CN107437339A - Variable information advices plate control method for coordinating and system under a kind of information guidance - Google Patents
Variable information advices plate control method for coordinating and system under a kind of information guidance Download PDFInfo
- Publication number
- CN107437339A CN107437339A CN201710467946.2A CN201710467946A CN107437339A CN 107437339 A CN107437339 A CN 107437339A CN 201710467946 A CN201710467946 A CN 201710467946A CN 107437339 A CN107437339 A CN 107437339A
- Authority
- CN
- China
- Prior art keywords
- mrow
- msub
- information
- msubsup
- value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/09—Arrangements for giving variable traffic instructions
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16Z—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS, NOT OTHERWISE PROVIDED FOR
- G16Z99/00—Subject matter not provided for in other main groups of this subclass
Abstract
The present invention discloses the variable information advices plate control method for coordinating under a kind of information guidance, and this method comprises the following steps:S1:Transport information parameter value in pickup area road network;S2:The TRANSFER MODEL of traffic flow modes in Regional Road Network is established using hidden Markov model;S3:EM Algorithm for Solving is used to the hidden Markov model of foundation, obtains road net traffic state information;S4:Guidance information is generated according to road net traffic state information;S5:The guidance information is issued by variable information advices plate;S6:The transport information of issue influences road network traffic flow situation.The invention also discloses the variable information advices plate coordinated control system under a kind of information guidance using this method, the system simulates the change of the road network state of subsequent time using dynamic effects TRANSFER MODEL, counted according to historical data and solve influence matrix, the state of adjacent segments under current event is estimated, and then appropriate message induction trip is issued by adjacent VMS.
Description
Technical field
The present invention relates to the issue of transport information and collection field.More particularly, to variable under a kind of information guidance
Information plate control method for coordinating and system.
Background technology
It can be found that the transport information of urban traffic management department issue can be during the traffic administration of reality
To a certain extent to having an impact (change for including its driving habit, even travel route) in way driver, thus directly
Or cause the dynamic change of road net traffic state indirectly.But because the information of traffic system issue is not 100% accurate, generally
In the presence of certain deviation, and selection of the driver to trip route is randomness, and information publisher can not grasp completely, institute
Not necessarily can strictly be deferred to driver traffic information terminal issue suggestion or instruction, but by its conventional driving experience with
What is currently obtained is merged in way transport information, to determine its behavior.This requires vehicle supervision department to hold road in time
The situation of change of net traffic behavior, to carry out Traffic flow guidance, road traffic control exactly, and it can be provided for traveler
The information services such as hourage forecast, travel route choice.In fact, traffic administration person can only also be issued by improving
The confidence level of transport information, it can just make more drivers confident and select path according to the suggestion or instruction of information issue.Cause
This, the changing rule of the road network traffic flow modes under the influence of transport information is weight of the intelligent transportation system in implementation process
Point and difficulties, this not only has influence on the horizontal accuracy of telecommunication flow information service (such as hourage forecast, navigation), more existed
In certain aspect, determine whether the formulation of traffic management measure is reasonable.
Real-time and accurately detection, predicted link traffic flow modes, are the basic and crucial rings for implementing effective traffic administration
Section.Transport information, including the information such as induce, prompt, (including it can be changed to being had an impact in way driver to a certain extent
Driving habit and driving path etc.), thus directly or indirectly trigger the change of certain road section traffic volume stream mode, and cause the section
Upstream and downstream relevant road segments and Regional Road Network in the traffic flow modes in other sections chain change also occurs.Therefore, in information
Under conditions of issue, collaborative variation and metastatic rule that open position is submitted in section in Regional Road Network are obtained exactly, is real-time
Detection, the important prerequisite of prediction road grid traffic stream state.In addition, traffic administration person can assess accordingly transport information satisfy the need it is netted
The disturbing influence of state, to formulate more reasonably traffic guidance strategy;Meanwhile also it is beneficial to obtain more accurately traffic flow modes
Predicted value, more meet actual road traffic control strategy to formulate.
Variable information advices plate (Variable Message Sign, VMS) be Traffic information demonstration important means it
One.VMS function is to provide road geometry information, road surface traffic information, section by composite signals such as text, image, numerals
The various information such as transport information and public's information on services, so that driver adjusts its driving behavior, reach alleviation traffic
The purpose block, reduce traffic accident, improved the freeway network traffic capacity.VMS has traffic sign and dynamic aobvious simultaneously
The characteristics of showing, the road signs information system of systematization is together form with static traffic mark, be that the orderly safety of traffic is smooth
Logical offer service.In recent years, VMS is obtained in many big cities and is widely applied, meanwhile, it is continuous with urbanization process
Deeply, the construction and application of VMS systems have all been carried out in many cities.VMS is the transport information commonly used in current big city
Issue terminal, in the information service chain to driver, play the effect to become more and more important.A kind of accordingly, it is desirable to provide letter
Variable information advices plate coordinated control system and method under breath guiding.
The content of the invention
, should it is an object of the present invention to provide the variable information advices plate control method for coordinating under a kind of information guidance
Method comprises the following steps:
S1:Transport information parameter value in pickup area road network, the transport information parameter value include traffic flow density, OK
Journey time, road average speed and roadway occupancy.
Preferably, transport information parameter value source includes:Section traffic flow data, the licence plate of through street detecting system collection
The section upstream crossing vehicle license information of identifying system collection and detection data message, the Coil Detector system of downstream road junction seizure
The section alert notice that the magnitude of traffic flow of collection of uniting and road average speed gather with roadway occupancy and 122 warning systems is believed
Breath.
S2:The TRANSFER MODEL of traffic flow modes in Regional Road Network is established using hidden Markov model.
The TRANSFER MODEL of traffic flow modes in Regional Road Network, the influence that the present invention establishes are established using hidden Markov model
Model be based on it is assumed hereinafter that:
Assuming that 1:When receiving transport information, driver determines according to the traffic behavior and its driving experience of current road segment
Whether Path selection is carried out;When being not received by transport information, the Path selection of driver is only determined by its experience.Big
During amount drives individual human according to maximization of utility principle progress Path selection, biography of the traffic flow modes in relevant road segments
Although certain random character can be showed by passing rule, under the hypothesis of driver's rationality, this randomness can use probability point
Cloth function is described.
Assuming that 2:The potential value of t traffic flow modes and its set of probability of occurrence, constitute the potential traffic of t
Stream mode space, and the parameter in traffic flow modes TRANSFER MODEL can be presented in the form of maximum likelihood probability function.
Assuming that 3:Certain linear combination of the potential traffic flow modes of t Regional Road Network can be found, with enough essences
Spend the true traffic flow modes approached on t+1 moment any section.
The TRANSFER MODEL of traffic flow modes in Regional Road Network is established using hidden Markov model, is shown below:
λ=(C, M, π, A, B)
In formula:λ is HMM symbol, and π is original state matrix, and A is state-transition matrix, and B turns for state
Probability is moved, C represents the number in section in Regional Road Network, while also illustrates that the maximum quantity of Traffic information demonstration terminal, and M is to draw
The maximum quantity of the traffic flow modes divided, the calculation formula of model parameter are as follows:
The initial value of i-th of state is in traffic flow modes sequence
In state-transition matrix A after training the i-th row j row element be
It is in the probability of j state observation to k after training
Ot=VkImplication be the observed value O that observes of ttEqual to Vk, wherein, k is observation set { V1,...VM}
In k-th of element VkFootmark;
All sneak condition values in C bars section on the Regional Road Network of tProbability of happening form probability to
Measure P (st):
Initialization vector π(c):
Original state matrix π:
To number the section for being c in t=mcMoment sneak condition valueProbable value equal to 1;
Influence matrix H:
P(st) sneak condition changes in distribution:
P(s1)=π
P(st+1)=P (st)×H
Observer stateValue be equal to 1 ..., mcProbable value be designated as a mc× 1 row vector:
For t observer stateValue be equal to 1 ..., mcProbable value composition a mc× 1 row vector;
They are connected into one 1 × (∑cmc) row vector bt, asProbability distribution;
Wherein,For forward variable of l-th of training sequence after ratio is handled,Passed through for l-th of training sequence
The backward variable crossed after ratio processing, πiIt is initial value,It is by variable πiValue obtained by after training, aijTurn for state before training
The element of the i-th row j row moved in matrix A;The observation probability that observed value for j-th of state of t+1 moment is l;Table
It is l to show t+1 moment observed value,Represent to produce the probability that observed value is l by model λ;Tl(1≤l≤L) is
L-th of information release terminal in road network, l are the sequence number of terminal, and L is the total number of information release terminal in Regional Road Network, and t is represented
Traffic flow modes observation gathers and the periodicity of the time of issue and Traffic information demonstration, wherein t ∈ N and 1≤t≤T, T are
The total degree of issuing traffic stream mode when t,For t when serial number l terminal where section
The potential value of traffic flow modes,The observation for the t traffic flow modes issued for serial number l terminal,For the potential value of the traffic flow modes in all sections in t road network, { 1 ..., mCFor can on C section
The total quantity of the potential traffic flow modes of division,For t section c sneak condition valueValue is i (1≤i
≤mC) probability of happening,For the vector of the potential traffic flow modes probability of happening of section c (1≤c≤C) t, i.e.,:Remember matrix DC×CFor the matrix of road network, each row sum of matrix is 1, noteMarkov matrix(1≤c1,c2≤ C) it is internal procedure state-transistion matrix, it is 1, H per a line sum
For a matrix in block form, row c1With row c2Submatrix beWherein each elementP(st)
A series of vector that is made up of associated data of probability distribution, D each row andEvery a line sum all be 1.
S3:EM Algorithm for Solving is used to the hidden Markov model of foundation, obtains road net traffic state information.
Hidden Markov model derivation algorithm uses EM algorithms, and solving model process is as follows:
Step1:Random generation initial matrix H and original state matrix π;
Step2:Structural regime matrix
mcRepresent sneak conditionA random variable values, b1It is 1 for t=1 moment each section observer state value ... mcProbable value series connection
Into one 1 × (∑cmc) row vector, here
btIt is 1 for each section observer state value of t ... mcOne 1 × (∑ being connected into of probable valuecmc) row vector;
Step3:Build probability space P (y):
To be preceding to parameter;
Step4:Pass through step solution procedure parameter backward forward;
m1The section sneak condition for being 1 for numberingA probable value;mCThe section sneak condition for being C for numbering
A probable value;The sneak condition in the section that the numbering for t construction is 1Value be i probability;For t
The sneak condition in the section that the numbering of moment construction is cValue be i probability;Numbering for t construction is C's
The sneak condition in sectionValue be i probability, wherein 1 < c < C;
Form a 1 × mcRow vectorAnd willSeries connection turns into one 1 × (∑cmc) row vector αt,
NtFor procedure parameter, and have
For backward parameter,For a mc×1
Column vectorAnd willSeries connection turns into (a ∑cmcThe column vector β of) × 1t;
γt=αt·diag[βt]εt-1→t=diag [αt-1]·H·diag[bt]·Nt·diag[βt]
For unilateral parameter,For a 1 × mcRow vectorAnd
WillSeries connection turns into one 1 × (Σcmc) row vector γt,
, will for Bilateral parameterComposition oneMatrixWherein,
ForThe i-th row j row element;
Step5:Standardize original state matrix and A, d, π, π(c)Meet in vector each number and for 1;Wherein, π(c)To be unilateral
ParameterNormalized form, A(i,j)For Bilateral parameterNormalized form, S is orthogonalization matrix, STTo be orthogonal
Change matrix S transposition, di,jBilateral parameterNormalized form after orthogonalization, For unilateral parameter,For a 1 × mcRow vector
Step6:By π(c)Bring into Step2, iterate.If P (y) keeps stable, solution terminates;If P (y) changes,
2~Step of Step 5 are repeated to iterate;
Wherein, for allNote:
It is c for numbering1Section sneak conditionA probable value,It is c for numbering2Section sneak conditionA probable value;I is setAn element;
To be preceding to parameter,Composition one 1 ×
mcRow vectorAnd willSeries connection turns into one 1 × (Σcmc) row vector αt;
For backward parameter,For a mc×1
Column vectorAnd willSeries connection turns into (a ΣcmcThe column vector β of) × 1t;
For unilateral parameter,For a 1 × mc's
Row vectorAnd willSeries connection turns into one 1 × (Σcmc) row vector γt;
, will for Bilateral parameter Composition oneMatrixWherein,ForThe i-th row j row member
Element.
S4:Guidance information is generated according to road net traffic state information.
S5:The guidance information is issued by variable information advices plate.
S6:The transport information of issue influences road network traffic flow situation.
It is another object of the present invention to provide the variable information information under a kind of information guidance of control method for coordinating
Plate coordinated control system, the coordinated control system include:
Data acquisition module, for gathering transport information parameter value, it is close that the transport information parameter value includes the magnitude of traffic flow
Degree, journey time, road average speed and roadway occupancy;
Data processing module, by establishing hidden Markov model, the data of collection are handled, obtain road grid traffic
Status information;
Information display module, treated transport information is issued by variable information advices plate, wherein variable
Information plate is one or more groups of, and the information of the variable information advices plate issue can be influenceed in the choosing of the path of way driver
Select.
Preferably, the data source of data acquisition module includes:Through street detecting system, plate recognition system, coil inspection
Examining system and 122 warning systems.Wherein through street detecting system is used to pass through remote microwave sensor detecting system, gathers and passes
Pass section traffic flow data;Plate recognition system is used to gather section upstream crossing vehicle license, the seizure detection of downstream road junction
Data;Coil detecting system is used to detect the magnitude of traffic flow and calculates road average speed and roadway occupancy;122 warning systems are used
In obtaining the alert notice in section.
Data processing module also includes induction outdoor screen Surveillance center, swap server, the monitoring point of multiple induction outdoor screens
Center and corresponding multiple tuning controllers, wherein induction outdoor screen Surveillance center will pass through processing by swap server
Transport information be transferred to multiple induction outdoor screens monitoring branch centers;Swap server enters row information issue based on induction strategies;
Treated transport information is transmitted by corresponding multiple tuning controllers multiple induction outdoor screen monitoring branch centers
To described information display module.
Preferably, information display module is variable information advices plate, and every group of variable information advices plate includes multiple induction rooms
External screen.
Preferably, traffic state information also includes the information being manually entered.
Preferably, the information transfer between data acquisition module, data processing module and information display module is special by ether
Net is carried out.
Beneficial effects of the present invention are as follows:
1st, the variable information advices plate control method for coordinating and system under information guidance proposed by the present invention, using probability point
Analyse driver Path Selection, make road network state the continuous moment state it is regular follow, pass through dynamic effects transmission
Model simulates the change of subsequent time road network state, can effectively analyze the hair in current variable information advices plate (VMS)
The state status of road network under cloth state.
2nd, based within certain time and spatial dimension, the same event net state that satisfies the need influences the principle with similitude,
, can be according to historical data using the variable information advices plate control method for coordinating and system under information guidance proposed by the present invention
Statistics solves influence matrix, estimates the state of the adjacent segments under current event.
3rd, according to the variable information advices plate control method for coordinating and system under information guidance proposed by the present invention, road network is worked as
In a certain region when event occurs, event information is transmitted by variable information advices plate (VMS) in road network, can be by adjacent
Variable information advices plate (VMS) issue rationally induction trip, specific publishing policy can be according to the case and experience of history
Formulate.
Brief description of the drawings
The embodiment of the present invention is described in further detail below in conjunction with the accompanying drawings.
Fig. 1 shows the variable information advices plate control method for coordinating under information guidance.
Fig. 2 shows the control system framework of the traffic induction outdoor screen based on distributed Ethernet.
Fig. 3 shows the variable information advices plate coordinated control system conceptual data flow graph under information guidance.
Fig. 4 shows process layer flow chart of data processing.
Fig. 5 shows display layer flow chart of data processing.
Fig. 6 shows potential structure influence process figure.
Fig. 7 shows the variable information advices plate coordinated control system framework under information guidance.
Fig. 8 shows the simple topology figure of the road network in case application section.
Fig. 9 shows the relevant road segments of road network.
Figure 10 shows Gary Assessment of Serviceability of Roads grading standard.
Figure 11 shows road network virtual condition distribution situation.
Figure 12 shows the road network state diagram of VMS issues.
Figure 13 shows probability space convergence graph.
Figure 14 shows influence matrix H.
Figure 15 shows original state figure.
Figure 16 shows the actual condition value that EM algorithms are drawn.
Embodiment
In order to illustrate more clearly of the present invention, the present invention is described further below in conjunction with the accompanying drawings.It is similar in accompanying drawing
Part be indicated with identical reference.It will be appreciated by those skilled in the art that specifically described content is below
It is illustrative and not restrictive, it should not be limited the scope of the invention with this.
Fig. 1 shows the variable information advices plate control method for coordinating under information guidance, and this method comprises the following steps:
S1:Transport information parameter value in pickup area road network, the transport information parameter value include traffic flow density, OK
Journey time, road average speed and roadway occupancy.
Preferably, transport information parameter value source includes:Section traffic flow data, the licence plate of through street detecting system collection
The section upstream crossing vehicle license information of identifying system collection and detection data message, the Coil Detector system of downstream road junction seizure
The section alert notice that the magnitude of traffic flow of collection of uniting and road average speed gather with roadway occupancy and 122 warning systems is believed
Breath.
S2:The TRANSFER MODEL of traffic flow modes in Regional Road Network is established using hidden Markov model.
The TRANSFER MODEL of traffic flow modes in Regional Road Network, the influence that the present invention establishes are established using hidden Markov model
Model be based on it is assumed hereinafter that:
Assuming that 1:When receiving transport information, driver determines according to the traffic behavior and its driving experience of current road segment
Whether Path selection is carried out;When being not received by transport information, the Path selection of driver is only determined by its experience.Big
During amount drives individual human according to maximization of utility principle progress Path selection, biography of the traffic flow modes in relevant road segments
Although certain random character can be showed by passing rule, under the hypothesis of driver's rationality, this randomness can use probability point
Cloth function is described.
Assuming that 2:The potential value of t traffic flow modes and its set of probability of occurrence, constitute the potential traffic of t
Stream mode space, and the parameter in traffic flow modes TRANSFER MODEL can be presented in the form of maximum likelihood probability function.
Assuming that 3:Certain linear combination of the potential traffic flow modes of t Regional Road Network can be found, with enough essences
Spend the true traffic flow modes approached on t+1 moment any section.
The TRANSFER MODEL of traffic flow modes in Regional Road Network is established using hidden Markov model, is shown below:
λ=(C, M, π, A, B)
In formula:λ is HMM symbol, and π is original state matrix, and A is state-transition matrix, and B turns for state
Probability is moved, C represents the number in section in Regional Road Network, while also illustrates that the maximum quantity of Traffic information demonstration terminal, and M is to draw
The maximum quantity of the traffic flow modes divided, the calculation formula of model parameter are as follows:
The initial value of i-th of state is in traffic flow modes sequence
In state-transition matrix A after training the i-th row j row element be
It is in the probability of j state observation to k after training
All sneak condition values in C bars section on the Regional Road Network of tProbability of happening form probability to
Measure P (st):
Initialization vector π(c):
Original state matrix π:
Influence matrix H:
P(st) sneak condition changes in distribution:
P(s1)=π
P(st+1)=P (st)×H
Observer stateValue be equal to 1 ..., mcProbable value be designated as a mc× 1 row vector:
They are connected into one 1 × (∑cmc) row vector bt, as yt (c)Probability distribution;
Wherein,For forward variable of l-th of training sequence after ratio is handled,Passed through for l-th of training sequence
The backward variable crossed after ratio processing, Tl(1≤l≤L) is l-th of information release terminal in road network, and l is the sequence number of terminal, and L is
The total number of information release terminal in Regional Road Network, t represent the collection of traffic flow modes observation and the time of issue and traffic letter
The periodicity of issue is ceased, wherein t ∈ N and 1≤t≤T, T are the total degree of issuing traffic stream mode when t,For t when serial number l terminal where section the potential value of traffic flow modes,For serial number
The observation of the t traffic flow modes of l terminal issue,For the friendship in all sections in t road network
The potential value of open position, { 1 ..., mCFor the total quantity for the potential traffic flow modes that can be divided on C section,
For t section c sneak condition valueValue is i (1≤i≤mC) probability of happening,For section c (1≤c≤C) t
The vector of moment potential traffic flow modes probability of happening, i.e.,:Remember matrix DC×C
For the matrix of road network, each row sum of matrix is 1, noteMarkov matrix(1≤c1,c2≤ C) it is inside
Process status transfer matrix, it is that 1, H is a matrix in block form per a line sum, row c1With row c2Submatrix beWherein each elementP(st) probability distribution be by a series of associated data groups
Into vector, D each row andEvery a line sum all be 1.
S3:EM Algorithm for Solving is used to the hidden Markov model of foundation, obtains road net traffic state information.
Hidden Markov model derivation algorithm uses EM algorithms, and solving model process is as follows:
Step1:Random generation initial matrix H and original state matrix π;
Step2:Structural regime matrix
Step3:Build probability space P (y):
Step4:Pass through step solution procedure parameter backward forward;
γt=αt·diag[βt]εt-1→t=diag [αt-1]·H·diag[bt]·Nt·diag[βt]
Step5:Standardize original state matrix and A, d, π, π(c)Meet in vector each number and for 1.
Step6:By π(c)Bring into Step2, iterate.If P (y) keeps stable, solution terminates;If P (y) changes,
2~Step of Step 5 are repeated to iterate;
Wherein, for allNote:
To be preceding to parameter,Composition one 1 ×
mcRow vectorAnd willSeries connection turns into one 1 × (∑cmc) row vector αt;
For backward parameter,For a mc×1
Column vectorAnd willSeries connection turns into (a ∑cmcThe column vector β of) × 1t;
For unilateral parameter,For a 1 × mc's
Row vectorAnd willSeries connection turns into one 1 × (∑cmc) row vector γt;
, will for Bilateral parameterComposition oneMatrixWherein,ForThe i-th row
The element of j row.
S4:Guidance information is generated according to road net traffic state information.
S5:The guidance information is issued by variable information advices plate.
S6:The transport information of issue influences road network traffic flow situation.
The present invention also provides the variable information advices plate coordinated control system under a kind of information guidance, and Fig. 7 shows that information is drawn
Variable information advices plate coordinated control system framework under leading, it includes data acquisition module, data processing module and information and shown
Show module.Data collecting module collected road grid traffic information parameter value, through data processing module structure and solving model, obtain road
Net traffic state information simultaneously generates guidance information, is issued finally by variable information advices plate, and then influences on way
The Path selection of driver.
Fig. 2 shows the control system framework of the traffic induction outdoor screen based on distributed Ethernet.Data processing module is also
Including induction outdoor screen Surveillance center, swap server, multiple induction outdoor screens monitoring branch center and corresponding multiple associations
Controller is adjusted, treated transport information is transferred to multiple lure by wherein induction outdoor screen Surveillance center by swap server
Lead outdoor screen monitoring branch center;Swap server enters row information issue based on induction strategies;In multiple induction outdoor screen monitoring point
The heart is transmitted treated transport information to information display module by corresponding multiple tuning controllers.Presentation of information
Module is variable information advices plate (VMS), and every group of variable information advices plate (VMS) includes multiple induction outdoor screens.Traffic behavior
Information also includes the information being manually entered.Information transfer between data acquisition module, data processing module and information display module
Carried out by Ethernet Private LAN.
The data source of data acquisition module includes:Through street detecting system, plate recognition system, coil detecting system and
122 warning systems, wherein through street detecting system are used to pass through remote microwave sensor detecting system, gather and transmit section friendship
Through-flow data;Plate recognition system is used to gather section upstream crossing vehicle license, the seizure detection data of downstream road junction;Coil
Detecting system is used to detect the magnitude of traffic flow and calculates road average speed and roadway occupancy;122 warning systems are used to obtain road
The alert notice of section.
Can transport information have an impact to driver's behavior, and then influence road grid traffic stream state, be mainly reflected in
In two class factors:When the traffic for the current road segment that driver observes, second, driver's experience (is presented as driver couple
The judgement of road traffic flow changed condition trend, and its trust and obedience degree to transport information).Thus, in transport information
Under the influence of, the optimizing paths of road network driver colony can be described with disaggregation theory, and drive individual human
Action selection meets maximization of utility it is assumed that as follows:
Uim=Vim+εim
Wherein, VimIt is the effectiveness triggered by current observable traffic flow modes;εimRepresent by potential traffic flow modes
And the effectiveness that driver's preference (such as O-D, driving habit) triggers.Conventional disaggregation theory all has IIA (Independence
Of Irrelevant Alternative) characteristic, that is, assume εimTo be separate.But the driver on different sections of highway can
Energy have similar preference, the optimizing paths under the conditions of being issued due to information, then the ε on different sections of highwayimBetween exist
Certain relevance, now disaggregation theory IIA characteristics assume just no longer set up.
The processing procedure of data can be divided into three aspects, be data Layer, process layer and display layer respectively.Fig. 3 shows letter
Variable information advices plate coordinated control system conceptual data flow graph under breath guiding.
(1) data Layer
The main function of this layer is that the initial data that other systems pass over is carried out verifying unpacking and pre-processed.It is former
Beginning data source is in following subsystem:Through street detecting system, i.e. remote microwave sensor detecting system, gather and transmit section
Traffic flow data;Plate recognition system, gather section upstream crossing vehicle license, the seizure detection data of downstream road junction.According to
The data that plate recognition system passes over, can calculate the average hourage in this section, and then obtain average traffic
Travel speed;Signal system, i.e. coil detecting system, in addition to detecting the magnitude of traffic flow, while road average speed can be calculated
With the data such as occupation rate;122 warning systems, obtain the alert notice and other systems in certain section.The traffic data of system above
Incoming vehicle supervision department, is on the one hand stored into database, on the other hand, is handled by process layer computing module.
(2) process layer
In process layer, corresponding model, such as wagon flow state classification computation model are called by data processing module, carried out
Data calculating is handled, and obtains traffic state information, such as, " certain section flow speeds is 20 kilometers/hour, and the duration is 7 points
The context information of clock " etc;Traffic state information is stored in system database.In addition, some emergent traffic incidents need people
Work information inputs and is stored in system database.
(3) display layer
Display layer is mainly handled traffic state information, forms guidance information.The generation of guidance information
Pass through two channels:Traffic state information in one side recalls information storehouse, by traffic guidance correlation analysis model, obtain
The issue expanded range of induction information and ageing parameter, then by prediction scheme storehouse, generate induction information.On the other hand, if
The information (such as control traffic message, duties information or advertisement information etc.) for needing to be manually entered then is carried out artificial by commanding
Editor, after generating programme, with reference to the prediction scheme storehouse in system database, together with above-mentioned induction information, generate induction information
Program, finally issued by variable information advices plate (VMS).
In whole data handling procedure, process layer and display layer process CIMS are complicated, employ a variety of model calculations,
And repeatedly stored in database and computing, it is necessary to which data processing and the effect of various databases are refined.
After the traffic data that data Layer verification unpacked reaches process layer, first pre-processed, rejecting is not being set
In the range of data, obtain valid data.State clustering is carried out afterwards, that is, makees state classification, and according to category of roads progress
Match somebody with somebody, finally carry out status predication, obtain traffic related information, store in traffic state information storehouse.Now, display layer can
Relevant information is obtained from the traffic state information storehouse of process layer and enters display layer, Fig. 4 shows process layer flow chart of data processing.
Display layer calls status information from traffic state information storehouse, generates induction information, with the generation of human-edited's information
Programme together, turns into induction information programme, deposit guidance information storehouse, while the induction information generated passes through variable letter
Advices plate (VMS) display is ceased, display layer flow chart of data processing is as shown in Figure 6.Wherein, a variety of meters are used in process layer and display layer
Model is calculated, process layer will pre-process to traffic data, be related to filtering algorithm, such as Kalman filtering, Wiener filtering etc.;Shape
Carry out state classification is first had in state cluster, the matching algorithm such as fuzzy clustering or wavelet analysis is then carried out according to category of roads,
Last to carry out status predication according to for different grades of road, forecast model has neutral net, time series models, nonparametric
Regression model etc., Fig. 5 show display layer flow chart of data processing.
With variable information advices plate (VMS) for information release terminal in the specific embodiment of the present invention, analysis is variable
Affecting laws of the transport information of information plate (VMS) issue to road grid traffic stream state.Choose Beijing North 2nd Ring Road and north
The outer street of drum tower between three rings is as case application region, and Fig. 8 shows the simple topology figure of the road network in case application section, figure
9 show the relevant road segments of road network, wherein No. 4 sections are target road section.
The traffic flow speed parameter in 6 sections is acquired by Floating Car, obtains the traffic flow modes conduct in section
Data set.Data acquisition time (does not choose weekend data, is the traffic because weekend for the 6:00 AM of Mon-Fri to 9 points
Properties of flow differs greatly with working day), collection period is 15 minutes, is sampled 13 times altogether daily, i.e. T=13.According to real data
Acquisition situation analysis assume:The acquisition precision of road traffic stream mode is 95%.
Combined type variable information plate (VMS), road section traffic volume stream mode where generally representing it with red, yellow, and green
Observation, Figure 12 shows the road network state diagram of VMS issues, represents that gray value is by shallow respectively with different gray values in present case
Crowded 2, unobstructed 1 and congestion 3 are represented successively to deep, then in hidden Markov modelFor sneak conditionThe judgement service level of road network experienced with traveler express.Figure 10 shows Gary Assessment of Serviceability of Roads etc.
The level criteria for classifying.Sneak conditionProbable value beVarious service levels are replaced with 1~6.
VMS issues the traffic flow modes information in section where it, if the section collection that VMS releases news be combined into 1,2 ...,
C }, the set in the corresponding section for possessing detector is also { 1,2 ..., C }.
It can be obtained by historical data, specific implication is as follows:
For 3x1 row vector, represent under certain sneak condition
The data distribution situation observed.For example,Represent t=2 moment numberings c section it is unobstructed, it is crowded,
The probability of blocking is respectively 0.2,0.3,0.5.
Represent the observation data before t
When determining, t numbering c section or the section sneak condition distribution situation of detector.For example, there is the letter of 2 detectors
Single channel net, T=3, Represent when releasing news known of t=1, during t=2,1
The Service level of road section that number section releases news is AAProbability be 0.2.
Represent the observation data after t
When determining, t numbering C VMS sneak condition distribution situation.For example, have the simple road network of 2 detectors,T=3, When representing to assume t=3 release news it is known under the conditions of, t=2
When No. 1 section Service level of road section that releases news be AAProbability be 0.2.
It is similar to the above, represent overall observation data
Information oneself when knowing, the sneak condition distribution situation in t numbering c section.
Represent overall observation data
Information oneself when knowing, c1Number sneak condition of the section at the t-1 moment is i and c2The joint that number section is j in the sneak condition of t
Probability distribution.This parameter reflects the state transmission of dynamic process.For example,Represent integrally observing data message
When oneself knows (under the overall observation data cases for having obtained detector issue), section where No. 1 VMS services water at the t=2 moment
Put down as AA, section is A in t=3 moment service level where No. 2 detectorsBProbability be 0.1.
By the acquisition of above-mentioned 4 kinds of data, all parameters in EM algorithms can obtain, and can then obtain state biography
The influence matrix H passed.
Figure 11 shows road network virtual condition distribution situation, and according to investigation result, the distribution situation of road network virtual condition is as schemed
Shown in 12, wherein:1) virtual condition can not be obtained accurately, and the data that Floating Car method obtains also have certain error;2) figure
The depth of color represents service level in 11, more deeply feels and shows that service level is lower, travel speed is lower.
Figure 12 shows the road network state diagram of VMS issues.In the information that the road traffic delay state detector of Beijing is obtained,
Travel speed is the main information of issue, the arithmetic for real-time traffic flow status display situation obtained on internet and the road of VMS issues
Traffic flow modes are identicals, and red representation speed is less than 20km/h, and yellow is 20~50km/h, and green then representation speed is more than
50km/h。
The deeper place of color is shown in red in fig. 12 in comparison diagram 10 and Figure 11, substantially figure 11 above, and color is more shallow
It is local shown in green, that is to say, that VMS issue information it is substantially accurate.Certainly, also place devious, such as No. 1
Section in t=12 period, changes, road network state does not change in t=6 period, No. 4 sections though releasing news
Become.On the one hand this phenomenon Producing reason is due to that information issue has certain time delay, be on the other hand because collection
The precision of information is relevant.
Figure 13 shows probability space convergence graph, and Figure 13 is the probability space of MATLAB generations as EM iterationses are gradually received
Figure is held back, abscissa is iterations, and ordinate is p (y).Observe the figure of EM algorithms generation, it is known that probability space is finally restrained
To stationary value, it can be deduced that influence matrix H and original state p (s at this moment1).Figure 14 shows influence matrix H, and influence matrix is
36 × 36 matrixes.Figure 15 shows original state figure.The depth of color correspond to the size of numerical value in Figure 15., will by EM algorithms
Road network sneak condition takes desired value, and Figure 15 shows virtual condition (s) value that EM algorithms are drawn
Figure 16 and Figure 11 similarities are higher, illustrate that the result that the present invention is drawn is more reliable.The actual shape that EM algorithms are drawn
Service level value where the real-time speed that state value is measured compared to Floating Car, mean error are about that 0.0995 (black is service water
Flat AF, white is AA, remaining linear transitions between black and white, error is the error of corresponding region color in two figures, that is, misses
Difference be less than color gamut 10%), corresponding velocity error is about in 6km/h.Comparison result shows that the inventive method can describe
The dynamic transmittance process of road network state, and to predicting that the road network state of subsequent time has certain applicability.
Obviously, the above embodiment of the present invention is only intended to clearly illustrate example of the present invention, and is not pair
The restriction of embodiments of the present invention, for those of ordinary skill in the field, may be used also on the basis of the above description
To make other changes in different forms, all embodiments can not be exhaustive here, it is every to belong to this hair
Row of the obvious changes or variations that bright technical scheme is extended out still in protection scope of the present invention.
Claims (10)
1. the variable information advices plate control method for coordinating under a kind of information guidance, it is characterised in that this method includes following step
Suddenly:
S1:Transport information parameter value in pickup area road network, when the transport information parameter value includes traffic flow density, stroke
Between, road average speed and roadway occupancy;
S2:The TRANSFER MODEL of traffic flow modes in Regional Road Network is established using hidden Markov model;
S3:EM Algorithm for Solving is used to the hidden Markov model of foundation, obtains road net traffic state information;
S4:Guidance information is generated according to road net traffic state information;
S5:The guidance information is issued by variable information advices plate;
S6:The transport information of issue influences road network traffic flow situation.
2. control method for coordinating according to claim 1, it is characterised in that transport information parameter value source in the step S1
Including:Section traffic flow data, the section upstream crossing vehicle board of plate recognition system collection of through street detecting system collection
Detection data message, the magnitude of traffic flow of coil detecting system collection and the road average speed caught according to information and downstream road junction with
Roadway occupancy and the section alert notice information of 122 warning systems collection.
3. control method for coordinating according to claim 1, it is characterised in that hidden Markov model is utilized in the step S2
The TRANSFER MODEL of traffic flow modes in Regional Road Network is established, is shown below:
λ=(C, M, π, A, B)
In formula:λ is HMM, and π is original state matrix, and A is state-transition matrix, and B is state transition probability, C
The number in section in Regional Road Network is represented, while also illustrates that the maximum quantity of Traffic information demonstration terminal, M is the traffic that can be divided
The maximum quantity of stream mode, the calculation formula of model parameter are as follows:
The initial value of i-th of state is in traffic flow modes sequence
<mrow>
<msub>
<mover>
<mi>&pi;</mi>
<mo>&OverBar;</mo>
</mover>
<mi>i</mi>
</msub>
<mo>=</mo>
<mfrac>
<mn>1</mn>
<mi>L</mi>
</mfrac>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>l</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>L</mi>
</munderover>
<msubsup>
<mi>&alpha;</mi>
<mn>1</mn>
<mrow>
<mo>*</mo>
<mrow>
<mo>(</mo>
<mi>l</mi>
<mo>)</mo>
</mrow>
</mrow>
</msubsup>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>)</mo>
</mrow>
<msubsup>
<mi>&beta;</mi>
<mn>1</mn>
<mrow>
<mo>*</mo>
<mrow>
<mo>(</mo>
<mi>l</mi>
<mo>)</mo>
</mrow>
</mrow>
</msubsup>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>)</mo>
</mrow>
<mo>,</mo>
<mn>1</mn>
<mo>&le;</mo>
<mi>i</mi>
<mo>&le;</mo>
<mi>C</mi>
</mrow>
In state-transition matrix A after training the i-th row j row element be
It is in the probability of j state observation to k after training
<mrow>
<msub>
<mover>
<mi>b</mi>
<mo>&OverBar;</mo>
</mover>
<mrow>
<mi>j</mi>
<mi>k</mi>
</mrow>
</msub>
<mo>=</mo>
<mfrac>
<mrow>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>l</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>L</mi>
</munderover>
<munderover>
<munder>
<mi>&Sigma;</mi>
<mrow>
<mi>t</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
</munder>
<mrow>
<msub>
<mi>O</mi>
<mi>t</mi>
</msub>
<mo>=</mo>
<msub>
<mi>V</mi>
<mi>k</mi>
</msub>
</mrow>
<msub>
<mi>T</mi>
<mi>l</mi>
</msub>
</munderover>
<msubsup>
<mi>&alpha;</mi>
<mi>t</mi>
<mrow>
<mo>*</mo>
<mrow>
<mo>(</mo>
<mi>l</mi>
<mo>)</mo>
</mrow>
</mrow>
</msubsup>
<mrow>
<mo>(</mo>
<mi>j</mi>
<mo>)</mo>
</mrow>
<msubsup>
<mi>&beta;</mi>
<mi>t</mi>
<mrow>
<mo>*</mo>
<mrow>
<mo>(</mo>
<mi>l</mi>
<mo>)</mo>
</mrow>
</mrow>
</msubsup>
<mrow>
<mo>(</mo>
<mi>j</mi>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>l</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>L</mi>
</munderover>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>t</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<msub>
<mi>T</mi>
<mi>l</mi>
</msub>
</munderover>
<msubsup>
<mi>&alpha;</mi>
<mi>t</mi>
<mrow>
<mo>*</mo>
<mrow>
<mo>(</mo>
<mi>l</mi>
<mo>)</mo>
</mrow>
</mrow>
</msubsup>
<mrow>
<mo>(</mo>
<mi>j</mi>
<mo>)</mo>
</mrow>
<msubsup>
<mi>&beta;</mi>
<mi>t</mi>
<mrow>
<mo>*</mo>
<mrow>
<mo>(</mo>
<mi>l</mi>
<mo>)</mo>
</mrow>
</mrow>
</msubsup>
<mrow>
<mo>(</mo>
<mi>j</mi>
<mo>)</mo>
</mrow>
</mrow>
</mfrac>
<mo>,</mo>
<mn>1</mn>
<mo>&le;</mo>
<mi>j</mi>
<mo>&le;</mo>
<mi>C</mi>
<mo>,</mo>
<mn>1</mn>
<mo>&le;</mo>
<mi>k</mi>
<mo>&le;</mo>
<mi>M</mi>
</mrow>
Ot=VkImplication be the observed value O that observes of ttEqual to Vk, wherein, k is observation set { V1,...VMIn kth
Individual element VkFootmark;
All sneak condition values in C bars section on the Regional Road Network of tProbability of happening composition probability vector P
(st):
<mfenced open = "" close = "">
<mtable>
<mtr>
<mtd>
<mrow>
<mi>P</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>s</mi>
<mi>t</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mrow>
<mo>(</mo>
<mi>P</mi>
<mo>(</mo>
<msubsup>
<mi>s</mi>
<mi>t</mi>
<mrow>
<mo>(</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
</msubsup>
<mo>)</mo>
</mrow>
<mo>,</mo>
<mn>...</mn>
<mo>,</mo>
<mi>P</mi>
<mrow>
<mo>(</mo>
<msubsup>
<mi>s</mi>
<mi>t</mi>
<mrow>
<mo>(</mo>
<mi>C</mi>
<mo>)</mo>
</mrow>
</msubsup>
<mo>)</mo>
</mrow>
<mo>)</mo>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mrow>
<mo>=</mo>
<mrow>
<mo>(</mo>
<mi>P</mi>
<mo>(</mo>
<msubsup>
<mi>s</mi>
<mi>t</mi>
<mrow>
<mo>(</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
</msubsup>
<mo>=</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
<mo>,</mo>
<mn>...</mn>
<mo>,</mo>
<mi>P</mi>
<mrow>
<mo>(</mo>
<msubsup>
<mi>s</mi>
<mi>t</mi>
<mrow>
<mo>(</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
</msubsup>
<mo>=</mo>
<msub>
<mi>m</mi>
<mn>1</mn>
</msub>
<mo>)</mo>
</mrow>
<mo>,</mo>
<mn>...</mn>
<mo>,</mo>
<mi>P</mi>
<mrow>
<mo>(</mo>
<msubsup>
<mi>s</mi>
<mi>t</mi>
<mrow>
<mo>(</mo>
<mi>c</mi>
<mo>)</mo>
</mrow>
</msubsup>
<mo>=</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
<mo>,</mo>
<mn>...</mn>
<mo>,</mo>
<mi>P</mi>
<mrow>
<mo>(</mo>
<msubsup>
<mi>s</mi>
<mi>t</mi>
<mrow>
<mo>(</mo>
<mi>c</mi>
<mo>)</mo>
</mrow>
</msubsup>
<mo>=</mo>
<msub>
<mi>m</mi>
<mi>C</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>)</mo>
</mrow>
</mtd>
</mtr>
</mtable>
</mfenced>
Initialization vector π(c):
<mrow>
<msup>
<mi>&pi;</mi>
<mrow>
<mo>(</mo>
<mi>c</mi>
<mo>)</mo>
</mrow>
</msup>
<mo>=</mo>
<mrow>
<mo>(</mo>
<msubsup>
<mi>&pi;</mi>
<mn>1</mn>
<mrow>
<mo>(</mo>
<mi>c</mi>
<mo>)</mo>
</mrow>
</msubsup>
<mo>,</mo>
<mo>...</mo>
<mo>,</mo>
<msubsup>
<mi>&pi;</mi>
<msub>
<mi>m</mi>
<mi>c</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>c</mi>
<mo>)</mo>
</mrow>
</msubsup>
<mo>)</mo>
</mrow>
</mrow>
Original state matrix π:
<mfenced open = "" close = "">
<mtable>
<mtr>
<mtd>
<mrow>
<mi>&pi;</mi>
<mo>=</mo>
<mrow>
<mo>(</mo>
<msup>
<mi>&pi;</mi>
<mrow>
<mo>(</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
</msup>
<mo>,</mo>
<mo>...</mo>
<mo>,</mo>
<msup>
<mi>&pi;</mi>
<mrow>
<mo>(</mo>
<mi>C</mi>
<mo>)</mo>
</mrow>
</msup>
<mo>)</mo>
</mrow>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mrow>
<mo>=</mo>
<mrow>
<mo>(</mo>
<msubsup>
<mi>&pi;</mi>
<mn>1</mn>
<mrow>
<mo>(</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
</msubsup>
<mo>,</mo>
<mn>...</mn>
<mo>,</mo>
<msubsup>
<mi>&pi;</mi>
<msub>
<mi>m</mi>
<mn>1</mn>
</msub>
<mrow>
<mo>(</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
</msubsup>
<mo>,</mo>
<msubsup>
<mi>&pi;</mi>
<mn>1</mn>
<mrow>
<mo>(</mo>
<mi>c</mi>
<mo>)</mo>
</mrow>
</msubsup>
<mo>,</mo>
<mn>...</mn>
<mo>,</mo>
<msubsup>
<mi>&pi;</mi>
<msub>
<mi>m</mi>
<mi>C</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>c</mi>
<mo>)</mo>
</mrow>
</msubsup>
<mo>)</mo>
</mrow>
</mrow>
</mtd>
</mtr>
</mtable>
</mfenced>
To number the section for being c in t=mcMoment sneak condition valueProbable value equal to 1;
Influence matrix H:
<mrow>
<mi>H</mi>
<mo>=</mo>
<mi>D</mi>
<mo>&CircleTimes;</mo>
<msub>
<mrow>
<mo>{</mo>
<msup>
<mi>A</mi>
<mrow>
<msub>
<mi>c</mi>
<mn>1</mn>
</msub>
<mo>,</mo>
<msub>
<mi>c</mi>
<mn>2</mn>
</msub>
</mrow>
</msup>
<mo>}</mo>
</mrow>
<mrow>
<mn>1</mn>
<mo>&le;</mo>
<msub>
<mi>c</mi>
<mn>1</mn>
</msub>
<mo>,</mo>
<msub>
<mi>c</mi>
<mn>2</mn>
</msub>
<mo>&le;</mo>
<mi>C</mi>
</mrow>
</msub>
<mo>=</mo>
<msub>
<mrow>
<mo>(</mo>
<msub>
<mi>d</mi>
<mrow>
<msub>
<mi>c</mi>
<mn>1</mn>
</msub>
<mo>,</mo>
<msub>
<mi>c</mi>
<mn>2</mn>
</msub>
</mrow>
</msub>
<msup>
<mi>A</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>c</mi>
<mn>1</mn>
</msub>
<mo>,</mo>
<msub>
<mi>c</mi>
<mn>2</mn>
</msub>
<mo>)</mo>
</mrow>
</msup>
<mo>)</mo>
</mrow>
<mrow>
<mn>1</mn>
<mo>&le;</mo>
<msub>
<mi>c</mi>
<mn>1</mn>
</msub>
<mo>,</mo>
<msub>
<mi>c</mi>
<mn>2</mn>
</msub>
<mo>&le;</mo>
<mi>C</mi>
</mrow>
</msub>
</mrow>
P(st) sneak condition changes in distribution:
P(s1)=π
P(st+1)=P (st)×H
Observer stateValue be equal to 1 ..., mcProbable value be designated as a mc× 1 row vector:
<mrow>
<msubsup>
<mi>b</mi>
<mi>t</mi>
<mrow>
<mo>(</mo>
<mi>c</mi>
<mo>)</mo>
</mrow>
</msubsup>
<mo>=</mo>
<mrow>
<mo>(</mo>
<mi>P</mi>
<mo>(</mo>
<mrow>
<msubsup>
<mi>y</mi>
<mi>t</mi>
<mrow>
<mo>(</mo>
<mi>c</mi>
<mo>)</mo>
</mrow>
</msubsup>
<mo>|</mo>
<msubsup>
<mi>s</mi>
<mi>t</mi>
<mrow>
<mo>(</mo>
<mi>c</mi>
<mo>)</mo>
</mrow>
</msubsup>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mo>)</mo>
<mo>,</mo>
<mo>...</mo>
<mo>,</mo>
<mi>P</mi>
<mo>(</mo>
<mrow>
<msubsup>
<mi>y</mi>
<mi>t</mi>
<mrow>
<mo>(</mo>
<mi>c</mi>
<mo>)</mo>
</mrow>
</msubsup>
<mo>|</mo>
<msubsup>
<mi>s</mi>
<mi>t</mi>
<mrow>
<mo>(</mo>
<mi>c</mi>
<mo>)</mo>
</mrow>
</msubsup>
<mo>=</mo>
<msub>
<mi>m</mi>
<mi>c</mi>
</msub>
</mrow>
<mo>)</mo>
<mo>)</mo>
</mrow>
</mrow>
For t observer stateValue be equal to 1 ..., mcProbable value composition a mc× 1 row vector;
They are connected into one 1 × (Σcmc) row vector bt, asProbability distribution;
Wherein,For forward variable of l-th of training sequence after ratio is handled,For l-th training sequence pass through than
Backward variable after example processing, πiIt is initial value,It is by variable πiValue obtained by after training, aijSquare is shifted for state before training
The element of the i-th row j row in battle array A;The observation probability that observed value for j-th of state of t+1 moment is l;Represent t+
1 moment observed value is l,Represent to produce the probability that observed value is l by model λ;Tl(1≤l≤L) is road network
In l-th of information release terminal, l be terminal sequence number, L be Regional Road Network in information release terminal total number, t represent traffic
The collection of stream mode observation and the time of issue and the periodicity of Traffic information demonstration, wherein t ∈ N and 1≤t≤T, T are cut-off
To the total degree of issuing traffic stream mode during t,For t when serial number l terminal where section traffic
The potential value of stream mode,The observation for the t traffic flow modes issued for serial number l terminal,For the potential value of the traffic flow modes in all sections in t road network, { 1 ..., mCFor can on C section
The total quantity of the potential traffic flow modes of division,For t section c sneak condition valueValue is i (1≤i
≤mC) probability of happening,For the vector of the potential traffic flow modes probability of happening of section c (1≤c≤C) t, i.e.,:Remember matrix DC×CFor the matrix of road network, each row sum of matrix is 1, noteMarkov matrixFor internal procedure state-transistion matrix, it is 1, H per a line sum
For a matrix in block form, row c1With row c2Submatrix beWherein each elementP(st)
A series of vector that is made up of associated data of probability distribution, D each row andEvery a line sum all be 1.
4. control method for coordinating according to claim 1, it is characterised in that hidden Markov model solves in the step S3
Algorithm uses EM algorithms, and solving model process is as follows:
Step1:Random generation initial effects matrix H and original state matrix π;
Step2:Structural regime matrix
mcRepresent sneak conditionA random variable values, b1It is 1 for t=1 moment each section observer state value ... mcProbable value
One 1 × (Σ being connected intocmc) row vector, here
btIt is 1 for each section observer state value of t ... mcOne 1 × (Σ being connected into of probable valuecmc) row vector;
Step3:Build probability space P (y):
<mrow>
<mi>P</mi>
<mrow>
<mo>(</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<munder>
<mo>&Pi;</mo>
<mrow>
<mi>t</mi>
<mo>,</mo>
<mi>c</mi>
</mrow>
</munder>
<mrow>
<mo>(</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<msub>
<mi>m</mi>
<mi>c</mi>
</msub>
</munderover>
<msubsup>
<mi>&alpha;</mi>
<mrow>
<mi>t</mi>
<mo>,</mo>
<mi>i</mi>
</mrow>
<mrow>
<mo>*</mo>
<mrow>
<mo>(</mo>
<mi>c</mi>
<mo>)</mo>
</mrow>
</mrow>
</msubsup>
<mo>)</mo>
</mrow>
</mrow>
To be preceding to parameter;
Step4:Pass through step solution procedure parameter backward forward;
m1The section sneak condition for being 1 for numberingA probable value;mCThe section sneak condition for being C for numberingOne
Individual probable value;The sneak condition in the section that the numbering for t construction is 1Value be i probability;For t
The sneak condition in the section that the numbering of construction is cValue be i probability;The section that numbering for t construction is C
Sneak conditionValue be i probability, wherein 1 < c < C;
<mrow>
<msub>
<mi>&alpha;</mi>
<mi>t</mi>
</msub>
<mo>=</mo>
<msubsup>
<mi>&alpha;</mi>
<mi>t</mi>
<mo>*</mo>
</msubsup>
<mo>&CenterDot;</mo>
<msub>
<mi>N</mi>
<mi>t</mi>
</msub>
</mrow>
Form a 1 × mcRow vectorAnd willSeries connection turns into one 1 × (Σcmc) row vector αt, NtFor
Procedure parameter, and have
<mrow>
<msub>
<mi>&beta;</mi>
<mi>t</mi>
</msub>
<mo>=</mo>
<mfenced open = "{" close = "">
<mtable>
<mtr>
<mtd>
<msub>
<mn>1</mn>
<mrow>
<msub>
<mi>&Sigma;m</mi>
<mi>c</mi>
</msub>
<mo>&times;</mo>
<mn>1</mn>
</mrow>
</msub>
</mtd>
<mtd>
<mrow>
<mi>t</mi>
<mo>=</mo>
<mi>T</mi>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mrow>
<mi>H</mi>
<mo>&CenterDot;</mo>
<mi>d</mi>
<mi>i</mi>
<mi>a</mi>
<mi>g</mi>
<mo>&lsqb;</mo>
<msub>
<mi>b</mi>
<mi>t</mi>
</msub>
<mo>&rsqb;</mo>
<mo>&CenterDot;</mo>
<msub>
<mi>N</mi>
<mrow>
<mi>t</mi>
<mo>+</mo>
<mn>1</mn>
</mrow>
</msub>
<mo>&CenterDot;</mo>
<msub>
<mi>&beta;</mi>
<mrow>
<mi>t</mi>
<mo>+</mo>
<mn>1</mn>
</mrow>
</msub>
</mrow>
</mtd>
<mtd>
<mrow>
<mi>t</mi>
<mo><</mo>
<mi>T</mi>
</mrow>
</mtd>
</mtr>
</mtable>
</mfenced>
</mrow>
For backward parameter,For a mc× 1 row
VectorAnd willSeries connection turns into (a ΣcmcThe column vector β of) × 1t;
γt=αt·diag[βt]εt-1→t=diag [αt-1]·H·diag[bt]·Nt·diag[βt]
For unilateral parameter,For a 1 × mcRow vector
And willSeries connection turns into one 1 × (∑cmc) row vector γt,
, will for Bilateral parameterComposition oneMatrixWherein,
ForThe i-th row j row element;
Step5:Standardize original state matrix and A, d, π, π(c)Meet in vector each number and for 1;
Wherein, π(c)For unilateral parameterNormalized form, A(i,j)For Bilateral parameterNormalized form, S is just
Friendshipization matrix, STFor orthogonalization matrix S transposition, di,jBilateral parameterNormalized form after orthogonalization,For unilateral parameter,For a 1 × mcRow vector
<mrow>
<msup>
<mi>A</mi>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>)</mo>
</mrow>
</msup>
<mo>=</mo>
<mi>n</mi>
<mi>o</mi>
<mi>r</mi>
<mi>m</mi>
<mi>a</mi>
<mi>l</mi>
<mi>i</mi>
<mi>z</mi>
<mi>e</mi>
<mo>&lsqb;</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>t</mi>
<mo>=</mo>
<mn>2</mn>
</mrow>
<mi>T</mi>
</munderover>
<msubsup>
<mi>&epsiv;</mi>
<mrow>
<mi>t</mi>
<mo>-</mo>
<mn>1</mn>
<mo>&RightArrow;</mo>
<mi>t</mi>
</mrow>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>)</mo>
</mrow>
</msubsup>
<mo>&rsqb;</mo>
</mrow>
<mrow>
<msub>
<mi>d</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
</mrow>
</msub>
<mo>=</mo>
<mi>n</mi>
<mi>o</mi>
<mi>r</mi>
<mi>m</mi>
<mi>a</mi>
<mi>l</mi>
<mi>i</mi>
<mi>z</mi>
<mi>e</mi>
<mo>&lsqb;</mo>
<mi>S</mi>
<mrow>
<mo>(</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>t</mi>
<mo>=</mo>
<mn>2</mn>
</mrow>
<mi>T</mi>
</munderover>
<msub>
<mi>&epsiv;</mi>
<mrow>
<mi>t</mi>
<mo>-</mo>
<mn>1</mn>
<mo>&RightArrow;</mo>
<mi>t</mi>
</mrow>
</msub>
<mo>)</mo>
</mrow>
<msup>
<mi>S</mi>
<mi>T</mi>
</msup>
<mo>&rsqb;</mo>
</mrow>
<mrow>
<msup>
<mi>&pi;</mi>
<mrow>
<mo>(</mo>
<mi>c</mi>
<mo>)</mo>
</mrow>
</msup>
<mo>=</mo>
<mi>n</mi>
<mi>o</mi>
<mi>r</mi>
<mi>m</mi>
<mi>a</mi>
<mi>l</mi>
<mi>i</mi>
<mi>z</mi>
<mi>e</mi>
<mo>&lsqb;</mo>
<msubsup>
<mi>&gamma;</mi>
<mn>1</mn>
<mrow>
<mo>(</mo>
<mi>c</mi>
<mo>)</mo>
</mrow>
</msubsup>
<mo>&rsqb;</mo>
<mo>;</mo>
</mrow>
Step6:By π(c)Bring into Step2, iterate;If P (y) keeps stable, solution terminates;If P (y) changes, repeat
2~Step of Step 5 iterate;
Wherein, for allNote:
It is c for numbering1Section sneak conditionA probable value,It is c for numbering2Section sneak condition
A probable value;I is setAn element;
To be preceding to parameter,Form a 1 × mc's
Row vectorAnd willSeries connection turns into one 1 × (∑cmc) row vector αt;
For backward parameter,For a mc× 1 row
VectorAnd willSeries connection turns into (a ∑cmcThe column vector β of) × 1t;
For unilateral parameter,For a 1 × mcRow to
AmountAnd willSeries connection turns into one 1 × (∑cmc) row vector γt;
, will for Bilateral parameter
Composition oneMatrixWherein,ForThe i-th row j row element.
5. usage right requires that the variable information advices plate in 1-4 under the information guidance of any control method for coordinating coordinates control
System, it is characterised in that the coordinated control system includes:
Data acquisition module, for gathering transport information parameter value, the transport information parameter value includes traffic flow density, OK
Journey time, road average speed and roadway occupancy;
Data processing module, by establishing hidden Markov model, the data of collection are handled, obtain road net traffic state
Information;
Information display module, treated transport information is issued by variable information advices plate, wherein variable information
Advices plate is one or more groups of, and the information of the variable information advices plate issue can influence the Path selection in way driver.
6. coordinated control system according to claim 5, it is characterised in that the data source of the data acquisition module
Including:Through street detecting system, plate recognition system, coil detecting system and 122 warning systems, wherein
The through street detecting system is used to pass through remote microwave sensor detecting system, gathers and transmits section traffic fluxion
According to;
The plate recognition system is used to gather section upstream crossing vehicle license, the seizure detection data of downstream road junction;
The coil detecting system is used to detect the magnitude of traffic flow and calculates road average speed and roadway occupancy;
122 warning system is used to obtain the alert notice in section.
7. coordinated control system according to claim 5, it is characterised in that the data processing module also includes outside induction room
Shield Surveillance center, swap server, multiple induction outdoor screens monitoring branch center and corresponding multiple tuning controllers, wherein
Treated transport information is transferred to multiple induction rooms by the induction outdoor screen Surveillance center by swap server
External screen monitors branch center;
The swap server enters row information issue based on induction strategies;
The multiple induction outdoor screen monitoring branch center is by corresponding multiple tuning controllers by treated traffic
Information transfer is to described information display module.
8. coordinated control system according to claim 5, it is characterised in that described information display module is variable information information
Plate, every group of variable information advices plate include multiple induction outdoor screens.
9. coordinated control system according to claim 5, it is characterised in that the traffic state information also includes being manually entered
Information.
10. coordinated control system according to claim 5, it is characterised in that data acquisition module, data processing module and letter
Information transfer between breath display module is carried out by Ethernet Private LAN.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710467946.2A CN107437339A (en) | 2017-06-20 | 2017-06-20 | Variable information advices plate control method for coordinating and system under a kind of information guidance |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710467946.2A CN107437339A (en) | 2017-06-20 | 2017-06-20 | Variable information advices plate control method for coordinating and system under a kind of information guidance |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107437339A true CN107437339A (en) | 2017-12-05 |
Family
ID=60458658
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710467946.2A Pending CN107437339A (en) | 2017-06-20 | 2017-06-20 | Variable information advices plate control method for coordinating and system under a kind of information guidance |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107437339A (en) |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108492374A (en) * | 2018-01-30 | 2018-09-04 | 青岛中兴智能交通有限公司 | The application process and device of a kind of AR on traffic guidance |
CN108694834A (en) * | 2018-07-24 | 2018-10-23 | 深圳市显科科技有限公司 | A kind of display methods and display system of guidance information |
CN109614066A (en) * | 2018-12-19 | 2019-04-12 | 北京南师信息技术有限公司 | Information display method and device |
CN109766642A (en) * | 2019-01-15 | 2019-05-17 | 电子科技大学 | One kind is from evolution traffic network topological modelling approach |
CN109859467A (en) * | 2019-01-30 | 2019-06-07 | 银江股份有限公司 | A kind of mining analysis method of Environmental Factors in traffic model |
CN110599769A (en) * | 2019-09-10 | 2019-12-20 | 南京城建隧桥经营管理有限责任公司 | Hierarchical ranking method for road importance in urban road network in time intervals |
CN111833630A (en) * | 2019-12-31 | 2020-10-27 | 北京嘀嘀无限科技发展有限公司 | Method and device for determining data release position, storage medium and electronic equipment |
CN111932899A (en) * | 2020-10-15 | 2020-11-13 | 江苏广宇协同科技发展研究院有限公司 | Traffic emergency control method and device based on traffic simulation |
CN113034904A (en) * | 2021-03-05 | 2021-06-25 | 交通运输部公路科学研究所 | ETC data-based traffic state estimation method and device |
CN113837421A (en) * | 2020-06-23 | 2021-12-24 | 济南市公安局交通警察支队 | Method for calculating information board release range |
CN113963559A (en) * | 2021-10-19 | 2022-01-21 | 北京中交国通智能交通系统技术有限公司 | Vehicle-road cooperative roadside device information publishing method based on event importance |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101281684A (en) * | 2008-01-30 | 2008-10-08 | 吉林大学 | Area traffic control system for synergism operation of inducement and zone control of display panel |
CN103413443A (en) * | 2013-07-03 | 2013-11-27 | 太原理工大学 | Short-term traffic flow forecasting method based on hidden Markov model |
CN104318792A (en) * | 2014-10-10 | 2015-01-28 | 同济大学 | Method for evaluating parking/driving guidance panel based on amount of effective information |
CN105303856A (en) * | 2015-11-11 | 2016-02-03 | 清华大学 | Variable message sign information release method of prediction model |
CN105513375A (en) * | 2015-09-21 | 2016-04-20 | 青岛智能产业技术研究院 | Regional safety traffic control system |
CN106816018A (en) * | 2017-02-16 | 2017-06-09 | 上海电科智能系统股份有限公司 | A kind of city changeable-message sign traffic above-ground induction section determines method |
-
2017
- 2017-06-20 CN CN201710467946.2A patent/CN107437339A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101281684A (en) * | 2008-01-30 | 2008-10-08 | 吉林大学 | Area traffic control system for synergism operation of inducement and zone control of display panel |
CN103413443A (en) * | 2013-07-03 | 2013-11-27 | 太原理工大学 | Short-term traffic flow forecasting method based on hidden Markov model |
CN104318792A (en) * | 2014-10-10 | 2015-01-28 | 同济大学 | Method for evaluating parking/driving guidance panel based on amount of effective information |
CN105513375A (en) * | 2015-09-21 | 2016-04-20 | 青岛智能产业技术研究院 | Regional safety traffic control system |
CN105303856A (en) * | 2015-11-11 | 2016-02-03 | 清华大学 | Variable message sign information release method of prediction model |
CN106816018A (en) * | 2017-02-16 | 2017-06-09 | 上海电科智能系统股份有限公司 | A kind of city changeable-message sign traffic above-ground induction section determines method |
Non-Patent Citations (1)
Title |
---|
GUANGYU ZHU等: "A traffic flowstatetransitionmodelforurbanroadnetworkbasedon Hidden Markov Model", 《NEUROCOMPUTING》 * |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108492374A (en) * | 2018-01-30 | 2018-09-04 | 青岛中兴智能交通有限公司 | The application process and device of a kind of AR on traffic guidance |
CN108492374B (en) * | 2018-01-30 | 2022-05-27 | 青岛中兴智能交通有限公司 | Application method and device of AR (augmented reality) in traffic guidance |
CN108694834A (en) * | 2018-07-24 | 2018-10-23 | 深圳市显科科技有限公司 | A kind of display methods and display system of guidance information |
CN109614066A (en) * | 2018-12-19 | 2019-04-12 | 北京南师信息技术有限公司 | Information display method and device |
CN109766642A (en) * | 2019-01-15 | 2019-05-17 | 电子科技大学 | One kind is from evolution traffic network topological modelling approach |
CN109859467A (en) * | 2019-01-30 | 2019-06-07 | 银江股份有限公司 | A kind of mining analysis method of Environmental Factors in traffic model |
CN110599769A (en) * | 2019-09-10 | 2019-12-20 | 南京城建隧桥经营管理有限责任公司 | Hierarchical ranking method for road importance in urban road network in time intervals |
CN111833630A (en) * | 2019-12-31 | 2020-10-27 | 北京嘀嘀无限科技发展有限公司 | Method and device for determining data release position, storage medium and electronic equipment |
CN113837421A (en) * | 2020-06-23 | 2021-12-24 | 济南市公安局交通警察支队 | Method for calculating information board release range |
CN111932899A (en) * | 2020-10-15 | 2020-11-13 | 江苏广宇协同科技发展研究院有限公司 | Traffic emergency control method and device based on traffic simulation |
CN113034904A (en) * | 2021-03-05 | 2021-06-25 | 交通运输部公路科学研究所 | ETC data-based traffic state estimation method and device |
CN113963559A (en) * | 2021-10-19 | 2022-01-21 | 北京中交国通智能交通系统技术有限公司 | Vehicle-road cooperative roadside device information publishing method based on event importance |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107437339A (en) | Variable information advices plate control method for coordinating and system under a kind of information guidance | |
CN109544932B (en) | Urban road network flow estimation method based on fusion of taxi GPS data and gate data | |
CN107180530B (en) | A kind of road network trend prediction method based on depth space-time convolution loop network | |
CN106384504B (en) | City road network congestion characteristic descriptive analysis method based on data visualization | |
CN104778834B (en) | Urban road traffic jam judging method based on vehicle GPS data | |
CN108492555A (en) | A kind of city road net traffic state evaluation method and device | |
CN103699775B (en) | Urban road traffic guidance strategy automatic generation method and system | |
CN109754597A (en) | A kind of urban road area congestion regulating strategy recommender system and method | |
CN105513370B (en) | The traffic zone division methods excavated based on sparse license plate identification data | |
CN106205114A (en) | A kind of Freeway Conditions information real time acquiring method based on data fusion | |
CN102346964A (en) | Real-time jam prediction and intelligent management system for road traffic network area | |
CN112613225B (en) | Intersection traffic state prediction method based on neural network cell transmission model | |
CN103824450B (en) | Based on the large-scale activity Special running layout of roads method of traffic behavior rule | |
CN109191849A (en) | A kind of traffic congestion Duration Prediction method based on multi-source data feature extraction | |
CN107886720A (en) | One kind closes festivals or holidays urban traffic blocking transfer analysis method based on section type-2 fuzzy sets | |
CN105139670A (en) | Video-based regional self-optimizing signal control method and apparatus | |
Sun et al. | Exploration on spatiotemporal data repairing of parking lots based on recurrent GANs | |
CN112884014A (en) | Traffic speed short-time prediction method based on road section topological structure classification | |
CN106384507A (en) | Travel time real-time estimation method based on sparse detector | |
CN116543552A (en) | Urban arterial road slow blocking strategy simulation evaluation method and system based on vehicle-road cooperation | |
Alam et al. | Intellegent traffic light control system for isolated intersection using fuzzy logic | |
CN105118310A (en) | Video-based single-point self-optimization signal control method and device | |
CN104966404A (en) | Single-point self-optimization signal control method and device based on array radars | |
CN110111561A (en) | A kind of road network method for dynamically partitioning based on SOM-PAM clustering algorithm | |
CN105070075A (en) | Trunk line self-optimization signal control method based on array radar and device |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20171205 |
|
RJ01 | Rejection of invention patent application after publication |