CN109272170B - A kind of traffic zone dividing system based on Louvain algorithm - Google Patents
A kind of traffic zone dividing system based on Louvain algorithm Download PDFInfo
- Publication number
- CN109272170B CN109272170B CN201811186046.1A CN201811186046A CN109272170B CN 109272170 B CN109272170 B CN 109272170B CN 201811186046 A CN201811186046 A CN 201811186046A CN 109272170 B CN109272170 B CN 109272170B
- Authority
- CN
- China
- Prior art keywords
- building
- groups
- data
- network
- road
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 claims abstract description 44
- 238000004364 calculation method Methods 0.000 claims abstract description 9
- 230000032823 cell division Effects 0.000 claims abstract description 6
- 239000011159 matrix material Substances 0.000 claims description 26
- 230000011218 segmentation Effects 0.000 claims description 20
- 230000008569 process Effects 0.000 claims description 18
- 238000009826 distribution Methods 0.000 claims description 16
- 238000004919 topotaxy Methods 0.000 claims description 12
- 238000000926 separation method Methods 0.000 claims description 5
- 230000004888 barrier function Effects 0.000 claims description 3
- 238000002955 isolation Methods 0.000 claims description 3
- 238000002360 preparation method Methods 0.000 claims description 3
- 238000012545 processing Methods 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 2
- 238000004458 analytical method Methods 0.000 abstract description 4
- 238000005457 optimization Methods 0.000 abstract description 4
- 230000000694 effects Effects 0.000 description 7
- 230000006399 behavior Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 230000006835 compression Effects 0.000 description 2
- 238000007906 compression Methods 0.000 description 2
- 230000002596 correlated effect Effects 0.000 description 2
- 230000000875 corresponding effect Effects 0.000 description 2
- 238000005520 cutting process Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 230000015572 biosynthetic process Effects 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000007429 general method Methods 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 238000004513 sizing Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
- G06Q10/043—Optimisation of two dimensional placement, e.g. cutting of clothes or wood
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
Landscapes
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Engineering & Computer Science (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- Physics & Mathematics (AREA)
- Development Economics (AREA)
- Theoretical Computer Science (AREA)
- Marketing (AREA)
- General Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Game Theory and Decision Science (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Entrepreneurship & Innovation (AREA)
- Educational Administration (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Traffic Control Systems (AREA)
Abstract
A kind of traffic zone dividing system based on Louvain algorithm, comprising: space topological relation discrimination module, groups of building divide category of roads computing module, groups of building topology distance computing module, complex network side right value computing module, Louvain algorithm execution module, cell division result display module.Traffic zone dividing system provided by the invention is by the way that in urban mass-transit system, the factor divided to traffic zone is analyzed, to reduce the complexity of traffic system.The present invention is minimum division unit with architecture ensemble, the abstract method of Discussion of City groups of building net, planned network structural parameters calculation method, constructs the complex network using groups of building as node, and the similar groups of building of traffic attribute are divided into same traffic zone based on the network.The division result of traffic zone will affect the accuracy of the work such as transport need analysis and prediction, and the division and optimization to traffic zone are significant in terms of the feasibility and accuracy that ensure traffic programme scheme.
Description
Technical field
The invention belongs to field of intelligent transportation technology, in particular to based on the traffic zone dividing system of Louvain algorithm.
Background technique
Before being divided to traffic zone, needs to carry out discretization to urban area, construct Geographic Unit, as
The initial clustering unit that traffic zone divides.It is exactly to use limited multiple geographical units to the essence of the discretization of urban area,
Continuous city space is replaced, city space is abstracted as comprising several with space attribute, spatial relationship basically
Manage unit set.The Main Morphology of Geographic Unit includes: street, demographics subregion, administrative area, polygonal mesh.Mesh
There are two types of the preceding discretization unit application to traffic zone is most: road network structure and planar object.
Discrete region method based on road structure: in this method, the road network of survey region is carried out first
The topological structure of road network is calculated in abstract modeling.Using section as geographical unit, the natural segmentation of urban area is realized.And
And can be split according to research needs, the specified section for using different category of roads, it can control the fining journey of segmentation
Degree.
Discrete region method based on planar object: in the discrete region method based on planar object, use is most wide
General method is " Thiessen polygon method ".The characteristic of this method is only to contain a discrete points data in each Thiessen polygon,
The distance of point to corresponding discrete point in Thiessen polygon is nearest, the point on Thiessen polygon side to its both sides discrete point
It is equidistant.
The advantages of discrete region method based on road structure is, due to road be Urban Natural segmentation as a result,
This dividing method is not suitable for subsequent than the space characteristics that the geographical unit of manual construction clusters more representative of urban microscopic
Research.In a larger sense, traffic zone refers to the set of the mikey that is mutually related with certain similitude, divides traffic
The essence of cell is to replace point with area, reduces the complexity of traffic system.
Summary of the invention
Technology of the invention solves the problems, such as: overcoming the deficiencies of the prior art and provide the traffic zone based on Louvain algorithm
Dividing system, for reducing the complexity of traffic system, the purpose for carrying out traffic zone division is for preferably analyzed area
To the influence power of bus station passenger flow, therefore, it is desirable to Intra-cell groups of building to be spatially closely connected, most on traffic attribute
It measures similar.The traffic zone that the present invention divides has above-mentioned two characteristic, while being able to solve the ignored problem in small community.
The present invention devises the traffic zone dividing system based on Louvain algorithm, as shown in Figure 1, comprising: space
Topological relation discrimination module, groups of building divide category of roads computing module, groups of building topology distance computing module, complex network side
Weight computing module, Louvain algorithm execution module and cell division result display module；
Space topological relation discrimination module: it after obtaining initial groups of building data, is pre-processed, preprocessing process
For according to the physical distance of groups of building (the distance between groups of building), main road segmentation (whether have advanced road between groups of building,
Such as loop, expressway), groups of building block and (whether have other buildings to block between groups of building) factor to generate initial groups of building empty
Between topological relation data, do data preparation for groups of building category of roads computing module and groups of building topology distance computing module；
Groups of building divide category of roads computing module: according to the initial groups of building spatial topotaxy data and road network of generation
Data (data such as the width of road, grade, position, i.e. city road network data in city) are differentiated by vector and obtain groups of building
Between segmentation road, obtain highest category of roads as final category of roads, generate groups of building divided roadway number of degrees
It is called accordingly for complex network side right value computing module；
Groups of building topology distance computing module: according to the initial groups of building spatial topotaxy data of generation, pass through calculating
Direction in space similarity obtains the topology distance before every two groups of building, the topology distance represent before groups of building between phase
Like relationship, groups of building topology distance data are generated according to topology distance；
Complex network side right value computing module: divide category of roads data and groups of building topology distance number according to groups of building
According to the company side progress weight assignment of groups of building complex network, formation weight matrix generates complex network weight matrix data；
Louvain algorithm execution module: according to complex network weight matrix data, traffic is carried out using Louvain algorithm
Small Division finds the division methods so that when modularity maximum by the modularity function of setting repeatedly, final using complicated
Groups of building are classified as different traffic zones by network weight matrix data, are called for cell division result display module；The calculation
Method is a kind of algorithm based on modularity, and the modularity the big, and the Intra-cell side right divided is again bigger, and Intra-cell connection is got over
Closely, the difference between cell is more obvious, using modularity as the Louvain community detection algorithm of evaluation criterion, in efficiency and effect
On fruit all performance preferably, and it can be found that hierarchy structure, and it is it can be found that lesser community；
Cell division result display module: executing the different traffic zones generated according to traffic zone division methods, and
It is visualized, marks different traffic zones with different colours on base map.
The space topological relation discrimination module the specific implementation process is as follows:
After obtaining initial groups of building data, judges whether there is even frontier juncture system in a network between two groups of building, differentiate
Process is as follows:
(1) judge the physical distance (physical distance factor) between two groups of building；
(2) judge whether to be separated by the high road of category of roads (expressway, the main roads such as loop) between two groups of building (big
Divide factor in road)；
(3) judge whether blocked by other groups of building between two groups of building, measurement coverage extent (groups of building block because
Element)；
(4) according to above-mentioned 3 differentiate as a result, if the physical distance between two groups of building less than 200 meters, no expressway,
Loop segmentation, is blocked without other groups of building, then judges that the two groups of building have even frontier juncture system in a network；
(5) step (4) are executed repeatedly, generates initial groups of building spatial topotaxy data as the calculating of groups of building category of roads
Module and groups of building topology distance computing module do data preparation.
The groups of building segmentation category of roads computing module realizes that process is as follows:
(1) initial groups of building spatial topotaxy data and city road network data are read；
(2) according to city road network data, all roads divided between two groups of building are obtained；
(3) to the road divided between two groups of building, the category of roads of every road is successively calculated.Specific method example
Are as follows: first judge whether selected section has divided the two groups of building, if groups of building B and groups of building C is separated by section 1 and section 2,
Set the central point of groups of building B then as Pb, the central point of groups of building C is Pc, line segment LbcFor point PbWith point PcLine, LbcWith section 1
Intersection point is generated with section 2, so that section 1 and section 2 become the separation section of groups of building B and C, then judges section 1 and line segment
LbcWhether intersect, groups of building B and groups of building C have been divided in section 1 if intersection.If the two buildings have been divided in selected section
Group establishes path search index table to groups of building, retrieves city then according to urban architecture distribution density and city road network road length
Then section concordance list in road net data inquires the category of roads in selected section；
(4) to the category of roads calculated in step (3), the road for taking out highest level in segmentation section is built as two
Divide the grade of road between group；
(5) groups of building segmentation category of roads data are generated to call for complex network side right value computing module.
The groups of building topology distance computing module realizes that process is as follows:
(1) initial groups of building spatial topotaxy data are read；
(2) the field groups of building of each groups of building are calculated, method particularly includes: for groups of building A, successively calculate other buildings
Group arrives the space physics distance of groups of building A, if the physical distance is less than the threshold value of setting, which belongs to the field of A
Building set, wherein threshold value carries out dynamic adjustment according to region site coverage；
(3) directional similarity square of all groups of building relative to groups of building A in the field groups of building set of groups of building is calculated
Battle array Skk, wherein k is the size of neighborhood；
(4) similarity matrix in processing step (3) takes out opposite building to each groups of building in similarity matrix
Group A has the groups of building set of close direction similarity, and specific method is the direction phase according to groups of building in (3) for groups of building A
Like property matrix Sij, wherein i represents i-th of groups of building, and j represents j-th of groups of building, if meeting Sij>Sset, wherein SsetTo set
Fixed threshold value is adjustable parameter, and value of the embodiment of the present invention is 0.75, and the groups of building for being set as obtaining when 0.75 are integrated into space
There is highest similarity, this is obtained by a large amount of repetition tests, when S is greater than S in relationshipsetWhen, it is believed that two groups of building
The same direction is in relative to groups of building A；
(5) for the groups of building set in step (4), by the distance-taxis of wherein all groups of building and groups of building A, sequence
The topology distance of first groups of building and groups of building A are 1 afterwards, are secondly 2, successively plus 1；
(6) groups of building topology distance data are generated according to the topology distance being calculated, for complex network side weight computing
Module uses.
The complex network side right value computing module realizes that process is as follows:
(1) groups of building segmentation category of roads data and groups of building topology distance data are read；
(2) space topological distance and road isolation information are combined, the frontier juncture system, company of nodes, nodes are established
viWith vjBetween similarity and node represented by land used attribute, road barrier, space length and topology between groups of building
Distance is all related, calculates node v in conjunction with land used attribute, road network and the geographical location information of groups of buildingiWith vjBetween
Weight, the weight between the two nodes are expressed as:
W(vi,vj)=F (Lij,SPDij,Rij,TDij)
LijIndicate node viWith vjBetween land used attributes similarity, SPDijIndicate node viWith vjBetween space physics
Distance, RijIndicate node viWith vjBetween divided roadway grade, TDijIndicate node viWith vjBetween topology distance, F indicate
Similarity calculation function；
(3) weight assignment is carried out according to company side of the similarity being calculated to groups of building complex network, generates complex web
Network weight matrix data, the weight matrix are called for Louvain algorithm execution module.
The Louvain algorithm execution module realizes that process is as follows:
(1) complex network weight matrix data are read, regard all groups of building as a complex network；
(2) each groups of building of complex network are seen as an independent traffic zone, at this time number of cells and groups of building
Number is identical；
(3) it for groups of building A each in current network, successively attempts for A to be assigned to the cell where its neighbours' groups of building,
Then the preceding modularity changes delta Q with network after distribution of distribution is calculated according to above-mentioned calculation method, finds Δ Q most after the completion of traversal
Distribution method when big, is denoted as max Δ Q；
(4) if max Δ Q > 0 in step (3), and plot area when Δ Q maximum where that neighbours' groups of building
Less than 0.5 square kilometre, then otherwise the node division is remained unchanged to the cell；
(5) step (3) and step (4) are repeated, until the affiliated subdistrict of all groups of building is no longer changed；
(6) complex network is compressed, all groups of building in the same cell are compressed into a new groups of building, small
The side right in section is converted into the side right weight between new building group again；
(7) result that step (6) obtains is transferred to step (2), continues to repeat step (3) and step (4), until entirely multiple
The modularity of miscellaneous network no longer changes；
(8) algorithm terminates, and all groups of building are divided into different traffic zones.
The traffic zone division result display module realizes that process is as follows:
(1) according to Louvain algorithm execution module obtain as a result, obtaining cell number belonging to each groups of building；
(2) map datum is obtained, and will be in groups of building Data Matching to map；
(3) according to step (1) and step (2), corresponding color is drawn to the groups of building of different numbers, completes to show.
The advantages of the present invention over the prior art are that:
The traffic zone dividing system that the present invention designs establishes Space expanding to groups of building first, then carries out topology
Distance calculates, and then assigns weight, then according to Louvain algorithm partition traffic zone.
There are Intra-cell groups of building to be spatially closely connected, is high on traffic attribute for the traffic zone that the present invention divides
Similar two characteristics are spent, while solving the problems, such as that small community is ignored.
Regard all groups of building as a complex network, the present invention can carry out hierarchy division to complex network, every time
Compression back and the number of node all reduce, and calculating speed is fast, divide that effect is good, complex network can be carried out one it is careful
It divides, obtains the traffic zone with obvious community structure.
It is exactly to reduce answering for traffic system to the critically important effect that urban area carries out division traffic zone
Miscellaneous degree.In public transport demand behaviors, traffic zone is the fundamental space unit for studying public traffic passenger flow demand and generating and being distributed,
The attributes such as geographical location, size and the distribution situation relative to road network of traffic zone can all influence transport need analysis
With the accuracy of the work such as prediction, division to traffic zone and optimization are in the feasibility and accuracy for ensuring traffic programme scheme
Aspect is significant.
Detailed description of the invention
Fig. 1 is the traffic zone dividing system architecture diagram provided in an embodiment of the present invention based on Louvain algorithm；
Fig. 2 is the schematic diagram of groups of building in the embodiment of the present invention；
Fig. 3 is the schematic diagram of city road network structure in the embodiment of the present invention；
Fig. 4 is the schematic illustration of groups of building Spatial Direction Relations in the embodiment of the present invention；
Fig. 5 is the flow chart of space topological relation discrimination module in the embodiment of the present invention；
Fig. 6 is the flow chart of groups of building topology distance computing module in the embodiment of the present invention；
Fig. 7 is the flow chart of groups of building divided roadway class computing module in present example；
Fig. 8 is the schematic illustration of Louvain algorithm execution module in the embodiment of the present invention；
Fig. 9 is the effect picture of small Division result display module in the invention patent.
Specific embodiment
To make the object, technical solutions and advantages of the present invention clearer, the present invention is made with reference to the accompanying drawing further
Detailed description.
First basic concepts are once illustrated.
Traffic zone dividing system provided by the invention is by urban mass-transit system, the factor that divides to traffic zone
It is analyzed, to reduce the complexity of traffic system.The present invention uses for reference Complex Networks Theory model, is minimum with architecture ensemble
Division unit, the abstract method of Discussion of City groups of building net, planned network structural parameters calculation method construct and are with groups of building
The similar groups of building of traffic attribute are divided into same traffic zone based on the network by the complex network of node.The present invention is to friendship
Before logical cell is divided, needs to carry out discretization to urban area, construct Geographic Unit, divided as traffic zone
Initial clustering unit.It is exactly using limited multiple geographical units, to replace continuously to the essence of the discretization of urban area
City space is abstracted as the Geographic Unit set comprising several with space attribute, spatial relationship by city space.City
City's traffic system is the especially big network system of complexity an of close coupling, nonlinear, strong randomness, is drawn to urban area
The critically important effect of one of point traffic zone is exactly the complexity in order to reduce traffic system.In public transport demand behaviors, hand over
Logical cell is the fundamental space unit for studying public traffic passenger flow demand and generating and being distributed, the geographical location of traffic zone, area
The attributes such as size and distribution situation relative to road network can all influence the accuracy of the work such as transport need analysis and prediction, right
The division and optimization of traffic zone are significant in terms of the feasibility and accuracy that ensure traffic programme scheme.
As shown in Fig. 2, groups of building refer to it is being made of adjacent buildings several in city, be closely connected in spatial organization
Building body set.Wherein minimum unit is single building body, by central point latitude and longitude coordinates and boundary point latitude and longitude coordinates
Composition, its boundary point of regular shape is made of 4 coordinates, and irregular shape is formed by then putting coordinates by 4 or more.Longitude and latitude
Coordinate is made of (x, y), and wherein x represents longitude, and value range is [0.00,180.00], y represent latitude value range as
[0.00,90.00].Multiple building body objects are arrived comprising one inside each groups of building.
As shown in figure 3, city road network is made of the road of a variety of grades.City road network model is represented by R=(I, S):
I={ I1, I2..., InIndicate intersection point set in road network, produced by all sections are intersected two-by-two in intersection point road network
Coordinate points, be made of (x, y), wherein x (0≤x≤180) represents longitude, and y represents latitude (0≤y≤90)；
S={ S1, S2..., SnIndicating that the section in road network is gathered, each section is indicated by ordered pair<s, e>composition
The starting point in section is s, and terminating point e, section is one section of directed line segment from s to e.Section 3 is expressed as < intersection point 1, intersection point 2
>；
As shown in figure 4, the main embodiment of space topological relation discrimination module are as follows: groups of building data are obtained,
In the case that node has been selected, in order to establish network, it is thus necessary to determine that whether have connection relationship between two nodes in network.It builds
It builds group and relies on road structure, it is intended to spatial stochastically distribution rule be presented, then can define two in conjunction with road network information
There are should meet when connection relationship between a groups of building:
(1) judge the physical distance (physical distance factor) between two groups of building；
(2) judge whether to be separated by the high road of category of roads (expressway, the main roads such as loop) between two groups of building (big
Divide factor in road)；
(3) judge between two groups of building whether by other groups of building to block (groups of building block factor)；
In above three condition, first condition makees a coarse sizing according to physical distance, and second condition makes algorithm
It can use road structure and primary cutting done to architecture ensemble group.In order to realize third condition, use space direction matrix
Direction in space similarity between groups of building is calculated, so that it is determined that whether there is connection relationship between groups of building, to obtain just
It establishes and builds group space topological relation data.
As shown in figure 5, the main embodiment of groups of building segmentation category of roads computing module are as follows: it is empty to read initial groups of building
Between topological relation data and city road network data, then calculate the category of roads of segmentation.Such as groups of building B and groups of building C are by road
Section 1 separates with section 2, it assumes that the central point of groups of building B is Pb, the central point of groups of building C is Pc, line segment LbcFor point PbWith point
PcLine, Cong Tuzhong is it can be found that LbcIntersection point is generated with section 1,2, so that section 1 and section 2 become groups of building B, C
Separation section.Section is by coordinate points to<S, E>composition, it is assumed that the start-stop point coordinate in section 1 is respectively Ps(Sx, Sy), Pe(Ex,
Ey), PbCoordinate be (Bx, By), PcCoordinate be (Cx, Cy), then section 1 and line segment L may determine that by vector calculatingBCWhether
Intersection.
Based on above-mentioned algorithm, two groups of building central point lines and section can be calculated with the presence or absence of intersection point, to identify
Separation section between groups of building out.But the section quantity in city road network data generally reaches 100,000 grades even million grades,
If carrying out calculated crosswise with system-wide network section when calculating, algorithm expense cannot be received.It is distributed according to urban architecture close
Groups of building are established path search index table, for each groups of building, stored in concordance list by degree and city road network road length
With the road section ID within the scope of circumference 300 meters, to greatly reduce calculation amount when calculating.By retrieving city road network data
In section concordance list, take out segmentation section in highest level road as segmentation road grade, generate groups of building separation
Category of roads data.
As shown in fig. 6, the main embodiment of groups of building topology distance computing module are as follows: read initial building group space and open up
Relation data is flutterred, for groups of building A, steps are as follows with the detailed calculating of the frontier juncture system, company of A in a network for the groups of building on periphery:
(1) the field groups of building of each groups of building are calculated, method particularly includes: for groups of building A, successively calculate other buildings
Group arrives the space physics distance of groups of building A, if the physical distance is less than the threshold value of setting, which belongs to the field of A
Building set, wherein threshold value carries out dynamic adjustment according to region site coverage；
(2) directional similarity square of all groups of building relative to groups of building A in the field groups of building set of groups of building is calculated
Battle array Skk, wherein k is the size of neighborhood；
(3) similarity matrix in processing step (2) takes out opposite building to each groups of building in similarity matrix
Group A has the groups of building set of close direction similarity, and specific method is the direction phase according to groups of building in (3) for groups of building A
Like property matrix Sij, wherein i represents i-th of groups of building, and j represents j-th of groups of building, if meeting Sij>Sset, wherein SsetTo set
Fixed threshold value is adjustable parameter, and value 0.75, the groups of building for being set as obtaining when 0.75, which are integrated into spatial relationship, to be had most
High similarity, when S is greater than SsetWhen, it is believed that two groups of building are in the same direction relative to groups of building A；
(4) for groups of building set obtained in step (3), by the distance-taxis of wherein groups of building and A, first after sequence
The topology distance of a element and A are 1, are secondly 2, successively plus 1；
(5) by the result in (4) as groups of building topology distance data.
As shown in fig. 7, the main embodiment of complex network side right value computing module are as follows:
(1) groups of building segmentation category of roads data and groups of building topology distance data are read；
(2) space topological distance and road isolation information are combined, the frontier juncture system, company of nodes, nodes are established
viWith vjBetween similarity and node represented by land used attribute, road barrier, space length and topology between groups of building
Distance is all related, calculates node v in conjunction with land used attribute, road network and the geographical location information of groups of buildingiWith vjBetween
The weight of weight, the two nodes is expressed as:
W(vi,vj)=F (Lij,SPDij,Rij,TDij)
LijIndicate node viWith vjBetween land used attributes similarity, SPDijIndicate node viWith vjBetween space physics
Distance, RijIndicate node viWith vjBetween divided roadway grade, TDijIndicate node viWith vjBetween topology distance, F indicate
Similarity calculation function.
The solution of function F has no optimal process, it is assumed that F be it is linear, it is solved.Attribute A, R and W are positively correlated, SPD,
TD and W is negatively correlated, therefore herein by F is defined as:
(3) weight assignment is carried out according to company side of the similarity being calculated to groups of building complex network, generates complex web
Network weight matrix data, the weight matrix are called for Louvain algorithm execution module.
As shown in figure 8, Louvain algorithm execution module realizes that process is as follows:
(1) complex network weight matrix data are read, regard all groups of building as a complex network；
(2) each groups of building of complex network are seen as an independent traffic zone, at this time number of cells and groups of building
Number is identical；
(3) it for groups of building A each in current network, successively attempts for A to be assigned to the cell where its neighbours' groups of building,
Then the preceding modularity changes delta Q with network after distribution of distribution is calculated according to above-mentioned calculation method, finds Δ Q most after the completion of traversal
Distribution method when big, is denoted as max Δ Q；
(4) if max Δ Q > 0 in (3), and plot area when Δ Q maximum where that neighbours' groups of building is less than
0.5 square kilometre, then otherwise the node division is remained unchanged to the cell；
(5) (3) and step (4) are repeated, until the affiliated subdistrict of all groups of building is no longer changed；
(6) complex network is compressed, all groups of building in the same cell are compressed into a new groups of building, small
The side right in section is converted into the side right weight between new building group again；
(7) result that step (6) obtains is transferred to step (2), continues to repeat step (3) and step (4), until entirely multiple
The modularity of miscellaneous network no longer changes；
(8) algorithm terminates, and all groups of building are divided into different traffic zones.
As shown in figure 9, the groups of building with same color belong to a traffic zone in figure as unit of groups of building.It should
Region divides traffic zone 18 altogether, and maximum traffic zone area is no more than 1 square kilometre.It is seen that the road between cell
Road distributed relation is obvious, and spatial distribution is also more uniform, and the difference of land used attribute is also more obvious.Lower left corner region in such as figure,
Dark brown part building is three community Yi Miao Wests, and purple part is friendship community, and RED sector is the Renmin University of China, model
Success is divided three different traffic zones.
The target that traffic zone divides is: using traffic zone as unit, city cutting being become multiple cells, these cells
Meet worldlet, cluster property characteristic cell, i.e., road spaced relation is clear between cell；Geographical position between Intra-cell groups of building
It sets closely coupled；Boundary is clearly demarcated between cell and cell；Intra-cell has similar traffic characteristics.
In short, there are Intra-cell groups of building to be spatially closely connected, in traffic category for the traffic zone that the present invention divides
Property on similar two characteristics of height, while solving the problems, such as that small community is ignored.
Regard all groups of building as a complex network, the present invention can carry out layer to the complex network of intelligent transportation field
Secondary property divides, and the number of every second compression back and node all reduces, and calculating speed is fast, and division effect is good, can be to complex network
A careful division is carried out, the traffic zone with obvious community structure is obtained.
It is exactly to reduce answering for traffic system to the critically important effect that urban area carries out division traffic zone
Miscellaneous degree.In public transport demand behaviors, traffic zone is the fundamental space unit for studying public traffic passenger flow demand and generating and being distributed,
The attributes such as geographical location, size and the distribution situation relative to road network of traffic zone can all influence transport need analysis
With the accuracy of the work such as prediction, division to traffic zone and optimization are in the feasibility and accuracy for ensuring traffic programme scheme
Aspect is significant.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the scope of the present invention.
Claims (8)
1. a kind of traffic zone dividing system based on Louvain algorithm, it is characterised in that: including space topological relation
Discrimination module, groups of building divide category of roads computing module, groups of building topology distance computing module, complex network side weight computing
Module, Louvain algorithm execution module and cell division result display module；
Space topological relation discrimination module: it after obtaining initial groups of building data, is pre-processed, preprocessing process is root
It blocks factor according to the physical distance of groups of building, main road segmentation, groups of building and generates initial groups of building spatial topotaxy data, to build
It builds group category of roads computing module and groups of building topology distance computing module does data preparation；
Groups of building divide category of roads computing module: according to the initial groups of building spatial topotaxy data and road network number of generation
According to, the segmentation road obtained between groups of building is differentiated by vector, obtains highest category of roads as final category of roads,
Groups of building divided roadway level data is generated for complex network side right value computing module calling；
Groups of building topology distance computing module: according to the initial groups of building spatial topotaxy data of generation, by calculating space
Direction similarity obtains the topology distance before every two groups of building, which represents the similar pass between groups of building
System generates groups of building topology distance data according to topology distance；
Complex network side right value computing module: dividing category of roads data and groups of building topology distance data according to groups of building, right
The company side of groups of building complex network carries out weight assignment, forms weight matrix, generates complex network weight matrix data；
Louvain algorithm execution module: according to complex network weight matrix data, traffic zone is carried out using Louvain algorithm
It divides, by the modularity function of setting, finds the division methods so that when modularity maximum repeatedly, finally utilize complex network
Groups of building are classified as different traffic zones by weight matrix data；
Cell division result display module: obtained different traffic zones are visualized and is marked.
2. the traffic zone dividing system according to claim 1 based on Louvain algorithm, it is characterised in that: described to build
Build spatial topotaxy discrimination module the specific implementation process is as follows:
After obtaining initial groups of building data, judges whether there is even frontier juncture system in a network between two groups of building, differentiate process
It is as follows:
(1) judge the physical distance between two groups of building；
(2) judge whether separated by the high road of category of roads between two groups of building；
(3) judge whether blocked by other groups of building between two groups of building, measure coverage extent；
(4) according to above-mentioned 3 differentiate as a result, if the physical distance between two groups of building less than 200 meters, no expressway, loop
Segmentation, is blocked without other groups of building, then judges that the two groups of building have even frontier juncture system in a network；
(5) step (4) are executed repeatedly, generates initial groups of building spatial topotaxy data.
3. the traffic zone dividing system according to claim 1 based on Louvain algorithm, it is characterised in that: described to build
It builds group's segmentation category of roads computing module and realizes that process is as follows:
(1) initial groups of building spatial topotaxy data and city road network data are read；
(2) according to city road network data, all roads divided between two groups of building are obtained；
(3) to the road divided between two groups of building, the category of roads of every road is successively calculated, method particularly includes: first sentence
Whether disconnected selected section has divided the two groups of building, if groups of building B and groups of building C is separated by section 1 and section 2, then sets and builds
The central point for building crowd B is Pb, the central point of groups of building C is Pc, line segment LbcFor point PbWith point PcLine, LbcWith section 1 and section
2 generate intersection point, so that section 1 and section 2 become the separation section of groups of building B and C, then judge section 1 and line segment LbcWhether
Intersection, groups of building B and groups of building C have been divided in section 1 if intersection；If the two groups of building, root have been divided in selected section
According to urban architecture distribution density and city road network road length, path search index table is established to groups of building, retrieves city road network number
Then section concordance list in inquires the category of roads in selected section；
(4) to category of roads calculated in step (3) take out highest level as between two groups of building divide road etc.
Grade.
4. the traffic zone dividing system according to claim 1 based on Louvain algorithm, it is characterised in that: described to build
It builds group topology distance computing module and realizes that process is as follows:
(1) initial groups of building spatial topotaxy data are read；
(2) the field groups of building of each groups of building are calculated, method particularly includes: for groups of building A, defining its neighborhood is
Neighbors={ N1,N2…Nk, wherein SPD (A, Ni)<SPDset, SPD (A, Ni) indicate groups of building A and NiSpace physics away from
From SPDsetFor the threshold value of setting, dynamic adjustment is carried out according to region site coverage；
(3) the field groups of building of each groups of building are calculated, method particularly includes: for groups of building A, successively calculates other groups of building and arrive
The space physics distance of groups of building A, if the physical distance is less than the threshold value of setting, which belongs to the field building of A
Set, wherein threshold value carries out dynamic adjustment according to region site coverage；
(4) directional similarity matrix S of all groups of building relative to groups of building A in the field groups of building set of groups of building is calculatedkk,
Wherein k is the size of neighborhood；
(5) similarity matrix in processing step (4), to each groups of building in similarity matrix, taking out opposite groups of building A has
The groups of building set of close direction similarity, specific method are the directional similarity square according to groups of building in (4) for groups of building A
Battle array Sij, wherein i represents i-th of groups of building, and j represents j-th of groups of building, if meeting Sij>Sset, wherein SsetFor the threshold of setting
Value, when S is greater than SsetWhen, it is believed that two groups of building are in the same direction relative to groups of building A；
(6) for groups of building set obtained in step (5), by the distance-taxis of wherein all groups of building and groups of building A, sequence
The topology distance of first groups of building and groups of building A are 1 afterwards, are secondly 2, successively plus 1；
(7) groups of building topology distance data are generated according to the topology distance being calculated.
5. the traffic zone dividing system according to claim 4 based on Louvain algorithm, it is characterised in that: the Sset
It is set as 0.75.
6. the traffic zone dividing system according to claim 1 based on Louvain algorithm, it is characterised in that: described multiple
Miscellaneous network edge weight computing module realizes that process is as follows:
(1) groups of building segmentation category of roads data and groups of building topology distance data are read；
(2) space topological distance and road isolation information are combined, the frontier juncture system, company of nodes, nodes v are establishediWith
vjBetween similarity and node represented by land used attribute, road barrier, space length and topology between groups of building away from
It is related from all, node v is calculated in conjunction with land used attribute, road network and the geographical location information of groups of buildingiWith vjBetween power
Value, then the weight between two nodes is expressed as:
W(vi,vj)=F (Lij,SPDij,Rij,TDij)
LijIndicate node viWith vjBetween land used attributes similarity, SPDijIndicate node viWith vjBetween space physics distance,
RijIndicate node viWith vjBetween divided roadway grade, TDijIndicate node viWith vjBetween topology distance, F indicate similarity
Calculate function；
(3) weight assignment is carried out according to company side of the similarity being calculated to groups of building complex network, generates complex network power
Value matrix data.
7. the traffic zone dividing system according to claim 1 based on Louvain algorithm, it is characterised in that: described
Louvain algorithm execution module realizes that process is as follows:
(1) complex network weight matrix data are read, regard all groups of building as a complex network；
(2) each groups of building of complex network are seen as an independent traffic zone, at this time number of cells and groups of building number
It is identical；
(3) it for groups of building A each in current network, successively attempts for groups of building A to be assigned to small where its neighbours' groups of building
Then area calculates the preceding modularity changes delta Q with network after distribution of distribution according to above-mentioned calculation method, finds Δ after the completion of traversal
Distribution method when Q maximum is denoted as max Δ Q；
(4) if max Δ Q > 0 in step (3), and plot area when Δ Q maximum where that neighbours' groups of building is less than
0.5 square kilometre, then otherwise the node division is remained unchanged to the cell；
(5) step (3) and step (4) are repeated, until the affiliated subdistrict of all groups of building is no longer changed；
(6) complex network is compressed, all groups of building in the same cell are compressed into a new groups of building, minizone
Side right be converted into the side right weight between new building group again；
(7) result that step (6) obtains is transferred to step (2), continues to repeat step (3) and step (4), until entire complex web
The modularity of network no longer changes, and all groups of building are divided into different traffic zones.
8. the traffic zone dividing system according to claim 1 based on Louvain algorithm, it is characterised in that: described small
Division result display module realizes that process is as follows:
(1) according to Louvain algorithm execution module obtain as a result, obtaining cell number belonging to each groups of building；
(2) map datum is obtained, and will be in groups of building Data Matching to map；
(3) it is based on step (1) and step (2), is drawn accordingly according to different cell numbers for different groups of building on map
Color, complete show.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811186046.1A CN109272170B (en) | 2018-10-11 | 2018-10-11 | A kind of traffic zone dividing system based on Louvain algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811186046.1A CN109272170B (en) | 2018-10-11 | 2018-10-11 | A kind of traffic zone dividing system based on Louvain algorithm |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109272170A CN109272170A (en) | 2019-01-25 |
| CN109272170B true CN109272170B (en) | 2019-08-30 |
Family
ID=65195808
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811186046.1A Active CN109272170B (en) | 2018-10-11 | 2018-10-11 | A kind of traffic zone dividing system based on Louvain algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109272170B (en) |
Families Citing this family (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109841063A (en) * | 2019-04-01 | 2019-06-04 | 武汉凤翥科技有限公司 | A kind of Division approach of traffic control work zone |
| CN110111567B (en) * | 2019-04-23 | 2021-05-18 | 刘畅 | Traffic control subarea division method and system based on modularity evaluation |
| CN110309248B (en) * | 2019-06-26 | 2022-04-29 | 东南大学 | Method for automatically dividing traffic cells of traffic road network based on Voronoi diagram |
| CN111738624B (en) * | 2020-07-20 | 2020-11-13 | 武大吉奥信息技术有限公司 | Region division method and system for solving supply and demand relation balance |
| CN112100723B (en) * | 2020-09-16 | 2024-01-30 | 浙大城市学院 | College teaching building plane learning space connection diversity assessment method and system |
| CN112214567A (en) * | 2020-10-28 | 2021-01-12 | 兰州交通大学 | Urban space structure discovery design scheme under support of complex graph theory |
| CN113284030B (en) * | 2021-06-28 | 2023-05-23 | 南京信息工程大学 | Urban traffic network community division method |
| CN113553357B (en) * | 2021-07-26 | 2022-11-11 | 中国电子科技集团公司第五十四研究所 | HW-Louvain-based urban public transport network divisible spatial community detection method |
| CN114386536B (en) * | 2022-03-22 | 2022-07-01 | 腾讯科技（深圳）有限公司 | Region determination method, device, computing equipment and storage medium |
| CN115277116B (en) * | 2022-07-06 | 2024-02-02 | 中能电力科技开发有限公司 | Network isolation method and device, storage medium and electronic equipment |
| CN116681215B (en) * | 2023-07-28 | 2023-11-03 | 四川元祉智慧科技有限公司 | Urban gas pipeline gridding minimum management unit division method |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105869401A (en) * | 2016-05-12 | 2016-08-17 | 华南理工大学 | Road network dynamic partitioning method based on different crowdedness degrees |
| CN108242203A (en) * | 2016-12-23 | 2018-07-03 | 中兴通讯股份有限公司 | A kind of region partitioning method and device |
| CN108280550A (en) * | 2018-01-30 | 2018-07-13 | 杭州电子科技大学 | A kind of visual analysis method that relatively public bicycles website community divides |
| CN108364465A (en) * | 2018-02-09 | 2018-08-03 | 太原理工大学 | The method for dynamically partitioning of city road network control work zone based on macroscopical parent map |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105225503B (en) * | 2015-11-09 | 2017-10-24 | 中山大学 | Traffic control sub-district optimizes and self-adapting regulation method |
| CN107808518B (en) * | 2017-10-26 | 2020-02-18 | 东南大学 | Traffic cell classification method based on multi-path spectral clustering theory |
| CN108400889A (en) * | 2018-01-30 | 2018-08-14 | 重庆邮电大学 | A kind of community discovery method based on suboptimization |
-
2018
- 2018-10-11 CN CN201811186046.1A patent/CN109272170B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105869401A (en) * | 2016-05-12 | 2016-08-17 | 华南理工大学 | Road network dynamic partitioning method based on different crowdedness degrees |
| CN108242203A (en) * | 2016-12-23 | 2018-07-03 | 中兴通讯股份有限公司 | A kind of region partitioning method and device |
| CN108280550A (en) * | 2018-01-30 | 2018-07-13 | 杭州电子科技大学 | A kind of visual analysis method that relatively public bicycles website community divides |
| CN108364465A (en) * | 2018-02-09 | 2018-08-03 | 太原理工大学 | The method for dynamically partitioning of city road network control work zone based on macroscopical parent map |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109272170A (en) | 2019-01-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109272170B (en) | A kind of traffic zone dividing system based on Louvain algorithm | |
| Wang et al. | Analysing the spatial configuration of urban bus networks based on the geospatial network analysis method | |
| Ding et al. | Detecting the urban traffic network structure dynamics through the growth and analysis of multi-layer networks | |
| CN106780089B (en) | Permanent basic farmland planning method based on neural network cellular automaton model | |
| CN109949574B (en) | Data-driven urban road network traffic cell multistage division method | |
| CN107247938A (en) | A kind of method of high-resolution remote sensing image City Building function classification | |
| CN108683448B (en) | Influence node identification method and system suitable for aviation network | |
| CN105206048A (en) | Urban resident traffic transfer mode discovery system and method based on urban traffic OD data | |
| CN110232398A (en) | A kind of road network sub-area division and its appraisal procedure based on Canopy+Kmeans cluster | |
| CN110781267A (en) | Multi-scale space analysis and evaluation method and system based on geographical national conditions | |
| CN108090624B (en) | Urban ecological safety simulation and prediction method for improving cellular automaton | |
| CN110689187B (en) | Automatic site selection method for transformer substation based on multi-condition constraint | |
| Ai et al. | Generation of constrained network Voronoi diagram using linear tessellation and expansion method | |
| CN106845536A (en) | A kind of parallel clustering method based on image scaling | |
| CN111274700B (en) | Habitat network construction method based on vector data | |
| CN114386536B (en) | Region determination method, device, computing equipment and storage medium | |
| Zhang et al. | Basic farmland zoning and protection under spatial constraints with a particle swarm optimisation multiobjective decision model: a case study of Yicheng, China | |
| Yan et al. | Dynamic partitioning of urban traffic network sub-regions with spatiotemporal evolution of traffic flow | |
| CN111008730B (en) | Crowd concentration prediction model construction method and device based on urban space structure | |
| Ma et al. | Urban traffic signal control network automatic partitioning using laplacian eigenvectors | |
| Jiang et al. | Partitioning of urban networks with polycentric congestion pattern for traffic management policies: Identifying protected networks | |
| CN115759883B (en) | Ecological management partitioning method based on network group characteristics | |
| CN111738527A (en) | Urban traffic zone dividing method based on hot spot detection model | |
| CN115730711A (en) | Space division method, device and storage medium | |
| CN113284030B (en) | Urban traffic network community division method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |