CN104732279A - Improved cellular automaton traffic flow simulation analysis method based on geographic information system - Google Patents

Improved cellular automaton traffic flow simulation analysis method based on geographic information system Download PDF

Info

Publication number
CN104732279A
CN104732279A CN201510130531.7A CN201510130531A CN104732279A CN 104732279 A CN104732279 A CN 104732279A CN 201510130531 A CN201510130531 A CN 201510130531A CN 104732279 A CN104732279 A CN 104732279A
Authority
CN
China
Prior art keywords
cellular
step
traffic flow
data
geographic information
Prior art date
Application number
CN201510130531.7A
Other languages
Chinese (zh)
Inventor
董文永
康岚兰
董学士
刘宇航
王豫峰
Original Assignee
武汉大学
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 武汉大学 filed Critical 武汉大学
Priority to CN201510130531.7A priority Critical patent/CN104732279A/en
Publication of CN104732279A publication Critical patent/CN104732279A/en

Links

Abstract

The invention discloses an improved cellular automaton traffic flow simulation analysis method based on a geographic information system. According to the method, multi-criteria judgment, logistic regression, an improved principal component analysis method and an improved decision-making tree and kernel-based learning machine method are comprehensively used for mining and extracting the transformational rule of a cellular automaton, then the improved cellular automaton is used for conducting simulation analysis of traffic flow, and corresponding traffic flow data, geographic information data and the like are used for conducting traffic flow modeling and analysis. The method is based on a geographic information system technology and a database technology to conduct unified control and management on traffic flow information, geographic information and social information in a studied area. The method achieves integrated simulation analysis and maintenance management with diagrams and texts, comprehensive inquiry, thematic map and spatial analysis and traffic flow simulation analysis, the intelligence level and the aid decision making capacity of traffic management are improved, and the method is of certain reference significance to work study in the related fields.

Description

Based on the improvement cellular automaton traffic flow and traffic analog analysing method of Geographic Information System

Technical field

The invention belongs to Geoprocessing technical field, relate to the improvement cellular automaton traffic flow and traffic analog analysing method based on Geographic Information System.

Background technology

Cellular automaton is due to the feature of itself, be well suited for the sunykatuib analysis of spatial data, existing principal component analysis (PCA), decision tree and core learning machine etc. are optimized cellular automaton and are applied it to the case of the reallocation of land now, but do not apply it to the case that traffic flow simulation is analyzed, first the present invention improves principal component analysis (PCA), decision tree and core learning machine scheduling algorithm, then the transformation rule extracting cellular automaton is excavated with the algorithm after improvement, carry out traffic flow simulation analysis again, do not have the research case of related fields so far.

Traditional cellular automaton traffic flow and traffic analog analysing method efficiency is lower, the multiple modified hydrothermal process of this patent optimizes cellular automaton, new model can improve the level that traffic flow simulation is analyzed, and by GIS application to wherein, improve the ability of the method spatial simulation and analysis to a certain extent.

Summary of the invention

The object of the present invention is to provide the improvement cellular automaton traffic flow and traffic analog analysing method based on Geographic Information System, this method solve traditional single method traffic flow simulation analytical approach efficiency lower, the shortcomings such as spatial analysis level is limited, by improve cellular automaton and GIS application wherein, improve the Simulation and analysis level of the method to a certain extent.

The technical solution adopted in the present invention is carried out according to following steps:

Step 1: obtain spatial data and attribute data, set up corresponding spatial database and attribute database;

Step 2: build traffic flow simulation analytical model;

401 multiple criteria judgment models;

402 Logic Regression Models;

403 principal component models;

404 decision-tree models;

405 core learning machine models;

Step 3: the structure of cellular Automation Model, comprises cellular space Sum fanction/transforming function transformation function two parts;

Step 4: application multiple criteria judgment models, Logic Regression Models, principal component model, decision-tree model, core learning machine model, extracts cellular automaton transformation rule;

Step 5: build traffic flow simulation and analyze Geographic Information System.

Further, utilize Geographic Information System (GIS) software to carry out digitized processing to raster data or map datum lack of standardization in described step 1, make a width digital map; Wherein attribute data comprises geographic information data, the social investigation data needed for traffic flow simulation analysis.

Further, in described step 2,

Step 403 principal component model;

The standardization of step 4031 original index data;

The standardized acquisition p of original index data ties up random vector x=(x 1, x 2..., x p) tn sample x i=(x i1, x i2..., x ip) t, i=1,2 ...., n, n>p, structure sample battle array, following standardized transformation is carried out to sample array element: Z ij = x ij - x j ‾ s j , i = 1,2 , . . . , n ; j = 1,2 , . . . , p Wherein x j ‾ = Σ i = 1 n x ij n , s j 2 = Σ i = 1 n ( x ij - x j ‾ ) 2 n - 1 , Obtain standardization battle array Z;

Step 4032 pair normalized matrix asks correlation matrix;

Correlation matrix is asked to standardization battle array Z wherein, r ij = Σ z kj · z kj n - 1 , i , j = 1,2 , . . . , p ;

Step 4033 separates the secular equation of sample correlation matrix;

Separate the secular equation of sample correlation matrix R | R-λ I p|=0 p characteristic root, determine major component by determine m value, make the utilization factor of information reach more than 85%, to each λ j, j=1,2 ..., m, solving equations Rb=λ jb obtains unit character vector

Target variable after standardization is converted to major component by step 4034;

Target variable after standardization is converted to major component u 1be called first principal component, U 2be called the second main composition ...., U pbe called p major component;

Step 4035 pair major component carries out comprehensive evaluation;

Be weighted summation to m major component, obtain final evaluation of estimate, flexible strategy are the variance contribution ratio of each major component;

Step 404 decision-tree model;

Be provided with a training data set D={D 1, D 2..., D n, inside have n to train example, wherein each trains routine D i=(x i1, x i2..., x im, a i) (i=1,2 ..., n), D ithere are m non-categorical property value and a categorical attribute value a iif, non-categorical community set X={X 1, X 2..., X m, the routine D of each training ifor attribute X j(j=1,2 ..., value m) is x ij, wherein X jget discrete or successive value, categorical attribute set is A, quantizes; Separately there is test data set T, be used for assessing the nicety of grading of decision tree generated;

Step 4041 inputs: D, X, A;

Step 4042 basic decision tree generation algorithm exports: basic decision is set;

The clipping algorithm that step 4043 is set exports: improvement decision tree;

Step 4044 extracts classifying rules;

Step 4045 inputs: T, X, A;

Step 4046 is pressed classifying rules classification and is estimated error in classification ε;

Step 4047 inputs data set to be sorted, obtains classification results.

Further, in described step 3,

Step 501 cellular space;

Cellular space: the cellular automaton A of a standard is made up of cellular, cellular state, neighborhood and state updating rule, with mathematical notation is:

A=(L,d,S,N,f)

Wherein L is cellular space; D is the dimension in cellular space in cellular automaton; S is limited, the discrete state set of cellular; N is the set of all cellulars in certain neighborhood; F is that local maps or local rule;

Step 502 rule/transforming function transformation function;

Any one n ties up cellular automaton can be defined as following four-tuple

C=(D n,S,N,f)

In formula, D nfor n ties up Euclidean space, S is finite state set, the case of r representation element cellular automaton, can be expressed as the state of the cellular on case r in t

S(r,t)={S 1(r,t),S 2(r,t),L,S k(r,t)}

S k(r, t) represents kth the state of the cellular on case r in t; N is the neighborhood of cellular centered by r, is D nlimited sequence subset

N={N 1,N 2,L,N q}

N qrepresent the position of q neighbours relative to r of cellular r; F is the transformation rule of S (r, t) → S (r, t+1)

f={f 1,f 2,L,f m}

F mrepresent m the transformation rule in the space of cellular.If the current state of cellular is S (r, t), so a jth transformation rule of its next state is

S(r,t+1)=f j(S(r+N 1,t),S(r+N 2,t),L,S(r+N q,t))j=1,2,L,m

That is, the state of cellular subsequent time is only relevant with the state of its current neighbours.

In domain, first regional carries out stress and strain model to data, on the basis of conventional cellular Automation Model, introduces zoning schemes, domain space is divided into 9 regions.

The invention has the beneficial effects as follows

1) integrated use Geographic Information System of the present invention, multiple improvement cellular automaton and techniques of spatial data analysis, the Method and Technology that research traffic flow simulation is analyzed, the theoretical method of circulation sunykatuib analysis research realizes innovation;

2) carried out sunykatuib analysis and the management of traffic flow by the present invention, obtain the situations such as the change in time and space of traffic flow, provide decision-making foundation for all departments formulate relevant policies;

3) the present invention utilize Geographic Information System powerful geodata administrative analysis, visual and scientific algorithm function, express-analysis and assessment traffic flow situation, monitor and simulate development and the change of regional traffic flow, for relevant departments provide good aid decision making data, improve the level of IT application and the decision service efficiency of traffic administration.

Accompanying drawing explanation

Fig. 1 is based on the process flow diagram of the improvement cellular automaton traffic flow and traffic analog analysing method of Geographic Information System;

The process flow diagram that Fig. 2 principal component analysis (PCA) is implemented;

The process flow diagram that Fig. 3 decision tree method is implemented;

The process flow diagram that Fig. 4 cellular automaton builds;

Embodiment

Below in conjunction with embodiment, the present invention is described in detail.

The present invention is geodata based on map data base and remotely-sensed data, utilize traffic flow data, geographic information data, investigation statistics data, based on geographic information system technology and database technology, the Design Mode adopting centralized management to safeguard, carries out unified control and management to the telecommunication flow information in research area, geography information, humane information.This invention realizes the function such as maintenance management, comprehensive inquiry, thematic maps, spatial analysis, traffic flow simulation analysis of ducoment with illusion integration, for traffic simulation analysis and management provide the aid decision making foundation of science.

The present invention for achieving the above object, adopts following technical scheme:

Step 1: obtain spatial data and attribute data, set up corresponding spatial database and attribute database;

Utilize corresponding Geographic Information System (GIS) software to carry out digitized processing to raster data or map datum lack of standardization, make the digital map that a width has actual application value.Attribute data comprises geographic information data, social investigation data etc. needed for traffic flow simulation analysis; Utilize geographic information system technology and database technology, according to traffic flow data, social investigation data, geographic information data after screening, statistical treatment, set up corresponding spatial database and attribute database;

Step 2: build traffic flow simulation analytical model;

The key of cellular automaton how to define transformation rule, but current extracted transformation rule is mostly implicit, is carry out display rule by mathematical formulae, how to determine that the parameter in formula is comparatively difficult.The methods such as the principal component analysis (PCA) improved, the decision tree of improvement and core learning machine extract cellular automaton transformation rule, are then applied to by the new model of structure among traffic flow simulation analysis.

401 multiple criteria judgment models;

402 Logic Regression Models;

403 principal component models;

404 decision-tree models;

405 core learning machine models;

Mainly introduce principal component analysis (PCA) and decision tree method two models and optimization method thereof below:

Step 403 principal component model;

The standardization of step 4031 original index data;

The standardized acquisition p of original index data ties up random vector x=(x 1, x 2..., x p) tn sample x i=(x i1, x i2..., x ip) t, i=1,2 ...., n, n>p, structure sample battle array, following standardized transformation is carried out to sample array element: Z ij = x ij - x j ‾ s j , i = 1,2 , . . . , n ; j = 1,2 , . . . , p Wherein x j ‾ = Σ i = 1 n x ij n , s j 2 = Σ i = 1 n ( x ij - x j ‾ ) 2 n - 1 , Obtain standardization battle array Z.

Step 4032 pair normalized matrix asks correlation matrix;

Correlation matrix is asked to standardization battle array Z wherein, r ij = Σ z kj · z kj n - 1 , i , j = 1,2 , . . . , p .

Step 4033 separates the secular equation of sample correlation matrix;

Separate the secular equation of sample correlation matrix R | R-λ I p|=0 p characteristic root, determine major component by determine m value, make the utilization factor of information reach more than 85%, to each λ j, j=1,2 ..., m, solving equations Rb=λ jb obtains unit character vector

Target variable after standardization is converted to major component by step 4034;

Target variable after standardization is converted to major component u 1be called first principal component, U 2be called the second main composition ...., U pbe called p major component.

Step 4035 pair major component carries out comprehensive evaluation.

Be weighted summation to m major component, obtain final evaluation of estimate, flexible strategy are the variance contribution ratio of each major component.

Principal component model is improved one's methods: the nondimensionalization method improving raw data, and averaging method is exactly wherein a kind of preferably, is provided with N number of object be evaluated, and P index, and raw data is (X ij) n*p, the average of each index is X j, equalization is exactly remove their corresponding raw data, i.e. Z by the average of each index ij=X ji/ X j, equalization process does not change the related coefficient between each index, and the full detail of correlation matrix is all reflected in corresponding covariance matrix.Covariance matrix after equalization process not only eliminates the impact of index dimension and the order of magnitude, can also comprise the full detail of raw data, therefore when using principal component analytical method, can apply equalization method and carrying out without tempering process.

Step 404 decision-tree model;

The overall framework of decision Tree algorithms, is provided with a training data set D={D 1, D 2..., D n, inside have n to train example, wherein each trains routine D i=(x i1, x i2..., x im, a i) (i=1,2 ..., n), D ithere are m non-categorical property value and a categorical attribute value a iif, non-categorical community set X={X 1, X 2..., X m, the routine D of each training ifor attribute X j(j=1,2 ..., value m) is x ij, wherein X jdesirable discrete or successive value, categorical attribute set is A, quantizes.Separately there is test data set T, be used for assessing the nicety of grading of decision tree generated, be illustrated in fig. 4 shown below.

Step 4041 inputs: D, X, A;

Step 4042 basic decision tree generation algorithm exports: basic decision is set;

The clipping algorithm that step 4043 is set exports: improvement decision tree;

Step 4044 extracts classifying rules;

Step 4045 inputs: T, X, A;

Step 4046 is pressed classifying rules classification and is estimated error in classification ε;

Step 4047 inputs data set to be sorted, obtains classification results.

The improvement of decision tree method:

Algorithm described above carries out when data are very good, and the data in reality can not meet the condition required by algorithm as a rule, so just can not directly apply the algorithm setting up decision tree.Therefore, decision Tree algorithms should be improved in the following aspects before practical application.

1) process of missing values situation

When setting up decision tree, in training sample, the situation of missing values can often occur.A kind of disposal route missing values is regarded as a kind of possible value of attribute.If clearly, this disposal route is very appropriate for the missing values situation of this attribute to a certain extent; If missing values is in confused situation aobvious, then need more complicated solution.Another kind of missing values disposal route usually can provide a lot of information by the example of missing values.If the attribute of missing values does not play a role in decision-making at all, so these examples do not have the example of missing values not have what difference with other.

2) beta pruning of decision tree

The beta pruning problem of decision tree is a part and parcel in decision tree technique.Because initial decision tree of setting up not is a best decision tree analyzing new data-objects.This is mainly due in the process setting up decision tree, and the training sample data of each branch process reduce rapidly along with the increase of the number of plies, and its data characteristics embodied is also just more concrete, and generality is poorer, even may occur the conclusion of some absurdities.The main method addressed this is that carries out necessary beta pruning to decision tree.

3) derivation rule from tree

The rule of decision tree represents with IF-THEN form.The method of generation rule is: first produce a rule for each leaf node, then merges from all conditions on this leaf node path, this creates the terminal a rule.Wherein, the conjunct of the IF part of each " attribute-value " formation rule, leaf node comprises class prediction, formation rule THEN part.Paper format, the sequencing of rule does not affect its execution result.

Step 3: the structure of cellular Automation Model, comprises cellular space Sum fanction/transforming function transformation function two parts;

Step 501 cellular space;

Cellular space: the cellular automaton (A) of a standard is made up of cellular, cellular state, neighborhood and state updating rule.With mathematical notation be:

A=(L,d,S,N,f)

Wherein L is cellular space; D is the dimension in cellular space in cellular automaton; S is limited, the discrete state set of cellular; N is the set of all cellulars in certain neighborhood; F is that local maps or local rule.

Cellular space is the set of the site, space that cellular distributes.Cellular space is upwards unlimited extension in each dimension in theory, in order to realize on computers, and define boundary condition, comprise preiodic type, reflection-type and constant value type.

A cellular only takes from a kind of state of a finite aggregate usually a moment, such as { 0,1}.Cellular state can represent individual attitude, feature, behavior etc.Spatially adjacent with cellular cell is called adjacent unit, all adjacent units composition neighborhood.

Step 502 rule/transforming function transformation function;

The key of cellular automaton how to define transformation rule, but current extracted transformation rule is mostly implicit, is carry out display rule by mathematical formulae, how to determine that the parameter in formula is comparatively difficult.

Any one n ties up cellular automaton can be defined as following four-tuple

C=(D n,S,N,f)

In formula, D nfor n ties up Euclidean space, S is finite state set, the case of r representation element cellular automaton, can be expressed as the state of the cellular on case r in t

S(r,t)={S 1(r,t),S 2(r,t),L,S k(r,t)}

S k(r, t) represents kth the state of the cellular on case r in t; N is the neighborhood of cellular centered by r, is D nlimited sequence subset

N={N 1,N 2,L,N q}

N qrepresent the position of q neighbours relative to r of cellular r; F is the transformation rule of S (r, t) → S (r, t+1)

f={f 1,f 2,L,f m}

F mrepresent m the transformation rule in the space of cellular.If the current state of cellular is S (r, t), so a jth transformation rule of its next state is

S(r,t+1)=f j(S(r+N 1,t),S(r+N 2,t),L,S(r+N q,t))j=1,2,L,m

That is, the state of cellular subsequent time is only relevant with the state of its current neighbours.

Be below the foundation of model: first regional carries out stress and strain model to data in domain, on the basis of conventional cellular Automation Model, introduce zoning schemes, domain space is divided into 9 regions.

Related notion based in cellular Automation Model:

1) cellular: the measured data point within the scope of the domain discussed can be individual, financial institution, enterprise etc.Contacting between cellular and neighbours shows the interactive relation of each cellular;

2) cellular space: consider two-dimensional space, i.e. D 2, so the state of cellular can be written as: S (r, t)=S (x, y, t), and x, y are cellular two-dimensional coordinate in space.Plane is divided into the grid of 100 × 100, each grid represents a cellular.The scale of grid can change along with the regional extent size of simulation.

3) neighbours' form: for convenience's sake, the Moore type neighbours of we selected typical, the neighbours of cellular are made up of 8 cellulars of surrounding;

4) cellular state space: the state S (x, y, t) of cellular represents the elevation of t position (x, y).

5) transformation rule of cellular:

S(x,y,t+1)=f(S(x,y,t),L,S(x,y,t-k),S(x-1,y-1,t),S(x-1,y,t),S(x-1,y+1,t)

S(x,y-1,t),S(x,y+1,t),S(x+1,y-1,t),S(x+1,y,t),S(x+1,y+1,t))

Wherein, function f (x) is transfer function to be learned, can be obtained by historical data study.

Step 4: 5 models of step 2, and the cellular automaton of step 3 combines;

The key of cellular automaton how to define transformation rule, but current extracted transformation rule is mostly implicit, is carry out display rule by mathematical formulae, how to determine that the parameter in formula is comparatively difficult.Improve principal component analysis (PCA), method such as improvement decision tree and core learning machine etc. extracts cellular automaton transformation rule.The transformation rule that this model extracts does not need to be expressed by mathematical formulae, can be more convenient and describe the complex relationship of occurring in nature exactly, and these rules are understood than the easier people of allowing of mathematical formulae.And by GIS application to wherein, thus improve the capability and qualification of the sunykatuib analysis of the method.

Step 5: build traffic flow simulation and analyze Geographic Information System;

After core algorithm of the present invention builds, adopt following steps 60a to step 60k to build traffic flow simulation and analyze Geographic Information System, the system the built cellular Automation Model improved and spatial analysis etc. are analyzed, and wherein spatial analysis comprises the part described in step 604.

Step 60a demand analysis;

Step 60b feasibility analysis and primary design;

Step 60c detailed design;

Step 60d database design;

Step 60e software development;

Step 60f building database;

Step 60g program composition;

Step 60k software test, debugging, examination.

Step 604 spatial analysis module;

The process of step 6041 Spatial data capture;

The spatialization of step 6042 attribute data;

Step 6043 data space spatial scaling;

Step 6044 spatial information Exploring Analysis;

Step 6045 geo-statistic;

The data analysis of step 6046 lattice;

The inverting of step 6047 information and forecast.

Advantage of the present invention:

1) the method integrated use multiple criteria judgement, logistic regression, principal component analysis (PCA), decision tree and core learning machine method excavate the transformation rule extracting cellular automaton, the cellular automaton after improving is utilized to carry out the sunykatuib analysis of traffic flow, the case also not having core learning machine cellular automaton, decision tree cellular automaton etc. to apply in traffic flow simulation at present, comprehensive above five kinds of methods improved are applied to traffic flow simulation analysis, thus realize innovation in traffic flow simulation analytical approach.

2) another one innovative point of the present invention is in traffic flow simulation analysis by GIS application, utilize the spatial information analysis of Geographic Information System, visual and scientific algorithm function, express-analysis and assessment traffic flow situation, monitor and simulate development and the change of regional traffic flow, thus enhance the analysing and decision function of traffic flow simulation analytic system, can be relevant departments and better aid decision making data are provided, improve the level of IT application and the decision service efficiency of traffic administration.

Claims (4)

1., based on the improvement cellular automaton traffic flow and traffic analog analysing method of Geographic Information System, it is characterized in that: carry out according to following steps:
Step 1: obtain spatial data and attribute data, set up corresponding spatial database and attribute database;
Step 2: build traffic flow simulation analytical model;
401 multiple criteria judgment models;
402 Logic Regression Models;
403 principal component models;
404 decision-tree models;
405 core learning machine models;
Step 3: the structure of cellular Automation Model, comprises cellular space Sum fanction/transforming function transformation function two parts;
Step 4: application multiple criteria judgment models, Logic Regression Models, principal component model, decision-tree model, core learning machine model, extracts cellular automaton transformation rule;
Step 5: build traffic flow simulation and analyze Geographic Information System.
2. according to the improvement cellular automaton traffic flow and traffic analog analysing method based on Geographic Information System described in claim 1, it is characterized in that: utilize Geographic Information System (GIS) software to carry out digitized processing to raster data or map datum lack of standardization in described step 1, make a width digital map; Wherein attribute data comprises geographic information data, the social investigation data needed for traffic flow simulation analysis.
3., according to the improvement cellular automaton traffic flow and traffic analog analysing method based on Geographic Information System described in claim 1, it is characterized in that: in described step 2,
Step 403 principal component model;
The standardization of step 4031 original index data;
The standardized acquisition p of original index data ties up random vector x=(x 1, x 2..., x p) tn sample x i=(x i1, x i2..., x ip) t, i=1,2 ...., n, n>p, structure sample battle array, following standardized transformation is carried out to sample array element: Z ij = x ij - x j ‾ s j , I=1,2 ..., n; J=1,2 ..., p is wherein x j ‾ = Σ i = 1 n x ij n , s j 2 = Σ i = 1 n ( x ij - x j ‾ ) 2 n - 1 , Obtain standardization battle array Z;
Step 4032 pair normalized matrix asks correlation matrix;
Correlation matrix is asked to standardization battle array Z wherein,
r ij = Σ z kj · z kj n - 1 , i , j = 1,2 , . . . , p ;
Step 4033 separates the secular equation of sample correlation matrix;
Separate the secular equation of sample correlation matrix R | R-λ I p|=0 p characteristic root, determine major component by determine m value, make the utilization factor of information reach more than 85%, to each λ j, j=1,2 ..., m, solving equations Rb=λ jb obtains unit character vector
Target variable after standardization is converted to major component by step 4034;
Target variable after standardization is converted to major component j=1,2 ...., m U 1be called first principal component, U 2be called the second main composition ...., U pbe called p major component;
Step 4035 pair major component carries out comprehensive evaluation;
Be weighted summation to m major component, obtain final evaluation of estimate, flexible strategy are the variance contribution ratio of each major component;
Step 404 decision-tree model;
Be provided with a training data set D={D 1, D 2..., D n, inside have n to train example, wherein each trains routine D i=(x i1, x i2..., x im, a i) (i=1,2 ..., n), D ithere are m non-categorical property value and a categorical attribute value a iif, non-categorical community set X={X 1, X 2..., X m, the routine D of each training ifor attribute X j(j=1,2 ..., value m) is x ij, wherein X jget discrete or successive value, categorical attribute set is A, quantizes; Separately there is test data set T, be used for assessing the nicety of grading of decision tree generated;
Step 4041 inputs: D, X, A;
Step 4042 basic decision tree generation algorithm exports: basic decision is set;
The clipping algorithm that step 4043 is set exports: improvement decision tree;
Step 4044 extracts classifying rules;
Step 4045 inputs: T, X, A;
Step 4046 is pressed classifying rules classification and is estimated error in classification ε;
Step 4047 inputs data set to be sorted, obtains classification results.
4., according to the improvement cellular automaton traffic flow and traffic analog analysing method based on Geographic Information System described in claim 1, it is characterized in that: in described step 3,
Step 501 cellular space;
Cellular space: the cellular automaton A of a standard is made up of cellular, cellular state, neighborhood and state updating rule, with mathematical notation is:
A=(L,d,S,N,f)
Wherein L is cellular space; D is the dimension in cellular space in cellular automaton; S is limited, the discrete state set of cellular; N is the set of all cellulars in certain neighborhood; F is that local maps or local rule;
Step 502 rule/transforming function transformation function;
Any one n ties up cellular automaton can be defined as following four-tuple
C=(D n,S,N,f)
In formula, D nfor n ties up Euclidean space, S is finite state set, the case of r representation element cellular automaton, can be expressed as the state of the cellular on case r in t
S(r,t)={S 1(r,t),S 2(r,t),L,S k(r,t)}
S k(r, t) represents kth the state of the cellular on case r in t; N is the neighborhood of cellular centered by r, is D nlimited sequence subset
N={N 1,N 2,L,N q}
N qrepresent the position of q neighbours relative to r of cellular r; F is the transformation rule of S (r, t) → S (r, t+1)
f={f 1,f 2,L,f m}
F mrepresent m the transformation rule in the space of cellular; If the current state of cellular is S (r, t), so a jth transformation rule of its next state is
S(r,t+1)=f j(S(r+N 1,t),S(r+N 2,t),L,S(r+N q,t))j=1,2,L,m
That is, the state of cellular subsequent time is only relevant with the state of its current neighbours;
In domain, first regional carries out stress and strain model to data, on the basis of conventional cellular Automation Model, introduces zoning schemes, domain space is divided into 9 regions.
CN201510130531.7A 2015-03-25 2015-03-25 Improved cellular automaton traffic flow simulation analysis method based on geographic information system CN104732279A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201510130531.7A CN104732279A (en) 2015-03-25 2015-03-25 Improved cellular automaton traffic flow simulation analysis method based on geographic information system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201510130531.7A CN104732279A (en) 2015-03-25 2015-03-25 Improved cellular automaton traffic flow simulation analysis method based on geographic information system

Publications (1)

Publication Number Publication Date
CN104732279A true CN104732279A (en) 2015-06-24

Family

ID=53456151

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201510130531.7A CN104732279A (en) 2015-03-25 2015-03-25 Improved cellular automaton traffic flow simulation analysis method based on geographic information system

Country Status (1)

Country Link
CN (1) CN104732279A (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105335752A (en) * 2015-09-18 2016-02-17 国网山东省电力公司菏泽供电公司 Principal component analysis multivariable decision-making tree-based connection manner identification method
CN105590240A (en) * 2015-12-30 2016-05-18 合一网络技术(北京)有限公司 Discrete calculating method of brand advertisement effect optimization
CN105608604A (en) * 2015-12-30 2016-05-25 合一网络技术(北京)有限公司 Continuous calculation method of brand advertisement effectiveness optimization
CN106530691A (en) * 2016-10-25 2017-03-22 中山大学 Hybrid vehicle model multilane cellular automaton model considering vehicle occupancy space

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050222751A1 (en) * 2004-04-06 2005-10-06 Honda Motor Co., Ltd Method for refining traffic flow data
CN101783075A (en) * 2010-02-05 2010-07-21 北京科技大学 System for forecasting traffic flow of urban ring-shaped roads
CN101853290A (en) * 2010-05-25 2010-10-06 南京信息工程大学 Meteorological service performance evaluation method based on geographical information system (GIS)

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050222751A1 (en) * 2004-04-06 2005-10-06 Honda Motor Co., Ltd Method for refining traffic flow data
CN101783075A (en) * 2010-02-05 2010-07-21 北京科技大学 System for forecasting traffic flow of urban ring-shaped roads
CN101853290A (en) * 2010-05-25 2010-10-06 南京信息工程大学 Meteorological service performance evaluation method based on geographical information system (GIS)

Non-Patent Citations (5)

* Cited by examiner, † Cited by third party
Title
周娟: "中国铅锌工业布局评价体系研究", 《中国优秀硕士学位论文全文数据库 经济与管理科学辑》 *
李学伟 等: "《实用元胞自动机导论》", 31 August 2013, 北京交通大学出版社 *
王旭红: "遥感影像数据挖掘技术研究", 《中国优秀博硕士学位论文全文数据库 (博士) 信息科技辑》 *
邹杰: "基于元胞自动机的交通流模型研究", 《中国博士学位论文全文数据库 工程科技Ⅱ辑》 *
龙瀛 等: "北京城市空间发展分析模型", 《城市与区域规划研究》 *

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105335752A (en) * 2015-09-18 2016-02-17 国网山东省电力公司菏泽供电公司 Principal component analysis multivariable decision-making tree-based connection manner identification method
CN105590240A (en) * 2015-12-30 2016-05-18 合一网络技术(北京)有限公司 Discrete calculating method of brand advertisement effect optimization
CN105608604A (en) * 2015-12-30 2016-05-25 合一网络技术(北京)有限公司 Continuous calculation method of brand advertisement effectiveness optimization
CN106530691A (en) * 2016-10-25 2017-03-22 中山大学 Hybrid vehicle model multilane cellular automaton model considering vehicle occupancy space

Similar Documents

Publication Publication Date Title
Feng et al. Modeling dynamic urban growth using cellular automata and particle swarm optimization rules
Barredo et al. Urban sustainability in developing countries’ megacities: modelling and predicting future urban growth in Lagos
Zhong et al. Detecting the dynamics of urban structure through spatial network analysis
Wellmann et al. Uncertainties have a meaning: Information entropy as a quality measure for 3-D geological models
Crooks et al. Key challenges in agent-based modelling for geo-spatial simulation
Shang et al. A unified framework for multicriteria evaluation of transportation projects
Levy Multiple criteria decision making and decision support systems for flood risk management
Dietzel et al. The effect of disaggregating land use categories in cellular automata during model calibration and forecasting
Jain et al. Data mining techniques: a survey paper
Süzen et al. A comparison of the GIS based landslide susceptibility assessment methods: multivariate versus bivariate
Rafiee et al. Simulating urban growth in Mashad City, Iran through the SLEUTH model (UGM)
Yang et al. Cellular automata for simulating land use changes based on support vector machines
Li et al. Data mining of cellular automata's transition rules
Hansen GIS-based multi-criteria analysis of wind farm development
Gorsevski et al. A group-based spatial decision support system for wind farm site selection in Northwest Ohio
Tsamboulas et al. Use of multicriteria methods for assessment of transport projects
Sánchez-Lozano et al. GIS-based photovoltaic solar farms site selection using ELECTRE-TRI: Evaluating the case for Torre Pacheco, Murcia, Southeast of Spain
Yeh et al. Errors and uncertainties in urban cellular automata
Berling-Wolff et al. Modeling urban landscape dynamics: A review
Ménard et al. Exploration of spatial scale sensitivity in geographic cellular automata
Brown et al. Stochastic simulation of land-cover change using geostatistics and generalized additive models
Zhang et al. Simulation and analysis of urban growth scenarios for the Greater Shanghai Area, China
Ferentinou et al. Computational intelligence tools for the prediction of slope performance
Griffith et al. Non-standard spatial statistics and spatial econometrics
Li et al. An extended cellular automaton using case‐based reasoning for simulating urban development in a large complex region

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
EXSB Decision made by sipo to initiate substantive examination
SE01 Entry into force of request for substantive examination
WD01 Invention patent application deemed withdrawn after publication
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20150624