CN101853294A - Multi-mode multi-standard path search method based on genetic algorithm - Google Patents
Multi-mode multi-standard path search method based on genetic algorithm Download PDFInfo
- Publication number
- CN101853294A CN101853294A CN 201010185830 CN201010185830A CN101853294A CN 101853294 A CN101853294 A CN 101853294A CN 201010185830 CN201010185830 CN 201010185830 CN 201010185830 A CN201010185830 A CN 201010185830A CN 101853294 A CN101853294 A CN 101853294A
- Authority
- CN
- China
- Prior art keywords
- pattern
- operator
- population
- mode
- individuality
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a multi-mode multi-standard path search method based on genetic algorithm, which is realized by computer. The method solves the multi-mode multi-standard traffic path search problem by defining genetic operator, matching corresponding strategy and utilizing the optimal theory. Under various trip standards, the method is used to solve, which does not have to set the weight value of each standard in advance, avoids no-solution risk brought by constraint method and reduces the influence of personal subjective factors. The path output result of the invention can satisfy various trip requirements and has the characteristic of various trip mode combinations, provides personalized and various trip projects for traveler and provides good technical support for the public trip information service.
Description
Technical field
The present invention relates to a kind of traffic route searching method, particularly a kind of computer implemented multi-mode based on genetic algorithm, many standards traffic route searching method are applicable to public trip service application.
Background technology
The public trip service system in city contains multiple traffic trip patterns such as motorbus, taxi, subway, light rail, bicycle, walking.The go on a journey selection of travel pattern of the public has greatly been enriched in the setting-up and development of multimode traffic system, has also excited personalized trip requirements simultaneously.Single evaluation standards such as distance is the shortest, the time is the shortest, transfer is minimum, expense is minimum can not satisfy diversified comprehensive travel demand.The research focus of the public trip service in service becoming city, many standard trips path of satisfying the multiple integration requirement of traveler is provided.
The route searching problem of considering multiple standards adopts two kinds of methods usually.A kind of is weighted method, and promptly transforming multiple standards is scalar, adopts methods such as linear weighted function summation, for various standards are determined a weighted value, obtains a scalar value after comprehensive, utilizes existing single canonical algorithm to calculate again; Another kind is a leash law, and k-1 the standard handovers that is about in k the standard is constraint condition, and surplus next standard is as the objective function of single criteria optimization problem.The former lacks scientific foundation on the determining of various standard weighted values, and can not obtain all optimum solutions on the uniform curved of non-convexity; The latter's result mainly depends on the single standard that is confirmed as objective function, relies on the personal experience equally, and may cause this list typical problem not had separating.
At present, along with the continuous development of intelligence computation method and perfect, utilize evolution algorithm especially genetic algorithm for solving multi-criteria optimization problem obtained some achievements in research.These achievements are for the invention provides theoretical the support.Yet the difficult point of using genetic algorithm for solving multi-mode multi-standard transfer path planning problem is how path code is become chromosome.On going result is carried out under the monotype environment mostly, and is only applicable to small scale network.And the urban network of real world mostly is on a large scale, the multi-mode complex environment, and existing achievement in research can't address this problem.In addition, genetic operation operator will be applied to the multimode network environment, and current achievement lacks between corresponding pattern and the interior genetic operation operator of pattern, therefore can not directly apply to the multi-criteria optimization route searching problem that solves under the multimode traffic system.
Summary of the invention
The invention provides a kind of computer implemented based on genetic algorithm solution multi-mode multi-standard traffic route searching method (traffic route, hereinafter referred is " path "), this method is expanded to adapt to the multimode traffic environment genetic algorithm, adopt genetic algorithm and optimum theory to handle the multi-criteria optimization route searching, can't provide the problem of multi-mode multi-standard path service to solve existing public trip service technology, for the service of public's trip information provides better technical support.
Technical solution of the present invention: a kind of based on genetic algorithm solution multi-mode multi-standard method for searching path, its feature comprises:
(1) input multimode traffic network model data
(2) multi-mode path code
(3) intersects between intersection in the pattern, mutation operator and pattern, mutation operator definition and operating
(4) carry out the many standard genetic algorithm that are used for the multi-mode route searching
(5) recommendation paths result set output
Described multimode traffic network model is according to the multiple travel pattern data in city, utilizes Geographic Information System spatial analysis technology, the network-in-dialing relation of city multimode traffic is carried out robotization handle, and sets up multi-mode integrated traffic network model.
Described multi-mode path code, promptly the individuality in the genetic algorithm (Individual) coding is expressed the path exactly with certain coding method.Wherein, the multi-mode path requires individual coding can embody multiple travel pattern combination, does not produce too much data redundancy simultaneously again; Genetic algorithm requires individual coding to be easy to the genetic operator operation.
Therefore, definition multi-mode individuality is encoded to: adopt the random length phenotype coding of band model sign, form by mode flag district and ID code area, shape as:
Indivicual＝{T
1，I
1，...，I
j，T
2，I
1，...，I
k，T
3，...，T
m，...}
In the following formula, T
iBe pattern identification, I
iBe the various ID values that go out under the row mode (as: motorbus, taxi, subway, light rail, bicycle, walking etc.).
Intersects between intersection in the described pattern, mutation operator and pattern, mutation operator definition and operating, be the evolutionary operator that is different from the redetermination of traditional genetic algorithm evolutionary operator.Since traditional genetic operator at individuality coding be (similar) of homogeneity, the gene that is to say each position is of equal importance.But in the multi-mode path, be non-homogeneous between the gene section between different mode.Therefore, require to redefine the genetic operator that satisfies multi-mode (polymorphic type) feature.Genetic operator generates new individuality by gene is intersected, makes a variation, and promptly obtains new route.
Therefore, definition crossover and mutation operator are respectively in the pattern and intersect and mutation operator, and promptly (as: in the public transport pattern that coexists) carried out genetic operator and operated in model identical.Definition hypercrossover is respectively with the hypermutation operator and intersects between pattern and mutation operator, and promptly (as: public transport pattern--＞walking pattern) carried out genetic operator and operated between different mode.
Intersection with the characteristics of mutation operator is in the above-mentioned pattern, does not introduce new travel pattern, i.e. the increase of maximization control number of transfer.
Intersection is that the operator operation can cause the increase on the travel pattern number of transfer with the characteristics of mutation operator between above-mentioned pattern.
The described many standard genetic algorithm that are used for the multi-mode route searching comprise:
(a) algorithm context initialization is changeed (b);
(b) initialization of population is changeed (c);
(c) initial population is carried out crossover operator computing between pattern, replenish and not exclusively separate, change (d) to separating fully;
(d) judge whether to reach maximum evolutionary generation, if "Yes" is changeed (i), if "No" is changeed (d);
(e) calculate evaluation of estimate, change (f);
(f) current population is carried out in the pattern between crossover operator, pattern mutation operator computing between mutation operator and pattern in crossover operator, the pattern, generate new individual record and go into population of future generation, change (g);
(g) current population is selected the operator computing, generate new individual record and go into population of future generation, change (h);
(h) new population replaces current population, and evolutionary generation adds 1, changes (d);
(i) calculate evaluation of estimate, finish.
(a) context initialization of algorithm described in comprises:
(i) set the trip requirements parameter, described parameter comprises starting point, terminal point, goes out row mode, goes out column criterion and travel time etc.
Wherein, going out row mode relates to but is not limited to motorbus, taxi, subway, light rail, bicycle, walking etc.; Go out that column criterion relates to but the shortest, total distance of the time that is not limited to is the shortest, walking distance is the shortest, landscape is best, turning is minimum, number of transfer is minimum, expense is minimum etc.; Travel time is the plan departure time, and the dynamic environment Time Calculation is provided with in order to cooperate accurately.
Going out column criterion and travel time understands influential to the result in path.Go out column criterion and determined to adopt what trip mode combinations; In time dependent dynamic environment, travel time difference, the traffic capacity in various patterns, each highway section are also different, can influence the result of final path, may be embodied on the different paths, also may be embodied in the different mode combination.
(ii) set genetic algorithm parameter, described parameter comprises make a variation between crossover probability, pattern between variation probability, pattern in crossover probability, the pattern in initial population number, the pattern probability, evolutionary generation, selection operator selection number etc.
Genetic algorithm parameter is very big to the influence of route result.The population size influences the space consumption and the time consumption of algorithm; Evolutionary generation is given algorithm end mark, crosses senior general and produces more repetition individuality, and the diversity that forfeiture is separated, time consumption also will increase simultaneously, and too small, it is insufficient to evolve, the result who can not get wanting; Intersection, mutation operator all are in order to produce new individuality, directly to influence individual route result; The excellent individual of selecting operator to keep some directly enters the next generation, crosses conference and causes being absorbed in too early local solution, and is too small, can not embody the operator meaning.
(b) population of initialization described in is characterized in:
Under each monotype environment, generate the initial individuality of some respectively, remerge the composition initial population, be characterized in that individuality is monotype and allows not exclusively to separate, promptly under single travel pattern to separating that given, stop (O-D) carry out that route searching obtains, this is separated may be the separating of sub-range O '-D ' of O-D, i.e. an O-D part of separating.
(c) described between pattern the crossover operator calculation step comprise:
Not exclusively separate individuality for one, select another to separate individuality fully at random, carry out crossover operator computing between pattern, separate the individual former individuality of not exclusively separating that replaces fully with newly-generated.
(e), calculate evaluation of estimate described in (i), that is, calculate each individual evaluation of estimate according to evaluation function.Be characterized in adopting the sort method of multiple goal (many standards) Optimum Theory to sort.
If evaluation vector is f=(C
1, C
2..., C
n), C wherein
iExpression is individual at the substandard calculated value (this value is calculated according to the respective standard computing formula) that expends of i kind, and then evaluation function is F=ParetoRank (f).That is, from small to large all individualities are sorted by quality according to Optimum Theory.
(f) the crossover operator computing comprises in the pattern described in:
To given two individualities, the point of crossing occurs in the same pattern, promptly seeks the model identical subbase and intersects because of the homologous genes in the section, and two of the back generations that intersect are new individual.Can adopt various cross methods to carry out.
The transfer slit is repaired and is adopted the walking pattern to repair usually.
(f) described between pattern the crossover operator computing comprise:
To given two individualities, the point of crossing occurs in the different mode: determine individual 1 and individual 2 cross-mode M earlier, N, in corresponding modes, seek and to intersect gene, the gene that can intersect is judged according to transformational relation between the multi-mode of setting up in the multimode network model, intersect according to the intersected gene pairs of choosing, answer the polishing pattern identification during intersection.After the intersection,, need change to the slit and survey and repair, adopt the walking pattern usually because of pattern changes (transfer increases).Can adopt various cross methods to carry out.
(f) the mutation operator computing comprises in the pattern described in:
For given individuality, variation occurs in the same pattern, and promptly definitive variation gene section in variation mode adopts path search algorithm to generate new gene section and replaces the protogene section.Can adopt various variation methods to carry out.
(f) described between pattern the mutation operator computing comprise:
Adopt the directed variation strategy, quicken the purpose that excellent individual produces to reach.For given individuality, variation occurs in the different mode: first definitive variation pattern M, according to variation mode M definitive variation target pattern N, but first detection can make a variation to the orientation mutant gene of N in variation mode M, but the orientation mutant gene is judged according to transformational relation between the multi-mode of setting up in the multimode network model, replace the protogene section but generate new gene section at pattern N, polishing pattern identification when generating new gene section according to the gene section between the orientation mutant gene of choosing.Can adopt various variation methods to carry out.
(g) select the operator computing to comprise described in:
Determine the individual amount X that the selection operator is selected according to current number of individuals G ' of new population and definition population size G;
X=G-G ', wherein G '=G '=G
Crossover+ G
Mutation+ G
Hypercrossover+ G
Hypermutation
The statistical appraisal value is 1 the individual number M of non-repetition:
If X＜M then, selects the non-repetition individuality of current evaluation of estimate to enter population of future generation at random;
If X＞M then, determines that the non-repetition individuality of current evaluation of estimate enters population of future generation, and adds up the individual number N of non-repetition of next evaluation of estimate, in this evaluation of estimate, select X-M Different Individual to enter population of future generation.Select the non-repetition individuality of current evaluation of estimate to enter population of future generation at random.
Described recommendation paths result set output is characterized in:
Export non-repetitive individuality in last generation evolution colony, and based on a certain standard, be arranged in order output, this results set is recommends the optimal path result.Each individuality in the output set is a kind of recommendation paths scheme, and this route scheme may comprise multiple travel pattern (instrument), and satisfies the optimization route result of multiple evaluation criterion (many standards).Output form is including, but not limited to text description, figure path, voice broadcast etc.
The present invention's advantage compared with prior art is: the present invention can really adapt to the trip service under the multimode traffic system, go out under the column criterion multiple, do not need to preestablish the weighted value of each standard, but employing genetic algorithm for solving, reduce the influence of artificial subjective factor, its path output result can either satisfy various trip requirements, have diversified trip mode combinations characteristics again, for traveler provides personalized and selects various trip scheme.This invention can be applied to have real operability and practicality in the complicated transportation network environment of extensive multi-mode.
Description of drawings
Fig. 1 is the process flow diagram of the inventive method
Fig. 2 is the individual coding of genetic algorithm among the present invention synoptic diagram
Fig. 3 is genetic algorithm calculation flow chart among the present invention
Embodiment
In order to make those skilled in the art person understand the scheme of the embodiment of the invention better, the embodiment of the invention is described in further detail below in conjunction with drawings and embodiments.
As shown in Figure 1, specific implementation step of the present invention is as follows:
(1) according to the multiple travel pattern data in city, utilizes Geographic Information System spatial analysis technology, the network-in-dialing relation of city multimode traffic is carried out robotization handle, set up multi-mode integrated traffic network model.
Specifically, on urban road network, public transport, track traffic and network data base basis, pavement, utilize Geographic Information System spatial analysis technology, set up the integrated multimode network of logic.This multimode network is that media is set up the transfer relation between various travel patterns with point-like key element such as bus stop, subway station, point of interest etc.
(2) definition is fit to the path code that comprises the multi-mode feature of genetic algorithm operation, and promptly the individuality in the genetic algorithm (Individual) coding is expressed the path exactly with certain coding method.Wherein, the multi-mode path requires individual coding can embody multiple travel pattern combination, does not produce too much data redundancy simultaneously again; Genetic algorithm requires individual coding to be easy to the genetic operator operation.
Therefore, definition multi-mode individuality is encoded to: adopt the random length phenotype coding of band model sign, form (Fig. 2) by mode flag district and ID code area, shape as:
Indivicual＝{T
1，I
1，...，I
j，T
2，I
1，...，I
k，T
3，...，T
m，...}
In the following formula, T
iBe pattern identification, I
iBe the various ID values that go out under the row mode (as: motorbus, taxi, subway, light rail, bicycle, walking etc.).Because of id field value type in the multimode data is non-negative integer, so available negative shaping markers.This example goes out row mode with four kinds: motorbus, subway, taxi and walking pattern are that example describes, wherein preceding two kinds of travel patterns adopt the some coded system for the public transport pattern, be bus station and subway station ID value, the lines coding is adopted for two kinds in the back, promptly record the line ID value (table 1) of process.
The individual coding of table 1
(3) evolutionary operator and the operation thereof of multi-mode path code are satisfied in definition, mainly comprise between intersection in the pattern, mutation operator and pattern intersect, mutation operator and operation thereof.
Above-mentioned operator is the evolutionary operator that is different from the redetermination of traditional genetic algorithm evolutionary operator.Since traditional genetic operator at individuality coding be (similar) of homogeneity, the gene that is to say each position is of equal importance.But in the multi-mode path, be non-homogeneous between the gene section between different mode.Therefore, require to redefine the genetic operator that satisfies multi-mode (polymorphic type) feature.Genetic operator generates new individuality by gene is intersected, makes a variation, and promptly obtains new route.
Therefore, definition crossover and mutation operator are respectively in the pattern and intersect and mutation operator, and promptly (as: in the public transport pattern that coexists) carried out genetic operator and operated in model identical.Definition hypercrossover is respectively with the hypermutation operator and intersects between pattern and mutation operator, and promptly (as: public transport pattern--＞walking pattern) carried out genetic operator and operated between different mode.
Intersection with the characteristics of mutation operator is in the above-mentioned pattern, does not introduce new travel pattern, i.e. the increase of maximization control number of transfer.
Intersection with the characteristics of mutation operator is between above-mentioned pattern, and the operator operation can cause the increase on the travel pattern number of transfer, therefore will take certain measure to control the optimization of other standards in concrete operations.
(4) carry out the many standard genetic algorithm that are used for the multi-mode route searching, this algorithm cooperates the operation of multi-mode evolutionary operator in conjunction with genetic algorithm and multiple goal (many standards) Optimum Theory, is used for the route searching of multi-mode environment.
As Fig. 3, may further comprise the steps:
(a) algorithm context initialization is changeed (b);
(b) initialization of population is changeed (c);
(c) initial population is carried out crossover operator computing between pattern, replenish and not exclusively separate, change (d) to separating fully;
(d) judge whether to reach maximum evolutionary generation, if "Yes" is changeed (i), if "No" is changeed (e);
(e) calculate evaluation of estimate, change (f);
(f) current population is carried out in the pattern between crossover operator, pattern mutation operator computing between mutation operator and pattern in crossover operator, the pattern, generate new individual record and go into population of future generation, change (g);
(g) current population is selected the operator computing, generate new individual record and go into population of future generation, change (h);
(h) new population replaces current population, and evolutionary generation adds 1, changes (d);
(i) calculate evaluation of estimate, finish.
(a) context initialization of algorithm described in comprises:
(i) set the trip requirements parameter, described parameter comprises starting point, terminal point, goes out row mode, goes out column criterion and travel time etc.
Going out column criterion and travel time understands influential to the result in path.Go out column criterion and determined to adopt what trip mode combinations; In time dependent dynamic environment, travel time difference, the traffic capacity in various patterns, each highway section are also different, can influence the result of final path, may be embodied on the different paths, also may be embodied in the different mode combination.
Going out row mode according to the multimode network data, is that example describes with motorbus (B), taxi (T), subway (S), walking (W)/bicycle etc. in the embodiment of the invention.Going out that column criterion is the shortest with walking distance, the time is minimum, expense is minimum, transfer is minimum etc. is that example describes.Travel time is used to set the traveler E. T. D., and this is provided with and cooperates Real-time Traffic Information can provide accurate journey time to calculate.
(ii) set genetic algorithm parameter, described parameter comprises make a variation between crossover probability, pattern between variation probability, pattern in crossover probability, the pattern in initial population number, the pattern probability, evolutionary generation, selection operator selection number etc.
Genetic algorithm parameter is very big to the influence of route result.The initial population number should be taken the quantity of travel pattern into account, and the population size influences the space consumption and the time consumption of algorithm; Evolutionary generation is given algorithm end mark, crosses senior general and produces more repetition individuality, and the diversity that forfeiture is separated, time consumption also will increase simultaneously, and too small, it is insufficient to evolve, the result who can not get wanting; Intersection, mutation operator all are in order to produce new individuality, directly to influence individual route result; The excellent individual of selecting operator to keep some directly enters the next generation, crosses conference and causes being absorbed in too early local solution, and is too small, can not embody the operator meaning.
The genetic algorithm parameter setting sees Table 2 in the example of the present invention.
Table 2 genetic algorithm parameter
(b) population of initialization described in is characterized in:
Under each monotype environment, generate the initial individuality of some respectively, remerge the composition initial population, be characterized in that individuality is monotype and allows not exclusively to separate.Individual production can be adopted any paths searching algorithm (as: Dijkstra, A* etc.).
(c) described between pattern the crossover operator calculation step comprise:
Not exclusively separate individuality for one, select another to separate individuality fully at random, carry out crossover operator computing between pattern, separate the individual former individuality of not exclusively separating that replaces fully with newly-generated.If intersect failure, select another to separate individuality fully, separate individuality fully until generating.
(e), calculate evaluation of estimate described in (i), that is, calculate each individual evaluation of estimate according to evaluation function.Be characterized in adopting the sort method of multiple goal (many standards) Optimum Theory to sort.
Evaluation vector is f=(C in this example
Wsp, C
St, C
Ltr, C
Lf), C wherein
Wsp, C
St, C
Ltr, C
LfExpression is individual respectively expends calculated value under the shortest walking distance, shortest time, minimum number of transfer and minimum expense standard.Have:
C
wsp＝C
wsp W
C
st＝C
st W+C
st S+C
st B+C
st T
C
ltr＝C
ltr W+C
ltr S+C
ltr B+C
ltr T
C
lf＝C
lf W+C
lf S+C
lf B+C
lf T
Wherein, subscript is illustrated in the calculated value under the different mode.
Evaluation function is F=ParetoRank (f).Can adopt various sort methods.Specifically the multiple goal sort method that adopts in the example of the present invention, promptly is labeled as order 1 to the Pareto front individuality in the population (non-bad individuality), removes these individualities then from competition; In the individuality of remainder, seek non-bad individuality and it is labeled as order 2, repeat this process up to finishing all orderings.
(f) the crossover operator computing comprises in the pattern described in:
For two individualities selecting at random:
1) surveys the two common mode flag, select a pattern at random;
2) survey all possible intersection gene pairs (genic value is identical), how a pair of as if existing to then selecting at random, then do not return 1 if do not exist) and select another pattern at random, if do not exist, then turn to crossover operator computing between pattern;
3) carrying out single-point intersects;
4) carry out loop checking and survey, and repair with the transfer slit.
The transfer slit is repaired and is adopted the walking pattern to repair usually.
(f) described between pattern the crossover operator computing comprise:
For two individualities selecting at random:
1) survey each individual mode flag respectively, and it is right to be combined into combined crosswise, selects a combination at random;
2) in this mode combinations, survey all possible intersection gene pairs,, then do not return 1 if do not exist if exist how a pair ofly to then selecting at random) select another combination at random;
3) carrying out single-point intersects;
4) carry out loop checking and survey, and repair with the transfer slit.
(f) the mutation operator computing comprises in the pattern described in:
For the individuality of selecting at random:
1) survey all mode flag, it is too short to ignore length, and selects one section at random;
2) survey variable gene,, otherwise return 1 if then select two variable genes (surpassing minimum length) at random more than two) select another section at random, if all get nowhere, turn to mutation operator computing between pattern.
3) under this pattern, produce new gene section sequence and replace the gene section between variable gene in the former sequence
4) carry out loop checking and survey, and repair with the transfer slit.
(f) described between pattern the mutation operator computing comprise:
Adopt the directed variation strategy, quicken the purpose that excellent individual produces to reach.The directed variation parameter sees Table 3, and wherein T, B, S, W represent taxi, motorbus, subway, walking pattern etc. respectively.
Table 3 directed variation parameter
For the individuality of selecting at random:
1) surveys all mode flag, and select a kind of pattern at random;
2) definitive variation target pattern;
3) under target pattern, generate new gene section and replace the protogene section;
4) merge transfer in the inessential continuous mode.
5) carry out loop checking and survey, and repair with the transfer slit.
(g) select the operator computing to comprise described in:
Determine the individual amount X that the selection operator is selected according to current number of individuals G ' of new population and definition population size G;
X=G-G ', wherein G '=G
Crossover+ G
Mutation+ G
Hypercrossover+ G
Hypermutation
Max{G
crossover}＝G×P
crossover
Max{G
mutation}＝G×P
mutation
Max{G
hypercrossover}＝G×P
hypercrossover
Max{G
hypermutation}＝G×P
hypermutation
G
Crossover, G
Mutation, G
HypercrossoverWith G
HypermutationThe new individual number of representing corresponding operator to generate respectively, its concrete value is relevant with operator computing success ratio, and therefore, the value of X may be respectively for different.
The statistical appraisal value is 1 the individual number M of non-repetition:
If X＜M then, selects the non-repetition individuality of current evaluation of estimate to enter population of future generation at random;
If X＞M then, determines that the non-repetition individuality of current evaluation of estimate enters population of future generation, and adds up the individual number N of non-repetition of next evaluation of estimate, in this evaluation of estimate, select X-M Different Individual to enter population of future generation.Select the non-repetition individuality of current evaluation of estimate to enter population of future generation at random.
(5) the non-bad individuality of storage, and output recommendation paths result set.
Reach evaluation of estimate in the population of maximum evolutionary generation after storage algorithm is finished and be 1 individuality, and export non-repetitive individuality in last generation evolution colony, and based on a certain standard, be arranged in order output, this results set is recommends the optimal path result.Each individuality in the output set is a kind of recommendation paths scheme, and this route scheme may comprise multiple travel pattern (instrument), and satisfies the optimization route result of multiple evaluation criterion (many standards).Output form is including, but not limited to text description, figure path, voice broadcast etc.
The present invention has improved traditional public and has gone on a journey in the service platform, and the problem of multi-mode multi-standard trip path service can't be provided.Cooperate corresponding strategy by the definition genetic operator, make genetic algorithm can be applied to the multi-mode route searching.Go out under the column criterion multiple, adopt genetic algorithm for solving, do not need to preestablish the weighted value of each standard, the nothing of also having avoided leash law to bring is separated risk, reduces the influence of artificial subjective factor, its path output result, can either satisfy various trip requirements, have diversified trip mode combinations characteristics again, for traveler provides personalized and selects various trip scheme, for the service of public's trip information provides the good technical support.This invention can be applied to have real operability and practicality in the complicated transportation network environment of extensive multi-mode.
More than the embodiment of the invention is described in detail, used embodiment herein the present invention set forth, the explanation of above embodiment just is used for help understanding device and method of the present invention; Simultaneously, for one of ordinary skill in the art, according to thought of the present invention, the part that all can change in specific embodiments and applications, in sum, this description should not be construed as limitation of the present invention.
Claims (7)
1. computer implemented multi-mode multi-standard traffic route searching method based on genetic algorithm is characterized in that mainly may further comprise the steps:
(1) input multimode traffic network model data;
(2) multimode traffic path code:
Wherein, define described being encoded to: adopt the random length phenotype coding of band model sign, form by mode flag district and ID code area, shape as:
Indivicual＝{T
1，I
1，...，I
j，T
2，I
1，...，I
k，T
3，...，T
m，...}
In the following formula, T
iBe pattern identification, I
iBe the ID value under the various traffic trip patterns (as: motorbus, taxi, subway, light rail, bicycle, walking etc.);
(3) intersects between intersection in the pattern, mutation operator and pattern, mutation operator definition and operating:
Definition crossover and mutation operator are respectively in the pattern and intersect and mutation operator, and promptly (as: in the public transport pattern that coexists) carried out genetic operator and operated in model identical.Definition hypercrossover is respectively with the hypermutation operator and intersects between pattern and mutation operator, and promptly (as: public transport pattern--＞walking pattern) carried out genetic operator and operated between different mode;
(4) carry out the many standard genetic algorithm that are used for the multi-mode route searching:
(a) algorithm context initialization is changeed (b);
(b) initialization of population is changeed (c);
(c) initial population is carried out crossover operator computing between pattern, replenish and not exclusively separate, change (d) to separating fully;
(d) judge whether to reach maximum evolutionary generation, if "Yes" is changeed (i), if "No" is changeed (d);
(e) calculate evaluation of estimate, change (f);
(f) current population is carried out in the pattern between crossover operator, pattern mutation operator computing between mutation operator and pattern in crossover operator, the pattern, generate new individual record and go into population of future generation, change (g);
(g) current population is selected the operator computing, generate new individual record and go into population of future generation, change (h);
(h) new population replaces current population, and evolutionary generation adds 1, changes (d);
(i) calculate evaluation of estimate, finish;
(5) recommendation paths result set output:
Export non-repetitive individuality in last generation evolution population, and based on a certain standard, be arranged in order output, this results set is recommends the optimal path result, and each individuality in the output set is a kind of recommendation paths scheme; Output form is including, but not limited to text description, figure path, voice broadcast etc.
2. calculating evaluation of estimate according to claim 1 is characterized in that: all individualities are sorted by quality according to Optimum Theory, wherein establishing evaluation vector is f=(C
1, C
2..., C
n), C wherein
iExpression is individual at the substandard calculated value that expends of i kind, and then evaluation function is F=ParetoRank (f).
3. initialization of population according to claim 2 is characterized in that: generate initial individuality respectively under each monotype environment, remerge the composition initial population, be characterized in that individuality is monotype and allows not exclusively to separate.
4. according to claim 3 replenishing not exclusively separated to separating fully, it is characterized in that: not exclusively separate individuality for one, select another to separate individuality fully at random, carry out crossover operator computing between pattern, replace the former individuality of not exclusively separating with the newly-generated individuality of separating fully.
5. crossover operator between pattern according to claim 4, it is characterized in that: the point of crossing occurs in the different mode, promptly determine individual 1 and individual 2 cross-mode M earlier, N, in corresponding modes, seek and to intersect gene, intersect according to the intersected gene pairs of choosing, answer the polishing pattern identification during intersection.
6. mutation operator between pattern according to claim 5 is characterized in that: change point occurs in the different mode, adopts the directed variation strategy, quickens the purpose that excellent individual produces to reach.
7. the described selection operator of claim 6 is characterized in that: determine the individual amount X that the selection operator is selected according to current number of individuals G ' of new population and definition population size G;
X=G-G ', wherein G '=G
Crossover+ G
Mutation+ G
Hypercrossover+ G
Hypermutation
Max{G
crossover}＝G×P
crossover
Max{G
mutation}＝G×P
mutation
Max{G
hypercrossover}＝G×P
hypercrossover
Max{G
hypermutation}＝G×P
hypermutation
G
Crossover, G
Mutation, G
HypercrossoverWith G
HypermutationThe new individual number of representing corresponding operator to generate respectively;
Determine in the population of evaluation of estimate minimum successively that then non-repetition is individual, until reaching the select target number.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2010101858308A CN101853294B (en) | 2010-05-21 | 2010-05-21 | Multi-mode multi-standard path search method based on genetic algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2010101858308A CN101853294B (en) | 2010-05-21 | 2010-05-21 | Multi-mode multi-standard path search method based on genetic algorithm |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101853294A true CN101853294A (en) | 2010-10-06 |
| CN101853294B CN101853294B (en) | 2011-12-14 |
Family
ID=42804786
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2010101858308A Expired - Fee Related CN101853294B (en) | 2010-05-21 | 2010-05-21 | Multi-mode multi-standard path search method based on genetic algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101853294B (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102081752A (en) * | 2011-01-27 | 2011-06-01 | 西北工业大学 | Dynamic flight path planning method based on adaptive mutation genetic algorithm |
| CN102158799A (en) * | 2011-01-24 | 2011-08-17 | 东软集团股份有限公司 | Method and system for determining recommended passage place sequence |
| CN103914976A (en) * | 2014-03-28 | 2014-07-09 | 北方工业大学 | Heterogeneous user-oriented system optimal traffic distribution model and distribution method |
| CN104281761A (en) * | 2013-07-01 | 2015-01-14 | 株式会社日立制作所 | Method and device for evaluating land deterioration |
| CN104504477A (en) * | 2014-12-30 | 2015-04-08 | 中山大学 | Route optimizing method based on evolutionary mechanism of birds |
| CN105678421A (en) * | 2016-01-07 | 2016-06-15 | 中山大学 | Genetic algorithm-based guide sign guidance reachability optimization method |
| CN107944611A (en) * | 2017-11-17 | 2018-04-20 | 北京工业大学 | Towards the cross-module formula combined travel planing method of personalized trip requirements |
| CN109102124A (en) * | 2018-08-24 | 2018-12-28 | 山东师范大学 | Dynamic multi-objective multipath abductive approach, system and storage medium based on decomposition |
| CN111582552A (en) * | 2020-04-16 | 2020-08-25 | 浙江大学城市学院 | Shared bicycle parking point distribution method based on multi-target genetic algorithm |
| CN112381284A (en) * | 2020-11-11 | 2021-02-19 | 成都信息工程大学 | Improved genetic algorithm for optimizing multi-station path of unmanned transfer vehicle |
| CN112562309A (en) * | 2021-03-01 | 2021-03-26 | 湖南师范大学 | Network taxi appointment scheduling method based on improved Dijkstra algorithm |
| CN117273380A (en) * | 2023-10-25 | 2023-12-22 | 交通运输部公路科学研究所 | Travel scheme planning and recommending method and system for different travel scenes |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11829909B2 (en) | 2020-11-06 | 2023-11-28 | Sap Se | Route finder for integrated planning |
| US11775925B2 (en) | 2020-11-06 | 2023-10-03 | Sap Se | Route planning engine using unique location identifiers |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030145110A1 (en) * | 1998-06-02 | 2003-07-31 | Makoto Ohnishi | Network topology design apparatus and network topology design method, and recording medium recorded with a network topology design program |
| CN101089913A (en) * | 2007-07-12 | 2007-12-19 | 中国科学院地理科学与资源研究所 | Method for computing cross turning ratio based on genetic calculating method and large entropy mould |
| CN101614551A (en) * | 2009-07-21 | 2009-12-30 | 中国科学院地理科学与资源研究所 | Sneakernet method for building up and device, method for searching path and device |
| CN101707000A (en) * | 2009-10-26 | 2010-05-12 | 北京交通大学 | Urban road traffic multiobjective optimization control method |
-
2010
- 2010-05-21 CN CN2010101858308A patent/CN101853294B/en not_active Expired - Fee Related
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030145110A1 (en) * | 1998-06-02 | 2003-07-31 | Makoto Ohnishi | Network topology design apparatus and network topology design method, and recording medium recorded with a network topology design program |
| CN101089913A (en) * | 2007-07-12 | 2007-12-19 | 中国科学院地理科学与资源研究所 | Method for computing cross turning ratio based on genetic calculating method and large entropy mould |
| CN101614551A (en) * | 2009-07-21 | 2009-12-30 | 中国科学院地理科学与资源研究所 | Sneakernet method for building up and device, method for searching path and device |
| CN101707000A (en) * | 2009-10-26 | 2010-05-12 | 北京交通大学 | Urban road traffic multiobjective optimization control method |
Cited By (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102158799A (en) * | 2011-01-24 | 2011-08-17 | 东软集团股份有限公司 | Method and system for determining recommended passage place sequence |
| CN102158799B (en) * | 2011-01-24 | 2013-07-31 | 东软集团股份有限公司 | Method and system for determining recommended passage place sequence |
| CN102081752A (en) * | 2011-01-27 | 2011-06-01 | 西北工业大学 | Dynamic flight path planning method based on adaptive mutation genetic algorithm |
| CN102081752B (en) * | 2011-01-27 | 2014-04-16 | 西北工业大学 | Dynamic flight path planning method based on adaptive mutation genetic algorithm |
| CN104281761A (en) * | 2013-07-01 | 2015-01-14 | 株式会社日立制作所 | Method and device for evaluating land deterioration |
| CN103914976A (en) * | 2014-03-28 | 2014-07-09 | 北方工业大学 | Heterogeneous user-oriented system optimal traffic distribution model and distribution method |
| CN103914976B (en) * | 2014-03-28 | 2016-04-06 | 北方工业大学 | Heterogeneous user-oriented system optimal traffic distribution model and distribution method |
| CN104504477A (en) * | 2014-12-30 | 2015-04-08 | 中山大学 | Route optimizing method based on evolutionary mechanism of birds |
| CN104504477B (en) * | 2014-12-30 | 2018-01-09 | 中山大学 | A kind of method for optimizing route based on birds spore mechanism |
| CN105678421A (en) * | 2016-01-07 | 2016-06-15 | 中山大学 | Genetic algorithm-based guide sign guidance reachability optimization method |
| CN107944611A (en) * | 2017-11-17 | 2018-04-20 | 北京工业大学 | Towards the cross-module formula combined travel planing method of personalized trip requirements |
| CN107944611B (en) * | 2017-11-17 | 2021-11-26 | 北京工业大学 | Cross-mode combined travel planning method for personalized travel demand |
| CN109102124A (en) * | 2018-08-24 | 2018-12-28 | 山东师范大学 | Dynamic multi-objective multipath abductive approach, system and storage medium based on decomposition |
| CN109102124B (en) * | 2018-08-24 | 2021-07-20 | 山东师范大学 | Dynamic multi-target multi-path induction method and system based on decomposition and storage medium |
| CN111582552A (en) * | 2020-04-16 | 2020-08-25 | 浙江大学城市学院 | Shared bicycle parking point distribution method based on multi-target genetic algorithm |
| CN111582552B (en) * | 2020-04-16 | 2023-04-25 | 浙江大学城市学院 | Shared bicycle parking point distribution method based on multi-target genetic algorithm |
| CN112381284A (en) * | 2020-11-11 | 2021-02-19 | 成都信息工程大学 | Improved genetic algorithm for optimizing multi-station path of unmanned transfer vehicle |
| CN112381284B (en) * | 2020-11-11 | 2023-12-01 | 成都信息工程大学 | Improved genetic algorithm for multi-station path optimization of unmanned connection vehicle |
| CN112562309A (en) * | 2021-03-01 | 2021-03-26 | 湖南师范大学 | Network taxi appointment scheduling method based on improved Dijkstra algorithm |
| CN117273380A (en) * | 2023-10-25 | 2023-12-22 | 交通运输部公路科学研究所 | Travel scheme planning and recommending method and system for different travel scenes |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101853294B (en) | 2011-12-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101853294B (en) | Multi-mode multi-standard path search method based on genetic algorithm | |
| CN108764555B (en) | Shared bicycle parking point site selection method based on Hadoop | |
| CN112836892B (en) | Multi-target vehicle distribution path determining method and system based on improved genetic algorithm | |
| CN104634343B (en) | A kind of scenic spot route automatic planning based on multiple-objection optimization | |
| CN109034465A (en) | Consider the charging station bi-level optimization method that charging station addressing is coupled with trip route | |
| CN104200263B (en) | Power distribution network route planning method based on tabu differential evolution and GIS (Geographic Information System) | |
| CN109035767A (en) | A kind of tide lane optimization method considering Traffic Control and Guidance collaboration | |
| CN112734176B (en) | Charging station building method and device, terminal equipment and readable storage medium | |
| CN102903038B (en) | Based on smallclothes Logistics Distribution Method and the system of user trajectory | |
| Verma et al. | Feeder bus routes generation within integrated mass transit planning framework | |
| CN109377048A (en) | A kind of comprehensive traffic network hub node selection method | |
| CN112347596B (en) | Urban public transport network optimization method | |
| CN106289287A (en) | A kind of vehicle-mounted end Multiobjective Shortest Path computational methods based on route selection experience | |
| Johar et al. | Transit network design and scheduling using genetic algorithm–a review | |
| CN107832957A (en) | A kind of public lease bicycle dispatching method based on the handling of polymorphic type bicycle | |
| CN115983755A (en) | Multi-type combined transport path optimization method | |
| Cao | An Optimal Round‐Trip Route Planning Method for Tourism Based on Improved Genetic Algorithm | |
| CN110210667A (en) | O-D spatial and temporal distributions prediction technique based on mutative scale Geographical Weighted Regression Model | |
| CN112884229A (en) | Large-scale public place people flow guiding path planning method based on differential evolution algorithm | |
| CN117273380A (en) | Travel scheme planning and recommending method and system for different travel scenes | |
| CN114880819B (en) | Space optimization method based on town node importance degree and traffic area bit line research | |
| CN105678421A (en) | Genetic algorithm-based guide sign guidance reachability optimization method | |
| Wang et al. | Optimization of Traveler Intermodal Transport Routes Considering Carbon Emissions | |
| Mirzahossein et al. | Land-Use Modification Based on Transit-Oriented Development Adjacent to Historical Context (Case Study: Qazvin City) | |
| CN114329783A (en) | Multi-target electric vehicle charging network planning method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20111214 Termination date: 20150521 |
|
| EXPY | Termination of patent right or utility model |