CN111598335A - Traffic area division method based on improved spectral clustering algorithm - Google Patents
Traffic area division method based on improved spectral clustering algorithm Download PDFInfo
- Publication number
- CN111598335A CN111598335A CN202010410157.7A CN202010410157A CN111598335A CN 111598335 A CN111598335 A CN 111598335A CN 202010410157 A CN202010410157 A CN 202010410157A CN 111598335 A CN111598335 A CN 111598335A
- Authority
- CN
- China
- Prior art keywords
- matrix
- road network
- traffic
- population
- algorithm
- 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
- 230000003595 spectral effect Effects 0.000 title claims abstract description 27
- 238000000034 method Methods 0.000 title claims abstract description 24
- 239000011159 matrix material Substances 0.000 claims abstract description 57
- 230000002068 genetic effect Effects 0.000 claims abstract description 23
- 239000013598 vector Substances 0.000 claims abstract description 17
- 238000005457 optimization Methods 0.000 claims abstract description 7
- 108090000623 proteins and genes Proteins 0.000 claims description 12
- 230000007704 transition Effects 0.000 claims description 10
- 238000004364 calculation method Methods 0.000 claims description 6
- 238000007667 floating Methods 0.000 claims description 6
- 238000013507 mapping Methods 0.000 claims description 5
- 210000000349 chromosome Anatomy 0.000 claims description 3
- 238000010276 construction Methods 0.000 claims description 3
- 230000008569 process Effects 0.000 claims description 3
- 238000010187 selection method Methods 0.000 claims description 3
- 101100256746 Mus musculus Setdb1 gene Proteins 0.000 claims 1
- 238000000638 solvent extraction Methods 0.000 abstract description 7
- 230000000694 effects Effects 0.000 abstract description 5
- 238000011156 evaluation Methods 0.000 description 7
- 230000008859 change Effects 0.000 description 5
- 238000010586 diagram Methods 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 3
- 230000003068 static effect Effects 0.000 description 3
- 239000006185 dispersion Substances 0.000 description 2
- 238000002474 experimental method Methods 0.000 description 2
- 238000012546 transfer Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Images
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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- 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—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Human Resources & Organizations (AREA)
- General Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Tourism & Hospitality (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Development Economics (AREA)
- Evolutionary Biology (AREA)
- General Business, Economics & Management (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Probability & Statistics with Applications (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Health & Medical Sciences (AREA)
- Quality & Reliability (AREA)
- Educational Administration (AREA)
Abstract
The invention relates to a traffic region partitioning method based on an improved spectral clustering algorithm, which relates to the technical field of intelligent traffic, and is a clustering algorithm based on a graph theory, wherein the principle is to convert the original clustering problem into the partitioning problem of a topological graph, design a similarity matrix by utilizing the similarity between nodes, calculate the first n characteristic vectors of the matrix, classify different data points, use a standardized Laplace matrix as a basis, then calculate the characteristic value and the characteristic vector of the matrix, and finally calculate a clustering result through a genetic algorithm; the invention has the advantages that the traditional spectral clustering algorithm is improved from two angles of road network structure information and a clustering center, the similar graph is reconstructed through a Markov chain, more complex road network information is considered, and then the global optimization searching capability is improved by combining a genetic algorithm, thereby effectively ensuring the homogeneity inside the sub-area of the road network and meeting the difference requirements among the sub-areas, and simultaneously having better clustering effect and effectively dividing the traffic area.
Description
Technical Field
The invention relates to the technical field of intelligent traffic, in particular to a traffic region dividing method based on an improved spectral clustering algorithm.
Background
Traffic jam is a problem which is commonly faced and needs to be solved all over the world at present, and the situation of the traffic jam generally has locality and uncertainty. The reasonable and efficient traffic area division can clearly show the traffic propagation rule and the road congestion condition of the area, and has practical guiding significance for traffic dispersion. The traffic area division determines an optimal control method of the intersection according to the form of the sub-area, and single-point control, trunk line coordination control and area coordination control are respectively executed on the isolated sub-area, the line control sub-area and the surface control sub-area. Particularly, in a drive-by-wire and surface control area, the internal intersections are organically coordinated by running a common coordination period and reasonably setting phase differences, so that traffic delay can be effectively reduced.
In 1969, a static subregion division technology is adopted in a TRANSYT system successfully developed in UK, and the system firstly divides a road network into different regions by considering factors such as natural and administrative divisions in a city and then divides each region into different control subregions by considering factors such as intersection connecting line length, traffic flow size, traffic jam state, traffic flow generation source and termination point in each region. Similar static subdivision techniques were still employed in the successfully developed SCOOT system in 1979.
Wallichus clearly provides a concept of sub-area division, and whether the traffic flow characteristics change remarkably or not is taken as a standard of the sub-area division, and the traffic sub-area division needs to consider the influence of factors such as phase difference errors, intersection saturation, remarkable change of road physical characteristics and the like. It is recommended in the united states federal highway administration for a manual of unified traffic control facilities (MUTCD) not to perform coordinated control by dividing 2 intersections more than 0.5 miles apart (about 800m) into the same subarea. Pline suggests that intersection spacing, traffic flow speed, signal timing, traffic flow, fleet dispersion, and road disturbance should be considered in sub-area division.
The traffic zone division technology can be divided into 2 types according to whether the traffic zone division changes with the traffic state: the method includes the steps of (1) static subregion dividing technology. And dividing the subareas according to historical traffic flow data, wherein the subarea scheme is fixed after being formed and does not change due to the change of the traffic state any more. ② dynamic subarea division technique. The method is characterized in that the sub-area scale is adjusted according to the dynamic change of the traffic state on the basis of traffic flow data detected in real time in a road network, so that the optimal sub-area division scheme is maintained, intersections which are closely connected are always divided into the same sub-area, and the intersections are guaranteed to execute the optimal control mode.
The road network structure information borne by the similar graph of the traditional spectral clustering structure is less, so that the road network structure information is difficult to effectively represent when the traffic area is divided by the similar matrix, the complex relevance among road network nodes is easy to ignore by the traditional spectral clustering Euclidean distance, and the traditional spectral clustering is sensitive to the selection of an initial value when a k-means algorithm is adopted, and the problem of local optimal solution is easy to cause.
Disclosure of Invention
In order to solve the technical problems, the invention provides a traffic region partitioning method based on an improved spectral clustering algorithm, the spectral clustering of the invention is a clustering algorithm based on a graph theory, the principle is to convert the original clustering problem into the partitioning problem of a topological graph, a similarity matrix is designed by utilizing the similarity between nodes, the first n eigenvectors of the matrix are calculated, thus different data points are classified, a standardized Laplace matrix is taken as a basis, then the eigenvalue and the eigenvector of the matrix are solved, and finally the clustering result is calculated by a genetic algorithm.
The technical scheme of the invention is as follows:
step S1: constructing a road network topological graph G (V, E) according to an actual road network;
step S2: constructing a similar matrix W according to the weight between two nodes of the road network;
step S3: calculating a Laplace matrix L and constructing a standardized Laplace matrix;
step S4: constructing a characteristic sample set, and mapping corresponding nodes to a k-dimensional space;
step S5: carrying out genetic optimization;
step S6: and dividing road networks according to the final result.
Preferably, in step S1, the road network topology G is constructed from the actual road network as (V, E), and the following steps are performed:
step S1-1: marking each intersection: mapping the data points to the studied road network undirected graph, wherein each intersection vertex data point xiCorresponding to a node x of graph GiE represents a set of edges in the graph, (v)i，vj) Representing point viTo point vjThe calculation formula is shown as follows:
E＝{(vi,vj)|(vi,vj)∈V×V,vivj}；
step S1-2: selecting flow characteristic parameters;
step S1-3: calculating the transition probability among the nodes in the region, and calculating a high-order transition probability matrix Y:
where t represents the length of the Markov chain, i represents the state of the Markov chain at time i, PiRepresenting the i-order transition matrix, wiIs PiThe weight of (2).
Preferably, step S2: constructing a similarity matrix W according to the weight between two nodes of the road network, and performing the following steps:
step S2-1: the whole undirected graph is represented by a non-negative similarity matrix W, the elements of which are WijRepresenting two nodes v in an undirected graphiAnd vjWeight between, and wij＝wjiObtaining a similar matrix W;
the calculation formula of the similarity matrix W is:
wherein: | xi-xj| | represents the euclidean distance between nodes i and j; σ is the standard deviation of the sample.
Preferably, step S3: calculating a Laplace matrix L and constructing a standardized Laplace matrix, wherein the Laplace matrix L is obtained by the following steps:
step S3-1: calculating the Laplace matrix L as D-W, D is a degree matrix, and the formula is as follows:
wherein the element wijRepresenting two nodes v in an undirected graphiAnd vjWeight in between;
step S3-2: construction of standardized post-pull matrices D-1/2AD-1/2And decomposing the L characteristic, and calculating the minimum characteristic value and the corresponding characteristic vector, wherein A is a matrix of the similarity matrix W.
Preferably, in step S4, a feature sample set is constructed, and the corresponding nodes are mapped to the k-dimensional space by finding the first k largest eigenvalues of the laplace matrix and the corresponding eigenvectors (ξ)1ξ2...ξk) Obtaining a characteristic vector matrix, calculating each row vector of the vector matrix, taking each row vector as a sample of a space to obtain k characteristic samples, and performing the following steps:
step S4-1, calculating the first k largest eigenvalues of L and corresponding eigenvectors (ξ)1ξ2...ξk) Obtaining a characteristic vector matrix X;
step S4-2: calculating the row vector of the matrix X to obtain a matrix Y, wherein the formula is as follows:
wherein XijRepresenting the ith row and jth column elements.
Preferably, step S5: genetic optimization was performed as follows:
step S5-1: coding the sample set and generating an initial population t;
step S5-2: calculating a fitness function of the initial population and setting iteration times;
step S5-3: carrying out genetic operation on the population of the ith generation by combining a traditional genetic algorithm to obtain a new generation of population;
step S5-4: and judging whether the iteration number is the maximum value of the iteration number, if so, i is i +1, returning to the step S5-3, otherwise, ending the iteration process, and taking the individual with the maximum fitness function in the population as the final result of the algorithm.
Preferably, step S5-1: encoding the sample set and generating an initial population t, wherein the specific operation mode is as follows:
generating an initial population by adopting a floating point number coding mode, giving a cluster number k, setting a calculated characteristic vector as an initial sample set X, setting an association weight between two nodes of a road network as a gene, namely a class center, carrying out chromosome coding on each data sample in the X, taking the coded sample as an individual, randomly generating a population, and executing the steps for N times to obtain a set of N different individuals and finish population initialization.
Preferably, step S5-3: and (3) carrying out genetic operation on the population of the ith generation by combining a traditional genetic algorithm to obtain a new generation population, wherein the specific operation mode is as follows:
and (4) combining the traditional genetic algorithm to carry out genetic operation on the population of the ith generation to obtain the population of the new generation. Selecting a roulette selection method, and determining whether each individual can enter the next generation or not according to the relative fitness of each individual. And then carrying out individual crossing, randomly exchanging genes between the two individuals, merging the same points if the filial generations have the same points after crossing, and determining whether the parents have the weight of the corresponding genes in the road network which is not zero or not if the filial generations have the same points, namely, a direct road network relation exists. If so, selecting the optimal gene pair in the filial generation to carry out cross operation.
The invention adopting the technical scheme can bring the following beneficial effects:
the invention provides a spectral clustering traffic region partitioning method based on combination of transition probability and genetic algorithm, which improves the traditional spectral clustering algorithm from two angles of road network structure information and clustering center, reconstructs a similar graph through a Markov chain, considers more complex road network information, combines the genetic algorithm to improve the global optimization capability, effectively ensures the homogeneity inside a road network sub-region, meets the difference requirement between the sub-regions, has better clustering effect and can effectively partition the traffic region.
Drawings
The invention is further described with reference to the following figures and detailed description.
Fig. 1 shows the general steps of traffic zone division.
Fig. 2 is a traffic zone division range.
FIG. 3 is a graph of a floating traffic flow profile over a period of time in a traffic zone.
Fig. 4 is a traffic zone division result diagram.
Fig. 5 is a schematic diagram of road network division evaluation indexes.
Detailed Description
The technical solution in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention.
Referring to fig. 1-5, the present invention provides a traffic region division method based on an improved spectral clustering algorithm, including the following steps:
step 1: determining and dividing the traffic area range, and the steps are as follows:
from the actual road network and the road GPS data, a topology structure diagram G of the road network can be obtained as (V, E), and the topology diagram is shown in fig. 2.
Marking individual intersections
xi∈RdThese data points are first mapped into the road network undirected graph G ═ (V, E) under study, where each intersection vertex data point x isiCorresponding to a node x of graph GiAnd E represents a set of edges in the graph, and the calculation formula is as follows:
E＝{(vi,vj)|(vi,vj)∈V×V,vivj}
step 2: and selecting flow characteristic parameters.
Floating traffic data in a certain time period of the traffic area is selected, and the GPS track data is matched with the actual road network through geographic coordinates by utilizing ArcGIS, as shown in figure 3.
And step 3: and calculating the transition probability among the nodes of the region.
Computing a high order transition probability matrix
Where t represents the length of the Markov chain, i represents the state of the Markov chain at time i, PiRepresenting the i-order transition matrix, wiIs PiThe weight of (2).
And 4, step 4: a region similarity matrix W is constructed.
The whole undirected graph is represented by a non-negative similarity matrix W, the elements of which are WijRepresenting two nodes v in an undirected graphiAnd vjWeight between, and wij＝wji. Then, a similarity matrix W is obtained, and the calculation formula of the similarity matrix W is as follows:
wherein: | xi-xj| | represents the euclidean distance between nodes i and j; σ is the standard deviation of the sample.
And 5: the laplacian matrix L ═ D-W is calculated. D is a degree matrix, and the formula is as follows:
wherein the element wijRepresenting two nodes v in an undirected graphiAnd vjWeight in between.
Step 6: construction of standardized post-pull matrices D-1/2AD-1/2And A is a matrix of the similarity matrix W.
Step 7, solving the first k maximum eigenvalues of L and corresponding eigenvectors (ξ)1ξ2...ξk) And obtaining a feature vector matrix X.
And 8: calculating the row vector of the matrix X to obtain a matrix Y, wherein the formula is as follows:
wherein: xijRepresenting the ith row and jth column elements.
And step 9: and mapping the corresponding nodes to a k-dimensional space to be used as a characteristic sample set.
Step 10: the sample set is encoded and an initial population is generated.
Generating an initial population by adopting a floating point number coding mode, giving a cluster number k, setting a calculated characteristic vector as an initial sample set X, setting an association weight between two nodes of a road network as a gene, namely a class center, carrying out chromosome coding on each data sample in the X, taking the coded sample as an individual, randomly generating a population, and executing the steps for N times to obtain a set of N different individuals and finish population initialization.
Step 11: and calculating a fitness function of the initial population and setting iteration times.
Step 12: and (4) combining the traditional genetic algorithm to carry out genetic operation on the population of the ith generation to obtain the population of the new generation.
Selecting a roulette selection method, and determining whether each individual can enter the next generation or not according to the relative fitness of each individual. And then carrying out individual crossing, randomly exchanging genes between the two individuals, merging the same points if the filial generations have the same points after crossing, and determining whether the parents have the weight of the corresponding genes in the road network which is not zero or not if the filial generations have the same points, namely, a direct road network relation exists. If so, selecting the optimal gene pair in the filial generation to carry out cross operation.
Step 13: and (6) iteration.
And judging whether the iteration times are the maximum value of the iteration times, if so, i is i +1, returning to the step 12, otherwise, ending the iteration process, and taking the individual with the maximum fitness function in the population as the final result of the algorithm.
Step 14: the road network is divided according to the final result, as shown in fig. 4.
The effects of the present invention are further illustrated by the following experimental data.
The result of dividing the sub-regions of the road network should satisfy the homogeneity inside the region and the difference between adjacent sub-regions, so the normalized total variance method (TV) is usedn) To evaluate the homogeneity, NCut Silhouette (NS), within the region of the algorithm of the inventionk) The difference of adjacent subintervals of the algorithm is represented, so that the rationality and the scientificity of the road network division result are judged.
In consideration of practical application, the floating traffic flow distribution diagram in a certain period of the traffic area is selected as a research object (as shown in fig. 2), road network division is realized by clustering the flows with similar characteristics, and relevant parameters required by experiments are set as shown in the following table 1.
TABLE 1 relevant parameter settings required for the experiment
| Parameter(s) | Value taking | Parameter(s) | Value taking |
| Size of population | 20 | Constant b | 1000 |
| Maximum number of iterations | 100 | Cluster | 6 |
| Probability of crossing | 0.75 | Standard deviation sigma | 0.9 |
| Probability of variation | 0.03 |
The algorithm of the invention is a transfer probability-based spectral clustering genetic algorithm (TPGASC), and effectiveness comparison is carried out on the partitioning results of the traditional spectral clustering algorithm (SC), the transfer probability-based spectral clustering algorithm (TPSC) and the genetic algorithm-based spectral clustering algorithm (GASC) under the same experimental environment, and the obtained sub-region partitioning evaluation results are shown in table 2 and fig. 5.
TABLE 2 road network division evaluation results
It can be seen from table 2 that the TPSC algorithm compares TV with the conventional SC algorithmnThe evaluation index is reduced by 9.16 percent, and NSkThe index is reduced by 2.8%, which shows that compared with the traditional spectral clustering algorithm, the TPSC algorithm considers the complexity between the road networks, enhances the robustness of the similar graph and further improves the clustering effect. The GASC algorithm compares TV with the traditional SC algorithmnThe evaluation index is reduced by 8.42 percent, and NSkThe index is reduced by 11.58%, which shows that on the basis of the traditional spectral clustering algorithm, the GASC algorithm is combined with the genetic algorithm, so that the clustering center can be effectively optimized, the global optimization capability is improved, and the clustering precision is improved. Compared with other three road network division algorithms, the algorithm provided by the invention is applied to TVnThe indexes are respectively reduced by 10.27 percent, 1.22 percent and 2.03 percent, and NSkThe indexes are respectively reduced by 13.23 percent, 10.73 percent and 1.87 percent,the TPGASC algorithm provided by the invention is superior to other algorithms, the complexity of a road network is considered, the optimization of a clustering center is considered, and the clustering effect is improved.
It can be seen from fig. 5 that the algorithm of the present invention compares TV with other algorithmsnEvaluation index and NSkThe evaluation index is relatively low, which shows that the algorithm of the invention not only has higher similarity in each subarea, but also can ensure the difference between subareas of the road network, has advantages in road network division, better accords with the actual traffic characteristics, and verifies the effectiveness of the algorithm of the invention.
While the invention has been shown and described with respect to the embodiments, it will be understood by those skilled in the art that various changes and modifications may be made without departing from the scope of the invention as defined in the following claims.
Claims (8)
1. A traffic region division method based on an improved spectral clustering algorithm is characterized by comprising the following steps:
step S1: constructing a road network topological graph G (V, E) according to an actual road network;
step S2: constructing a similar matrix W according to the weight between two nodes of the road network;
step S3: calculating a Laplace matrix L and constructing a standardized Laplace matrix;
step S4: constructing a characteristic sample set, and mapping corresponding nodes to a k-dimensional space;
step S5: carrying out genetic optimization;
step S6: and dividing road networks according to the final result.
2. The method for dividing the traffic zone based on the improved spectral clustering algorithm as claimed in claim 1, wherein the step S1 is performed as follows:
step S1-1, marking each intersection:
mapping the data points to the studied road network undirected graph, wherein each intersection vertex data point xiCorresponding to a node x of graph GiGeneration ESet of edges in the Table map, (v)i，vj) Representing point viTo point vjThe calculation formula is shown as follows:
E＝{(vi,vj)|(vi,vj)∈V×V,vivj}；
step S1-2, selecting flow characteristic parameters;
step S1-3, calculating the transition probability among the region nodes, and calculating a high-order transition probability matrix Y:
where t represents the length of the Markov chain, i represents the state of the Markov chain at time i, PiRepresenting the i-order transition matrix, wiIs PiThe weight of (2).
3. The method for dividing the traffic zone based on the improved spectral clustering algorithm as claimed in claim 1, wherein the step S2 is performed as follows:
step S2-1: the calculation formula of the similarity matrix W is:
wherein: | xi-xjAnd | | represents the Euclidean distance between the nodes i and j, and σ is the standard deviation of the samples.
4. The method for dividing the traffic zone based on the improved spectral clustering algorithm as claimed in claim 1, wherein the step S3 is performed as follows:
step S3-1: calculating the Laplace matrix L as D-W, D is a degree matrix, and the formula is as follows:
wherein, YuanElement wijRepresenting the weight between two nodes vi and vj in the undirected graph;
step S3-2: construction of standardized post-pull matrices D-1/2AD-1/2And decomposing the L characteristic, and calculating the minimum characteristic value and the corresponding characteristic vector, wherein A is a matrix of the similarity matrix W.
5. The method for dividing traffic areas based on improved spectral clustering algorithm according to claim 4, wherein step S4 is performed as follows:
step S4-1, calculating the first k largest eigenvalues of L and corresponding eigenvectors (ξ)1ξ2...ξk) Obtaining a characteristic vector matrix X;
step S4-2: calculating the row vector of the matrix X to obtain a matrix Y, wherein the formula is as follows:
wherein, XijRepresenting the ith row and jth column elements.
6. The method for dividing the traffic zone based on the improved spectral clustering algorithm as claimed in claim 1, wherein the step S5 is performed as follows:
step S5-1: coding the sample set and generating an initial population t;
step S5-2: calculating a fitness function of the initial population and setting iteration times;
step S5-3: combining a traditional genetic algorithm to carry out a genetic mode on the population of the ith generation to obtain a new generation of population;
step S5-4: and judging whether the iteration number is the maximum value of the iteration number, if so, i is i +1, returning to the step S5-3, otherwise, ending the iteration process, and taking the individual with the maximum fitness function in the population as the final result of the algorithm.
7. The method for dividing the traffic zone based on the improved spectral clustering algorithm as claimed in claim 6, wherein the step S5-1 is performed as follows:
generating an initial population by adopting a floating point number coding mode, giving a cluster number k, setting a calculated characteristic vector as an initial sample set X, setting an association weight between two nodes of a road network as a gene, namely a class center, carrying out chromosome coding on each data sample in the X, taking the coded sample as an individual, randomly generating a population, and executing the steps for N times to obtain a set of N different individuals and finish population initialization.
8. The method for dividing traffic zones based on improved spectral clustering algorithm according to claim 6, wherein step S5-3 is performed as follows:
combining a traditional genetic algorithm to carry out a genetic mode on the population of the ith generation to obtain a new generation of population; selecting a roulette selection method, wherein whether each individual can enter the next generation or not is determined by the relative fitness of each individual; then, carrying out individual crossing, randomly exchanging genes for two individuals, merging the same points if the filial generations have the same points after crossing, and determining whether the parents have the weight of the corresponding genes in the road network which is not zero or not if the filial generations do not have the same points, namely, a direct road network relation exists; if so, selecting the optimal gene pair in the filial generation to carry out a crossing mode.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010410157.7A CN111598335A (en) | 2020-05-15 | 2020-05-15 | Traffic area division method based on improved spectral clustering algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010410157.7A CN111598335A (en) | 2020-05-15 | 2020-05-15 | Traffic area division method based on improved spectral clustering algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111598335A true CN111598335A (en) | 2020-08-28 |
Family
ID=72182420
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010410157.7A Pending CN111598335A (en) | 2020-05-15 | 2020-05-15 | Traffic area division method based on improved spectral clustering algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111598335A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112508225A (en) * | 2020-10-27 | 2021-03-16 | 宁波工程学院 | Multi-detail traffic cell partitioning method and system based on spectral clustering algorithm |
| CN113160556A (en) * | 2021-03-12 | 2021-07-23 | 北京邮电大学 | Urban road network dynamic division method and device, computer equipment and storage medium |
| CN113724892A (en) * | 2021-08-31 | 2021-11-30 | 平安国际智慧城市科技股份有限公司 | Method and device for analyzing population mobility, electronic equipment and storage medium |
| CN114360239A (en) * | 2021-12-03 | 2022-04-15 | 武汉工程大学 | Traffic prediction method and system for multilayer space-time traffic knowledge map reconstruction |
| CN115277350A (en) * | 2022-07-26 | 2022-11-01 | 湘潭大学 | Multi-controller deployment method based on improved spectral clustering algorithm in SDN environment |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103106279A (en) * | 2013-02-21 | 2013-05-15 | 浙江大学 | Clustering method simultaneously based on node attribute and structural relationship similarity |
| CN105959270A (en) * | 2016-04-25 | 2016-09-21 | 盐城工学院 | Network attack detection method based on spectral clustering algorithm |
| CN106251625A (en) * | 2016-08-18 | 2016-12-21 | 上海交通大学 | Three-dimensional urban road network global state Forecasting Methodology under big data environment |
| 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 |
| CN108734187A (en) * | 2017-04-20 | 2018-11-02 | 中山大学 | A kind of multiple view spectral clustering based on tensor singular value decomposition |
| CN109117872A (en) * | 2018-07-24 | 2019-01-01 | 贵州电网有限责任公司信息中心 | A kind of user power utilization behavior analysis method based on automatic Optimal Clustering |
| CN109887297A (en) * | 2019-04-23 | 2019-06-14 | 太原理工大学 | The division methods of urban traffic control sub-district based on quick global K-means spectral clustering |
| CN110210509A (en) * | 2019-03-04 | 2019-09-06 | 广东交通职业技术学院 | A kind of road net traffic state method of discrimination based on MFD+ spectral clustering+SVM |
| CN110336768A (en) * | 2019-01-22 | 2019-10-15 | 西北大学 | A kind of Tendency Prediction method based on joint hidden Markov model and genetic algorithm |
| CN110415523A (en) * | 2019-08-13 | 2019-11-05 | 东南大学 | A kind of signal control work zone division methods based on vehicle driving track data |
| CN110610186A (en) * | 2019-04-24 | 2019-12-24 | 广东交通职业技术学院 | Road network partition based on ACO-FCM clustering algorithm and evaluation method thereof |
-
2020
- 2020-05-15 CN CN202010410157.7A patent/CN111598335A/en active Pending
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103106279A (en) * | 2013-02-21 | 2013-05-15 | 浙江大学 | Clustering method simultaneously based on node attribute and structural relationship similarity |
| CN105959270A (en) * | 2016-04-25 | 2016-09-21 | 盐城工学院 | Network attack detection method based on spectral clustering algorithm |
| CN106251625A (en) * | 2016-08-18 | 2016-12-21 | 上海交通大学 | Three-dimensional urban road network global state Forecasting Methodology under big data environment |
| CN108734187A (en) * | 2017-04-20 | 2018-11-02 | 中山大学 | A kind of multiple view spectral clustering based on tensor singular value decomposition |
| 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 |
| CN109117872A (en) * | 2018-07-24 | 2019-01-01 | 贵州电网有限责任公司信息中心 | A kind of user power utilization behavior analysis method based on automatic Optimal Clustering |
| CN110336768A (en) * | 2019-01-22 | 2019-10-15 | 西北大学 | A kind of Tendency Prediction method based on joint hidden Markov model and genetic algorithm |
| CN110210509A (en) * | 2019-03-04 | 2019-09-06 | 广东交通职业技术学院 | A kind of road net traffic state method of discrimination based on MFD+ spectral clustering+SVM |
| CN109887297A (en) * | 2019-04-23 | 2019-06-14 | 太原理工大学 | The division methods of urban traffic control sub-district based on quick global K-means spectral clustering |
| CN110610186A (en) * | 2019-04-24 | 2019-12-24 | 广东交通职业技术学院 | Road network partition based on ACO-FCM clustering algorithm and evaluation method thereof |
| CN110415523A (en) * | 2019-08-13 | 2019-11-05 | 东南大学 | A kind of signal control work zone division methods based on vehicle driving track data |
Non-Patent Citations (9)
| Title |
|---|
| 任淑霞等: "基于DCBM的马尔可夫谱聚类社区发现算法", 《四川大学学报(自然科学版)》 * |
| 何东晓等: "复杂网络社区挖掘―基于聚类融合的遗传算法", 《自动化学报》 * |
| 刘俊等: "谱聚类在给水管网分区优化中的应用", 《土木建筑与环境工程》 * |
| 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》 * |
| 戴晓明等: "粗粒度并行遗传算法收敛性分析及优化运算", 《上海交通大学学报》 * |
| 林树宽等: "基于用户移动行为相似性聚类的Markov位置预测", 《东北大学学报(自然科学版)》 * |
| 王会青等: "遗传优化的谱聚类方法研究", 《计算机工程与应用》 * |
| 王鹏等: "基于GA和FCM的岩体结构面的混合聚类方法", 《北京科技大学学报》 * |
| 苏琳等: "自适应谱聚类算法并行实现及优化", 《科研信息化技术与应用》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112508225A (en) * | 2020-10-27 | 2021-03-16 | 宁波工程学院 | Multi-detail traffic cell partitioning method and system based on spectral clustering algorithm |
| CN113160556A (en) * | 2021-03-12 | 2021-07-23 | 北京邮电大学 | Urban road network dynamic division method and device, computer equipment and storage medium |
| CN113724892A (en) * | 2021-08-31 | 2021-11-30 | 平安国际智慧城市科技股份有限公司 | Method and device for analyzing population mobility, electronic equipment and storage medium |
| CN114360239A (en) * | 2021-12-03 | 2022-04-15 | 武汉工程大学 | Traffic prediction method and system for multilayer space-time traffic knowledge map reconstruction |
| CN115277350A (en) * | 2022-07-26 | 2022-11-01 | 湘潭大学 | Multi-controller deployment method based on improved spectral clustering algorithm in SDN environment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111598335A (en) | Traffic area division method based on improved spectral clustering algorithm | |
| Ji et al. | On the spatial partitioning of urban transportation networks | |
| Huang et al. | Automatic generation of road maps from low quality GPS trajectory data via structure learning | |
| Kriegel et al. | Hierarchical graph embedding for efficient query processing in very large traffic networks | |
| Wang et al. | Preference-based spatial co-location pattern mining | |
| CN101996245A (en) | Form feature describing and indexing method of image object | |
| Wang et al. | Multi-attribute graph convolution network for regional traffic flow prediction | |
| Neethu et al. | Review of spatial clustering methods | |
| Yu et al. | Automatic segmentation of parallel drainage patterns supported by a graph convolution neural network | |
| Huang et al. | A time-dependent attention convolutional LSTM method for traffic flow prediction | |
| Hu et al. | Spatio-temporal trajectory similarity measures: A comprehensive survey and quantitative study | |
| Zhang et al. | GATC and DeepCut: Deep spatiotemporal feature extraction and clustering for large-scale transportation network partition | |
| CN116166975A (en) | Multi-graph hierarchical road surface characterization method based on graph neural network | |
| Jiang et al. | Partitioning of urban networks with polycentric congestion pattern for traffic management policies: Identifying protected networks | |
| Liu et al. | Identification of drainage patterns using a graph convolutional neural network | |
| He et al. | Spatio-temporal trajectory anomaly detection based on common sub-sequence | |
| Yang | Landscape Preference of Rural Cultural Tourism Tourists Based on GIS Spatial Statistical Technology and Emotion Analysis | |
| CN112508225A (en) | Multi-detail traffic cell partitioning method and system based on spectral clustering algorithm | |
| Cao et al. | Trajectory outlier detection on trajectory data streams | |
| Ying et al. | Probabilistic matching of map objects in multi-scale space | |
| Wang et al. | Trajectory distance algorithm based on segment transformation distance | |
| Jia et al. | Human Mobility Prediction Based on Trend Iteration of Spectral Clustering | |
| Zhuang et al. | Traffic Knowledge Graph Based Trajectory Destination Prediction | |
| Liu et al. | A New Retraceable Clustering Algorithm | |
| Ding et al. | Analyzing sensitive information leakage in trajectory embedding models |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200828 |