CN110322117A - A kind of traffic control sub-area division method, apparatus, electronic equipment and storage medium - Google Patents
A kind of traffic control sub-area division method, apparatus, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN110322117A CN110322117A CN201910487157.4A CN201910487157A CN110322117A CN 110322117 A CN110322117 A CN 110322117A CN 201910487157 A CN201910487157 A CN 201910487157A CN 110322117 A CN110322117 A CN 110322117A
- Authority
- CN
- China
- Prior art keywords
- crossing
- matrix
- similarity
- crossings
- section
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 56
- 239000011159 matrix material Substances 0.000 claims abstract description 261
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 32
- 239000000203 mixture Substances 0.000 claims description 23
- 238000004891 communication Methods 0.000 claims description 19
- 238000010606 normalization Methods 0.000 claims description 10
- 238000004590 computer program Methods 0.000 claims description 9
- 230000006870 function Effects 0.000 claims description 9
- 238000010586 diagram Methods 0.000 description 14
- 238000012545 processing Methods 0.000 description 8
- 238000012986 modification Methods 0.000 description 5
- 230000004048 modification Effects 0.000 description 5
- 230000003068 static effect Effects 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 239000006185 dispersion Substances 0.000 description 2
- 230000016507 interphase Effects 0.000 description 2
- 230000002093 peripheral effect Effects 0.000 description 2
- 238000013500 data storage Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000000802 evaporation-induced self-assembly Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005498 polishing Methods 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- 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
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0637—Strategic management or analysis, e.g. setting a goal or target of an organisation; Planning actions based on goals; Analysis or evaluation of effectiveness of goals
-
- 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
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0125—Traffic data processing
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Educational Administration (AREA)
- Tourism & Hospitality (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Entrepreneurship & Innovation (AREA)
- General Business, Economics & Management (AREA)
- Development Economics (AREA)
- Marketing (AREA)
- Game Theory and Decision Science (AREA)
- Artificial Intelligence (AREA)
- General Health & Medical Sciences (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Primary Health Care (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a kind of traffic control sub-area division method, apparatus, electronic equipment and storage mediums, method includes: for any two crossing, according to two crossings geographically whether be adjacent crossing and two crossings difference in flow, determine the similarity between two crossings；Each sample is clustered using every a line in eigenvectors matrix as a sample using clustering algorithm, and determines the corresponding first mean profile coefficient of cluster result；Judge whether the first mean profile coefficient is greater than the silhouette coefficient threshold value currently saved, if not, the corresponding crossing of sample that classification each in cluster result includes is divided into a sub-district.The present invention only needs to carry out the division of traffic control sub-district based on the subflow amount in each section, it can dynamically be divided according to subflow amount, and in sub-area division, by realizing the division accurately, flexibly to traffic control sub-district by the various vector matrixs of flow and using clustering algorithm.
Description
Technical field
The present invention relates to technical field more particularly to a kind of traffic control sub-area division method, apparatus, electronic equipment and deposit
Storage media.
Background technique
Conventional traffic control work zone divides, mainly by the statistical history magnitude of traffic flow and Combining with terrain and administrative region with
And city point of interest closeness carrys out comprehensive descision, is a kind of sub-area division mode of static state, not in view of dynamic adjusts.However
It goes on a journey the today become increasingly active in urban transportation, road network traffic flow stochastic and dynamic variation characteristic is further significant, through after a period of time
After need again to be adjusted sub-district scale, guarantee to be divided into same son always by the stronger intersection of correlation or for crossing
Area.Road network generally comprises hundreds and thousands of crossings, if repartitioning sub-district, will generate great workload.
By analyze above it is found that at present traffic control sub-area division method systematicness, it is scientific, in terms of also
There are some problems, lack the modeling method of the theoretical frame and science at system, excessive dependence engineering experience, and be somebody's turn to do
Division mode be it is static, not consider flow dynamic change, cause divide sub-district be inaccurate.
Summary of the invention
The embodiment of the invention provides a kind of traffic control sub-area division method, apparatus, electronic equipment and storage mediums, use
By solve be in a manner of sub-area division in the prior art it is static, do not consider the dynamic change of flow, lead to the sub-district divided not
Accurate problem.
The embodiment of the invention provides a kind of traffic control sub-area division methods, which comprises
For any crossing in the road network topology structural relation obtained in advance, there is connection relationship according to the crossing
The subflow amount in each section determines the flow at the crossing, and the road between the adjacent crossing of any two of them constitutes section；
It whether is geographically adjacent crossing and this two according to two crossings for any two crossing
The difference in flow at crossing determines the similarity between two crossings；And similarity neighbour is determined according to the similarity at any two crossing
Matrix is connect, wherein each row and column is each crossing in the matrix；
According to the similarity adjacency matrix and the diagonal matrix of the similarity adjacency matrix, Laplce's square is determined
Battle array and the corresponding feature vector of the Laplacian Matrix；According to the size of the corresponding characteristic value of each feature vector, choose special
The corresponding target feature vector composition characteristic vector matrix of characteristic value of the lesser setting quantity of value indicative, wherein each target signature
Vector is described eigenvector matrix column vector；
Each sample is carried out using every a line in described eigenvector matrix as a sample using clustering algorithm
Cluster obtains cluster result, and determines the corresponding first mean profile coefficient of the cluster result；
Judge whether the first mean profile coefficient is greater than the silhouette coefficient threshold value currently saved, if not, will be described
The corresponding crossing of the sample that each classification includes in cluster result is divided into a sub-district.
Further, described to be directed to any two crossing, it whether is geographically adjacent according to two crossings
The difference in flow at crossing and two crossings determines that the similarity between two crossings includes:
Using following formula:
Determine this two
Similarity between a crossing, w (xi,xj) be crossing i and crossing j similarity, wherein xiRepresent the flow of crossing i, xjRepresent road
The flow of mouth j, σ are parameter preset, and exp is the exponential function using natural constant e the bottom of as, | | xi-xj| | for crossing i flow with
The difference of the flow of crossing j.
Further, the basis has the subflow amount in each section of connection relationship with the crossing, determines the stream at the crossing
Before amount, the method also includes:
Judge whether to obtain the subflow amount in all sections；
If not, determining two crossings for constituting the target road section for the target road section of subflow amount missing, being directed to respectively
Each crossing determines the subflow amount in each section of crossing connection；According to each crossing connection each section subflow amount,
Determine the subflow amount of the target road section.
Further, the subflow amount in each section according to the connection of each crossing, determines the flow of the target road section
Include:
Determine the absolute value of the sub- flow difference in wantonly one or two of section of two crossings connection；
By the average value of the subflow amount in corresponding two sections of the minimum value of the absolute value, the son of the target road section is determined
Flow.
Further, described according to the similarity adjacency matrix and the diagonal matrix of the similarity adjacency matrix, really
After determining Laplacian Matrix, before determining the corresponding feature vector of the Laplacian Matrix, further includes:
Using following formula:Laplacian Matrix L after determining standardizationsys, wherein D is institute
Similarity adjacency matrix is stated, L is the Laplacian Matrix,For the evolution of the inverse matrix of D.
Further, it is determined that the diagonal matrix of the similarity adjacency matrix includes:
Using following formula:It determines in the diagonal matrix positioned at the element of diagonal positions
It is worth, wherein wijCorresponding value is arranged for the i-th row jth of the similarity adjacency matrix, the diagonal matrix is diN*n pairs of composition
Angular moment battle array.
Further, the size according to the corresponding characteristic value of each feature vector, selected characteristic are worth lesser setting
After the corresponding target feature vector composition characteristic vector matrix of the characteristic value of quantity, the method also includes:
Using following formula:Described eigenvector matrix is normalized, is returned
One change after eigenvectors matrix, wherein u_normijRefer to u in described eigenvector matrixijNumerical value after normalization.
Further, if the mean profile coefficient is greater than preset initial value, the method also includes:
The silhouette coefficient threshold value currently saved is updated according to the first mean profile coefficient, and by sub-district quantity
Increase default value, clustering algorithm is used again, using every a line in described eigenvector matrix as a sample, to each
Sample is clustered, and obtains cluster result, and determines the corresponding second mean profile coefficient of the cluster result, judges described the
Whether two mean profile coefficients are greater than the updated silhouette coefficient threshold value currently saved, until the second mean profile coefficient
No more than the updated silhouette coefficient threshold value currently saved.
The embodiment of the invention provides a kind of traffic control sub-area division device, described device includes:
First determining module, for for any crossing in the road network topology structural relation that obtains in advance, according to this
Crossing has the subflow amount in each section of connection relationship, determines the flow at the crossing, between the adjacent crossing of any two of them
Road constitute section；
Whether the second determining module is geographically phase according to two crossings for being directed to any two crossing
The difference in flow at adjacent crossing and two crossings, determines the similarity between two crossings；And according to the phase at any two crossing
Similarity adjacency matrix is determined like spending, and wherein each row and column is each crossing in the matrix；
Module is chosen, for the diagonal matrix according to the similarity adjacency matrix and the similarity adjacency matrix, really
Determine Laplacian Matrix and the corresponding feature vector of the Laplacian Matrix；According to the corresponding characteristic value of each feature vector
Size, selected characteristic be worth it is lesser setting quantity the corresponding target feature vector composition characteristic vector matrix of characteristic value,
In each target feature vector be described eigenvector matrix column vector；
Third determining module, for using clustering algorithm, using every a line in described eigenvector matrix as a sample
This, clusters each sample, obtains cluster result, and determines the corresponding first mean profile coefficient of the cluster result；
Judgment module, for judging whether the first mean profile coefficient is greater than the silhouette coefficient threshold value currently saved,
If not, the corresponding crossing of sample that classification each in the cluster result includes is divided into a sub-district.
The embodiment of the invention provides a kind of electronic equipment, comprising: processor, communication interface, memory and communication bus,
Wherein, processor, communication interface, memory complete mutual communication by communication bus；
It is stored with computer program in the memory, when described program is executed by the processor, so that the place
Manage the step of device executes any of the above-described the method.
The embodiment of the invention provides a kind of computer readable storage medium, it is stored with the meter that can be executed by electronic equipment
Calculation machine program, when described program is run on the electronic equipment, so that the electronic equipment executes described in any of the above-described
The step of method.
The embodiment of the invention provides a kind of traffic control sub-area division method, apparatus, electronic equipment and storage medium, institutes
The method of stating includes: according to the crossing there is connection to close for any crossing in the road network topology structural relation obtained in advance
The subflow amount in each section of system, determines the flow at the crossing, and the road between the adjacent crossing of any two of them constitutes section；
For any two crossing, according to two crossings geographically whether be adjacent crossing and two crossings stream
It is poor to measure, and determines the similarity between two crossings；And similarity adjacency matrix is determined according to the similarity at any two crossing,
In in the matrix each row and column be each crossing；According to the similarity adjacency matrix and the similarity adjacency matrix to angular moment
Battle array, determines Laplacian Matrix and the corresponding feature vector of the Laplacian Matrix；According to the corresponding spy of each feature vector
The size of value indicative, selected characteristic are worth the corresponding target feature vector composition characteristic moment of a vector of characteristic value of lesser setting quantity
Battle array, wherein each target feature vector is described eigenvector matrix column vector；Using clustering algorithm, by described eigenvector
Every a line in matrix clusters each sample as a sample, obtains cluster result, and determine the cluster result
Corresponding first mean profile coefficient；Judge whether the first mean profile coefficient is greater than the silhouette coefficient threshold currently saved
Value, if not, the corresponding crossing of sample that classification each in the cluster result includes is divided into a sub-district.Due at this
In inventive embodiments, it is only necessary to the division that traffic control sub-district is carried out based on the subflow amount in each section, because subflow amount exists
Different time sections may be different, so, it can dynamically be divided, and in sub-area division, be led to according to subflow amount
It crosses by the various vector matrixs of flow and using clustering algorithm, realizes the division accurately, flexibly to traffic control sub-district.
Detailed description of the invention
To describe the technical solutions in the embodiments of the present invention more clearly, make required in being described below to embodiment
Attached drawing is briefly introduced, it should be apparent that, drawings in the following description are only some embodiments of the invention, for this
For the those of ordinary skill in field, without creative efforts, it can also be obtained according to these attached drawings other
Attached drawing.
Fig. 1 is a kind of flow diagram of traffic control sub-area division method provided in an embodiment of the present invention；
Fig. 2 is a kind of schematic diagram of simple road network topology structural relation provided in an embodiment of the present invention；
Fig. 3 is a kind of flow diagram of traffic control sub-area division method provided in an embodiment of the present invention；
Fig. 4 is a kind of result schematic diagram of traffic control sub-area division device provided in an embodiment of the present invention；
Fig. 5 is a kind of electronic equipment provided in an embodiment of the present invention.
Specific embodiment
The present invention will be describe below in further detail with reference to the accompanying drawings, it is clear that described embodiment is only this
Invention a part of the embodiment, instead of all the embodiments.Based on the embodiments of the present invention, those of ordinary skill in the art exist
All other embodiment obtained under the premise of creative work is not made, shall fall within the protection scope of the present invention.
Embodiment 1:
Fig. 1 is a kind of flow diagram of traffic control sub-area division method provided in an embodiment of the present invention, the process packet
Include following steps:
S101: for any crossing in the road network topology structural relation obtained in advance, there is connection according to the crossing
The subflow amount in each section of relationship determines the flow at the crossing, and the road between the adjacent crossing of any two of them constitutes road
Section.
This method provided in an embodiment of the present invention is applied to be divided into corresponding sub-district each crossing in city road network
Scene, include crossing and section in road network topology structural relation, wherein section refers to the road between two adjacent crossings,
Crossing is also referred to as intersection, is the node connected between section.
According to real-time traffic statistics, the subflow amount in available each section, because crossing is positioned at section both ends
Node, and same crossing can connect different sections, therefore the flow at crossing can be the subflow amount in each section of its connection
In maximum stream flow or subflow amount average value, preferably, the flow at crossing for its connection each section subflow amount sum.
It can be indicated using following formula:Wherein xiRepresent the stream of crossing i
Amount, m indicate the quantity in the section of crossing i connection, w_roadijIndicate the subflow amount in j-th of section of crossing i connection.
It is exemplary for the subflow amount in each section of real-time statistics, one letter can be installed in each section in advance
Number machine, the semaphore are used to detect the subflow amount in the section where the semaphore, and subflow amount can be vehicle flowrate and/or the stream of people
Amount, preferably, the subflow amount can refer to that vehicle flowrate, vehicle flowrate refer to the vehicle number in the unit time by section.Semaphore
The subflow amount in the section that will test, is uploaded to server, when terminal needs the sub- flow information using each section, from
The subflow amount in each section needed for server downloading.
Whether S102: being directed to any two crossing, be geographically adjacent crossing according to two crossings, and should
The difference in flow at two crossings determines the similarity between two crossings；And it is determined according to the similarity at any two crossing similar
Adjacency matrix is spent, wherein each row and column corresponds to each crossing in the similarity adjacency matrix.
When two crossings are connect with the same section respectively, then it can determine that two crossings are geographically adjacent,
That is it is adjacent on position indoor for constituting two crossings in a section.For example, one kind as shown in Figure 2 is simple
The schematic diagram of road network topology structural relation, including 7 crossings, respectively crossing a, b, c, d, e, f and g, including 6 roads
Section, respectively section 1,2,3,4,5 and 6, wherein the section of crossing a, c, f, d, g connection is more than in Fig. 2 in practical situation
Shown in section, be also connected with other sections, no longer draw herein.Wherein crossing a is connected with section 1 with crossing b, then crossing a
Be geographically adjacent two crossings with crossing b, similarly, crossing b and crossing c are two adjacent crossings, crossing b with
Crossing d is two adjacent crossings, and two crossing b adjacent with crossing e crossings, crossing e and crossing f are two adjacent roads
Mouthful, crossing e and crossing g are two adjacent crossings.
In order to realize sub-area division, it is thus necessary to determine that similarity adjacency matrix, specifically according to similar between every two crossing
Degree, determines the similarity adjacency matrix.Due to a possibility that two non-conterminous crossings are divided into same sub-district very little, and phase
A possibility that adjacent crossing is divided into a sub-district is comparatively bigger, therefore, can according to whether adjacent between crossing,
Determining the similarity at two crossings, it is assumed that the value range of similarity is [0,1], and two crossings are geographically closer,
Then similarity is higher, then value is higher.For example, if two crossing direct neighbors, it may be considered that two crossing similarities compared with
Height sets 1 for its corresponding similarity, if two crossing interphases are every a crossing, the similarity at two crossings
It is set as 0.8, if two crossing interphases, every two or three crossings, the similarity at two crossings is set as 0.6, etc.
Deng.
It, in embodiments of the present invention can be according to two in order to more accurately embody the similarity between any two crossing
Whether a crossing adjacent and the difference in flow at two crossings determines the similarity at two crossings, if two crossings are in physical bit
It sets adjacent, then according to the difference in flow at two crossings, the similarity between two crossings is determined, particularly according to this two
The flow absolute value of the difference at crossing, determines the similarity between two crossings.For example, by each flow absolute value of the difference and all
The ratio of maximum value in difference in flow absolute value is as the similarity between two crossings.If two crossings are geographically
It is non-conterminous, then 0 is set by its corresponding similarity.
After having determined the similarity between any two crossing, according to all similarities, similarity adjacency matrix is determined,
Wherein in the matrix each row and column each crossing of element representation, the value of element is the flow at crossing.Also it is illustrated with Fig. 2,
Such as 1-6 corresponding subflow amount in section is respectively 20,25,14,40,15,30.Then the flow of crossing b is 20+25+14+40=
The flow of 99, crossing e are 40+15+30=85, and so on, the flow of crossing a, c, d, f, g is respectively 20,25,14,15,
30.The flow absolute value of the difference for successively acquiring crossing a and b, b and c, b and d, b and e, e and f, e and g is respectively 77,74,85,
14,70,55, wherein the maximum value in flow absolute value of the difference is 85, then successively acquires crossing a and b, b and c, b and d, b and e, e
Similarity with f, e and g is respectively1、It is non-conterminous
Any two crossing between similarity to be divided into be 0, then similarity adjacency matrix is
Wherein each column in the similarity adjacency matrix is followed successively by a, b, c, d, e, f, g respectively, and each column is successively a, b, c, d, e, f, g,
Wij is the similarity that the corresponding crossing of the i-th row arranges corresponding crossing with jth in similarity adjacency matrix.
S103: according to the similarity adjacency matrix and the diagonal matrix of the similarity adjacency matrix, La Pula is determined
This matrix and the corresponding feature vector of the Laplacian Matrix；According to the size of the corresponding characteristic value of each feature vector, choosing
The corresponding target feature vector composition characteristic vector matrix of characteristic value for taking the lesser setting quantity of characteristic value, wherein each target
Feature vector is described eigenvector matrix column vector.
After obtaining the similarity adjacency matrix, according to the similarity adjacency matrix, the similarity adjacency matrix pair is determined
The diagonal matrix answered because the similarity adjacency matrix be a n*n matrix, the diagonal matrix of the similarity adjacency matrix,
It can be the matrix that the element on the diagonal line of the similarity adjacency matrix is constituted, such as will be non-right in the similarity adjacency matrix
Element on linea angulata is all set to 0, to obtain the diagonal matrix of the similarity adjacency matrix.
After obtaining the diagonal matrix, according to the similarity adjacency matrix and the diagonal matrix, Laplacian Matrix is calculated
L uses following formula: L=D-W, determine that Laplacian Matrix L, D are diagonal matrix, W is similarity adjacency matrix.
After obtaining Laplacian Matrix, the characteristic value of the Laplacian Matrix is calculated, by characteristic value according to from small to large
Sequence is ranked up, and takes the preceding setting quantity characteristic value of sequence, and wherein this sets quantity as k, and calculates preceding k feature
The target feature vector u of value1,u2,…,uk, wherein each target feature vector is the column vector of a n*1, and k is greater than 1
And it is less than the integer of n.Wherein the process of the characteristic value of calculating matrix and feature vector is the prior art, and therefore not to repeat here.
By k target feature vector composition characteristic vector matrix U={ u obtained above1,u2,…,uk},U∈Rn*k,
Middle U is characterized vector matrix, and each target feature vector is as the matrix column vector.
S104: clustering algorithm is used, using every a line in described eigenvector matrix as a sample, to each sample
It is clustered, obtains cluster result, and determine the corresponding first mean profile coefficient of the cluster result.
After obtaining eigenvectors matrix, using every a line in eigenvectors matrix as a sample, even sample yi∈Rk
It is the vector of the i-th row of U, wherein i=1,2 ..., n cluster each sample, clustered using any clustering algorithm
As a result.Wherein the clustering algorithm can for it is following any one: K-means algorithm, mean shift clustering algorithm, DBSCAN cluster
Algorithm etc..
Preferably, in embodiments of the present invention, it will be new using Agglomerative Clustering hierarchical clustering algorithm
Sample point Y={ y1,y2,…,ynCluster cluster C1,C2,…,CL, wherein L is divided sub-district quantity, wherein at the beginning of the quantity
The default value of beginning can be set to 2.
After obtaining cluster result, according to the cluster result, the corresponding mean profile coefficient of the cluster result is determined.Example
Such as, for the sample i in certain cluster, the average distance a of calculating sample i to same other samples of clusteri。aiIt is smaller, illustrate that sample i is got over
It should be clustered the cluster, by aiReferred to as dissmilarity degree in the cluster of sample i；Calculate sample i to other clusters all samples put down
Distance bi, biIt is bigger, illustrate that sample i is more not belonging to other clusters, by biReferred to as dissmilarity degree between the cluster of sample i.According to sample i
Cluster in dissmilarity degree aiThe dissmilarity degree b between clusteri, define the silhouette coefficient S of sample ii, whereinOr Then, according to the silhouette coefficient of each sample, according to following formula: S=mean { si,
I=1,2 ..., n }, determine the first mean profile coefficient.For the first mean profile coefficient, codomain is distributed as [- 1,1],
Therefore the initial value of the silhouette coefficient threshold value currently saved can be set to -1.Meanwhile can be seen that by above-mentioned formula, S is closer
In 1, classifying quality is better.
S105: judging whether the first mean profile coefficient is greater than the silhouette coefficient threshold value currently saved, if not, will
The corresponding crossing of the sample that each classification includes in the cluster result is divided into a sub-district.
Determine whether current first mean profile coefficient is greater than the silhouette coefficient threshold value currently saved, if being not more than, really
Cluster result before settled is best, and the corresponding crossing of sample that classification each in the cluster result includes is divided into a son
Area.
Method in the embodiment of the present invention by calculating the first mean profile coefficient, for automatically selecting control of division
The number in area；This method determines to divide the superiority and inferiority of number, phase by calculating the mean profile coefficient of feature vector clusters result
Compared with other by way of final sub-area division result judgement, more directly effectively, and it is more efficient.
Due in embodiments of the present invention, it is only necessary to carry out drawing for traffic control sub-district based on the subflow amount in each section
Point, because subflow amount may be different in different time sections, it can dynamically be divided according to subflow amount, and
In sub-area division, by realizing accurately, flexibly to traffic control by the various vector matrixs of flow and using clustering algorithm
The division in system area.
Embodiment 2:
For the similarity being determined more accurately between two crossings, on the basis of the above embodiments, of the invention real
Apply in example, it is described to be directed to any two crossing, according to two crossings geographically whether be adjacent crossing and this two
The difference in flow at a crossing determines that the similarity between two crossings includes:
Using following formula:
Determine this two
Similarity between a crossing, w (xi,xj) be crossing i and crossing j similarity, wherein xiRepresent the flow of crossing i, xjRepresent road
The flow of mouth j, σ are parameter preset, and exp is the exponential function using natural constant e the bottom of as, | | xi-xj| | for crossing i flow with
The difference of the flow of crossing j, wherein σ is parameter preset, and value is bigger, and dispersion ratio is bigger, and dispersion ratio is bigger in the embodiment of the present invention
In be presented as divide sub-district more disperse, the numerical values recited of σ is generally between 0~10.
If two crossings are geographically adjacent crossings, when determining the similarity between two crossings,
If only using the absolute value of the difference of the maximum value in the difference in flow and all difference in flow between two crossings as this two
Similarity between crossing, it is determined that similarity value may it is larger, be not easy to subsequent calculating.It is therefore advantageous to, the present invention is real
Applying example uses above-mentioned formula to determine the similarity between two crossings.
Embodiment 3:
In order to keep the subregion divided more reasonable, on the basis of the various embodiments described above, in embodiments of the present invention, institute
The subflow amount according to each section with the crossing with connection relationship is stated, before the flow for determining the crossing, the method is also wrapped
It includes:
Judge whether to obtain the subflow amount in all sections；
If not, determining two crossings for constituting the target road section for the target road section of subflow amount missing, being directed to respectively
Each crossing determines the subflow amount in each section of crossing connection；According to each crossing connection each section subflow amount,
Determine the subflow amount of the target road section.
When obtaining the subflow amount in each section, it is understood that there may be the semaphore where certain a road section generates failure, or at this
When subflow amount to the server of semaphore uploading detection, network speed is slower, therefore when obtaining the subflow amount in each section, cannot
The subflow amount in the section is timely obtained, that is, obtains the subflow amount less than the section.In order to accurately calculate the stream at each crossing
Amount, it is therefore desirable to get the subflow amount in all sections.
In embodiments of the present invention, if the subflow amount of certain target road section lacks, the crossing at the target road section both ends is found,
According to the subflow amount in each section of two crossings connection, the subflow amount of the target road section is determined.A kind of possible embodiment party
Formula be using the maximum value in the subflow amount in each section as the subflow amount of the target road section, be also illustrated with Fig. 3, wherein
The subflow amount in section 4 lacks, then by comparing the subflow amount in each section connected with crossing e crossing b, wherein section 1,2,3,
5,6 corresponding subflow amounts are respectively 10,20,25,35,30,28, maximum value 35, then the subflow amount in section 4 is 35；It is another
Possible embodiment is, using the average value of the subflow amount in each section as the subflow amount of the target road section, also to be carried out with Fig. 3
Illustrate, wherein the subflow amount missing in section 4, then calculate the average value of the subflow amount in each section that crossing b is connected with crossing e,
Wherein the corresponding subflow amount in section 1,2,3,5,6 is respectively 10,20,25,35,30,28, then the subflow amount in section 4 is 25.
If the target road section of subflow amount missing is relatively more, in the subflow amount for determining each target road section, can first look into
The target road section that the subflow amount positioned at each of the crossing connection at the target road section both ends other sections does not lack is looked for, is first counted
The subflow amount of the target road section is calculated, successively the subflow amount of each target road section of polishing.
Embodiment 4:
In order to improve the robustness for carrying out sub-area division, on the basis of the various embodiments described above, in embodiments of the present invention,
The subflow amount in each section according to the connection of each crossing, determines that the flow of the target road section includes:
Determine the absolute value of the sub- flow difference in wantonly one or two of section of two crossings connection；
By the average value of the subflow amount in corresponding two sections of the minimum value of the absolute value, the son of the target road section is determined
Flow.
In order to be determined more accurately Deleting mutants flow target road section subflow amount, in embodiments of the present invention, first
The absolute value for determining the difference of the subflow amount in wantonly one or two of section of two crossings connection, obtains the smallest two roads of absolute value
Section, using the average value of the subflow amount in two sections as the subflow amount of the target road section.
Specifically, if the data on flows of certain target road section lacks, two crossings of the target road section be respectively the crossing A and
Minimal difference between the subflow amount in all sections of the crossing B, the subflow amount in all sections for asking the crossing A to connect and the connection of the crossing B:
min{‖Ai-Bj‖|AiEach section subflow amount of the crossing ∈ A connection, BjEach section subflow amount of the crossing ∈ B connection }
Corresponding minimal difference is section A_min respectivelyiWith section B_minj, and calculate section A_miniSubflow amount with
Section B_minjSubflow amount average value:Wherein w_road is the mesh acquired
Mark the subflow amount in section.Although the subflow amount is not the true subflow amount of the target road section, the target road section acquired
Subflow amount, the case where can also preferably reacting the target road section.
Embodiment 5:
Operation is carried out in order to facilitate terminal, on the basis of the various embodiments described above, in embodiments of the present invention, the basis
The diagonal matrix of the similarity adjacency matrix and the similarity adjacency matrix, after determining Laplacian Matrix, in determination
Before the corresponding feature vector of the Laplacian Matrix, further includes:
Using following formula:Laplacian Matrix L after determining standardizationsys, wherein D is institute
Similarity adjacency matrix is stated, L is the Laplacian Matrix,For the evolution of the inverse matrix of D.
Specifically,Refer to the inverse matrix for taking similarity adjacency matrix, inverse matrix is opened into quadratic power operation.
Operation is carried out in order to facilitate terminal, after carrying out operation to each matrix, is required to the obtained square after operation
Battle array is standardized.It, can be using standardization in subsequent operating process after determining the Laplacian Matrix after standardization
Laplacian Matrix afterwards.
In order to preferably determine that diagonal matrix in embodiments of the present invention, determines institute on the basis of the various embodiments described above
The diagonal matrix for stating similarity adjacency matrix includes:
Using following formula:It determines in the diagonal matrix positioned at the element of diagonal positions
It is worth, wherein wijCorresponding value is arranged for the i-th row jth of the similarity adjacency matrix, the diagonal matrix is diN*n pairs of composition
Angular moment battle array.
The sum of the numerical value of every a line of similarity adjacency matrix W is sought first, that is, uses following formula:
Wherein wijFor the similarity between crossing i and crossing j, diagonal matrix D is diThe n*n diagonal matrix of composition, wherein n is all
The quantity at crossing.
Operation is carried out in order to facilitate terminal, on the basis of the various embodiments described above, in embodiments of the present invention, the basis
The size of the corresponding characteristic value of each feature vector, selected characteristic are worth the corresponding target signature of characteristic value of lesser setting quantity
After vector forms eigenvectors matrix, the method also includes:
Using following formula:Described eigenvector matrix is normalized, is returned
One change after eigenvectors matrix, wherein u_normijRefer to u in described eigenvector matrixijNumerical value after normalization.
After obtaining feature vector, and clustering algorithm is being used, using every a line in this feature vector matrix as one
Sample before clustering to each sample, carries out subsequent arithmetic according to this feature vector for the ease of terminal, uses following public affairs
Formula normalizes this feature vector matrix U, the eigenvectors matrix U_norm ∈ R after obtaining new normalizationn*k,Eigenvectors matrix U_norm after normalization is u_normijThe n*k matrix of composition.In determination
After eigenvectors matrix after normalization, subsequent operating process can be using the eigenvectors matrix after normalization.
Embodiment 6:
In order to accurately realize the division to control work zone, on the basis of the various embodiments described above, in the embodiment of the present invention
In, if the first mean profile coefficient is greater than the silhouette coefficient threshold value currently saved, the method also includes:
The silhouette coefficient threshold value currently saved is updated according to the first mean profile coefficient, and by sub-district quantity
Increase default value, clustering algorithm is used again, using every a line in described eigenvector matrix as a sample, to each
Sample is clustered, and obtains cluster result, and determines the corresponding second mean profile coefficient of the cluster result, judges described the
Whether two mean profile coefficients are greater than the updated silhouette coefficient threshold value currently saved, until the second mean profile coefficient
No more than the updated silhouette coefficient threshold value currently saved.
Clustering algorithm, which needs to preassign, needs to divide how many a sub-districts, that is, a sub-district quantity is preset, using poly-
When class algorithm clusters each sample, then it can be divided into the sub-district of the quantity, that is, be divided into the cluster of the quantity, or point
Class quantity is the quantity, and wherein the sub-district quantity can be set in advance as a default value, such as the default value is 2, should when using
When clustering algorithm clusters each sample, available 2 sub-districts；If the default value is 3, using the clustering algorithm pair
When each sample is clustered, available 3 sub-districts.
After the corresponding first mean profile coefficient of cluster result has been determined, determine whether the first mean profile coefficient is greater than
The silhouette coefficient threshold value currently saved, if more than current class result C is then saved1,C2,…,Ck, while using current average wheel
Wide coefficient is updated the silhouette coefficient threshold value currently saved, and is updated to sub-district quantity, in specific implementation process
In, can be sub-district quantity plus default value, the preset value can for it is following any one: 1,2,3 etc., preferably this is pre-
If numerical value is 1.Later, it is re-execute the steps S104, until the mean profile coefficient determined is not more than the profile system currently saved
Number threshold value.
For example, Fig. 3 is a kind of flow diagram of traffic control sub-area division method provided in an embodiment of the present invention, mainly
The following steps are included:
According to the topological relation of road network, the subflow amount in each section is successively filled, obtains link flow matrix W _ road.
If the subflow amount in certain section lacks, using the son of the target road section of the method completion Deleting mutants flow of the description of above-described embodiment 3
Flow.It whether is geographically adjacent crossing and two crossings according to two crossings for any two crossing
Difference in flow, determine the similarity between two crossings；And the adjacent square of similarity is determined according to the similarity at any two crossing
Battle array.According to the similarity adjacency matrix and the diagonal matrix of the similarity adjacency matrix, Laplacian Matrix is determined, and use
The methodological standardization of the description of above-described embodiment 5 Laplacian Matrix.Calculate the corresponding feature vector of Laplacian Matrix；According to
The size of the corresponding characteristic value of each feature vector, selected characteristic are worth the corresponding target signature of characteristic value of lesser setting quantity
Vector forms eigenvectors matrix, later, carries out normalizing to this feature vector matrix using the method that above-described embodiment 5 describes
Change.Using every a line in this feature vector matrix as a sample, using AgglomerativeClustering hierarchical clustering
Algorithm clusters each sample, obtains cluster result, and determines the corresponding first mean profile coefficient of the cluster result.Sentence
Whether the first mean profile coefficient that breaks is greater than the silhouette coefficient threshold value currently saved, if not, by each in the cluster result
The corresponding crossing of the sample that classification includes is divided into a sub-district, if it is, saving current cluster result, while using current
First mean profile coefficient is updated the silhouette coefficient threshold value currently saved, and sub-district quantity adds default value, compared with
The good default value be 1, then again using AgglomerativeClustering hierarchical clustering algorithm to each sample into
Row cluster, and the step of after progress, therefore not to repeat here, until the mean profile coefficient obtained according to cluster result is not more than
The silhouette coefficient threshold value currently saved.
Embodiment 7:
On the basis of the various embodiments described above, in embodiments of the present invention, a kind of short message text sorter is additionally provided,
Fig. 4 is a kind of structural schematic diagram of traffic control sub-area division device provided in an embodiment of the present invention, which includes:
First determining module 401, for for any crossing in the road network topology structural relation that obtains in advance, according to
The crossing has the subflow amount in each section of connection relationship, determines the flow at the crossing, the adjacent crossing of any two of them it
Between road constitute section；
Second determining module 402, for geographically whether being according to two crossings for any two crossing
The difference in flow at adjacent crossing and two crossings determines the similarity between two crossings；And according to any two crossing
Similarity determines similarity adjacency matrix, and wherein each row and column is each crossing in the matrix；
Module 403 is chosen, for the diagonal matrix according to the similarity adjacency matrix and the similarity adjacency matrix,
Determine Laplacian Matrix and the corresponding feature vector of the Laplacian Matrix；According to the corresponding feature of each feature vector
The size of value, selected characteristic are worth the corresponding target feature vector composition characteristic vector matrix of characteristic value of lesser setting quantity,
Wherein each target feature vector is described eigenvector matrix column vector；
Third determining module 404, for using clustering algorithm, using every a line in described eigenvector matrix as one
Sample clusters each sample, obtains cluster result, and determine cluster result corresponding first mean profile system
Number；
Judgment module 405, for judging whether the first mean profile coefficient is greater than the silhouette coefficient threshold currently saved
Value, if not, the corresponding crossing of sample that classification each in the cluster result includes is divided into a sub-district.
Further, the second determining module 402 is specifically used for using following formula:
Determine this two
Similarity between a crossing, w (xi,xj) be crossing i and crossing j similarity, wherein xiRepresent the flow of crossing i, xjRepresent road
The flow of mouth j, σ are parameter preset, and exp is the exponential function using natural constant e the bottom of as, | | xi-xj| | for crossing i flow with
The difference of the flow of crossing j.
Further, first determining module 401 is also used to have each of connection relationship in the basis and the crossing
The subflow amount in section before the flow for determining the crossing, judges whether the subflow amount for obtaining all sections；
If not, determining two crossings for constituting the target road section for the target road section of subflow amount missing, being directed to respectively
Each crossing determines the subflow amount in each section of crossing connection；According to each crossing connection each section subflow amount,
Determine the subflow amount of the target road section.
Further, first determining module 401, specifically for determining wantonly one or two of section of two crossings connection
The absolute value of sub- flow difference；
By the average value of the subflow amount in corresponding two sections of the minimum value of the absolute value, the son of the target road section is determined
Flow.
Further, the selection module 403 is also used to described according to the similarity adjacency matrix and described similar
Spend adjacency matrix diagonal matrix, after determining Laplacian Matrix, determine the corresponding feature of the Laplacian Matrix to
Before amount, using following formula:Laplacian Matrix L after determining standardizationsys, wherein D is institute
Similarity adjacency matrix is stated, L is the Laplacian Matrix,For the evolution of the inverse matrix of D.
Further, the selection module 403 is specifically used for using following formula:Determine institute
The value for being located at the element of diagonal positions in diagonal matrix is stated, wherein wijFor the i-th row jth column pair of the similarity adjacency matrix
The value answered, the diagonal matrix are diThe n*n diagonal matrix of composition.
Further, the selection module 403 is also used to described according to the big of the corresponding characteristic value of each feature vector
Small, selected characteristic is worth after the corresponding target feature vector composition characteristic vector matrix of characteristic value of lesser setting quantity, adopts
With following formula:Described eigenvector matrix is normalized, the spy after being normalized
Vector matrix is levied, wherein u_normijRefer to u in described eigenvector matrixijNumerical value after normalization.
Further, the judgment module 405 currently saves if being also used to the first mean profile coefficient and being greater than
Silhouette coefficient threshold value is updated the silhouette coefficient threshold value currently saved according to the first mean profile coefficient, and will be sub
Area's quantity increases default value, uses clustering algorithm again, using every a line in described eigenvector matrix as a sample,
Each sample is clustered, cluster result is obtained, and determines the corresponding second mean profile coefficient of the cluster result, judgement
Whether the second mean profile coefficient is greater than the updated silhouette coefficient threshold value currently saved, until the described second average wheel
Wide coefficient is not more than the updated silhouette coefficient threshold value currently saved.
Embodiment 8:
On the basis of the various embodiments described above, the embodiment of the invention also provides a kind of electronic equipment 500, as shown in figure 5,
It include: processor 501, communication interface 502, memory 503 and communication bus 504, wherein processor 501, communication interface 502,
Memory 503 completes mutual communication by communication bus 504；
It is stored with computer program in the memory 503, when described program is executed by the processor 501, so that
The processor 501 executes following steps:
For any crossing in the road network topology structural relation obtained in advance, there is connection relationship according to the crossing
The subflow amount in each section determines the flow at the crossing, and the road between the adjacent crossing of any two of them constitutes section；
It whether is geographically adjacent crossing and this two according to two crossings for any two crossing
The difference in flow at crossing determines the similarity between two crossings；And similarity neighbour is determined according to the similarity at any two crossing
Matrix is connect, wherein each row and column is each crossing in the matrix；
According to the similarity adjacency matrix and the diagonal matrix of the similarity adjacency matrix, Laplce's square is determined
Battle array and the corresponding feature vector of the Laplacian Matrix；According to the size of the corresponding characteristic value of each feature vector, choose special
The corresponding target feature vector composition characteristic vector matrix of characteristic value of the lesser setting quantity of value indicative, wherein each target signature
Vector is described eigenvector matrix column vector；
Each sample is carried out using every a line in described eigenvector matrix as a sample using clustering algorithm
Cluster obtains cluster result, and determines the corresponding first mean profile coefficient of the cluster result；
Judge whether the first mean profile coefficient is greater than the silhouette coefficient threshold value currently saved, if not, will be described
The corresponding crossing of the sample that each classification includes in cluster result is divided into a sub-district.
Further, described to be directed to any two crossing, it whether is geographically adjacent according to two crossings
The difference in flow at crossing and two crossings determines that the similarity between two crossings includes:
Using following formula:
Determine this two
Similarity between a crossing, w (xi,xj) be crossing i and crossing j similarity, wherein xiRepresent the flow of crossing i, xjRepresent road
The flow of mouth j, σ are parameter preset, and exp is the exponential function using natural constant e the bottom of as, | | xi-xj| | for crossing i flow with
The difference of the flow of crossing j.
Further, the basis has the subflow amount in each section of connection relationship with the crossing, determines the stream at the crossing
Before amount, the method also includes:
Judge whether to obtain the subflow amount in all sections；
If not, determining two crossings for constituting the target road section for the target road section of subflow amount missing, being directed to respectively
Each crossing determines the subflow amount in each section of crossing connection；According to each crossing connection each section subflow amount,
Determine the subflow amount of the target road section.
Further, the subflow amount in each section according to the connection of each crossing, determines the flow of the target road section
Include:
Determine the absolute value of the sub- flow difference in wantonly one or two of section of two crossings connection；
By the average value of the subflow amount in corresponding two sections of the minimum value of the absolute value, the son of the target road section is determined
Flow.
Further, described according to the similarity adjacency matrix and the diagonal matrix of the similarity adjacency matrix, really
After determining Laplacian Matrix, before determining the corresponding feature vector of the Laplacian Matrix, further includes:
Using following formula:Laplacian Matrix L after determining standardizationsys, wherein D is institute
Similarity adjacency matrix is stated, L is the Laplacian Matrix,For the evolution of the inverse matrix of D.
Further, it is determined that the diagonal matrix of the similarity adjacency matrix includes:
Using following formula:It determines in the diagonal matrix positioned at the element of diagonal positions
It is worth, wherein wijCorresponding value is arranged for the i-th row jth of the similarity adjacency matrix, the diagonal matrix is diN*n pairs of composition
Angular moment battle array.
Further, the size according to the corresponding characteristic value of each feature vector, selected characteristic are worth lesser setting
After the corresponding target feature vector composition characteristic vector matrix of the characteristic value of quantity, the method also includes:
Using following formula:Described eigenvector matrix is normalized, is returned
One change after eigenvectors matrix, wherein u_normijRefer to u in described eigenvector matrixijNumerical value after normalization.
Further, if the first mean profile coefficient is greater than the silhouette coefficient threshold value currently saved, the method
Further include:
The silhouette coefficient threshold value currently saved is updated according to the first mean profile coefficient, and by sub-district quantity
Increase default value, clustering algorithm is used again, using every a line in described eigenvector matrix as a sample, to each
Sample is clustered, and obtains cluster result, and determines the corresponding second mean profile coefficient of the cluster result, judges described the
Whether two mean profile coefficients are greater than the updated silhouette coefficient threshold value currently saved, until the second mean profile coefficient
No more than the updated silhouette coefficient threshold value currently saved.
The communication bus that above-mentioned electronic equipment is mentioned can be Peripheral Component Interconnect standard (Peripheral Component
Interconnect, PCI) bus or expanding the industrial standard structure (Extended Industry Standard
Architecture, EISA) bus etc..The communication bus can be divided into address bus, data/address bus, control bus etc..For just
It is only indicated with a thick line in expression, figure, it is not intended that an only bus or a type of bus.
Communication interface 502 is for the communication between above-mentioned electronic equipment and other equipment.
Memory may include random access memory (Random Access Memory, RAM), also may include non-easy
The property lost memory (Non-Volatile Memory, NVM), for example, at least a magnetic disk storage.Optionally, memory may be used also
To be storage device that at least one is located remotely from aforementioned processor.
Above-mentioned processor can be general processor, including central processing unit, network processing unit (Network
Processor, NP) etc.；It can also be digital command processor (Digital Signal Processing, DSP), dedicated collection
At circuit, field programmable gate array or other programmable logic device, discrete gate or transistor logic, discrete hard
Part component etc..
Embodiment 9:
On the basis of the various embodiments described above, the embodiment of the invention also provides a kind of computer readable storage medium, institutes
The computer program for being stored with and being executed in computer readable storage medium by electronic equipment is stated, when described program is in the electronics
When being run in equipment, so that the electronic equipment realizes following steps when executing:
For any crossing in the road network topology structural relation obtained in advance, there is connection relationship according to the crossing
The subflow amount in each section determines the flow at the crossing, and the road between the adjacent crossing of any two of them constitutes section；
It whether is geographically adjacent crossing and this two according to two crossings for any two crossing
The difference in flow at crossing determines the similarity between two crossings；And similarity neighbour is determined according to the similarity at any two crossing
Matrix is connect, wherein each row and column is each crossing in the matrix；
According to the similarity adjacency matrix and the diagonal matrix of the similarity adjacency matrix, Laplce's square is determined
Battle array and the corresponding feature vector of the Laplacian Matrix；According to the size of the corresponding characteristic value of each feature vector, choose special
The corresponding target feature vector composition characteristic vector matrix of characteristic value of the lesser setting quantity of value indicative, wherein each target signature
Vector is described eigenvector matrix column vector；
Each sample is carried out using every a line in described eigenvector matrix as a sample using clustering algorithm
Cluster obtains cluster result, and determines the corresponding first mean profile coefficient of the cluster result；
Judge whether the first mean profile coefficient is greater than the silhouette coefficient threshold value currently saved, if not, will be described
The corresponding crossing of the sample that each classification includes in cluster result is divided into a sub-district.
Further, described to be directed to any two crossing, it whether is geographically adjacent according to two crossings
The difference in flow at crossing and two crossings determines that the similarity between two crossings includes:
Using following formula:
Determine this two
Similarity between a crossing, w (xi,xj) be crossing i and crossing j similarity, wherein xiRepresent the flow of crossing i, xjRepresent road
The flow of mouth j, σ are parameter preset, and exp is the exponential function using natural constant e the bottom of as, | | xi-xj| | for crossing i flow with
The difference of the flow of crossing j.
Further, the basis has the subflow amount in each section of connection relationship with the crossing, determines the stream at the crossing
Before amount, the method also includes:
Judge whether to obtain the subflow amount in all sections；
If not, determining two crossings for constituting the target road section for the target road section of subflow amount missing, being directed to respectively
Each crossing determines the subflow amount in each section of crossing connection；According to each crossing connection each section subflow amount,
Determine the subflow amount of the target road section.
Further, the subflow amount in each section according to the connection of each crossing, determines the flow of the target road section
Include:
Determine the absolute value of the sub- flow difference in wantonly one or two of section of two crossings connection；
By the average value of the subflow amount in corresponding two sections of the minimum value of the absolute value, the son of the target road section is determined
Flow.
Further, described according to the similarity adjacency matrix and the diagonal matrix of the similarity adjacency matrix, really
After determining Laplacian Matrix, before determining the corresponding feature vector of the Laplacian Matrix, further includes:
Using following formula:Laplacian Matrix L after determining standardizationsys, wherein D is institute
Similarity adjacency matrix is stated, L is the Laplacian Matrix,For the evolution of the inverse matrix of D.
Further, it is determined that the diagonal matrix of the similarity adjacency matrix includes:
Using following formula:It determines in the diagonal matrix positioned at the element of diagonal positions
It is worth, wherein wijCorresponding value is arranged for the i-th row jth of the similarity adjacency matrix, the diagonal matrix is diN*n pairs of composition
Angular moment battle array.
Further, the size according to the corresponding characteristic value of each feature vector, selected characteristic are worth lesser setting
After the corresponding target feature vector composition characteristic vector matrix of the characteristic value of quantity, the method also includes:
Using following formula:Described eigenvector matrix is normalized, is returned
One change after eigenvectors matrix, wherein u_normijRefer to u in described eigenvector matrixijNumerical value after normalization.
Further, if the first mean profile coefficient is greater than the silhouette coefficient threshold value currently saved, the method
Further include:
The silhouette coefficient threshold value currently saved is updated according to the first mean profile coefficient, and by sub-district quantity
Increase default value, clustering algorithm is used again, using every a line in described eigenvector matrix as a sample, to each
Sample is clustered, and obtains cluster result, and determines the corresponding second mean profile coefficient of the cluster result, judges described the
Whether two mean profile coefficients are greater than the updated silhouette coefficient threshold value currently saved, until the second mean profile coefficient
No more than the updated silhouette coefficient threshold value currently saved.
Above-mentioned computer readable storage medium can be any usable medium that the processor in electronic equipment can access
Or data storage device, including but not limited to magnetic storage such as floppy disk, hard disk, tape, magneto-optic disk (MO) etc., optical memory
Such as CD, DVD, BD, HVD and semiconductor memory such as ROM, EPROM, EEPROM, nonvolatile memory (NAND
FLASH), solid state hard disk (SSD) etc..
The present invention be referring to according to the method for the embodiment of the present invention, the process of equipment (system) and computer program product
Figure and/or block diagram describe.It should be understood that every one stream in flowchart and/or the block diagram can be realized by computer program instructions
The combination of process and/or box in journey and/or box and flowchart and/or the block diagram.It can provide these computer programs
Instruct the processor of general purpose computer, special purpose computer, Embedded Processor or other programmable data processing devices to produce
A raw machine, so that being generated by the instruction that computer or the processor of other programmable data processing devices execute for real
The device for the function of being specified in present one or more flows of the flowchart and/or one or more blocks of the block diagram.
These computer program instructions, which may also be stored in, is able to guide computer or other programmable data processing devices with spy
Determine in the computer-readable memory that mode works, so that it includes referring to that instruction stored in the computer readable memory, which generates,
Enable the manufacture of device, the command device realize in one box of one or more flows of the flowchart and/or block diagram or
The function of being specified in multiple boxes.
These computer program instructions also can be loaded onto a computer or other programmable data processing device, so that counting
Series of operation steps are executed on calculation machine or other programmable devices to generate computer implemented processing, thus in computer or
The instruction executed on other programmable devices is provided for realizing in one or more flows of the flowchart and/or block diagram one
The step of function of being specified in a box or multiple boxes.
Although preferred embodiments of the present invention have been described, it is created once a person skilled in the art knows basic
Property concept, then additional changes and modifications may be made to these embodiments.So it includes excellent that the following claims are intended to be interpreted as
It selects embodiment and falls into all change and modification of the scope of the invention.
Obviously, various changes and modifications can be made to the invention without departing from essence of the invention by those skilled in the art
Mind and range.In this way, if these modifications and changes of the present invention belongs to the range of the claims in the present invention and its equivalent technologies
Within, then the present invention is also intended to include these modifications and variations.
Claims (10)
1. a kind of traffic control sub-area division method, which is characterized in that the described method includes:
For any crossing in the road network topology structural relation obtained in advance, according to each road with the crossing with connection relationship
The subflow amount of section, determines the flow at the crossing, and the road between the adjacent crossing of any two of them constitutes section；
It whether is geographically adjacent crossing and two crossings according to two crossings for any two crossing
Difference in flow, determine the similarity between two crossings；And the adjacent square of similarity is determined according to the similarity at any two crossing
Battle array, wherein each row and column is each crossing in the matrix；
According to the similarity adjacency matrix and the diagonal matrix of the similarity adjacency matrix, Laplacian Matrix is determined, and
The corresponding feature vector of the Laplacian Matrix；According to the size of the corresponding characteristic value of each feature vector, selected characteristic value
The corresponding target feature vector composition characteristic vector matrix of characteristic value of lesser setting quantity, wherein each target feature vector
For described eigenvector matrix column vector；
Each sample is clustered using every a line in described eigenvector matrix as a sample using clustering algorithm,
Cluster result is obtained, and determines the corresponding first mean profile coefficient of the cluster result；
Judge whether the first mean profile coefficient is greater than the silhouette coefficient threshold value currently saved, if not, by the cluster
As a result the corresponding crossing of the sample that each classification includes in is divided into a sub-district.
2. the method as described in claim 1, which is characterized in that it is described to be directed to any two crossing, existed according to two crossings
Physically whether be adjacent crossing and two crossings difference in flow, determine that the similarity between two crossings includes:
Using following formula:
Determine the similarity between two crossings, w (xi,xj) be crossing i and crossing j similarity, wherein xiRepresent crossing i's
Flow, xjThe flow of crossing j is represented, σ is parameter preset, and exp is the exponential function using natural constant e the bottom of as, | | xi-xj| | it is
The difference of the flow of the flow and crossing j of crossing i.
3. the method as described in claim 1, which is characterized in that each section of the basis with the crossing with connection relationship
Subflow amount, before the flow for determining the crossing, the method also includes:
Judge whether to obtain the subflow amount in all sections；
If not, two crossings for constituting the target road section are determined for the target road section of subflow amount missing, respectively for each
Crossing determines the subflow amount in each section of crossing connection；According to the subflow amount in each section of each crossing connection, determine
The subflow amount of the target road section.
4. method as claimed in claim 3, which is characterized in that the subflow in each section according to the connection of each crossing
Amount, determines that the flow of the target road section includes:
Determine the absolute value of the sub- flow difference in wantonly one or two of section of two crossings connection；
By the average value of the subflow amount in corresponding two sections of the minimum value of the absolute value, the subflow of the target road section is determined
Amount.
5. the method as described in claim 1, which is characterized in that described according to the similarity adjacency matrix and the similarity
The diagonal matrix of adjacency matrix after determining Laplacian Matrix, is determining the corresponding feature vector of the Laplacian Matrix
Before, further includes:
Using following formula:Laplacian Matrix L after determining standardizationsys, wherein D is the phase
Like degree adjacency matrix, L is the Laplacian Matrix,For the evolution of the inverse matrix of D.
6. the method as described in claim 1, which is characterized in that the diagonal matrix for determining the similarity adjacency matrix includes:
Using following formula:Determine the value for being located at the element of diagonal positions in the diagonal matrix,
Wherein wijCorresponding value is arranged for the i-th row jth of the similarity adjacency matrix, the diagonal matrix is diThe n*n of composition is diagonal
Matrix.
7. the method as described in claim 1, which is characterized in that described according to the big of the corresponding characteristic value of each feature vector
Small, selected characteristic is worth after the corresponding target feature vector composition characteristic vector matrix of characteristic value of lesser setting quantity, institute
State method further include:
Using following formula:Described eigenvector matrix is normalized, is normalized
Eigenvectors matrix afterwards, wherein u_normijRefer to u in described eigenvector matrixijNumerical value after normalization.
8. the method as described in claim 1, which is characterized in that if the mean profile coefficient is greater than the profile currently saved
Coefficient threshold, the method also includes:
The silhouette coefficient threshold value currently saved is updated according to the first mean profile coefficient, and sub-district quantity is increased
Default value uses clustering algorithm, using every a line in described eigenvector matrix as a sample, to each sample again
It is clustered, obtains cluster result, and determine the corresponding second mean profile coefficient of the cluster result, judge that described second is flat
Whether equal silhouette coefficient is greater than the updated silhouette coefficient threshold value currently saved, until the second mean profile coefficient is little
In the updated silhouette coefficient threshold value currently saved.
9. a kind of traffic control sub-area division device, which is characterized in that described device includes:
First determining module, for for any crossing in the road network topology structural relation that obtains in advance, according to the crossing
The subflow amount in each section with connection relationship determines the flow at the crossing, the road between the adjacent crossing of any two of them
Road constitutes section；
Whether the second determining module is geographically adjacent according to two crossings for being directed to any two crossing
The difference in flow at crossing and two crossings determines the similarity between two crossings；And according to the similarity at any two crossing
Determine similarity adjacency matrix, wherein each row and column is each crossing in the matrix；
Module is chosen, for the diagonal matrix according to the similarity adjacency matrix and the similarity adjacency matrix, determines and draws
This matrix of pula and the corresponding feature vector of the Laplacian Matrix；According to the big of the corresponding characteristic value of each feature vector
Small, selected characteristic is worth the corresponding target feature vector composition characteristic vector matrix of characteristic value of lesser setting quantity, wherein often
A target feature vector is described eigenvector matrix column vector；
Third determining module is right using every a line in described eigenvector matrix as a sample for using clustering algorithm
Each sample is clustered, and cluster result is obtained, and determines the corresponding first mean profile coefficient of the cluster result；
Judgment module, for judging whether the first mean profile coefficient is greater than the silhouette coefficient threshold value currently saved, if
It is no, the corresponding crossing of sample that classification each in the cluster result includes is divided into a sub-district.
10. a kind of electronic equipment characterized by comprising processor, communication interface, memory and communication bus, wherein place
Device, communication interface are managed, memory completes mutual communication by communication bus；
It is stored with computer program in the memory, when described program is executed by the processor, so that the processor
Perform claim requires the step of any one of 1-8 the method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910487157.4A CN110322117A (en) | 2019-06-05 | 2019-06-05 | A kind of traffic control sub-area division method, apparatus, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910487157.4A CN110322117A (en) | 2019-06-05 | 2019-06-05 | A kind of traffic control sub-area division method, apparatus, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110322117A true CN110322117A (en) | 2019-10-11 |
Family
ID=68120778
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910487157.4A Pending CN110322117A (en) | 2019-06-05 | 2019-06-05 | A kind of traffic control sub-area division method, apparatus, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110322117A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111210625A (en) * | 2020-01-10 | 2020-05-29 | 阿里巴巴集团控股有限公司 | Traffic control method and device and electronic equipment |
| CN111325979A (en) * | 2020-02-28 | 2020-06-23 | 海信集团有限公司 | Method and device for dividing traffic control multistage subareas |
| CN111462490A (en) * | 2020-04-03 | 2020-07-28 | 海信集团有限公司 | Road network visualization method and device based on multistage subregion division |
| CN111462486A (en) * | 2020-03-31 | 2020-07-28 | 连云港杰瑞电子有限公司 | Intersection similarity measurement method based on traffic signal control |
| CN111625435A (en) * | 2020-05-21 | 2020-09-04 | 苏州浪潮智能科技有限公司 | Server analysis method, device and equipment and computer readable storage medium |
| CN112907985A (en) * | 2019-11-19 | 2021-06-04 | 杭州海康威视数字技术股份有限公司 | Method and device for dividing traffic control area |
| CN112907939A (en) * | 2019-12-04 | 2021-06-04 | 杭州海康威视数字技术股份有限公司 | Traffic control subarea dividing method and device |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107808518A (en) * | 2017-10-26 | 2018-03-16 | 东南大学 | A kind of traffic zone division methods theoretical based on multi-path spectral clustering |
| CN108320511A (en) * | 2018-03-30 | 2018-07-24 | 江苏智通交通科技有限公司 | Urban highway traffic sub-area division method based on spectral clustering |
| CN108920678A (en) * | 2018-07-10 | 2018-11-30 | 福州大学 | A kind of overlapping community discovery method based on spectral clustering with fuzzy set |
-
2019
- 2019-06-05 CN CN201910487157.4A patent/CN110322117A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107808518A (en) * | 2017-10-26 | 2018-03-16 | 东南大学 | A kind of traffic zone division methods theoretical based on multi-path spectral clustering |
| CN108320511A (en) * | 2018-03-30 | 2018-07-24 | 江苏智通交通科技有限公司 | Urban highway traffic sub-area division method based on spectral clustering |
| CN108920678A (en) * | 2018-07-10 | 2018-11-30 | 福州大学 | A kind of overlapping community discovery method based on spectral clustering with fuzzy set |
Non-Patent Citations (7)
| Title |
|---|
| WEIFU CHEN, GUOCAN FENG: "Spectral clustering with discriminant cuts", 《KNOWLEDGE-BASED SYSTEMS》 * |
| WEIFU CHEN, GUOCAN FENG: "Spectral clustering with discriminant cuts", 《KNOWLEDGE-BASED SYSTEMS》, 25 November 2011 (2011-11-25), pages 27 - 37 * |
| 尹洪英 等: "基于谱聚类算法的城市路网动态分区研究", 《交通信息与安全》, pages 16 - 20 * |
| 张晓琴 等: "基于谱聚类的二分网络社区发现算法", 《计算机科学》 * |
| 张晓琴 等: "基于谱聚类的二分网络社区发现算法", 《计算机科学》, vol. 46, no. 4, 30 April 2019 (2019-04-30), pages 2 - 3 * |
| 鄢小文: "城市路网连贯流信号协调控制理论与方法研究", 《中国博士学位论文全文数据库 工程科技II辑》 * |
| 鄢小文: "城市路网连贯流信号协调控制理论与方法研究", 《中国博士学位论文全文数据库 工程科技II辑》, 15 December 2018 (2018-12-15), pages 3 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112907985A (en) * | 2019-11-19 | 2021-06-04 | 杭州海康威视数字技术股份有限公司 | Method and device for dividing traffic control area |
| CN112907939A (en) * | 2019-12-04 | 2021-06-04 | 杭州海康威视数字技术股份有限公司 | Traffic control subarea dividing method and device |
| CN112907939B (en) * | 2019-12-04 | 2022-05-20 | 杭州海康威视数字技术股份有限公司 | Traffic control subarea dividing method and device |
| CN111210625A (en) * | 2020-01-10 | 2020-05-29 | 阿里巴巴集团控股有限公司 | Traffic control method and device and electronic equipment |
| CN111325979A (en) * | 2020-02-28 | 2020-06-23 | 海信集团有限公司 | Method and device for dividing traffic control multistage subareas |
| CN111462486A (en) * | 2020-03-31 | 2020-07-28 | 连云港杰瑞电子有限公司 | Intersection similarity measurement method based on traffic signal control |
| CN111462486B (en) * | 2020-03-31 | 2022-05-31 | 连云港杰瑞电子有限公司 | Intersection similarity measurement method based on traffic signal control |
| CN111462490A (en) * | 2020-04-03 | 2020-07-28 | 海信集团有限公司 | Road network visualization method and device based on multistage subregion division |
| CN111462490B (en) * | 2020-04-03 | 2022-06-21 | 海信集团有限公司 | Road network visualization method and device based on multistage subregion division |
| CN111625435A (en) * | 2020-05-21 | 2020-09-04 | 苏州浪潮智能科技有限公司 | Server analysis method, device and equipment and computer readable storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110322117A (en) | A kind of traffic control sub-area division method, apparatus, electronic equipment and storage medium | |
| US20210365823A1 (en) | Capacity evaluation method and device based on historical capacity similarity characteristic | |
| CN109597856A (en) | A kind of data processing method, device, electronic equipment and storage medium | |
| CN105446988B (en) | The method and apparatus for predicting classification | |
| CN111581046A (en) | Data anomaly detection method and device, electronic equipment and storage medium | |
| CN110322694A (en) | A kind of method and device of urban traffic control piece Division | |
| CN109995611B (en) | Traffic classification model establishing and traffic classification method, device, equipment and server | |
| CN110579359A (en) | Optimization method and system of automatic driving failure scene library, server and medium | |
| WO2022228344A1 (en) | Traffic prediction method and device, and storage medium | |
| CN114944053A (en) | Traffic flow prediction method based on spatio-temporal hypergraph neural network | |
| CN111274887A (en) | Crop identification method and device based on machine learning, storage medium and terminal | |
| Dimitriou et al. | Dynamic partitioning of urban road networks based on their topological and operational characteristics | |
| CN108898244A (en) | A kind of digital signage position recommended method coupling multi-source element | |
| US11410020B2 (en) | Machine learning models for predicting time in traffic | |
| CN113268929B (en) | Short-term load interval prediction method and device | |
| CN111462490B (en) | Road network visualization method and device based on multistage subregion division | |
| CN114584406A (en) | Industrial big data privacy protection system and method for federated learning | |
| Sadahiro | Analysis of the relations among spatial tessellations | |
| CN108932839A (en) | A kind of colleague's vehicle judgment method and device | |
| Calka et al. | The Application of Geoinformation Theory in Housing Mass Appraisal | |
| Nadoushani et al. | A fuzzy hybrid clustering method for identifying hydrologic homogeneous regions | |
| CN113656625B (en) | Method and device for determining human body space domain and electronic equipment | |
| CN112164223A (en) | Intelligent traffic information processing method and device based on cloud platform | |
| CN111984637B (en) | Missing value processing method and device in data modeling, equipment and storage medium | |
| CN114821617A (en) | Door and window hole identification method, device, equipment and readable storage medium |
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 |