CN105225503A - Traffic control subarea is optimized and self-adapting regulation method - Google Patents

Abstract

The invention discloses the optimization of a kind of traffic control sub-area division and self-adapting regulation method, traffic control sub-area division is carried out based on the category of roads of static state and dynamic arithmetic for real-time traffic flow, determine the traffic control subarea needing to carry out signal lamp interlock management and control and manual intervention according to saturation degree index, and traffic control community is upgraded automatically.The comprehensive category of roads of the present invention and arithmetic for real-time traffic flow carry out static state to traffic subarea and dynamically divide, and effectively improve sub-area division efficiency; Determine dissimilar traffic control subarea according to saturation degree, be conducive to building rational signal lamp linkage tube prosecutor case and manual intervention scheme; According to arithmetic for real-time traffic flow change, repartition for some traffic control communities changed, avoid the redundant computation that generation repartitioned by system-wide net.

Description

Traffic control subarea is optimized and self-adapting regulation method

Technical field

The present invention relates to urban traffic control sub-area division method, particularly relate to a kind of traffic control sub-area division in conjunction with category of roads and arithmetic for real-time traffic flow and region adaptivity method of adjustment.

Background technology

Sub-area division carries out the control of regional coordination signal and determines the prerequisite of reasonable traffic control measure and scope.Huge, complicated road network is divided into several according to certain Principles and ways and evaluates subarea by it, thus the relation of transportation supplies and transport need in acquisition subarea, and then determine the signal coordinated control strategy or traffic control measure etc. of different traffic control community.

In current existing sub-area division research, mainly contain static state and dynamically divide two kinds of methods.Static division, owing to can not carry out Reasonable adjustment according to real-time traffic situation in time, divides conversion to the dynamic traffic control community based on arithmetic for real-time traffic flow just gradually, but dynamic traffic community divides counting yield to be often difficult to meet current demand.How static division is combined with dynamically dividing, improve the real-time of sub-area division and to raise the efficiency be a direction being worth breakthrough.Meanwhile, different category of roads is due to the difference of road structure, stream characteristics, and generally speaking, road relevance that is identical or adjacent road grade is stronger.But category of roads is included in the consideration of traffic control community division and is gone by rarer research.In addition, signal control work zone partition problem is a dynamic problem, traffic flow change in city road network has very strong period characteristic, original signal control work zone can be made not support well, and coordinating signal controls, when signal control work zone no longer meets division principle, need to carry out dynamic conditioning according to the real-time change situation of traffic flow to control work zone.Existing signal lamp dynamic conditioning mostly is manually determines interval sometime, and division is recalculated in unification, but owing to disturbing by the time strong in part traffic control community, does not need to divide, result in redundant computation.

Summary of the invention

For overcoming above-mentioned the deficiencies in the prior art, the invention provides a kind of traffic control subarea to optimize and self-adapting regulation method, for improving sub-network division efficiency, accuracy and real-time, take different traffic dispersion strategies for differently traffic control cell type better.

The present invention has carried out the integration of system to " traffic zone based on arithmetic for real-time traffic flow divides, the real-time update of traffic control community ", and proposed before dynamic partitioning traffic control community, first based on category of roads, signal lamp crossing is carried out to the method for static traffic community division, and according to the traffic flow real-time change of dynamic traffic control community, targetedly again division is upgraded to signal lamp crossing, local.

To achieve these goals, the present invention proposes a kind of traffic control subarea and optimizes and self-adapting regulation method, comprises the following steps:

S1. build road network and traffic flow data storehouse, specify road network topology relation and category of roads;

S2. according to road network topology relation and category of roads, Preliminary division is carried out to signal lamp intersection type, calculate the physical similarity between two between signal lamp crossing, signal lamp crossing is divided into one or more static traffic according to physical similarity and controls community;

S3. the static traffic obtained with step S2 controls community for benchmark, in conjunction with arithmetic for real-time traffic flow, the traffic flow degree of association of further calculating between two between signal lamp crossing, is divided into multiple dynamic traffic control subarea according to the traffic flow degree of association by each static traffic control work zone;

S4. calculate each signal lamp crossing present period saturation degree, on this basis, calculate the saturation degree of each dynamic traffic control subarea at present period; Wagon flow is divided into by dynamic traffic control microzonation to pass unimpeded community, signal lamp Collaborative Control community and Collaborative Control community three types in many ways according to dynamic traffic control subarea saturation degree;

S5. judge whether current time reaches the default time apart from the moment of dynamic traffic control last time subarea adjustment, original dynamic traffic control microzonation offshoot program is maintained constant when not reaching, then recalculate each dynamic traffic control cell signal lamp intersection saturation degree when reaching, and according to each dynamic traffic control cell signal lamp intersection saturation degree, cell type is divided; If meet one of following two conditions, then community is controlled to the static traffic at the place, dynamic traffic control community changed and repartition: (1) exists dynamic traffic control cell type and changes; (2) the signal lamp crossing existed in dynamic traffic control community corresponding to saturation degree maximal value changes.

Compared with prior art, the invention has the beneficial effects as follows: consider category of roads and arithmetic for real-time traffic flow carries out static state to traffic subarea and dynamically divides, effectively improve sub-area division efficiency; Determine dissimilar traffic control subarea according to saturation degree, be conducive to building rational signal lamp linkage tube prosecutor case and manual intervention scheme; According to arithmetic for real-time traffic flow change, repartition for some traffic control communities changed, avoid the redundant computation that generation repartitioned by system-wide net.Method overcomes original traffic control sub-area division static and dynamic Status sub-area division and combines the problems such as not enough, control work zone traffic plan specific aim is strong, counting yield is on the low side and adaptive ability is not enough.

Accompanying drawing explanation

Fig. 1 is the process flow diagram of the inventive method.

Fig. 2 is six type signal lamp crossing schematic diagram.

Fig. 3 be based on category of roads static traffic control microzonation divide process flow diagram.

Fig. 4 divides process flow diagram based on the dynamic traffic control microzonation of arithmetic for real-time traffic flow.

Fig. 5 is the traffic control cell classification process flow diagram based on signal lamp intersection saturation degree.

Fig. 6 is that traffic control community adjusts process flow diagram in real time.

Fig. 7 is road network schematic diagram.

Embodiment

Below in conjunction with accompanying drawing, the present invention is further described.

Techniqueflow of the present invention as shown in Figure 1, comprises database sharing, controls community divide, divide based on the dynamic traffic control community of arithmetic for real-time traffic flow, adjust five steps based on the signal lamp traffic control cell classification of signal lamp intersection saturation degree, traffic control subdistrict self-adaptive based on the static traffic of category of roads.

1.1 database sharing:

Build road network and traffic flow data storehouse, specify road network topology relation and category of roads; Road network topology relation main manifestations is the syntople between signal lamp crossing, and category of roads is divided into major urban arterial highway, city subsidiary road, Urban Branch Road three major types.

1.2 based on category of roads static traffic control community divide:

1) road net is represented, wherein V={v with G=(V, E) ₁, v ₂..., v _n, v _irepresent i-th signal control signal lamp crossing, V is all signal control signal lamp crossings set in network; E={e ₁₂, e ₂₃..., e _ij, e _ijrepresent the section between connection signal control signal lamp crossing i and j, its weight w _ijrepresent the relevance of signal lamp crossing i and j, E is the set in all sections in network.

Different according to category of roads, signal lamp crossing is divided into major trunk roads and major trunk roads, major trunk roads and subsidiary road, major trunk roads and branch road, subsidiary road and subsidiary road, subsidiary road and branch road, branch road and branch road six type; And to six class signal lamp intersection type assignment respectively:

R _i＝θ，1≤θ≤6，θ∈n

Wherein, R _i=1 is major trunk roads and arterial-type signal lamp crossing, R _i=2 is major trunk roads and subsidiary road type signal lamp crossing, R _i=3 is major trunk roads and tributary type signals lamp crossing, R _i=4 is subsidiary road and subsidiary road type signal lamp crossing, R _i=5 is subsidiary road and tributary type signals lamp crossing, R _i=6 is branch road and tributary type signals lamp crossing;

The physical similarity in road network G between adjacent communications signals lamp crossing is made to be w _{(i, j)}represent the relevance of signal lamp crossing i and j, R _i, R _jbe respectively the types value of signal lamp crossing i, j, d _ijrepresent the distance of signal lamp crossing i and signal lamp crossing j; When signal lamp crossing i is not communicated with signal lamp crossing j, corresponding w _{(i, j)}be 0, e be natural constant;

2) adjacency matrix of road network G is H, when signal lamp crossing i in road network G is communicated with signal lamp crossing j is oriented, namely there is a section e _ijwhen being connected, the element a in H _ij=1, when signal lamp crossing i and signal lamp crossing j does not have a section e _ijwhen being connected, a _ij=0; As i=j, a _ij=0;

The weighted adjacency matrix of road network G is the element in W, W:

$w_{ij}=\left\{\begin{array}{cc}w(i,j),& a_{ij}=1\\ 0,& a_{ij}=0\end{array}\right.$

Diagonal matrix D=diag{d _i, d _i=∑ _jw (i, j)

3) a symmetric matrix L having n node, have the Laplace matrix of the non-directed graph of weight to be n × n dimension.If have limit to connect between these two nodes, then L _ijfor negative, otherwise be 0.Therefore matrix L can be expressed as L=D-W, wherein, D is a diagonal matrix, and the element on its diagonal line is with regard to the degree of each node corresponding, and W is then the weighted adjacency matrix of this network.

4) the Laplacian little eigenwert of matrix second and Fiedler vector is calculated.Traffic control community partition problem can be converted into the segmentation problem to figure G=(V, E).

All row of L matrix and row and be all 0, therefore, this matrix always has an eigenwert 0, then all elements of its characteristic of correspondence vector is all 1.If G is UNICOM, so the second little eigenvalue λ ₂for positive number, and in its characteristic of correspondence vector and Fiedler vector, the numerical values recited of each element (comprising positive and negative two kinds of elements) reflects the mutual relationship of its corresponding vertex.

5) according to Fiedler proper vector F=(f ₁, f ₂..., f _n) in the numerical value of each element, adopt dichotomy to split each key element of Fiedler vector, a subregion be divided into two.If the number of partitions reaches K, stop Fiedler segmentation, if the total number of partitions of road network does not reach K, return step 3), select the subregion that signal lamp crossing number is maximum, set up the Laplace matrix that this subregion is corresponding, and repeat step 4) and 5), solve Fiedler vector and carry out the segmentation of Fiedler vector; Concrete dividing method is: select critical value S=0, split, by f element each in Fiedler proper vector F _ia subregion is assigned to, all the other f in the summit of>=0 _ianother subregion is assigned to, i=1,2 in the summit of < 0 ..., n.

1.3 divide based on the dynamic traffic control community of arithmetic for real-time traffic flow:

1) divide according to static traffic control work zone, obtain K static traffic control work zone altogether, use G _k=(V _k, E _k) represent the road net of a corresponding kth static traffic control work zone, 1≤k≤K; Wherein V _k={ v _k1, v _k2..., v _kp..., v _kprepresent kth static traffic control work zone p signal lamp crossing, V _kit is all signal lamp crossings set in a kth static traffic control work zone; E _k={ e _k12, e _k23..., e _kpq, e _kpqrepresent the section between connection signal lamp crossing p and q, E _kfor the set in all sections in a kth static traffic control work zone;

2) in a kth static traffic control work zone, the traffic flow degree of association ρ of signal lamp crossing p and the q be connected _{k (p, q)}the computing method of recommending in U.S.'s " traffic control system handbook " are adopted to calculate:

${\mathrm{\&rho;}}_{k(p,q)}=\frac{0.5}{1+t}\&lsqb;\frac{n\&times;Q_{\max }}{{\mathrm{\&Sigma;}}_{i=1}^n{Q}_i}\&rsqb;-(N-2)$

Wherein, n is point number sailed into from the vehicle of stream signal lamp crossing; Q _ibe the volume of traffic that i-th branch arrives downstream signal lamp crossing, Q _maxfor arriving the volume of traffic maximal value of downstream signal lamp crossing in branch; for arriving the volume of traffic sum total of downstream signal lamp crossing; T is the running time of vehicle in two signal lamp crossings; N is the number of track-lines that downstream is driven towards in upstream; Due to signal lamp crossing p and signal lamp crossing q upstream and downstream signal lamp crossing each other, therefore, it is possible to obtain two ρ _{k (p, q)}value, gets the traffic flow degree of association ρ of both mean values as final signal lamp crossing p and q _{k (p, q)};

3) each static traffic control work zone road net G is calculated _kadjacency matrix be H _k, as road network G _kwhen middle signal lamp crossing p is communicated with signal lamp crossing q, H _kin element a _kpq=1, when signal lamp crossing p is not communicated with signal lamp crossing q, a _kpq=0; As p=q, a _kpq=0;

4) each static traffic control work zone road net G is calculated _kweighted adjacency matrix W _k, and diagonal angle adjacency matrix D _k, wherein, weighted adjacency matrix W _kin element w _kpqfor:

$w_{kpq}=\left\{\begin{array}{cc}{\mathrm{\&rho;}}_{k(p,q)},& a_{pq}=1\\ 0,& a_{pq}=0\end{array}\right.$

D _k＝diag{d _kp}，d _kp＝∑ _qρ _k(p,q)

5) road net signal lamp crossing, each static subarea Laplacian matrix L is calculated _k, wherein L _k=D _k-W _k;

6) according to the Laplacian matrix L built _k, solve the proper vector corresponding to the little eigenwert of this matrix second, i.e. Fiedler vector: F _k=(f _k1, f _k2..., f _km);

7) according to Fiedler proper vector F _k=(f _k1, f _k2..., f _km) in the numerical value of each element, adopt dichotomy to split each key element of Fiedler vector, a subregion be divided into two.If signal lamp crossing number maximal value is less than Z in subregion, stop Fiedler segmentation; If exist wherein in any one subregion signalized intersections number be more than or equal to Z, return step 5), select the subregion that signal lamp crossing number is maximum, set up the Laplace matrix that this subregion is corresponding, and repeat step 6) and 7), solve Fiedler vector and carry out the segmentation of Fiedler vector; Concrete dividing method is: select critical value S=0, to Fiedler proper vector F _kin each element split, by f _kpa subregion is assigned to, all the other f in the summit of>=0 _kpthe summit of < 0 assign to another subregion (p=1,2 ..., m); A final kth static traffic controls community will be divided into L dynamic traffic control community;

8) according to dynamic traffic control sub-area division, a kth static traffic control work zone correspondence obtains L dynamic traffic control subarea.Use G _kl=(V _kl, E _kl) represent l the dynamic traffic control subarea that a corresponding kth static traffic control work zone is corresponding, wherein 1 < k≤K, 1 < l≤L;

1.4 based on the traffic control cell classification of signal lamp intersection saturation degree:

1) each dynamic traffic control subarea G is calculated _klthe saturation degree of T period wherein k is that a kth static traffic controls cell number, and l is dynamic traffic control cell number, wherein, and 1 < k≤K, 1 < l≤L; Each signal lamp intersection saturation degree computing formula is as follows:

$S_{kl\mathrm{\&alpha;}}^T=\frac{V_{kl\mathrm{\&alpha;}}^T}{C_{kl\mathrm{\&alpha;}}^T}$

Wherein, α represents that signal lamp crossing is numbered, for dynamic traffic control subarea G _klsignal lamp crossing sum; for dynamic traffic control community G _klα signal lamp crossing at the volume of traffic of T period; for representing dynamic traffic control community G _klα signal lamp crossing in the maximum traffic capacity of T period;

2) the maximum traffic capacity computing formula as follows:

$C_{kl\mathrm{\&alpha;}}^T=\{\underset{\mathrm{\&gamma;}=1}{\overset{{\mathrm{\&Gamma;}}_{kl\mathrm{\&alpha;}1}^T}{\&Sigma;}}{C}_{kl\mathrm{\&alpha;}\mathrm{\&beta;}\mathrm{\&gamma;}}^T.\underset{\mathrm{\&gamma;}=1}{\overset{{\mathrm{\&Gamma;}}_{kl\mathrm{\&alpha;}2}^T}{\&Sigma;}}{C}_{kl\mathrm{\&alpha;}\mathrm{\&beta;}\mathrm{\&gamma;}}^T,\mathrm{...},\underset{\mathrm{\&gamma;}=1}{\overset{{\mathrm{\&Gamma;}}_{kl\mathrm{\&alpha;}\mathrm{\&beta;}}^T}{\&Sigma;}}{C}_{kl\mathrm{\&alpha;}\mathrm{\&beta;}\mathrm{\&gamma;}}^T,\mathrm{...}\}$

The maximum traffic capacity of each import in each element respective signal lamp crossing; Wherein, α is signal lamp crossing numbering, and β is α signal lamp crossing inlet numbering, and γ is β import lane number, and 0 < α≤A, A are dynamic traffic control community, place G _klsignal lamp crossing quantity; 0 < β≤B, B are the import volume of signal lamp crossing, place α; 0 < γ≤Γ; Γ is the track quantity of place import β; the dynamic traffic control community G in the T period _klthe traffic capacity in γ track of α β the import in signal lamp crossing, for the dynamic traffic control community G in the T period _klβ the import numbering import track sum α signal lamp crossing;

3) volume of traffic get the data on flows that each import records at the ground induction coil of T period, formula is as follows:

$V_{kl\mathrm{\&alpha;}}^T=\{\underset{\mathrm{\&gamma;}=1}{\overset{{\mathrm{\&Gamma;}}_{kl\mathrm{\&alpha;}1}^T}{\&Sigma;}}{V}_{kl\mathrm{\&alpha;}\mathrm{\&beta;}\mathrm{\&gamma;}}^T.\underset{\mathrm{\&gamma;}=1}{\overset{{\mathrm{\&Gamma;}}_{kl\mathrm{\&alpha;}2}^T}{\&Sigma;}}{V}_{kl\mathrm{\&alpha;}\mathrm{\&beta;}\mathrm{\&gamma;}}^T,\mathrm{...},\underset{\mathrm{\&gamma;}=1}{\overset{{\mathrm{\&Gamma;}}_{kl\mathrm{\&alpha;}\mathrm{\&beta;}}^T}{\&Sigma;}}{V}_{kl\mathrm{\&alpha;}\mathrm{\&beta;}\mathrm{\&gamma;}}^T,\mathrm{...}\}$

Each element is at the volume of traffic of each import in T period respective signal lamp crossing; Wherein, α is signal lamp crossing numbering, and β is α signal lamp crossing inlet numbering, and γ is β import lane number, and 0 < α≤A, A are dynamic traffic control community, place G _klsignal lamp crossing quantity; 0 < β≤B, B are the import volume of signal lamp crossing, place α; 0 < γ≤Γ, Γ are the track quantity of place import β; the dynamic traffic control community G in the T period _klthe volume of traffic in γ track of α β the import in signal lamp crossing, for the dynamic traffic control community G in the T period _klβ the import numbering import track sum α signal lamp crossing; 4) β import in the saturation degree of T period is:

$S_{kl\mathrm{\&alpha;}\mathrm{\&beta;}}^T=\frac{V_{kl\mathrm{\&alpha;}\mathrm{\&beta;}}^T}{C_{kl\mathrm{\&alpha;}\mathrm{\&beta;}}^T}=\frac{{\&Sigma;}_{\mathrm{\&gamma;}=1}^{{\mathrm{\&Gamma;}}_{\mathrm{\&beta;}}^T}{V}_{kl\mathrm{\&alpha;}\mathrm{\&beta;}\mathrm{\&gamma;}}^T}{{\&Sigma;}_{\mathrm{\&gamma;}=1}^{{\mathrm{\&Gamma;}}_{\mathrm{\&beta;}}^T}{C}_{kl\mathrm{\&alpha;}\mathrm{\&beta;}\mathrm{\&gamma;}}^T}$

5) import β saturation degree maximal value in the α of signal lamp crossing of winning the confidence is the saturation degree in the T period of this signal lamp crossing, that is:

$S_{kl\mathrm{\&alpha;}}^T=\frac{V_{kl\mathrm{\&alpha;}}^T}{C_{kl\mathrm{\&alpha;}}^T}=\max \left\{S_{kl\mathrm{\&alpha;}\mathrm{\&beta;}}^T\right\}=\max \left\{\frac{{\&Sigma;}_{\mathrm{\&gamma;}=1}^{{\mathrm{\&Gamma;}}_{\mathrm{\&beta;}}^T}{V}_{kl\mathrm{\&alpha;}\mathrm{\&beta;}\mathrm{\&gamma;}}^T}{{\&Sigma;}_{\mathrm{\&gamma;}=1}^{{\mathrm{\&Gamma;}}_{\mathrm{\&beta;}}^T}{C}_{kl\mathrm{\&alpha;}\mathrm{\&beta;}\mathrm{\&gamma;}}^T}\right\}$

6) saturation degree of dynamic traffic control community in the T period is calculated computing formula is:

$S_{kl}^T\frac{1}{\mathrm{\&alpha;}}\underset{\mathrm{\&alpha;}=1}{\overset{A_{kl}^T}{\&Sigma;}}{S}_{kl\mathrm{\&alpha;}}^T$

Wherein, k represents that static traffic control work zone is numbered, and l represents the dynamic traffic control subarea numbering in static traffic control work zone, and α is signal lamp crossing numbering in dynamic traffic control subarea, for the signal lamp intersection saturation degree of T period, for the dynamic traffic control subarea G in the T period _klthe number of middle signal lamp crossing;

7) according to dynamic traffic control community saturation degree namely $S_{kl}^T\&le;0.5,0.5<S_{kl}^T\&le;0.8,S_{kl}^T>0.8,$ Dynamic traffic control community is divided into respectively wagon flow to pass unimpeded and control community, signal lamp Collaborative Control community and Collaborative Control community in many ways.

Wherein, wagon flow community of passing unimpeded does not need to take signal lamp linkage tube prosecutor case and manual intervention, multi-party signal lamp Collaborative Control community needs to enable signal lamp linkage tube prosecutor case, in many ways Collaborative Control community needs, except signal lamp interlock management and control, to need to strengthen other all available forces equally and dredge traffic congestion section.

The real-time adjustment of 1.5 traffic control communities:

Calculate T+1 period each signal lamp intersection saturation degree calculating T+1 period each signal lamp intersection saturation degree basis on, calculate dynamic traffic control subarea saturation degree contrast with whether corresponding crossing α exists change, and whether the type in two period same dynamic traffic control subareas changes, if meet wherein a kind of change, then repartition dynamic traffic control subarea, static traffic control work zone remains unchanged; Otherwise, without the need to carrying out dynamic traffic control sub-area division.By above method, self-adaptative adjustment division can be carried out to according to road network vehicle flowrate to traffic control community, for signal lamp interlock management and control, traffic dispersion provide scope more accurately to determine.

The above-mentioned traffic control subdistrict self-adaptive division methods based on category of roads and arithmetic for real-time traffic flow, combine the aspect such as self-adaptative adjustment of the traffic control community static division of category of roads, the traffic control district dynamic division of arithmetic for real-time traffic flow, traffic control cell classification, traffic control community, form a perfect traffic control subdistrict self-adaptive adjustment System, can traffic control community be obtained divide adjustment situation in dynamic high-efficiency ground in time.

Compared with method in the past, the invention has the beneficial effects as follows: consider category of roads and arithmetic for real-time traffic flow carries out static state to traffic subarea and dynamically divides, effectively improve sub-area division efficiency; Determine dissimilar traffic control subarea according to saturation degree, be conducive to building rational signal lamp linkage tube prosecutor case and manual intervention scheme; According to arithmetic for real-time traffic flow change, repartition for some traffic control communities changed, avoid the redundant computation that generation repartitioned by system-wide net.Method overcomes original traffic control sub-area division static and dynamic Status sub-area division and combines the problems such as not enough, control work zone traffic plan specific aim is strong, counting yield is on the low side and adaptive ability is not enough.

Based on above feature, the traffic control subdistrict self-adaptive division methods based on category of roads and arithmetic for real-time traffic flow that the present invention announces also is played a great role as the smart city power-assisted of building in traffic flow real-time estimate, road condition assessment etc.

The above embodiment only have expressed the possible embodiment of the present invention, and it describes comparatively concrete and detailed, but therefore can not be interpreted as the restriction to the scope of the claims of the present invention.It should be pointed out that for the person of ordinary skill of the art, without departing from the inventive concept of the premise, can also make some distortion and improvement, these all belong to protection scope of the present invention.Therefore, the protection domain of patent of the present invention should be as the criterion with claims.

Claims (6)

1. traffic control sub-area division is optimized and a self-adapting regulation method, it is characterized in that, comprises the following steps:

S1. build road network and traffic flow data storehouse, specify road network topology relation and category of roads;

S2. according to road network topology relation and category of roads, Preliminary division is carried out to signal lamp intersection type, calculate the physical similarity between two between signal lamp crossing, signal lamp crossing is divided into one or more static traffic according to physical similarity and controls community;

S3. the static traffic obtained with step S2 controls community for benchmark, in conjunction with arithmetic for real-time traffic flow, the traffic flow degree of association of further calculating between two between signal lamp crossing, is divided into multiple dynamic traffic control subarea according to the traffic flow degree of association by each static traffic control work zone;

S4. calculate each signal lamp crossing present period saturation degree, on this basis, calculate the saturation degree of each dynamic traffic control subarea at present period; Wagon flow is divided into by dynamic traffic control microzonation to pass unimpeded community, signal lamp Collaborative Control community and Collaborative Control community three types in many ways according to dynamic traffic control subarea saturation degree;

S5. judge whether current time reaches the default time apart from the moment of dynamic traffic control last time subarea adjustment, original dynamic traffic control microzonation offshoot program is maintained constant when not reaching, then recalculate each dynamic traffic control cell signal lamp intersection saturation degree when reaching, and according to each dynamic traffic control cell signal lamp intersection saturation degree, cell type is divided; If meet one of following two conditions, then community is controlled to the static traffic at the place, dynamic traffic control community changed and repartition: (1) exists dynamic traffic control cell type and changes; (2) the signal lamp crossing existed in dynamic traffic control community corresponding to saturation degree maximal value changes.

2. traffic control sub-area division according to claim 1 is optimized and self-adapting regulation method, it is characterized in that, in step S1, road network topology relation main manifestations is the syntople between signal lamp crossing, and category of roads is divided into major urban arterial highway, city subsidiary road, Urban Branch Road three major types.

3. traffic control sub-area division according to claim 2 is optimized and self-adapting regulation method, it is characterized in that, the division methods that step S2 static traffic controls community is:

21) road network is represented, wherein V={v with G=(V, E) ₁, v ₂..., v _i..., v _irepresent i-th signal lamp crossing, V is all signal lamp crossings set in road network; E={e ₁₂, e ₂₃..., e _ij..., e _ijrepresent the section between connection signal lamp crossing i and j, E is the set in all sections in road network;

22) different according to category of roads, signal lamp crossing is divided into major trunk roads and major trunk roads, major trunk roads and subsidiary road, major trunk roads and branch road, subsidiary road and subsidiary road, subsidiary road and branch road, branch road and branch road six type; And to six class signal lamp intersection type assignment respectively:

R _i＝θ，1≤θ≤6，θ∈n

Wherein, R _i=1 is major trunk roads and arterial-type signal lamp crossing, R _i=2 is major trunk roads and subsidiary road type signal lamp crossing, R _i=3 is major trunk roads and tributary type signals lamp crossing, R _i=4 is subsidiary road and subsidiary road type signal lamp crossing, R _i=5 is subsidiary road and tributary type signals lamp crossing, R _i=6 is branch road and tributary type signals lamp crossing;

23) physical similarity in road network G between adjacent communications signals lamp crossing is made to be w _{(i, j)}represent the relevance of signal lamp crossing i and j, R _i, R _jbe respectively the types value of signal lamp crossing i, j, d _ijrepresent the distance of signal lamp crossing i and signal lamp crossing j; When signal lamp crossing i is not communicated with signal lamp crossing j, corresponding w _{(i, j)}be 0, e be natural constant;

24) calculate the adjacency matrix H of road network G, when signal lamp crossing i in road network G is communicated with signal lamp crossing j is oriented, namely there is a section e _ijwhen being connected, the element a in H _ij=1, when signal lamp crossing i and signal lamp crossing j does not have a section e _ijwhen being connected, a _ij=0; As i=j, a _ij=0;

25) weighted adjacency matrix W and the diagonal angle adjacency matrix D of road network G is calculated;

Element w wherein in weighted adjacency matrix W _ijvalue is:

$w_{ij}=\left\{\begin{array}{cc}w(i,j),& a_{ij}=1\\ 0,& a_{ij}=0\end{array}\right.$

D＝diag{d _i}，d _i＝∑ _jw(i,j)；

26) road network signal lamp crossing Laplacian matrix L is calculated, wherein L=D-W;

27) based on Laplacian matrix L, the proper vector corresponding to the little eigenwert of this matrix second is solved, i.e. Fiedler proper vector: F=(f ₁, f ₂..., f _n), wherein, the element f of proper vector F _inumerical value corresponding with signal lamp crossing i;

28) according to Fiedler proper vector F=(f ₁, f ₂..., f _n) in the numerical value of each element, adopt dichotomy to split each key element of Fiedler vector, a subregion be divided into two; If the number of partitions reaches K, stop Fiedler segmentation, if the total number of partitions of road network does not reach K, return step 26), select the subregion that signal lamp crossing number is maximum, set up the Laplace matrix that this subregion is corresponding, and repeat step 27) and 28), solve Fiedler vector and carry out the segmentation of Fiedler vector;

Concrete dividing method is: select critical value S=0, split, by f element each in Fiedler proper vector F _ia subregion is assigned to, all the other f in the summit of>=0 _ianother subregion is assigned to, i=1,2 in the summit of <0 ..., n.

4. traffic control sub-area division according to claim 3 is optimized and self-adapting regulation method, and it is characterized in that, the division methods in step S3 dynamic traffic control subarea is:

31) divide according to static traffic control work zone, obtain K static traffic control work zone altogether, use G _k=(V _k, E _k) represent the road net of a corresponding kth static traffic control work zone, 1≤k≤K; Wherein V _k={ v _k1, v _k2..., v _kp..., v _kprepresent kth static traffic control work zone p signal lamp crossing, V _kit is all signal lamp crossings set in a kth static traffic control work zone; E _k={ e _k12, e _k23..., e _kpq, e _kpqrepresent the section between connection signal lamp crossing p and q, E _kfor the set in all sections in a kth static traffic control work zone;

32) determine in a kth static traffic control work zone, the traffic flow degree of association ρ of signal lamp crossing p and the q be connected _{k (p, q)}for:

${\mathrm{\&rho;}}_{k(p,q)}=\frac{0.5}{1+t}\&lsqb;\frac{n\&times;Q_{max}}{{\mathrm{\&Sigma;}}_{i=1}^n{Q}_i}\&rsqb;-(N-2)$

Wherein, n is point number sailed into from the vehicle of stream signal lamp crossing; Q _ibe the volume of traffic that i-th branch arrives downstream signal lamp crossing, Q _maxfor arriving the volume of traffic maximal value of downstream signal lamp crossing in branch; for arriving the volume of traffic sum total of downstream signal lamp crossing; T is the running time of vehicle in two signal lamp crossings; N is the number of track-lines that downstream is driven towards in upstream; Due to signal lamp crossing p and signal lamp crossing q upstream and downstream signal lamp crossing each other, therefore, it is possible to obtain two ρ _{k (p, q)}value, gets the traffic flow degree of association ρ of both mean values as final signal lamp crossing p and q _{k (p, q)};

33) each static traffic control work zone road net G is calculated _kadjacency matrix be H _k, as road network G _kwhen middle signal lamp crossing p is communicated with signal lamp crossing q, H _kin element a _kpq=1, when signal lamp crossing p is not communicated with signal lamp crossing q, a _kpq=0; As p=q, a _kpq=0;

34) each static traffic control work zone road net G is calculated _kweighted adjacency matrix W _k, and diagonal angle adjacency matrix D _k, wherein, weighted adjacency matrix W _kin element w _kpqfor:

$w_{kpq}=\left\{\begin{array}{cc}{\mathrm{\&rho;}}_{k(p,q)},& a_{pq}=1\\ 0,& a_{pq}=0\end{array}\right.$

D _k＝diag{d _kp}，d _kp＝∑ _qρ _k(p,q)

35) each static traffic control work zone road net signal lamp crossing Laplacian matrix L is calculated _k, wherein L _k=D _k-W _k;

36) according to Laplacian matrix L _k, solve the proper vector corresponding to the little eigenwert of this matrix second, i.e. Fiedler vector: F _k=(f _k1, f _k2..., f _km);

37) according to Fiedler proper vector F _k=(f _k1, f _k2..., f _km) in the numerical value of each element, adopt dichotomy to split each key element of Fiedler vector, a subregion be divided into two; If signal lamp crossing number maximal value is less than Z in subregion, stop Fiedler segmentation; If exist wherein in any one subregion signalized intersections number be more than or equal to Z, return step 35), select the subregion that signal lamp crossing number is maximum, set up the Laplace matrix that this subregion is corresponding, and repeat step 36) and 37), solve Fiedler vector and carry out the segmentation of Fiedler vector;

Concrete dividing method is: select critical value S=0, to Fiedler proper vector F _kin each element split, by f _kpa subregion is assigned to, all the other f in the summit of>=0 _kpanother subregion is assigned to, p=1,2 in the summit of <0 ..., m; A final kth static traffic controls community will be divided into L dynamic traffic control community;

38) according to dynamic traffic control sub-area division, a kth static traffic control work zone correspondence obtains L dynamic traffic control subarea; Use G _kl=(V _kl, E _kl) represent l the dynamic traffic control subarea that a corresponding kth static traffic control work zone is corresponding, wherein 1 < k≤K, 1 < l≤L.

5. traffic control sub-area division according to claim 4 is optimized and self-adapting regulation method, it is characterized in that, in step S4 based on the dynamic traffic control cell categories of signal lamp intersection saturation degree mode of delimiting is:

41) each dynamic traffic control subarea G is calculated _klthe saturation degree of T period wherein k is that a kth static traffic controls cell number, and l is dynamic traffic control cell number, wherein, and 1 < k≤K, 1 < l≤L; Each signal lamp intersection saturation degree computing formula is as follows:

$S_{kl\mathrm{\&alpha;}}^T=\frac{V_{kl\mathrm{\&alpha;}}^T}{C_{kl\mathrm{\&alpha;}}}$

Wherein, α represents that signal lamp crossing is numbered, for dynamic traffic control subarea G _klsignal lamp crossing sum; for dynamic traffic control community G _klα signal lamp crossing at the volume of traffic of T period; for representing dynamic traffic control community G _klα signal lamp crossing in the maximum traffic capacity of T period;

42) the maximum traffic capacity computing formula as follows:

$C_{kl\mathrm{\&alpha;}}^T=\{\underset{\mathrm{\&gamma;}=1}{\overset{{\mathrm{\&Gamma;}}_{kl\mathrm{\&alpha;}1}^T}{\&Sigma;}}{C}_{kl\mathrm{\&alpha;}\mathrm{\&beta;}\mathrm{\&gamma;}}^T,\underset{\mathrm{\&gamma;}=1}{\overset{{\mathrm{\&Gamma;}}_{kl\mathrm{\&alpha;}2}^T}{\&Sigma;}}{C}_{kl\mathrm{\&alpha;}\mathrm{\&beta;}\mathrm{\&gamma;}}^T,...,\underset{\mathrm{\&gamma;}=1}{\overset{{\mathrm{\&Gamma;}}_{kl\mathrm{\&alpha;}\mathrm{\&beta;}}^T}{\&Sigma;}}{C}_{kl\mathrm{\&alpha;}\mathrm{\&beta;}\mathrm{\&gamma;}}^T,...\}$

The maximum traffic capacity of each import in each element respective signal lamp crossing; Wherein, α is signal lamp crossing numbering, and β is α signal lamp crossing inlet numbering, and γ is β import lane number, and 0 < α≤A, A are dynamic traffic control community, place G _klsignal lamp crossing quantity; 0 < β≤B, B are the import volume of signal lamp crossing, place α; 0 < γ≤Γ; Γ is the track quantity of place import β; the dynamic traffic control community G in the T period _klthe traffic capacity in γ track of α β the import in signal lamp crossing, for the dynamic traffic control community G in the T period _klβ the import numbering import track sum α signal lamp crossing;

43) volume of traffic get the data on flows that each import records at the ground induction coil of T period, formula is as follows:

$V_{kl\mathrm{\&alpha;}}^T=\{\underset{\mathrm{\&gamma;}=1}{\overset{{\mathrm{\&Gamma;}}_{kl\mathrm{\&alpha;}1}^T}{\&Sigma;}}{V}_{kl\mathrm{\&alpha;}\mathrm{\&beta;}\mathrm{\&gamma;}}^T,\underset{\mathrm{\&gamma;}=1}{\overset{{\mathrm{\&Gamma;}}_{kl\mathrm{\&alpha;}2}^T}{\&Sigma;}}{V}_{kl\mathrm{\&alpha;}\mathrm{\&beta;}\mathrm{\&gamma;}}^T,...,\underset{\mathrm{\&gamma;}=1}{\overset{{\mathrm{\&Gamma;}}_{kl\mathrm{\&alpha;}\mathrm{\&beta;}}^T}{\&Sigma;}}{V}_{kl\mathrm{\&alpha;}\mathrm{\&beta;}\mathrm{\&gamma;}}^T,...\}$

Each element is at the volume of traffic of each import in T period respective signal lamp crossing; Wherein, α is signal lamp crossing numbering, and β is α signal lamp crossing inlet numbering, and γ is β import lane number, and 0 < α≤A, A are dynamic traffic control community, place G _klsignal lamp crossing quantity; 0 < β≤B, B are the import volume of signal lamp crossing, place α; 0 < γ≤Γ, Γ are the track quantity of place import β; the dynamic traffic control community G in the T period _klthe volume of traffic in γ track of α β the import in signal lamp crossing, for the dynamic traffic control community G in the T period _klβ the import numbering import track sum α signal lamp crossing;

44) β import in the saturation degree of T period is:

$S_{kl\mathrm{\&alpha;}\mathrm{\&beta;}}^T=\frac{V_{kl\mathrm{\&alpha;}\mathrm{\&beta;}}^T}{C_{kl\mathrm{\&alpha;}\mathrm{\&beta;}}^T}=\frac{{\mathrm{\&Sigma;}}_{\mathrm{\&gamma;}=1}^{{\mathrm{\&Gamma;}}_{\mathrm{\&beta;}}^T}{V}_{kl\mathrm{\&alpha;}\mathrm{\&beta;}\mathrm{\&gamma;}}^T}{{\mathrm{\&Sigma;}}_{\mathrm{\&gamma;}=1}^{{\mathrm{\&Gamma;}}_{\mathrm{\&beta;}}^T}{C}_{kl\mathrm{\&alpha;}\mathrm{\&beta;}\mathrm{\&gamma;}}^T}$

45) import β saturation degree maximal value in the α of signal lamp crossing of winning the confidence is the saturation degree in the T period of this signal lamp crossing, that is:

$S_{kl\mathrm{\&alpha;}}^T=\frac{V_{kl\mathrm{\&alpha;}}^T}{C_{kl\mathrm{\&alpha;}}^T}=\max \left\{S_{kl\mathrm{\&alpha;}\mathrm{\&beta;}}^T\right\}=\max \left\{\frac{{\mathrm{\&Sigma;}}_{\mathrm{\&gamma;}=1}^{{\mathrm{\&Gamma;}}_{\mathrm{\&beta;}}^T}{V}_{kl\mathrm{\&alpha;}\mathrm{\&beta;}\mathrm{\&gamma;}}^T}{{\mathrm{\&Sigma;}}_{\mathrm{\&gamma;}=1}^{{\mathrm{\&Gamma;}}_{\mathrm{\&beta;}}^T}{C}_{kl\mathrm{\&alpha;}\mathrm{\&beta;}\mathrm{\&gamma;}}^T}\right\}$

46) saturation degree of dynamic traffic control community in the T period is calculated computing formula is:

$S_{kl}^T=\frac{1}{\mathrm{\&alpha;}}\underset{\mathrm{\&alpha;}=1}{\overset{A_{kl}^T}{\&Sigma;}}{S}_{kl\mathrm{\&alpha;}}^T$

Wherein, k represents that static traffic control work zone is numbered, and l represents the dynamic traffic control subarea numbering in static traffic control work zone, and α is signal lamp crossing numbering in dynamic traffic control subarea, for the signal lamp intersection saturation degree of T period, for the dynamic traffic control subarea G in the T period _klthe number of middle signal lamp crossing;

47) according to T period dynamic traffic control community saturation degree namely > 0.8, is divided into wagon flow respectively and passes unimpeded and control community, signal lamp Collaborative Control community and Collaborative Control community in many ways by dynamic traffic control community.

6. traffic control sub-area division according to claim 5 is optimized and self-adapting regulation method, and it is characterized in that, the real-time adjustment of traffic control community refers to: calculate T+1 period each signal lamp intersection saturation degree calculating T+1 period each signal lamp intersection saturation degree basis on, calculate dynamic traffic control subarea saturation degree contrast with whether corresponding crossing α exists change, and whether the type in two period same dynamic traffic control subareas changes, if meet wherein a kind of change, then repartition dynamic traffic control subarea, static traffic control work zone remains unchanged; Otherwise, without the need to carrying out dynamic traffic control sub-area division.