation is as shown in fig. 4 a, it is assumed that chromosome is:

Chromosome*1:S₁-4-5-10-5-6-11-16-D/S₂-14-15-16-D

Chromosome*2:S₁-4-9-14-15-10-11-16-D/S₂-14-15-16-D

Clearly there are two the same nodes in child chromosome chromosome*1, therefore needs to correct, cancellation loop. Result is as shown in Figure 4 b:

Chromosome*1:S1-4-5-6-11-16-D/S2-14-15-16-D

Chromosome*2:S1-4-9-14-15-10-11-16-D/S2-14-15-16-D

Step 6: the guided path poor by probability selection fitness carries out mutation genetic operation; Make a variation unrelated with Population Size, change some gene on chromosome with only small random chance, give good gene for change. Instantiation is as shown in Figure 5 a. Variation prochromosome is S₁-4-5-10-9-14-15-16-D/S₂-14-15-16-D, mutation operation method is to randomly select a gene (node) in item chromosome as mutant gene (node), from node (entrance) constant to the path of mutant gene, gene after mutant gene then starts to randomly choose in didactic mode from mutant gene until terminal (destination). Assuming that chromosome obtains mutant gene is 10, then after variation new chromosome as shown in Figure 5 b, for S₁-4-5-10-11-16-D/S₂-14-15-16-D。

Step 7: obtain after optimal path through genetic manipulation, does not have on optimal path on the crossing that destination is guided, sets up director information.

Last, shown in Fig. 6, to be embodied as Case retrieval Guangzhou College City. Choosing Zhongshan University is destination's node, and in figure, 1,2,3 is the starting point guiding infeasible paths. After utilizing the fingerpost based on genetic algorithm to guide the optimization of accessibility optimization method, can showing that position shown in the figure needs to set up director information, need altogether to set up 4 director informations, guided path overall length is 10835.6m.

The present invention be with infeasible paths inaccessible in road network is optimized for up to guided path for target, by designing consideration guide length and setting up the fitness function of fingerpost quantity, the genetic manipulation utilize genetic algorithm to select, to intersect-correct, making a variation searches optimal solution.According to result of calculation, obtain the optimum fingerpost layout scheme of road network. The fingerpost layout scheme optimized by the present invention is a kind of scientific method, and engineer applied is had directive significance.

The corresponding same or analogous parts of same or analogous label;

Position relationship described in accompanying drawing be used for the explanation of being merely cited for property, it is impossible to be interpreted as the restriction to this patent;

Obviously, the above embodiment of the present invention is only for clearly demonstrating example of the present invention, and is not the restriction to embodiments of the present invention. For those of ordinary skill in the field, can also make other changes in different forms on the basis of the above description. Here without also cannot all of embodiment be given exhaustive. All any amendment, equivalent replacement and improvement etc. made within the spirit and principles in the present invention, should be included within the protection domain of the claims in the present invention.

Claims (7)

1. the multi-source fingerpost based on genetic algorithm guides accessibility optimization method, it is characterised in that comprise the following steps:

S1. design considers the length of guided path and needs the fitness function of the fingerpost director information number set up;

S2. to guiding inaccessible multi-source coordinates measurement initial population;

S3. population chromosome is encoded by the mode of path code;

S4. calculating in each population chromosomal adaptive optimal control angle value between chromosomal fitness value and population, selective staining body carries out intersecting-correct operation;

S5. the genetic manipulation made a variation is carried out by probability selection chromosome;

S6. repeat to select-intersect-genetic manipulation of variation until stopping evolving after arriving the evolutionary generation specified, obtain population and can optimize inaccessible infeasible paths, it is achieved multi-source infeasible paths is to the global optimization of destination.

2. method according to claim 1, it is characterized in that, in described step S1, the fitness function of the fingerpost director information number that the length of design consideration guided path and needs are set up, specifically: by the director information number (N) of the fingerpost that the length (L) of optimizing index consideration guided path is set up with needs; Wherein the length of guided path formula (1) represents:

Wherein F₁Represent the length function of guided path; F (d_i) representing the i-th paths length, n represents the quantity in path.

Owing to consider that existing fingerpost lays situation, therefore when calculating needs to set up fingerpost director information, existing fingerpost information in guided path need to be considered; Represent with formula (2):

F₂=∑ P (2)

Wherein F₂What represent director information sets up quantity function; ∑ P represents the sum referring to set up guide label entry;

There is the difference of dimension in two big index L and N, index need to carry out dimension unified, all convert two big indexs to spent Financial cost;

The totle drilling cost of loss is needed to be about C if travelling 1km₁Ten thousand, laying a sign board needs to spend about C₂Ten thousand; Adopting Y-factor method Y to define fitness function, its function expression is with shown in formula (3):

3. method according to claim 2, it is characterised in that based on multi-source infeasible paths, when all of infeasible paths globality is optimized, can produce a plurality of optimization guided path, thus there will be the situation that guided path repeats and fingerpost is public; For this calculate total when setting up fingerpost Information Number, it is necessary to remove the part repeated, with reference to formula (3) combinations of definitions fitness function F_com, its function expression formula (4) represents:

Wherein:Represent public fingerpost number.

4. method according to claim 1, it is characterised in that in described step S2, to guiding inaccessible multi-source coordinates measurement initial population; Concrete mode is: obtain the fingerpost information on all indicative purposes ground, guide and default rule searching route according to label entry, when destination unreachable, then being called infeasible paths from pathfinding beginning to the path of fingerpost loss of learning, distal point is the crossing node of fingerpost loss of learning; In road network, a plurality of infeasible paths be would be likely to occur for same destination.

5. method according to claim 1, it is characterised in that in described step S3, encodes population chromosome by the mode of path code, and concrete mode is: the population of genetic algorithm is the chromosome congression of each multi-source infeasible paths; The chromosome of genetic algorithm is the ordered queue that multi-source infeasible paths forms from the node of origin-to-destination; Wherein starting point refers to the distal point of road network entrance or infeasible paths, and terminal refers to that destination, node refer to the gene in chromosome; The starting point of chromosomal first gene some infeasible paths i.e.;

For chromosome, it is stipulated that the gene in every chromosome is not allow for the gene of repeated encoding, chromosomal adjustable length, but its length needs to meet the node sum less than road network.

6. method according to claim 1, it is characterised in that in described step S4, calculates in each population chromosomal adaptive optimal control angle value between chromosomal fitness value and population, and selective staining body carries out intersecting-correct operation, specifically includes:

Operation is selected to determine that the individuality of intersection; First calculate all chromosomal fitness function value F, be ranked up according to the height of adaptedness; Secondly by selective staining body, carrying out next step intersection operation, wherein ideal adaptation degree its selected probability more big is also more high; The chromosomal quantity produced after selecting and the chromosome that the size of population is constant and in population, nothing repeats;

The operation that intersects is two individual part-structure phase interchangeable selection obtained, and generates new individual; Specifically include: select all chromosomal common node in population, intermediate node identical with head and the tail node different chromosome be homologous chromosome; In homologous chromosome, randomly choose a common node as cross point, exchange the node after cross point;

Correct operation is to produce loop in the new individuality in order to avoid producing after intersecting; If producing loop, its chromosome is corrected operation, cancellation loop.

7. method according to claim 1, it is characterised in that in described step S5, carries out the genetic manipulation made a variation by the guided path that probability selection fitness is poor; Specifically include: in a chromosome chosen, randomly choose a gene as mutant gene, this gene has not been node and destination node, from node constant to the path of mutant gene, gene after mutant gene then starts in the way of being similar to chromosome coding from mutant gene, and random didactic selection is until terminal.