CN109410577A - Adaptive traffic control sub-area division method based on Spatial Data Mining - Google Patents
Adaptive traffic control sub-area division method based on Spatial Data Mining Download PDFInfo
- Publication number
- CN109410577A CN109410577A CN201811332450.5A CN201811332450A CN109410577A CN 109410577 A CN109410577 A CN 109410577A CN 201811332450 A CN201811332450 A CN 201811332450A CN 109410577 A CN109410577 A CN 109410577A
- Authority
- CN
- China
- Prior art keywords
- feature
- traffic
- intersection
- flow
- traffic flow
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- 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
- G08G1/0137—Measuring and analyzing of parameters relative to traffic conditions for specific applications
- G08G1/0145—Measuring and analyzing of parameters relative to traffic conditions for specific applications for active traffic flow control
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/07—Controlling traffic signals
- G08G1/08—Controlling traffic signals according to detected number or speed of vehicles
Abstract
Adaptive traffic control sub-area division method based on Spatial Data Mining, belongs to technical field of transportation.The present invention is lower in the presence of line efficiency is gone out in order to solve the problems, such as the control zone that existing control work zone division methods are divided.The present invention carries out the feature extraction of high offpeak period Yu high, medium and low flow threshold first, and choosing influences the traffic information feature that control work zone divides；Then the influence power weight calculation model based on space-time characteristic is established, the weight for being characterized r is obtained, the division of traffic control sub-district is finally realized based on the cum rights community discovery algorithm that dynamic analog lumpiness divides.The present invention is suitable for the division of traffic control sub-district.
Description
Technical field
The invention belongs to technical field of transportation, and in particular to a kind of traffic control sub-area division method.
Background technique
Increasingly developed with traffic, traffic congestion has become the common phenomenon in daily trip, seriously affects
The living standard and quality of the people develops important problem urgently to be resolved so alleviating traffic congestion and becoming wisdom traffic.?
In current existing control theory and algorithm, signal control is to alleviate the common technology means of congestion, is had most apparent
Improvement effect and highest cost performance are the current important means for improving traffic control level and managing out line efficiency.
Control work zone, which divides, mainly uses the methods of related degree model, fuzzy control, machine learning and Optimum Theory,
According to the dynamic traffic characteristic at each signal cross crossing and static geographic properties, Regional Road Network is divided into different control
The process in area.The method for wherein extracting traffic information feature mostly uses Mining Method of Spatial Data.
Traffic control sub-area division based on fuzzy control.The magnitude of traffic flow, crossing spacing, period, traffic are comprehensively considered
Stream constitutes and the influence of 5 factors such as traffic flow discreteness, it is determined that the fuzzy reasoning method of cooperation index, and give benefit
The judgment basis of control work zone division is carried out with cooperation index.On the basis of the calculation of relationship degree method of fuzzy logic, building
Using key crossing and the crossing degree of association as the traffic control sub-area division model of judgment basis, and mould is improved using genetic algorithm
Type.
Traffic control sub-area division based on machine learning.Divide based on static division and dynamically the control work zone combined
Method for dynamically partitioning, and crossing spacing maximum-minimum principle is improved, control is improved by incorporating Forecasting Short-term Traffic
The real-time of system Division.It is proposed by fusion fuzzy control technology and neural network model based on flow, road section length
With the traffic control sub-area division model of traffic light cycle, the adjacent intersection of fuzzy reasoning logic calculation is determined by neural network
Cooperation index, according to cooperation index size divide control work zone.
Traffic control sub-area division method based on related degree model.Road synthetic feature based on motorcade dispersion model is closed
Join metrization method, considers the factors such as the magnitude of traffic flow, intersection distance, motorcade dispersion characteristic, establish index correction formula
Decision support is provided for whether traffic control sub-district merges control.Road network based on the intersection degree of association and the section degree of association is drawn
Divide method, by the analysis of the influence factor to traffic control and the road network region division of traffic guidance, considers traffic flow phase
Traffic state judging index is introduced while closing property, realizes that the road network degree of association cooperateed with towards induction with control models, Jin Ershi
Now cooperate with the division of traffic sub-district.More intersection combinations are given on the basis of adjacent intersection calculation of relationship degree model
The calculation formula of the degree of association.By defining disaggregation space, constraint condition and the interpretational criteria of control work zone splitting scheme, establish
Coordinated control sub-area division model.
The extraction of traffic information feature is handled mainly in track data.Space tracking is by geographical space
The track that moving object generates usually is indicated by a series of points being sequentially arranged, wherein each point is sat by geographical space
Mark collection and timestamp composition.By obtaining practical problem of the information to solve using different space digging technologies to track data
It is more and more extensive.Traffic congestion prediction model based on Spatial Data Mining Technique and Semantic Web technology, by extract weather,
The integrated informations such as road engineering and Urban Event come improve traffic jam prediction accuracy and consistency.Pass through background image point
Analysis method obtains road background characteristics to continuous multiple frames image to extract traffic congestion event, analyzes road using spatiotemporal data structure method
Corner features to detect traffic jam.
Existing research has been achieved for certain in the research for extracting traffic information feature and control work zone division methods
Achievement, but since the property complicated and changeable of city road network, the factors such as lengthy and jumbled property for the traffic information data amount that can be excavated influence, hand over
Logical control work zone dynamic division technology is difficult with general applicability.And the friendship chosen when calculating adjacent intersection cooperation index
Logical information characteristics have ignored the topological structure characteristic of transportation network.
Summary of the invention
The present invention is lower in the presence of line efficiency is gone out in order to solve the control zone that existing control work zone division methods are divided
Problem.
Adaptive traffic control sub-area division method based on Spatial Data Mining, comprising the following steps:
Step 1, the feature extraction for carrying out high offpeak period and high, medium and low flow threshold；
Step 2 chooses the traffic information feature for influencing control work zone division；
The traffic information feature includes: influence on traffic flow feature TF, intersection distance influence feature IF, traffic flow structure
At influence feature CI, cycle influences feature CE；
Step 3 establishes the influence power weight calculation model based on space-time characteristic, obtains the weight L' for being characterized rr, r ∈
[Isosorbide-5-Nitrae] be traffic information feature in feature label, indicate influence on traffic flow feature TF, intersection distance influence feature IF,
Traffic flow, which is constituted, influences feature CI, cycle influences feature CE；
Step 4, the division that traffic control sub-district is realized based on the cum rights community discovery algorithm that dynamic analog lumpiness divides, specifically
Process the following steps are included:
Step 4.1, by above-mentioned each feature calculated of the influence power weight calculation model based on space-time characteristic
Weight carries out product accumulation and obtains calculating cooperation index CC between intersection two-by-twoijAs the weight in the section, wherein R is respectively
The value of IF, CI, CE, TF；
When step 4.2, initialization community's matrix, each node in figure is an independent community, society in original state
Area's number is consistent with node number, indicates the non-directed graph of having the right with adjacency matrix, calculate each node degree weight and；
Step 4.3, each node and neighborhood of nodes trial merge community's operation, obtain the preceding mould with after distribution of distribution
The value of lumpiness changing value Δ Q, record Δ Q value maximum and Δ Q are greater than 0 community's merging phase, and node is assigned to the adjacent segments
Community where point, otherwise remains unchanged；
Step 4.4 is less than threshold value, i.e. mould until the community structure of weighted-graph changes caused modularity changing value
The value of lumpiness is maximum, completes a union operation；
Step 4.5 compresses figure according to the merging community state of record, and all nodes in the same community reconstruct
For a new node, new communities' matrix is created, the weight on side is converted into the weight of new node ring, side between community between community's interior nodes
Weight be converted into the side right weight between new node；
Step 4.6, the calculating process that step 4.2 to step 4.5 is repeated, until the modularity changing value of entire figure is small
In threshold value, the final division for realizing traffic control sub-district.
Further, the process packet of the feature extraction of the high offpeak period of progress and high, medium and low flow threshold described in step 1
Include following steps:
Step 1.1, calculating are calculated by the magnitude of traffic flow of the different direction of travel of intersection from two road intersections
On a road enter behind intersection begin to different directions road the magnitude of traffic flow；
Step 1.2 divides high offpeak period according to the corresponding magnitude of traffic flow of each time interval:
After the magnitude of traffic flow for calculating intersection difference direction of travel, on the basis of high ebb Time segments division, building is solved
The mathematical model of high, medium and low flow threshold；Design is set according to the real-time magnitude of traffic flow for different high slack hours sections
Dynamic Weights parameter sj, define weight formula sj
sj=α/γ
Wherein, α is the magnitude of traffic flow of current peak period intersection, and γ is all peak period traffic flows in current crossing
Measure summation；
Then establish the magnitude of traffic flow of single crossing peak period:
Wherein, j ∈ (peak period), the direction of travel of i ∈ [1,12] and vehicle, num indicate peak period number；
It establishes: the element group representation of the high and low peak discharge in period of time at single crossing, the peak period flow tuple D at single crossingth
For
Dth=[dth1,dth2,...,dth12]
The offpeak period flow tuple D at single crossing can similarly be soughttlFor
Dtl=[dtl1,dtl2,...,dtl12]
The unidirectional magnitude of traffic flow of single intersection is ranked up by step 1.3, then the stream of each intersection
Moment matrix is
Wherein, qIJIndicate that the traffic flow magnitude after the magnitude of traffic flow sorts, I ∈ [1,12] indicate that direction of travel, J indicate
Time interval, time interval are denoted as the period；
Step 1.4, according to DthAnd DtlIn d on identical direction of travelthiAnd dtliBy phase in the traffic matrix Q after sequence
Time interval with direction of travel corresponds to magnitude of traffic flow division low flow volume region SL, middle flow region SM, high flow volume region SH:
The length of three sequences is denoted as l1, l2, l3 respectively；For i direction of travel, S is taken respectivelyL, SM50 quartiles as i
The middle low discharge typical value r of direction of travelliWith middle flow typical value rmi, take SHHigh flow capacity generation of 70 quartiles as the i-th direction
Tabular value rhi:
Flow threshold matrix is expressed as R (I ∈ [1,12])；
Further, step 1.1 detailed process the following steps are included:
Step 1.1.1, two road intersections of intersection are divided into and are denoted as the first road and the second road, the first road
With the respective both sides of the edge of the second road as boundary；Two road intersections are divided into 5 regions by intersection, are respectively intersected
Crossing, the first road junction leading portion, the first road junction back segment, the second road junction leading portion, the second road
Intersection back segment；
Step 1.1.2, judge the original driving direction of vehicle, i.e. vehicle enters intersection from which direction；
Then according to the subsequent course bearing of current vehicle, judge the direction of travel of vehicle；
Step 1.1.3, according to the time tag of current trace points, judge time interval belonging to current vehicle；
Step 1.1.4, current vehicle is added up to magnitude of traffic flow matrix (μit, λ) in, i expression direction of travel, i ∈ [1,
12]；T indicates the serial number of time interval, μitIndicate the magnitude of traffic flow of i direction of travel t time interval；λ is crossing mark.
Further, influence on traffic flow feature described in step 2 is as follows:
Influence on traffic flow feature TF:
Wherein, q (t) is t time interval flow, and σ is one directional flow of crossing t time interval, and λ is Different Traffic Flows side
To weight factor；The introducing of λ is actually that the TF value calculated is adjusted between 0 to 100.
Further, it is as follows to influence feature for the distance of intersection described in step 2:
Adjacent intersection spacing is set as L；
Maximum-minimum spacing principle is as follows:
Maximum spacing principle:
ρ=avg (C1(t),C2(t))
In maximum spacing principle, RminThe average value of high flow capacity typical value, N in adjacent two intersectioninThe current green light period
Interior travelable direction of traffic；C1(t)、C2(t) previous intersection and the latter in adjacent two intersection is respectively indicated to hand over
The signal period of cross road mouth；TsampleIt is the length of chronomere, ρ is the average value in the signal period of adjacent intersection,
LmaxIndicate adjacent intersection maximum spacing, hcarIt is average length of car, NlaneIt is number of track-lines, β vehicle discrete distance, F traffic flow
Coefficient of dispersion；
Minimum spacing principle:
ρ=max (C1(t),C2(t))
In minimum spacing principle, RmaxThe larger value of high flow capacity typical value, T in adjacent two intersectionsampleIt is time list
The length of position, ρ is the larger value in the signal period of adjacent intersection, and μ is the anti-congestion factor, LminIndicate adjacent intersection most
Small spacing, hcarIt is average length of car；
At crossing on the basis of maximum-minimum principle, establishing intersection distance influences feature IF:
Wherein, χ is balance factor；L is the length of current road segment.
Further, it is as follows to constitute influence feature for traffic flow described in step 2:
Traffic flow composition is the important parameter of traffic information feature, is downstream road junction through vehicles middle and upper reaches crossing vehicle
Accounting determines the discreteness size of wagon flow and the harmony size at upstream and downstream crossing, and establishing traffic flow and constituting influences feature CI:
CI=O*F
Wherein, O is that the vehicle of upstream straight trip continues the ratio of straight trip in downstream, and value range [0,1], F wanders about as a refugee for traffic
Dissipate coefficient.
Further, cycle influences feature described in step 2 is as follows:
Establish cycle influences feature CE:
Wherein, T1For the green time that the time in green time in two adjacent intersections is long, T2For two adjacent intersections
Time short green time in middle green time.
Further, the influence power weight calculation model based on space-time characteristic is established described in step 3, obtains being characterized r's
Weight L'rProcess it is as follows:
Step 3.1 takes all samples of the feature r in traffic information feature to constitute set Xr:
Xr=[x1,x2,...,xn]T
Wherein n is sample total number, and r ∈ [Isosorbide-5-Nitrae] is the feature label in traffic information feature, represents influence on traffic flow spy
Sign, intersection distance influence feature, traffic flow constitutes influence feature, cycle influences feature；x1,x2,...,xnFeature r respectively
The sample representation of corresponding characteristic value；
If a sample point xi'It is another sample point xj'One of closest k sample point, then xi'And xj'Between deposit
In a line, the feature non-directed graph of each feature is established respectively using the rule；
Step 3.2, the weight W for calculating each edge in feature non-directed graphi'j', constitute weight matrix Wr, form an adjacent square
Battle array；
Step 3.3, the variance for calculating all samples of this feature；DrIt is diagonal matrix, diagonal entry value is each sample
The out-degree weight of this point and, and be based on DrIn element value obtain intermediate variable X 'r:
Dr=diag (Wr* 1) 1=(1,1 ..., 1)T
Thus the variance size of all samples of this feature is calculated
Var(Xr)=X 'rDrX′r
Step 3.4, known WrFor weight matrix, the influence power weight calculation of space-time characteristic is carried out to this feature, then is calculated
It can obtain
Lr=Dr-Wr
xri'、xrj'Respectively indicate the corresponding sample point x of feature ri'、xj'；
So the weight of feature r are as follows:
Further, the k=5.
The present invention has the effect that
The running time of 5 strokes is optimized to varying degrees in traffic control sub-area division scheme of the invention, and
And the distance of stroke is longer, effect of optimization is more obvious.The optimum results of comprehensive a plurality of stroke, average each stroke shorten
22.4% travel time, improve out line efficiency.Journey time, delay time, parking are based on by evaluation document is obtained
Four indexs of number and running speed are it can be concluded that traffic control sub-area division mathematical model proposed by the invention is optimizing
There is good effect in terms of vehicle driving, effectively reduces the running time of vehicle traveling over long distances, shorten vehicle row
Delay time in sailing increases the saturation degree and utilization rate of road, and under identical trip distance, model of the present invention is effectively reduced
Stop frequency, and improve running speed to a certain extent, complex optimum go out line efficiency and road conditions.
The control work zone that control work zone division methods of the invention simultaneously can be applicable in various road conditions corresponding region is drawn
Point, relevance factor is higher.
Detailed description of the invention
Fig. 1 is traffic flow direction region division and judgement schematic diagram；
Fig. 2 is one week magnitude of traffic flow distribution map；
Fig. 3 is one day magnitude of traffic flow distribution map；
Fig. 4 is track of vehicle figure；
Fig. 5 is true map；
Fig. 6 is crossroad administrative division map；
Fig. 7 is traffic characteristic weight coefficient figure；
Fig. 8 is control work zone division result figure；
Fig. 9 is the Louvain calculating process figure that dynamic analog lumpiness divides；
Figure 10 is road simulation and practical crossing position versus figure；
Figure 11 is signal period configuration diagram.
Specific embodiment
Specific embodiment 1:
Adaptive traffic control sub-area division method based on Spatial Data Mining, comprising the following steps:
The feature extraction of step 1, high offpeak period and high, medium and low flow threshold:
High, medium and low flow threshold is the important division limits of determining intersection different directions present period traffic behavior,
It is the important parameter that the principle of flow in later period traffic control sub-district model is realized.The precondition for calculating high, medium and low flow is
Need to realize the traffic statistics of intersection different directions under different time intervals using data digging method.
Step 1.1, calculating are calculated by the magnitude of traffic flow of the different direction of travel of intersection from two road intersections
On a road enter behind intersection begin to different directions road the magnitude of traffic flow:
Step 1.1.1, two road intersections of intersection are divided into and are denoted as the first road and the second road, the first road
With the respective both sides of the edge of the second road as boundary, the boundary of road is marked as L1, L2, L3, L4；Intersection is by two phases
Dealings road is divided into 5 regions, respectively intersection, the first road junction leading portion, the first road junction back segment,
Two road junction leading portions, the second road junction back segment are labeled as 1. -5. number region, as shown in Figure 1；
Step 1.1.2, judge the original driving direction of vehicle, i.e. vehicle enters from which direction and hands over shown in 5. number region
Cross road mouth；
Then according to the subsequent course bearing of current vehicle, judge the direction of travel of vehicle；Friendship of the vehicle in every road
Cross road mouth leading portion drives into behind intersection the subsequent course bearing having respectively to the left, forward, to the right, so every road is corresponding with
Three subsequent course bearings, the four direction road of intersection connection is total 12 direction of travel, such as the direction of travel of Fig. 1
Shown in 1-12 (arrow)；
Deterministic process can according to vehicle every road the corresponding positioning coordinate Loc of intersection leading portion (xi ',
Yi ') the corresponding position judgement of the corresponding positioning coordinate in corresponding position and subsequent track, illustrate in conjunction with attached drawing 1,
Wherein, Loc (xj,xj) it is Loc (xi,xi) successor node.
It similarly may determine that the traffic flow data in remaining direction；
Step 1.1.3, according to the time tag of current trace points, judge time interval belonging to current vehicle；This reality
Some time section will be divided into total time by time interval of 15min by applying mode, count the magnitude of traffic flow in each time interval,
As shown in Figures 2 and 3；
Step 1.1.4, current vehicle is added up to magnitude of traffic flow matrix (μit, λ) in, i expression direction of travel, i ∈ [1,
12]；T indicates the serial number of time interval, μitIndicate the magnitude of traffic flow of i direction of travel t time interval；λ is crossing mark, such as
(μit, 1) indicate No. 1 crossing magnitude of traffic flow matrix；
Step 1.2 divides high offpeak period according to the corresponding magnitude of traffic flow of each time interval；It can be several using calculating
The average value of the magnitude of traffic flow in a period divides high offpeak period, this embodiment party as the high ebb judgment criteria of the magnitude of traffic flow
The average value of the traffic in one week is calculated in formula as the high ebb judgment criteria of the magnitude of traffic flow, and daily time interval is divided into
High offpeak period；
After the magnitude of traffic flow for calculating intersection difference direction of travel, on the basis of high ebb Time segments division, building is solved
The mathematical model of high, medium and low flow threshold.In view of the gap between historical data and real time traffic data, peak may cause
There are errors for the selection of period, and in order to embody the dynamic property of model, this modelling is according to the real-time magnitude of traffic flow, for difference
High slack hours section set Dynamic Weights parameter sj, define weight formula sj
sj=α/γ
Wherein, α is the magnitude of traffic flow of current peak period (being a time interval) intersection, and γ is current crossing institute
There is peak period magnitude of traffic flow summation；
Then establish the magnitude of traffic flow of single crossing peak period:
Wherein, j ∈ (peak period), the direction of travel of i ∈ [1,12] and vehicle, num indicate peak period number, μijFor
Element in traffic matrix above；Therefore by foundation defined above: the element group representation of the high and low peak discharge in period of time at single crossing
(or vector expression), the peak period flow tuple D at single crossingthFor
Dth=[dth1,dth2,...,dth12]
The offpeak period flow tuple D at single crossing can similarly be soughttlFor
Dtl=[dtl1,dtl2,...,dtl12]
The unidirectional magnitude of traffic flow of single intersection is ranked up by step 1.3, then the stream of each intersection
Moment matrix is
Wherein, qIJIndicate that the traffic flow magnitude after the magnitude of traffic flow sorts, I ∈ [1,12] indicate that direction of travel, J indicate
Time interval, time interval are denoted as the period；In present embodiment, using daily 6 points to 23 points as a cycle, when according to 15min
Between interval the time of a cycle is divided into 72 time intervals, J ∈ [1,72]；
Step 1.4, according to DthAnd DtlIn d on identical direction of travelthiAnd dtliBy phase in the traffic matrix Q after sequence
Time interval with direction of travel corresponds to magnitude of traffic flow division low flow volume region SL, middle flow region SM, high flow volume region SH:
The length of three sequences is denoted as l1, l2, l3 respectively；For i direction of travel, S is taken respectivelyL, SM50 quartiles as i
The middle low discharge typical value r of direction of travelliWith middle flow typical value rmi, take SHHigh flow capacity generation of 70 quartiles as the i-th direction
Tabular value rhi:
Flow threshold matrix is expressed as R (I ∈ [1,12])；
Step 2 chooses the traffic information feature for influencing control work zone division, and the traffic information feature includes: traffic flow
Influence feature, intersection distance influences feature, traffic flow constitutes influence feature, cycle influences feature；
(1) force analysis model is influenced
The magnitude of traffic flow influences model when being that the magnitude of traffic flow when crossing is too small, the effect of coordinated control be it is unconspicuous, build
Vertical influence on traffic flow feature TF (Traffic flow influence coefficient):
Wherein, q (t) is t time interval flow, and σ is one directional flow of crossing t time interval, and λ is Different Traffic Flows side
To weight factor；The introducing of λ is actually that the TF value calculated is adjusted between 0 to 100；
(2) the influence model of intersection distance
The influence model of intersection distance is the important static state that adjacent intersection is traffic control sub-area division
The factor commonly uses division principle and uses crossing spacing minimax principle, sets adjacent intersection spacing as L；
Spacing maximum principle: upstream vehicle can cause during driving towards downstream road junction with the increase of crossing spacing
Discrete type enhancing, and make the coordinated control demand between crossing smaller and smaller.Traffic engineering experience is pointed out, vehicle dispersing is distinguished
Whether the crossing gap length within the scope of coordinated control is 800m.
Spacing minimum principle: any moment crossing dredging ability, which will meet discharge, is all circulated to current crossing
Otherwise vehicle will lead to vehicle queue and secondary queuing cause traffic congestion.
The partitioning standards of traffic control sub-district are according only to obtained by traffic experience in existing minimax spacing principle, and all
It is defined as static parameter, has ignored the dynamic traffic feature of current road network, so that the principle is being directed to variety classes, different periods
Road network carry out control work zone division during lose versatility.It is proposed by the present invention to be changed most based on dynamic flow
Greatly-minimum spacing principle can make up this defect, since road section length determines the traffic saturation of adjacent intersection, institute
It is to define road section length can successfully be exported the vehicle of the introducing of adjacent intersection and keep the discrete type between Adjacent vehicles
The important evidence of division rule.
Improved maximum-minimum spacing the principle of the present invention is as follows:
Maximum spacing principle:
ρ=avg (C1(t),C2(t))
In maximum spacing principle, RminThe average value of high flow capacity typical value, N in adjacent two intersectioninThe current green light period
Interior travelable direction of traffic；C1(t)、C2(t) previous intersection and the latter in adjacent two intersection is respectively indicated to hand over
The signal period of cross road mouth；TsampleIt is the length of chronomere, ρ is the average value in the signal period of adjacent intersection,
LmaxIndicate adjacent intersection maximum spacing, hcarIt is average length of car, NlaneIt is number of track-lines, β vehicle discrete distance, F traffic flow
Coefficient of dispersion；
Minimum spacing principle:
ρ=max (C1(t),C2(t))
In minimum spacing principle, RmaxThe larger value of high flow capacity typical value, T in adjacent two intersectionsampleIt is time list
The length of position, ρ is the larger value in the signal period of adjacent intersection, and μ is the anti-congestion factor, LminIndicate adjacent intersection most
Small spacing, hcarIt is average length of car；
At crossing on the basis of maximum-minimum principle, establishing intersection distance influences feature IF (Intersection
Distance influence coefficient):
Wherein, χ is balance factor；L is the length of current road segment；
(3) traffic flow, which is constituted, influences Modulus Model
If coming and going between two crossings without vehicle, coordinated control demand is not present between two crossings, conversely, if two crossings it is past
Come the main composition part that vehicle is traffic flow each other, then when only carrying out coordinated control to two crossings, just not will form traffic
Congestion.So traffic flow composition is the important parameter of traffic information feature, it is downstream road junction through vehicles middle and upper reaches crossing vehicle
Accounting, determine the discreteness size of wagon flow and the harmony size at upstream and downstream crossing, establish traffic flow constitute influence feature
CI (Compositional influence):
CI=O*F
Wherein, O is that the vehicle of upstream straight trip continues the ratio of straight trip in downstream, and value range [0,1], F wanders about as a refugee for traffic
Dissipate coefficient；
(4) cycle influences model
Cycle influences model is the road when the cycle length between adjacent intersection has big difference, and can not form integer ratio
The necessity coordinated between mouthful will reduce, and establish cycle influences feature CE (Cycle effect):
Wherein, T1For the green time that the time in green time in two adjacent intersections is long, T2For two adjacent intersections
Time short green time in middle green time；
Step 3 establishes the influence power weight calculation model based on space-time characteristic, obtains the weight L' for being characterized rr:
Step 3.1 takes all samples of the feature r in traffic information feature to constitute set Xr, that is, feature
The set that all different time intervals obtain:
Xr=[x1,x2,...,xn]T
Wherein n is sample total number, and r ∈ [Isosorbide-5-Nitrae] is the feature label in traffic information feature, represents influence on traffic flow spy
Sign, intersection distance influence feature, traffic flow constitutes influence feature, cycle influences feature；x1,x2,...,xnFeature r respectively
The sample representation of corresponding characteristic value；
If a sample point xi'It is another sample point xj'One of closest k sample point, k=under normal conditions
5, then xi'And xj'Between there are a lines, establish the feature non-directed graph of each feature respectively using the rule；
Step 3.2, the weight W for calculating each edge in feature non-directed graphi'j', constitute weight matrix Wr, form an adjacent square
Battle array；
Step 3.3, the variance for calculating all samples of this feature；DrIt is diagonal matrix, diagonal entry value is each sample
The out-degree weight of this point and, and be based on DrIn element value obtain intermediate variable X 'r:
Dr=diag (Wr* 1) 1=(1,1 ..., 1)T
Out-degree weight and be a complete noun, calculating is cumulative from the weight on all sides of a node
With；
Thus the variance size of all samples of this feature is calculated
Var(Xr)=X 'rDrX′r
Step 3.4, known WrFor weight matrix, the influence power weight calculation of space-time characteristic is carried out to this feature, then is calculated
It can obtain
Lr=Dr-Wr
xri'、xrj'Respectively indicate the corresponding sample point x of feature ri'、xj'；
So the weight of feature r are as follows:
The cum rights community discovery algorithm (the Louvain algorithm that dynamic analog lumpiness divides) that step 4, dynamic analog lumpiness divide:
Louvain algorithm is optimized in the present invention, and it is excessive to improve its number of iterations during merging community
The phenomenon that, the present invention is not that real calculate can make module angle value maximum community structure during merging and calculating,
But by defining threshold value e, the number of iterations is controlled, when the true value of Δ Q is less than e, it may be determined that the community of the Weighted Directed Graph
Structure has tended towards stability, and does not need to carry out community's merging again.
Under conditions of the definition of known module degree, the cum rights community discovery algorithm that dynamic analog lumpiness divides in the present invention is realized
The division of traffic control sub-district, detailed process the following steps are included:
Step 4.1, by above-mentioned each feature calculated of the influence power weight calculation model based on space-time characteristic
Weight carries out product accumulation and obtains calculating cooperation index CC between intersection two-by-twoij(Coordination coefficient) makees
For the weight in the section, wherein R is respectively the value of IF, CI, CE, TF；
When step 4.2, initialization community's matrix, each node in figure is an independent community, society in original state
Area's number is consistent with node number, indicates the non-directed graph of having the right with adjacency matrix, calculate each node degree weight and；
Step 4.3, each node and neighborhood of nodes trial merge community's operation, obtain the preceding mould with after distribution of distribution
The value of lumpiness changing value Δ Q, record Δ Q value maximum and Δ Q are greater than 0 community's merging phase, and node is assigned to the adjacent segments
Community where point, otherwise remains unchanged；
Step 4.4 is less than threshold value, i.e. mould until the community structure of weighted-graph changes caused modularity changing value
The value of lumpiness is maximum, completes a union operation；
Step 4.5 compresses figure according to the merging community state of record, and all nodes in the same community reconstruct
For a new node, new communities' matrix is created, the weight on side is converted into the weight of new node ring, side between community between community's interior nodes
Weight be converted into the side right weight between new node；
Step 4.6, the calculating process that step 4.2 to step 4.5 is repeated, until the modularity changing value of entire figure is small
In threshold value, the final division for realizing traffic control sub-district.
Embodiment
1. data prediction
The data set that the present invention uses is that the GPS of 1.4 ten thousand, Chengdu taxi is recorded, the time from August in 2014 3 days to
August 30 days, and zero point is had ignored to the data of 6 points of this periods of morning, and wash duplicate and abnormal note
It records, includes taxi ID, dimension, longitude and timestamp in every record, the time interval of every record is 60s.The shape of data
Formula is as follows:
1,30.624806,104.136604,1,2014/8/3 21:18:46
1,30.624809,104.136612,1,2014/8/3 21:18:15
1,30.624811,104.136587,1,2014/8/3 21:20:17
1,30.624811,104.136596,1,2014/8/3 21:19:16
When a large amount of tracing points are shown in chronological order, the form of traffic, the i.e. shape of road network can be shown.So real
It tests by the way that longitude and latitude is mapped as plane coordinates, coordinate is connected in timestamp order and forms track of vehicle, it is true by showing
The route map of track shows road network, can compare the shape of track road network Yu true map as shown in Figure 4, Figure 5.It can be seen that
Not having very big difference between true road network is shown to the time sequencing of tracing point, it was demonstrated that the accuracy of track point data and true
Reality, it is ensured that based on value and reliability that tracing point is tested, demonstrate data without biggish noise data.
Because of the similarity for focusing on crossing of research, then the type at crossing must be unified, the crossing that the present invention studies
Type is unified for crossroad.The number of crossroad, the geographical coordinate of crossroad, crossroad can pass through direction number and
Connectivity between crossroad be all it needs to be determined that characteristic point.Crossroad clearly apparent region is selected, matlab is utilized
In ginput function obtain the true coordinate of crossroad, the present invention has chosen 35 crossroads altogether and studies, such as Fig. 6
It is shown.
2. experimental result and analysis
According to determining intersection regional scope, statistics commutes the vehicle number in the region and by each in every 15 minutes
The vehicle flowrate average value of period determines the high offpeak period of traffic flow, to calculate the high, medium and low flow threshold of traffic flow, thus
The dynamic flow threshold value of available each intersection realizes that threshold value is synchronous with traffic flow variation.High, medium and low flow threshold
For determining the coefficient of determination of model, and then dynamic calculates cooperation index of each intersection based on traffic flow character.Each ten
Word crossing has 12 different wagon flow direction of travel, for a crossroad, using 15min as time sampling interval, utilizes
Magnitude of traffic flow computation model counts the traffic flow data of each direction of travel in same intersection.Substituting into the magnitude of traffic flow influences mould
Type calculates the impact factor size based on flow of each intersection, similarly using each model calculate road network road section length,
Traffic flow composition ratio calculates the impact factor of each traffic characteristic.
The influence degree of each traffic characteristic in the area is different, and is needed through influence power weight calculation model root
The weight coefficient of this feature element is determined according to the data correlation inside each characteristic element.Present invention determine that traffic flow, section are long
Degree, traffic components, signal period totally four traffic characteristics share 3816 sample elements inside each feature, are substituting to influence
The weight coefficient of traffic characteristic is obtained in power weight calculation model, as shown in Figure 7.
The magnitude of traffic flow, road section length, friendship in the figure to be calculated by space-time characteristic influence power weight calculation model
The weight number that logical period and traffic flow are constituted, respectively 30.37%, 13.46%, 22.88% and 33.29%.
Using traffic characteristic in conjunction with weight coefficient after the characteristic value that is calculated as the weight in each section, be updated to dynamic
In the Louvain algorithm that modularity divides, during the topological features of road network are dissolved into traffic control sub-area division.
Road network is indicated in the form of adjacency list, as the input data for the Louvain algorithm that dynamic analog lumpiness divides, to road network
Community's division is constantly carried out, until the changing value of modularity is less than threshold value e, i.e. the community of road network combines and do not have for modularity size
Have an impact.It is as shown in Figure 8 that control work zone after calculating divides state.
35 right-angled intersections are divided into 5 traffic control sub-districts, wherein the intersection in identical shadow region belongs to
In same traffic control sub-district, i.e. (1,2,3,4,5,13,14,15), (6,7,8,19,20,21,27,28), (9,10,11,12,
22,23,24,33,34,35), (16,17,18,25,26,29,30), (31,32).Each control work zone needs are individually believed
It is as shown in Figure 9 to have carried out community's combination, bonding state twice in community's partition process altogether for number coordinated control.
3. emulation and verifying
VISSIM simulation software is the minitype simulation modeling tool based on time section and driving behavior, for realizing city
With the modeling of public transport operation.It analyzes under various transportation condition assembled states, such as the setting of lane quantizating index, traffic components ginseng
Number adjustment, traffic signalization adjustment etc., the operation conditions in city and public transport are that evaluation engineering designs realisation and city
Plan the emulation tool of validity.Software inhouse includes traffic simulator and signal condition generator, is realized and is detected by interface
Device data and signal status information it is shared.VISSIM can export various statistical data offline, such as: journey time is lined up length
Degree etc..
The present invention is carried out using the mathematical model that VISSIM4.3 microscopic traffic simulation software divides control work zone to the present invention
Verifying.VISSIM simulation software is run in 7 operating system of Windows as shown in Figure 10, and true road network is imported into software
In.
Every section of road section length determines that all roads all have two lanes in this experiment, according to friendship with real section length
Logical periodic model inputs the traffic light cycle of each intersection, and traffic signals are using two phase place form, signal period in experiment
120s, by the signal time distributing conception for each intersection that dynamic traffic periodic model is calculated.
The traffic flow modes that will acquire are input in VISSIM simulation software, and stroke is arranged in road network, each for recording
Vehicle enters and leaves the time span of road network, when acquisition target such as the data acquisition, queue length, stroke of test document is arranged
Between and delay time at stop etc..
It can be obtained under former traffic control state according to test document, the travel condition of vehicle and the operation feelings of traffic
Condition.The control work zone partitioning model that the Louvain algorithm divided using dynamic analog lumpiness is calculated imports VISSIM emulation
In software, to the crossroad Uniform Traffic period in same control work zone, do not changing remaining traffic characteristic and test environment
Under conditions of, carry out analogue simulation.It is 15 minutes, i.e. 900s that it is consistent with the duration of chronomere, which to emulate duration, in this experiment.This
The road network, which is calculated, altogether by dynamic analog lumpiness Louvain algorithm in experiment need to divide 5 traffic control sub-districts, by dynamic traffic
The signal time distributing conception for each control work zone that periodic model is calculated is as shown in figure 11, and the period of same traffic control sub-district matches
When scheme it is identical.Amber light duration is arranged by Amber in the total duration for defining traffic light cycles in Figure 11 by Cycle Time,
The green light end time is set by GreenEnd, the red light end time is arranged by RedEnd, carries out complete two-phase traffic lights
Time cycle setting.
After simulation run, common property raw .rsr .vlr .mes .stz totally four files, record respectively journey time,
Delay time at stop, data acquire and the size of queue length.This experiment has selected 5 stroke routes altogether, and all strokes cover institute
Some intensive sections of vehicle and different traffic control sub-districts divide the effective of mathematical model for control work zone more of the present invention
Property, the present invention is compared in terms of average travel time and vehicle traveling information two.
1 journey time of table compares
As shown in Table 1, traffic control sub-area division scheme of the invention optimizes the traveling of 5 strokes to varying degrees
Time, and the distance of stroke is longer, and effect of optimization is more obvious the optimum results of the comprehensive a plurality of stroke of, average each stroke contracting
Short 22.4% travel time, improve out line efficiency.
The comprehensive driving parameters of 2 vehicle of table compare
As shown in Table 2, during traveling, the average queue length of each stroke reduces 30%, the contracting of mean delay time
Short by 44.6%, average overall travel speed increases by a small margin, accelerates 7.2%, and average stop frequency reduces 30%.Of the invention
Traffic control sub-area division scheme increases on 4 driving parameters, the trip experience of perfect people.
By obtained four based on journey time, delay time, the stop frequency and running speed index of evaluation document
It can be concluded that traffic control sub-area division mathematical model proposed by the invention has good effect in terms of optimizing vehicle driving
Fruit effectively reduces the running time of vehicle traveling over long distances, shortens the delay time in vehicle driving, increase road
Saturation degree and utilization rate, under identical trip distance, model of the present invention effectively reduces stop frequency, and to a certain extent
Improve running speed, complex optimum goes out line efficiency and road conditions.
Claims (9)
1. the adaptive traffic control sub-area division method based on Spatial Data Mining, which comprises the following steps:
Step 1, the feature extraction for carrying out high offpeak period and high, medium and low flow threshold；
Step 2 chooses the traffic information feature for influencing control work zone division；
The traffic information feature includes: influence on traffic flow feature TF, intersection distance influences feature IF, traffic flow constitutes shadow
Ring feature CI, cycle influences feature CE；
Step 3 establishes the influence power weight calculation model based on space-time characteristic, obtains the weight L' for being characterized rr, r ∈ [Isosorbide-5-Nitrae] is
Feature label in traffic information feature indicates that influence on traffic flow feature TF, intersection distance influence feature IF, traffic flow structure
At influence feature CI, cycle influences feature CE；
Step 4, the division that traffic control sub-district is realized based on the cum rights community discovery algorithm that dynamic analog lumpiness divides, detailed process
The following steps are included:
Step 4.1, by above-mentioned each feature weight calculated of the influence power weight calculation model based on space-time characteristic
Product accumulation is carried out to obtain calculating cooperation index CC between intersection two-by-twoijAs the weight in the section, wherein R be respectively IF,
The value of CI, CE, TF；
When step 4.2, initialization community's matrix, each node in figure is an independent community, community's number in original state
Mesh is consistent with node number, indicates the non-directed graph of having the right with adjacency matrix, calculate each node degree weight and；
Step 4.3, each node and neighborhood of nodes trial merge community's operation, obtain the preceding modularity with after distribution of distribution
The value of changing value Δ Q, record Δ Q value maximum and Δ Q are greater than 0 community's merging phase, and node is assigned to the adjacent node institute
Community, otherwise remain unchanged；
Step 4.4 is less than threshold value, i.e. modularity until the community structure of weighted-graph changes caused modularity changing value
Value it is maximum, complete a union operation；
Step 4.5 compresses figure according to the merging community state of record, and all nodes in the same community are reconstructed into one
A new node creates new communities' matrix, and the weight on side is converted into the weight of new node ring between community's interior nodes, the power on side between community
The side right weight being converted between new node again；
Step 4.6, the calculating process that step 4.2 to step 4.5 is repeated, until the modularity changing value of entire figure is less than threshold
Value, the final division for realizing traffic control sub-district.
2. the adaptive traffic control sub-area division method according to claim 1 based on Spatial Data Mining, feature
Be, the process of the feature extraction of the high offpeak period of progress and high, medium and low flow threshold described in step 1 the following steps are included:
Step 1.1, calculating are calculated from two road intersections by the magnitude of traffic flow of the different direction of travel of intersection
One road enter behind intersection begin to different directions road the magnitude of traffic flow；
Step 1.2 divides high offpeak period according to the corresponding magnitude of traffic flow of each time interval:
After the magnitude of traffic flow for calculating intersection difference direction of travel, on the basis of high ebb Time segments division, building solution height,
In, the mathematical model of low discharge threshold value；Design sets dynamic according to the real-time magnitude of traffic flow, for different high slack hours sections
Weighting parameter sj, define weight formula sj
sj=α/γ
Wherein, α is the magnitude of traffic flow of current peak period intersection, and γ is that all peak period magnitudes of traffic flow in current crossing are total
With；
Then establish the magnitude of traffic flow of single crossing peak period:
Wherein, j ∈ (peak period), the direction of travel of i ∈ [1,12] and vehicle, num indicate peak period number；
It establishes: the element group representation of the high and low peak discharge in period of time at single crossing, the peak period flow tuple D at single crossingthFor Dth
=[dth1,dth2,...,dth12]
The offpeak period flow tuple D at single crossing can similarly be soughttlFor
Dtl=[dtl1,dtl2,...,dtl12]
The unidirectional magnitude of traffic flow of single intersection is ranked up by step 1.3, then the flow square of each intersection
Battle array be
Wherein, qIJIndicate that the traffic flow magnitude after the magnitude of traffic flow sorts, I ∈ [1,12] indicate that direction of travel, J indicate the time
Section, time interval are denoted as the period；
Step 1.4, according to DthAnd DtlIn d on identical direction of travelthiAnd dtliBy identical traveling in the traffic matrix Q after sequence
The time interval in direction corresponds to the magnitude of traffic flow and divides low flow volume region SL, middle flow region SM, high flow volume region SH:
The length of three sequences is denoted as l1, l2, l3 respectively；For i direction of travel, S is taken respectivelyL, SM50 quartiles as i advance
The middle low discharge typical value r in directionliWith middle flow typical value rmi, take SHHigh flow capacity typical value of 70 quartiles as the i-th direction
rhi:
Flow threshold matrix is expressed as R (I ∈ [1,12])；
3. the adaptive traffic control sub-area division method according to claim 2 based on Spatial Data Mining, feature
Be, the detailed process of step 1.1 the following steps are included:
Step 1.1.1, two road intersections of intersection are divided into and are denoted as the first road and the second road, the first road and
The respective both sides of the edge of two roads are as boundary；Two road intersections are divided into 5 regions, respectively crossroad by intersection
Mouth, the first road junction leading portion, the first road junction back segment, the second road junction leading portion, the second road are handed over
Cross road mouth back segment；
Step 1.1.2, judge the original driving direction of vehicle, i.e. vehicle enters intersection from which direction；
Then according to the subsequent course bearing of current vehicle, judge the direction of travel of vehicle；
Step 1.1.3, according to the time tag of current trace points, judge time interval belonging to current vehicle；
Step 1.1.4, current vehicle is added up to magnitude of traffic flow matrix (μit, λ) in, i indicates direction of travel, i ∈ [1,12]；t
Indicate the serial number of time interval, μitIndicate the magnitude of traffic flow of i direction of travel t time interval；λ is crossing mark.
4. the adaptive traffic control sub-area division method according to claim 1 based on Spatial Data Mining, feature
It is, influence on traffic flow feature described in step 2 is as follows:
Influence on traffic flow feature TF:
Wherein, q (t) is t time interval flow, and σ is one directional flow of crossing t time interval, and λ is Different Traffic Flows direction
Weight factor；The introducing of λ is actually that the TF value calculated is adjusted between 0 to 100.
5. the adaptive traffic control sub-area division method according to claim 1 based on Spatial Data Mining, feature
It is, it is as follows that the distance of intersection described in step 2 influences feature:
Adjacent intersection spacing is set as L；
Maximum-minimum spacing principle is as follows:
Maximum spacing principle:
ρ=avg (C1(t),C2(t))
In maximum spacing principle, RminThe average value of high flow capacity typical value, N in adjacent two intersectioninIt can in the current green light period
The direction of traffic of traveling；C1(t)、C2(t) previous intersection and the latter crossroad in adjacent two intersection are respectively indicated
The signal period of mouth；TsampleIt is the length of chronomere, ρ is the average value in the signal period of adjacent intersection, LmaxTable
Show adjacent intersection maximum spacing, hcarIt is average length of car, NlaneIt is number of track-lines, β vehicle discrete distance, the discrete system of F traffic flow
Number；
Minimum spacing principle:
ρ=max (C1(t),C2(t))
In minimum spacing principle, RmaxThe larger value of high flow capacity typical value, T in adjacent two intersectionsampleIt is chronomere
Length, ρ are the larger value in the signal period of adjacent intersection, and μ is the anti-congestion factor, LminBetween indicating that adjacent intersection is minimum
Away from hcarIt is average length of car；
At crossing on the basis of maximum-minimum principle, establishing intersection distance influences feature IF:
Wherein, χ is balance factor；L is the length of current road segment.
6. the adaptive traffic control sub-area division method according to claim 1 based on Spatial Data Mining, feature
It is, it is as follows that traffic flow described in step 2 constitutes influence feature:
Traffic flow composition is the important parameter of traffic information feature, for accounting for for downstream road junction through vehicles middle and upper reaches crossing vehicle
Than determining the discreteness size of wagon flow and the harmony size at upstream and downstream crossing, establishing traffic flow and constituting influences feature CI:
CI=O*F
Wherein, O is that the vehicle of upstream straight trip continues the ratio of straight trip, value range [0,1] in downstream, and F is the discrete system of traffic flow
Number.
7. the adaptive traffic control sub-area division method according to claim 1 based on Spatial Data Mining, feature
It is, cycle influences feature described in step 2 is as follows:
Establish cycle influences feature CE:
Wherein, T1For the green time that the time in green time in two adjacent intersections is long, T2It is green in two adjacent intersections
Time short green time in the lamp time.
8. according to claim 1 to the adaptive traffic control sub-area division method described in one of 7 based on Spatial Data Mining,
It is characterized in that, establishing the influence power weight calculation model based on space-time characteristic described in step 3, the weight L' for being characterized r is obtainedr
Process it is as follows:
Step 3.1 takes all samples of the feature r in traffic information feature to constitute set Xr:
Xr=[x1,x2,...,xn]T
Wherein n be sample total number, r ∈ [Isosorbide-5-Nitrae] be traffic information feature in feature label, represent influence on traffic flow feature,
Intersection distance influences feature, traffic flow constitutes influence feature, cycle influences feature；x1,x2,...,xnFeature r is corresponding respectively
Characteristic value sample representation；
If a sample point xi'It is another sample point xj'One of closest k sample point, then xi'And xj'Between there are one
Side, the feature non-directed graph of each feature is established using the rule respectively；
Step 3.2, the weight W for calculating each edge in feature non-directed graphi'j', constitute weight matrix Wr, form an adjacency matrix；
Step 3.3, the variance for calculating all samples of this feature；DrIt is diagonal matrix, diagonal entry value is each sample point
Out-degree weight and, and be based on DrIn element value obtain intermediate variable Xr':
Dr=diag (Wr* 1) 1=(1,1 ..., 1)T
Thus the variance size of all samples of this feature is calculated
Var(Xr)=Xr′DrXr′
Step 3.4, known WrFor weight matrix, the influence power weight calculation of space-time characteristic is carried out to this feature, then can be calculated
Lr=Dr-Wr
xri'、xrj' respectively indicate the corresponding sample point x of feature ri'、xj'；
So the weight of feature r are as follows:
9. the adaptive traffic control sub-area division method according to claim 8 based on Spatial Data Mining, feature
It is, the k=5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811332450.5A CN109410577B (en) | 2018-11-09 | 2018-11-09 | Self-adaptive traffic control subarea division method based on space data mining |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811332450.5A CN109410577B (en) | 2018-11-09 | 2018-11-09 | Self-adaptive traffic control subarea division method based on space data mining |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109410577A true CN109410577A (en) | 2019-03-01 |
| CN109410577B CN109410577B (en) | 2020-10-09 |
Family
ID=65472438
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811332450.5A Active CN109410577B (en) | 2018-11-09 | 2018-11-09 | Self-adaptive traffic control subarea division method based on space data mining |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109410577B (en) |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109934850A (en) * | 2019-03-21 | 2019-06-25 | 北京沃东天骏信息技术有限公司 | The methods, devices and systems that moving target counts |
| CN110033613A (en) * | 2019-03-07 | 2019-07-19 | 吉林建筑大学 | Smart city management method and system based on regional traffic synchronism |
| CN110111567A (en) * | 2019-04-23 | 2019-08-09 | 刘畅 | A kind of traffic control sub-area division method and system based on modularity assessment |
| 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 |
| CN110634287A (en) * | 2019-08-26 | 2019-12-31 | 上海电科智能系统股份有限公司 | Urban traffic state refined discrimination method based on edge calculation |
| CN111145548A (en) * | 2019-12-27 | 2020-05-12 | 银江股份有限公司 | Important intersection identification and subregion division method based on data field and node compression |
| CN111192457A (en) * | 2020-03-20 | 2020-05-22 | 青岛海信网络科技股份有限公司 | Method and device for controlling urban-level integral traffic signals |
| 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 |
| CN111275964A (en) * | 2020-01-14 | 2020-06-12 | 浙江浙大中控信息技术有限公司 | Road section correlation matrix calculation method based on checkpoint data |
| 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 |
| CN111915904A (en) * | 2019-05-07 | 2020-11-10 | 阿里巴巴集团控股有限公司 | Track processing method and device and electronic equipment |
| CN112187499A (en) * | 2019-07-03 | 2021-01-05 | 四川大学 | Device partition management and division method in device network |
| CN113129614A (en) * | 2020-01-10 | 2021-07-16 | 阿里巴巴集团控股有限公司 | Traffic control method and device and electronic equipment |
| CN114973699A (en) * | 2022-05-10 | 2022-08-30 | 阿波罗智联(北京)科技有限公司 | Traffic control signal generation method, edge calculation unit and road side unit |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2002159048A (en) * | 2000-11-22 | 2002-05-31 | Yrp Mobile Telecommunications Key Tech Res Lab Co Ltd | Cdma mobile communication system |
| KR20050051956A (en) * | 2003-11-28 | 2005-06-02 | 주식회사 비츠로시스 | Control system and methdod for local divisional traffic signal |
| CN101751777A (en) * | 2008-12-02 | 2010-06-23 | 同济大学 | Dynamic urban road network traffic zone partitioning method based on space cluster analysis |
| CN105825690A (en) * | 2016-06-15 | 2016-08-03 | 北京航空航天大学 | Coordinated control oriented trunk line crossing correlation analysis and division method |
| CN105869401A (en) * | 2016-05-12 | 2016-08-17 | 华南理工大学 | Road network dynamic partitioning method based on different crowdedness degrees |
| 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 |
| CN108388961A (en) * | 2018-02-06 | 2018-08-10 | 华东师范大学 | Self-adapting random neighbours' community detecting algorithm based on modularity optimization |
-
2018
- 2018-11-09 CN CN201811332450.5A patent/CN109410577B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2002159048A (en) * | 2000-11-22 | 2002-05-31 | Yrp Mobile Telecommunications Key Tech Res Lab Co Ltd | Cdma mobile communication system |
| KR20050051956A (en) * | 2003-11-28 | 2005-06-02 | 주식회사 비츠로시스 | Control system and methdod for local divisional traffic signal |
| CN101751777A (en) * | 2008-12-02 | 2010-06-23 | 同济大学 | Dynamic urban road network traffic zone partitioning method based on space cluster analysis |
| CN105869401A (en) * | 2016-05-12 | 2016-08-17 | 华南理工大学 | Road network dynamic partitioning method based on different crowdedness degrees |
| CN105825690A (en) * | 2016-06-15 | 2016-08-03 | 北京航空航天大学 | Coordinated control oriented trunk line crossing correlation analysis and division method |
| CN108388961A (en) * | 2018-02-06 | 2018-08-10 | 华东师范大学 | Self-adapting random neighbours' community detecting algorithm based on modularity optimization |
| 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 |
Non-Patent Citations (5)
| Title |
|---|
| 朱芸: "基于交通流预测的控制子区交通状态识别技术研究", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 * |
| 李瑞敏等: "交通信号控制子区模糊动态划分方法研究", 《武汉理工大学学报》 * |
| 王霖青: "交通小区划分问题的整数规划建模与优化算法研究", 《中国博士学位论文全文数据库 工程科技Ⅱ辑》 * |
| 赵伟明: "面向交通控制的时段划分与子区划分", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 * |
| 郭海锋等: "短时交通状态预测下交通控制子区自动划分方法", 《系统科学与数学》 * |
Cited By (24)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110033613B (en) * | 2019-03-07 | 2020-07-10 | 吉林建筑大学 | Smart city management method and system based on regional traffic synchronism |
| CN110033613A (en) * | 2019-03-07 | 2019-07-19 | 吉林建筑大学 | Smart city management method and system based on regional traffic synchronism |
| CN109934850A (en) * | 2019-03-21 | 2019-06-25 | 北京沃东天骏信息技术有限公司 | The methods, devices and systems that moving target counts |
| CN109934850B (en) * | 2019-03-21 | 2021-04-30 | 北京沃东天骏信息技术有限公司 | Method, device and system for counting moving objects |
| CN110111567B (en) * | 2019-04-23 | 2021-05-18 | 刘畅 | Traffic control subarea division method and system based on modularity evaluation |
| CN110111567A (en) * | 2019-04-23 | 2019-08-09 | 刘畅 | A kind of traffic control sub-area division method and system based on modularity assessment |
| CN111915904A (en) * | 2019-05-07 | 2020-11-10 | 阿里巴巴集团控股有限公司 | Track processing method and device and electronic equipment |
| CN112187499B (en) * | 2019-07-03 | 2021-12-03 | 四川大学 | Device partition management and division method in device network |
| CN112187499A (en) * | 2019-07-03 | 2021-01-05 | 四川大学 | Device partition management and division method in device network |
| 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 |
| CN110415523B (en) * | 2019-08-13 | 2021-07-30 | 东南大学 | Signal control subarea division method based on vehicle travel track data |
| CN110634287A (en) * | 2019-08-26 | 2019-12-31 | 上海电科智能系统股份有限公司 | Urban traffic state refined discrimination method based on edge calculation |
| WO2021036278A1 (en) * | 2019-08-26 | 2021-03-04 | 上海电科智能系统股份有限公司 | Edge computing-based fine determination method for urban traffic state |
| CN110634287B (en) * | 2019-08-26 | 2021-08-17 | 上海电科智能系统股份有限公司 | Urban traffic state refined discrimination method based on edge calculation |
| US11361658B1 (en) | 2019-08-26 | 2022-06-14 | Shanghai Seari Intelligent System Co., Ltd. | Edge computing-based method for fine determination of urban traffic state |
| CN111145548A (en) * | 2019-12-27 | 2020-05-12 | 银江股份有限公司 | Important intersection identification and subregion division method based on data field and node compression |
| CN111145548B (en) * | 2019-12-27 | 2021-06-01 | 银江股份有限公司 | Important intersection identification and subregion division method based on data field and node compression |
| 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 |
| CN113129614A (en) * | 2020-01-10 | 2021-07-16 | 阿里巴巴集团控股有限公司 | Traffic control method and device and electronic equipment |
| CN111275964A (en) * | 2020-01-14 | 2020-06-12 | 浙江浙大中控信息技术有限公司 | Road section correlation matrix calculation method based on checkpoint data |
| CN111325979B (en) * | 2020-02-28 | 2021-07-16 | 海信集团有限公司 | Method and device for dividing traffic control multistage subareas |
| 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 |
| CN111192457A (en) * | 2020-03-20 | 2020-05-22 | 青岛海信网络科技股份有限公司 | Method and device for controlling urban-level integral traffic signals |
| CN114973699A (en) * | 2022-05-10 | 2022-08-30 | 阿波罗智联(北京)科技有限公司 | Traffic control signal generation method, edge calculation unit and road side unit |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109410577B (en) | 2020-10-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109410577A (en) | Adaptive traffic control sub-area division method based on Spatial Data Mining | |
| CN104778834B (en) | Urban road traffic jam judging method based on vehicle GPS data | |
| Zhan et al. | Citywide traffic volume estimation using trajectory data | |
| Fu et al. | Empirical analysis of large-scale multimodal traffic with multi-sensor data | |
| GB2599765A (en) | Vehicle traffic flow prediction method with missing data | |
| CN106228808B (en) | City expressway travel time prediction method based on Floating Car space-time grid data | |
| CN111275965B (en) | Real-time traffic simulation analysis system and method based on internet big data | |
| CN110709908A (en) | Computer system and method for state prediction of traffic system | |
| WO2019047905A1 (en) | Road traffic analysis system, method and apparatus | |
| CN110111574B (en) | Urban traffic imbalance evaluation method based on flow tree analysis | |
| CN112489426B (en) | Urban traffic flow space-time prediction scheme based on graph convolution neural network | |
| EP2590151A1 (en) | A framework for the systematic study of vehicular mobility and the analysis of city dynamics using public web cameras | |
| CN102842219B (en) | Forecasting method and system | |
| Yao et al. | An optimization model for arterial coordination control based on sampled vehicle trajectories: The STREAM model | |
| CN106157624B (en) | More granularity roads based on traffic location data shunt visual analysis method | |
| Mahut et al. | Calibration and application of a simulation-based dynamic traffic assignment model | |
| CN113806419B (en) | Urban area function recognition model and recognition method based on space-time big data | |
| CN108648445A (en) | Dynamic traffic Tendency Prediction method based on traffic big data | |
| CN110070720B (en) | Calculation method for improving fitting degree of traffic capacity model of intersection road occupation construction area | |
| CN113868492A (en) | Visual OD (origin-destination) analysis method based on electric police and checkpoint data and application | |
| CN111009140B (en) | Intelligent traffic signal control method based on open-source road condition information | |
| Wei et al. | A two-layer network dynamic congestion pricing based on macroscopic fundamental diagram | |
| Hu et al. | Traffic jams prediction method based on two-dimension cellular automata model | |
| Yu et al. | A review of the link traffic time estimation of urban traffic | |
| CN111275961A (en) | Urban traffic running state feature calculation method based on floating car data |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |