CN113380027B - Intersection traffic state parameter estimation method and system based on multi-source data - Google Patents

Intersection traffic state parameter estimation method and system based on multi-source data Download PDF

Info

Publication number
CN113380027B
CN113380027B CN202110598416.8A CN202110598416A CN113380027B CN 113380027 B CN113380027 B CN 113380027B CN 202110598416 A CN202110598416 A CN 202110598416A CN 113380027 B CN113380027 B CN 113380027B
Authority
CN
China
Prior art keywords
vehicle
determining
traffic
flow
road section
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202110598416.8A
Other languages
Chinese (zh)
Other versions
CN113380027A (en
Inventor
黄玮
胡洋
胡晶
胡芙瑜
张轩宇
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Sun Yat Sen University
Original Assignee
Sun Yat Sen University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Sun Yat Sen University filed Critical Sun Yat Sen University
Priority to CN202110598416.8A priority Critical patent/CN113380027B/en
Publication of CN113380027A publication Critical patent/CN113380027A/en
Application granted granted Critical
Publication of CN113380027B publication Critical patent/CN113380027B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G1/00—Traffic control systems for road vehicles
    • G08G1/01—Detecting movement of traffic to be counted or controlled
    • G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
    • G08G1/0125—Traffic data processing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G1/00—Traffic control systems for road vehicles
    • G08G1/01—Detecting movement of traffic to be counted or controlled
    • G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G1/00—Traffic control systems for road vehicles
    • G08G1/07—Controlling traffic signals
    • G08G1/08—Controlling traffic signals according to detected number or speed of vehicles
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T10/00—Road transport of goods or passengers
    • Y02T10/10—Internal combustion engine [ICE] based vehicles
    • Y02T10/40—Engine management systems

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Analytical Chemistry (AREA)
  • Economics (AREA)
  • Human Resources & Organizations (AREA)
  • Strategic Management (AREA)
  • Chemical & Material Sciences (AREA)
  • Game Theory and Decision Science (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • Operations Research (AREA)
  • General Business, Economics & Management (AREA)
  • Marketing (AREA)
  • Theoretical Computer Science (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Development Economics (AREA)
  • Traffic Control Systems (AREA)

Abstract

The invention discloses an intersection traffic state parameter estimation method and system based on multi-source data, wherein the method comprises the following steps: establishing a vehicle delay model according to a traffic wave theory, and establishing a flow direction-level vehicle travel time model according to the speed distribution characteristics of vehicles; detecting and acquiring the flow of a road section entrance by a fixed point detector, and determining the number of arriving vehicles at the road section entrance; determining the traffic flow steering ratio of the road section by setting a virtual travel route according to the vehicle track obtained by the movement detector; determining Bayesian networks in different flow directions according to the vehicle travel time model, the number of arriving vehicles and the traffic flow steering ratio; solving the Bayesian network and determining the maximum queuing lengths and the vehicle average travel time of different flow directions. According to the embodiment of the application, vehicle track data and fixed point detector data are combined, a parameter estimation method integrating traffic wave theory and statistical analysis is established, and the accuracy of traffic state parameter estimation is effectively improved.

Description

Intersection traffic state parameter estimation method and system based on multi-source data
Technical Field
The application relates to the field of traffic control, in particular to an intersection traffic state parameter estimation method and system based on multi-source data.
Background
Traffic jam is a representative 'urban disease' problem in the novel urbanization construction process of China, and causes huge loss on the development of socioeconomic performance. Optimization of the traffic condition of the intersection is a key for preventing and relieving traffic jam, and estimation of traffic state parameters of the intersection is an indispensable precondition for traffic optimization. In a traditional traffic signal system, a fixed point detector is mostly used for acquiring and estimating traffic state parameters, but factors such as the arrangement position and the detection frequency of the fixed point detector all affect the estimation precision of the traffic state parameters, and the maintenance cost of the fixed point detector is high.
With the rapid development of information technology, motion detectors are beginning to emerge. By taking an intelligent networked automobile as an example, high-precision vehicle track data can be acquired through wireless communication between the vehicle-mounted unit and the road side unit, and traffic state estimation is performed through multidimensional information contained in the vehicle track data. In the related art, the traffic state parameter methods based on vehicle trajectory data may be generally classified into a traffic flow model-based method and a machine learning-based method: the method based on the traffic flow model generally utilizes a sample data fitting model to analyze and obtain a deterministic traffic state, and often ignores the random fluctuation of actual traffic conditions; although the random characteristic is considered in the method based on machine learning, the process completely depending on statistical analysis generally only establishes a simple probability correlation function, ignores the physical characteristics of the traffic flow, and is difficult to see the causal nature of a physical system through data. That is, both methods are superior and inferior. In addition, at the present stage, the popularity of the movement detector is not high, the vehicle track data still has the problem of low permeability, and the insufficiency of the vehicle track data also has adverse effects on the accuracy of the estimation of the traffic state parameters.
Disclosure of Invention
The present application is directed to solving, at least to some extent, one of the technical problems in the related art. Therefore, the method and the system for estimating the intersection traffic state parameters based on the multi-source data can be combined with data of the fixed point detector and the mobile detector and a method based on a traffic flow model and machine learning, and the robustness, the interpretability and the estimation precision of the method for estimating the traffic state parameters are effectively improved.
In a first aspect, an embodiment of the present application provides an intersection traffic state parameter estimation method based on multi-source data, including: establishing a vehicle delay model according to a traffic wave theory; determining a vehicle travel time model of a flow direction level according to the vehicle delay model and the speed distribution characteristics of vehicles in different flow directions; determining the number of arriving vehicles at the road section entrance according to the road section entrance flow acquired by the fixed point detector; determining the traffic flow steering ratio of a road section by setting a virtual travel line according to the vehicle track acquired by the movement detector; determining Bayesian networks in different flow directions according to the vehicle travel time model, the number of arriving vehicles and the traffic steering ratio; and determining the estimated values of the maximum queuing lengths in different flow directions according to the traffic wave theory and the solution of the Bayesian network, and determining the estimated value of the average travel time of the vehicle.
Optionally, the step of establishing the vehicle delay model specifically includes: determining a first moment when a current flowing vehicle reaches a road section inlet in a current signal period; if the vehicle in the current flow direction is not saturated in the current signal period, determining a second moment when the vehicle reaches the vehicle entrance after the red light delay, and determining the vehicle delay model according to the first moment and the second moment; and if the current flowing vehicle is oversaturated in the current signal period, determining a third moment when the vehicle reaches the vehicle entrance after the red light delay, and determining the vehicle delay model according to the first moment and the third moment.
Optionally, the current flowing direction of the vehicle is not saturated by: in the current signal period, all vehicles flowing to the current intersection pass through the intersection, and the initial queuing length does not exist at the current intersection in the next signal period; the current direction of vehicle over-saturation means: in the current signal period, the part of the vehicle which flows to the current intersection passes through the intersection, and the initial queuing length exists at the current intersection in the next signal period.
Optionally, the speed distribution characteristics of the vehicles in different flow directions are specifically: : the target reciprocal obeys normal distribution, and the speed distribution characteristics comprise a target mean value and a target standard deviation; wherein the target reciprocal is the reciprocal of the free flow speed of the vehicle in the current flow direction; wherein the target mean is a mean of the target reciprocal, and the target standard deviation is a standard deviation of the target reciprocal.
Optionally, the determining, according to the vehicle track obtained by the movement detector, the traffic flow turning ratio of the road segment by setting a virtual route line includes: acquiring the vehicle tracks in different flow directions in different signal periods, and acquiring track data of a designated section in the vehicle tracks according to a set virtual travel line; when the vehicle passes through the virtual travel line, acquiring the current moment, the travel line ID and the vehicle ID; matching the downstream vehicle ID acquired by the downstream road section virtual travel line with the upstream vehicle ID acquired by the upstream road section virtual travel line; when the downstream vehicle ID is matched with the upstream vehicle ID, the counter of the downstream road section counts; when the current analysis interval is finished, determining the traffic flow steering ratio of the road section according to the counter value of the downstream road section corresponding to the current flow direction and the counter values of all the downstream road sections; wherein the analysis interval comprises a number of signal periods.
Optionally, the bayesian network is a three-layer structure, a first-layer parameter of the bayesian network is a poisson distribution parameter, a second-layer parameter includes a target mean value, a target standard deviation, the number of arriving vehicles and the traffic steering ratio, and a third-layer parameter includes the vehicle trajectory.
Optionally, the constructing step of the bayesian network comprises: determining a first probability distribution model between a first layer parameter and a second layer parameter of the Bayesian network according to the Poisson distribution parameters and the number of arriving vehicles; determining a second probability distribution model between the second layer parameter and the third layer parameter of the Bayesian network according to the speed distribution characteristics, the number of arriving vehicles, the traffic flow steering ratio and the vehicle track; determining a joint probability distribution model from the first probability distribution model and the second probability distribution model.
Optionally, the bayesian network is solved by using maximum likelihood estimation, and the specific step of solving the bayesian network includes: converting the joint probability distribution model into a log-likelihood function; according to a log-likelihood function, respectively carrying out derivation on the target mean value, the target standard deviation and the Poisson distribution parameters, and determining a solution of the Bayesian network; wherein the solution to the Bayesian network comprises an estimate of the target mean, an estimate of the target standard deviation, and an estimate of the Poisson distribution parameter.
Optionally, the determining the estimated values of the maximum queuing lengths in different flow directions and the determining the estimated value of the vehicle mean travel time according to the traffic wave theory and the solution of the bayesian network comprises: acquiring an arrival flow estimation value of a road section inlet and an initial queuing length estimation value in the current flow direction; in the current signal period, determining the estimated value of the maximum queuing length in the current flow direction according to the estimated value of the arrival flow and the estimated value of the initial queuing length; determining a total delay estimation value according to the estimation value of the maximum queuing length; and determining the estimated value of the average travel time of the vehicle according to the total delay estimated value.
In a second aspect, an embodiment of the present application provides an intersection traffic state parameter estimation system based on multi-source data, including: the delay calculation module is used for establishing a vehicle delay model according to a traffic wave theory; the travel time calculation module is used for determining a vehicle travel time model of a flow direction level according to the vehicle delay model and the speed distribution characteristics of vehicles with different flow directions; the flow calculation module is used for determining the number of arriving vehicles at the road section inlet according to the road section inlet flow acquired by the fixed point detector; the steering ratio calculation module is used for determining the traffic flow steering ratio of the road section by setting a virtual travel line according to the vehicle track acquired by the movement detector; the statistical analysis module is used for determining Bayesian networks with different flow directions according to the vehicle travel time model, the number of arriving vehicles and the traffic flow steering ratio; and the traffic state parameter estimation module is used for determining the maximum queuing lengths and the vehicle average travel time in different flow directions according to the traffic wave theory and the solution of the Bayesian network.
The beneficial effects of the embodiment of the application are as follows: the embodiment of the application combines the vehicle track data and the fixed point detector data, and can still provide a more reliable traffic state parameter estimation result under the condition of lower vehicle track data sampling rate; secondly, the method for estimating the parameters by fusing the traffic wave theory and the statistical analysis is established, a vehicle delay model is established based on the traffic wave theory, and a statistical relationship between observation data and an analysis model is established through a Bayesian network, so that the random volatility of actual traffic conditions is considered, the physical characteristics of a traffic system are combined, and the accuracy of the estimation of the traffic state parameters is effectively improved; in addition, the method and the device can realize joint estimation of the intersection free flow speed and the traffic state variable, and further improve the intersection traffic running state monitoring level.
Drawings
The accompanying drawings are included to provide a further understanding of the claimed subject matter and are incorporated in and constitute a part of this specification, illustrate embodiments of the subject matter and together with the description serve to explain the principles of the subject matter and not to limit the subject matter.
FIG. 1 is a method for estimating intersection traffic state parameters based on multi-source data according to an embodiment of the present application;
FIG. 2 is a basic macroscopic view of traffic flow provided by embodiments of the present application;
FIG. 3 is a flowchart illustrating steps for modeling vehicle delays according to an embodiment of the present application;
fig. 4 is a time-space diagram of a vehicle trajectory and a propagation process of a traffic wave in a single intersection flowing to the z-th signal cycle in an unsaturated state according to an embodiment of the present application;
FIG. 5 is a traffic wave propagation process and vehicle trajectory space-time diagram for a single intersection in the flow direction z-th signal cycle under oversaturation conditions as provided by an embodiment of the present application;
fig. 6 is a schematic diagram of traffic conditions at an intersection according to an embodiment of the present application;
FIG. 7 is a flowchart of a step of calculating a road traffic steering ratio according to an embodiment of the present application;
FIG. 8 is a flowchart of steps provided in an embodiment of the present application for constructing a Bayesian network;
FIG. 9 is a flowchart illustrating steps for solving a Bayesian network as provided by an embodiment of the present application;
FIG. 10 is a flowchart illustrating steps for traffic state parameter estimation provided by an embodiment of the present application;
fig. 11 is a schematic diagram of an intersection traffic state parameter estimation system based on multi-source data according to an embodiment of the present application.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
It should be noted that although functional block divisions are provided in the system drawings and logical orders are shown in the flowcharts, in some cases, the steps shown and described may be performed in different orders than the block divisions in the systems or in the flowcharts. The terms first, second and the like in the description and in the claims, and the drawings described above, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order.
For the purpose of facilitating an understanding of the embodiments of the present application, a brief introduction to the related concepts related to the present application will be provided below:
flow direction: a signalized intersection may have several entrance lanes, and a vehicle may turn left, go straight, or turn right after entering the intersection, and the combination of the entrance lane and the turn that the vehicle enters at the current intersection is referred to as the direction of flow of the vehicle. For example, a ═ {1, 2., m } represents a number set of entrance lanes, and H ═ { Le, St, Ri } represents a turn set of entrance lanes, where m represents the number of entrance lanes, Le represents a left turn, St represents a straight run, Ri represents a right turn, and a combination (a, H) of entrance lanes and turns (a e a, H e H) represents one flow direction (a turn of one entrance lane, possibly including multiple lanes), denoted as z, and for convenience of description, z is also used below to represent a vehicle flow direction z.
Signal phase: at a signalized intersection, each control state of the intersection (i.e., the combination of different light colors displayed for different turns of each approach) is referred to as a signal phase, which may include one or more flow directions. For convenience of explanation, the expressions "signalized intersection" and "signalized intersection" in the whole text are the same meaning and refer to "signalized intersection".
Signal period: a phase is the time that elapses from the start of one red light time until the start of the next red light time. In the embodiment of the present application, an analysis interval includes several signal periods.
Free flow: when the traffic flow density is small, the driver can drive according to the driving characteristics, vehicle conditions and road conditions, and the traffic flow state of the expected speed is maintained without or with little influence of other users on the road.
Saturation flow rate: a continuous fleet of vehicles on an approach may be able to pass the maximum flow of the approach stop line during a continuous green light signal time.
In the related art, fixed point detectors such as a loop coil, a microwave radar and a video detector are used for acquiring and estimating traffic state parameters of an intersection, however, the use of the fixed point detectors for estimating the traffic state parameters has certain limitations, on one hand, factors such as the arrangement position, reliability and detection frequency of the detectors can affect the estimation precision, and on the other hand, the installation, maintenance and operation costs of the fixed point detectors are relatively high. Therefore, it is an important development trend to utilize, for example, floating cars, intelligent networked cars, etc. to perform movement detection on vehicles, a movement detector can acquire multi-dimensional and high-precision vehicle trajectory data, a traffic state parameter method based on the vehicle trajectory data can be generally divided into a method based on a traffic flow model and a method based on machine learning, while a method based on a traffic flow model alone often ignores random fluctuation of actual traffic conditions, and a method based on machine learning ignores physical characteristics of traffic flow itself, and both methods have disadvantages. In addition, the current mobile detector is not popular enough, the urban road is still in a stage of coexistence of the fixed detector and the mobile detector for a long time, and the problems of low vehicle track data sampling rate and insufficient data also become a barrier for improving the estimation accuracy of the traffic state parameters.
Based on the method and the system, the intersection traffic state parameter estimation method and the system based on the multi-source data are provided, the vehicle track data and the fixed point detector data are combined, and a reliable traffic state parameter estimation result can be provided under the condition that the sampling rate of the vehicle track data is low; secondly, the method for estimating the parameters by fusing the traffic wave theory and the statistical analysis is established, a vehicle delay model is established based on the traffic wave theory, and a statistical relationship between observation data and an analysis model is established through a Bayesian network, so that the random volatility of actual traffic conditions is considered, the physical characteristics of a traffic system are combined, and the accuracy of the estimation of the traffic state parameters is effectively improved; in addition, the method and the device can realize joint estimation of the intersection free flow speed and the traffic state variable, and further improve the intersection traffic running state monitoring level.
The embodiments of the present application will be further explained with reference to the drawings.
Referring to fig. 1, fig. 1 is a method for estimating intersection traffic state parameters based on multi-source data according to an embodiment of the present application, where the method includes, but is not limited to, steps S100 to S150:
s100, establishing a vehicle delay model according to a traffic wave theory;
specifically, traffic wave theory is first briefly explained. The traffic wave theory specifically comprises the analysis of a macroscopic traffic flow basic diagram and a vehicle track space-time diagram. Referring to fig. 2, fig. 2 is a basic macroscopic traffic flow diagram provided in the embodiment of the present application, as shown in fig. 2, a horizontal axis represents vehicle density k, a vertical axis represents vehicle flow q, and the basic macroscopic traffic flow diagram is used to describe a basic relationship among the flow q, the density k, and a spatial average velocity v, and the basic relationship may be represented as: q ═ kv. A triangular base map may be constructed from the saturation flow rates q in FIG. 2cFree flow velocity vfAnd the blocking density kjThree parameters are determined, and further, other parameters in fig. 2 can be derived through the three parameters, and the specific derivation is disclosed as follows:
kc＝qc/vf,
ka,i＝qa,i/vf,
w＝qc/(kj-kc),
wa,i＝qa,i/(kj-ka.i)
wherein k iscIs critical density, ka,iFor intersection arrival density in the ith signal cycle, qa,iThe intersection arrival flow in the ith signal cycle, w is the evanescent wave velocity, wa,iAnd queuing the wave speed for the intersection in the ith signal period.
And the vehicle track space-time diagram is used for describing the flow-direction stage queuing evolution process and the delay mode analysis. In step S100, a vehicle delay model is established based on the traffic wave theory, and it should be noted that the vehicle delay model is a traffic-flow-level vehicle delay model and represents delay conditions of vehicles flowing in different directions at an intersection.
Referring to fig. 3, fig. 3 is a flowchart illustrating steps of building a vehicle delay model according to an embodiment of the present application, where the steps of the method include, but are not limited to, steps S101-S104:
s101, determining a first moment when a current flowing vehicle reaches a road section inlet in a current signal period;
specifically, in the embodiment of the present application, delay condition analysis is performed on vehicles in different flow directions in different signal periods, and therefore, it is necessary to project each time when a vehicle actually travels in a vehicle track to a time in a signal period, so as to facilitate analysis in the signal period, for example, the time when the vehicle reaches a signal control intersection, and the like.
Referring to fig. 4, fig. 4 is a traffic wave propagation process and a vehicle track space-time diagram in a single intersection flowing to z ith signal cycle in an unsaturated state, where a current flowing direction is z and a current signal cycle is ith signal cycle, a time when a vehicle flowing to z arrives at a signalized intersection in the ith signal cycle is projected to a time when the vehicle arrives at a signalized intersection section entrance, where the time is called a first time and is represented by ti,zTo indicate, determine ti,zThe steps are as follows:
firstly, the time t of starting red light of the ith signal period flowing to z is startedR,i,zAnd the time t when the red light starts in the (i + 1) th signal cycleR,i+1,zAre respectively projected to ts,j,zAnd ts,j+1,z，ts,j,zThe calculation formula of (a) is as follows:
ts,i,z＝tR,i,z-l·pf,z+Qi,0,z/qc
ts,i,zthe time indicated is the speed v of the vehicle in free flow when the vehicle is moving in the direction zf,zThrough the road entrance, the vehicle will not experience the delay caused by the red light time before the ith signal period, but will experience the full red light time of the ith signal periodThe time point of the delay, l being the distance from the entrance of the road section to the stop line, pf,zVehicle free flow velocity v in order of direction zf,zReciprocal of (2), Qi,0,zFor the initial queue length in the ith signal period flowing to z, Q is the initial queue length in the embodiment of this applicationi,0,zIn particular the number of vehicles in line. In addition, the evolution process of initial queuing to different signal periods of z conforms to the law of conservation of flow, so that Qi+1,0,zThe calculation formula of (a) is as follows:
Qi+1,0,z＝max{0,Qi,0,z+qa,i,z(tc,i,z-ts,i,z)-TG,i,zqc}
wherein q isa,i,zFor intersection arrival flow in the ith signal cycle flowing to z, TG,i,zFor duration of green light, t, in the ith signal period flowing to zc,i,zThe second time when the vehicle reaches the entrance of the road section in the flow direction z can be determined by the following step S103
Then, the vehicle will flow to z at tR,i,zAnd tR,i+1,zTime between passing through the stop line is projected to ts,j,zAnd ts,j+1,zFirst moment t of passing through the section of road entrancei,z. After the projection is completed, it should be noted that all the signal periods mentioned in the following refer to the signal periods after the projection, i.e. from ts,j,zTo ts,j+1,zThe elapsed time.
S102, judging the current flowing vehicle state in the current signal period;
specifically, the vehicle conditions are classified as unsaturated and oversaturated depending on whether the queue in the ith signal period of flow z can be emptied, see FIG. 4, due to Qi,0,z+qa,i,z(tc,i,z-ts,i,z)-TG,i,zqcAnd if the value is less than or equal to 0, the queuing in the signal period can be emptied, the initial queuing length does not exist in the next signal period, and the vehicle state belongs to the unsaturated condition.
Referring to fig. 5, fig. 5 is a traffic wave propagation process and a vehicle trajectory space-time diagram in the ith signal period of a single intersection flowing to z in an oversaturated state provided by the embodiment of the present application, where the current flowing direction is z and the current signal is zThe signal period is the ith signal period. Referring to FIG. 5, due to Qi,0,z+qa,i,z(tc,i,z-ts,i,z)-TG,i,zqc>0, it means that the queue in the signal period cannot be completely emptied, and the initial queue length will be generated in the next signal period, so that the current vehicle status belongs to the over-saturation condition.
S103, if the vehicle in the current flow direction is not saturated in the current signal period, determining a second moment when the vehicle reaches the entrance of the road section after the red light delay, and determining a vehicle delay model according to the first moment and the second moment;
specifically, in the unsaturated condition, a second time when the vehicle reaches the road section entrance is calculated according to whether the vehicle in the flow direction z experiences delay caused by red light in the ith signal cycle, and the second time is tc,i,zDenotes, tc,i,zThe calculation formula is as follows:
Figure GDA0003505065980000071
wherein, TR,i,zThe duration of the red light in the ith signal period flowing to z. After the second time is determined, t is respectively calculatedi,z≤tc,i,zAnd ti,z>tc,i,zIn the case of a vehicle delay model of flow direction z, the vehicle delay model being Di,zRepresents, calculates Di,zThe steps are as follows:
first, referring to FIG. 4, when t isi,z≤tc,i,zThe vehicle flowing to z experiences a delay caused by the red light in the ith signal cycle, and thus Di,zThe calculation formula of (a) is as follows:
Figure GDA0003505065980000081
and, referring to FIG. 4, when t isi,z>tc,i,zThe vehicle moving to z travels through the intersection at free speed, i.e. the vehicle does not experience the delay caused by the red light in the ith signal cycle, so Di,z＝0。
S104, if the current flowing vehicle is over-saturated in the current signal period, determining a third moment when the vehicle reaches the entrance of the road section after the red light delay, and determining a vehicle delay model according to the first moment and the third moment;
in particular, in case of over-saturation, a third time of arrival of the vehicle in flow direction z at the entrance of the stretch is calculated, with t, depending on whether the vehicle in flow direction z experiences a delay caused by the red light of the (i + 1) th signal cycleb,i,zTo indicate. Referring to FIG. 5, according to the law of conservation of flow, the flow direction z is the initial queue length of the ith signal period, plus ts,j,zAnd tb,i,zThe number of vehicles in between, equal to the number of vehicles passing the stop line at the saturation flow rate for the duration of the green light of the current signal period, the mathematical formula is expressed as follows:
qa.i.z(tb,i,z-ts,i,z)+Qi,0,z＝TG,i,zqc
transforming the above equation to obtain tb,i,zThe calculation formula of (c) is as follows:
Figure GDA0003505065980000082
as shown in fig. 5, in the case of over-saturation, the second time tc,i,zAnd ts,j+1,zIs the same point in time.
After the third time is determined, t is respectively calculatedi,z≤tb,i,zAnd ti,z>tb,i,zIn the case of the vehicle delay model flowing to z, as described above, the vehicle delay model is Di,zIndicating that D is calculated in the case of supersaturationi,zComprises the following steps:
first, referring to FIG. 5, when ti,z≤tb,i,zThen it means that the vehicle experiencing a delay in flow z caused by the red light in the ith signal cycle and passing the stop line during the green light in the current signal cycle, then Di,zAnd t in the case of no saturationi,z≤tc,i,zThe same, namely the following formula:
Figure GDA0003505065980000083
when t isi,z>tb,i,zReferring to fig. 5, after the vehicle flowing to z experiences the delay caused by the red light of the ith signal period, the vehicle fails to pass through the stop line during the green light of the current signal period, and further experiences the delay caused by all the red lights of the (i + 1) th signal period, at this time, the vehicle delays the model Di,zThe calculation formula of (c) is as follows:
Figure GDA0003505065980000091
through steps S101-S104, a vehicle delay model is determined, beginning with the description of step S110 in FIG. 1.
S110, determining a vehicle travel time model of a flow direction level according to a vehicle delay model and speed distribution characteristics of vehicles in different flow directions;
specifically, a vehicle travel time model of the flow direction level is determined according to a vehicle delay model and speed distribution characteristics of vehicles with different flow directions. In the embodiment of the present application, the speed distribution of the vehicle is specifically characterized by: setting the reciprocal p of the free flow speed of the vehicle in the direction zf,zObeying a normal distribution, called pf,zIs the target reciprocal, the target reciprocal pf,zAs shown in the following formula:
Figure GDA0003505065980000092
wherein, N represents a normal distribution,
Figure GDA0003505065980000093
is pf,zMean value of
Figure GDA0003505065980000094
Is a target mean value;
Figure GDA0003505065980000095
is pf,zStandard deviation of (1), title
Figure GDA0003505065980000096
Target mean to target standard deviation
Figure GDA0003505065980000097
And target standard deviation
Figure GDA0003505065980000098
Is a speed profile characteristic of the vehicle. The process of establishing a model of the travel time of the vehicle for the flow direction stage is set forth below.
Firstly, establishing an analytical expression of a vehicle travel time model of a flow direction stage, wherein the expression is as follows:
Ti,z＝Di,z+lzpf,z
wherein, Ti,zFor the time of travel of the vehicle in the ith signal period, lzIs the distance between the entrance of the upstream section and the entrance of the downstream section in the flow direction z.
Then, combining the speed distribution characteristics of vehicles with different flow directions and the analytic expressions, establishing a vehicle travel time model of the flow direction stage, wherein the expressions are as follows:
Figure GDA0003505065980000099
assuming that the timing information and saturation flow rate of the signalized intersection are known parameters, Di,zCan be regarded as a parameter
Figure GDA00035050659800000910
The mathematical expression is as follows:
Figure GDA00035050659800000911
when the current signal flows to the arriving vehicle at the entrance of the signal control intersection road section in the ith signal periodNumber of vehicles ni,zWhen q is greater than qz,i,zThe calculation formula of (c) is as follows:
Figure GDA0003505065980000101
wherein, betazIs the turning ratio of the traffic flowing to z in an analysis interval.
To sum up, Ti,zAlso obey by parameters
Figure GDA0003505065980000102
The determined normal distribution is expressed as follows:
Figure GDA0003505065980000103
wherein Q isi,0,zN can be determined according to the above step S100i,zCan be obtained by the fixed point detector in the following step S120, betai,zAnd ti,zCan be determined directly or indirectly according to the acquired vehicle track data,
Figure GDA0003505065980000104
and
Figure GDA0003505065980000105
is the parameter to be estimated.
S120, determining the number of arriving vehicles at the road section entrance according to the road section entrance flow acquired by the fixed point detector;
specifically, the fixed point detector includes, but is not limited to, a loop coil, a microwave radar, and a video detector, and the embodiments of the present application exemplify a high frequency coil detector arranged at an entrance of each road section and having a sampling frequency of 1s, which provides flow and occupancy data per second of the section entrance cross section. According to the road section entrance flow acquired by the fixed point detector, the number of arriving vehicles at the road section entrance can be determined, and the number of arriving vehicles specifically is as follows: the number of vehicles arriving at the entrance of the signal control intersection in the ith signal cycle of the flow direction z is ni,zIt is shown that,for convenience of description, n is also used hereinafteri,zIndicating the number of arriving vehicles.
S130, determining the traffic flow steering ratio of the road section by setting a virtual travel route according to the vehicle track obtained by the movement detector;
specifically, the Virtual Trip Line (VTL) is specifically: a geographic marker indicating where the networked vehicle should provide information updates. In the embodiment of the application, the vehicle track data is acquired, and then the track data of the designated section is extracted according to the set virtual travel line, wherein the designated section can be an upstream or downstream entrance of each road section. And providing the current time point, the travel route ID and the vehicle ID information every time the networked vehicle passes through the virtual travel route. In the embodiment of the application, the virtual travel line of the upstream road section of the signal control intersection associated with the flow direction z is set as VTLup,zAnd the virtual travel line of the downstream road section of the signal control intersection associated with the flow direction z is VTLdown,z. In addition, because the virtual travel line only extracts the track data of the designated section in the embodiment of the application, the track of the vehicle can be protected to a certain extent.
Referring to fig. 6, fig. 6 is a schematic diagram of traffic conditions at an intersection according to the embodiment of the present application, and as shown in fig. 6, a vehicle travels from left to right, an entrance of a left road segment is set as an upstream virtual travel line, and the vehicle enters a different lane to turn when turning green and travels to a different downstream virtual travel line. The number n of arriving vehicles mentioned in the above step S120i,zFor the flow to the z-th signal cycle, the road-level vehicle set count data needs to be converted into the flow-level vehicle set count data, in this embodiment, by converting the road-level vehicle set count data ni,zMultiplied by the steering ratio beta in the direction zzObtaining the flow direction level vehicle set counting data ni,zβz。
Referring to fig. 7, fig. 7 is a flowchart illustrating steps of calculating a road traffic steering ratio according to an embodiment of the present application, where the steps include, but are not limited to, steps S131 to S134:
s131, when the vehicle passes through the virtual travel line, acquiring the current moment, the travel line ID and the vehicle ID;
specifically, referring to FIG. 6, the vehicle travels an upstream road segment VTLup,zThen, the current time, the travel route ID and the vehicle ID are obtained, and the VTL passes through the downstream road sectiondown,zAnd similarly, according to the time when the vehicle passes through the virtual travel line, the vehicle track data matrix in different signal periods with different flow directions can be determined, and the vehicle track data matrix is expressed as STi,z，STi,zThe mathematical expression of (a) is as follows:
Figure GDA0003505065980000111
wherein k isi,zFor the total number of networked vehicles flowing in the z-th signal cycle,
Figure GDA0003505065980000112
for j-th internet connection vehicle passing through upstream road section VTL in ith signal period flowing to zup,zAt the point in time of (a) in time,
Figure GDA0003505065980000113
for the j net vehicle in the ith signal period flowing to z to get from the upstream section VTLup,zTo a downstream road segment VTLdown,zThe travel time of (c).
S132, matching the downstream vehicle ID acquired by the downstream road section virtual travel line with the upstream vehicle ID acquired by the upstream road section virtual travel line;
in particular, VTL for the downstream road sectiondown,zAllocating a counter, the value of which is represented by Cdown,zSince the information of the current time point, the trip line ID, and the vehicle ID is provided every time the internet-connected vehicle passes through the virtual trip line, the downstream vehicle ID acquired on the downstream section of the virtual trip line is matched with the upstream vehicle ID acquired on the upstream section.
S133, when the downstream vehicle ID is matched with the upstream vehicle ID, the counter of the downstream road section counts;
specifically, when the downstream vehicle ID matches the upstream vehicle ID, indicating that the vehicle driving on the upstream link has reached the current downstream link, the corresponding downstream link counterCdown,z＝Cdown,z+1. As mentioned above, the vehicle may turn left, go straight, or turn right after entering the intersection, and by matching the downstream vehicle ID with the upstream vehicle ID, the turning of the vehicle, i.e., whether the vehicle is turning left, going straight, or turning right, may be determined.
S134, when the current analysis interval is finished, determining the traffic flow steering ratio of the road section according to the counter value of the downstream road section corresponding to the current flow direction and the counter values of all the downstream road sections;
specifically, in an analysis interval, the traffic steering ratio β of the current road section can be determined according to the counter value of the downstream road section corresponding to the current flow direction and the counter values of all the downstream road sectionsz。
Figure GDA0003505065980000114
After the traffic flow steering ratio of the current analysis interval is determined, all the downstream road segment counters C are useddown,zAnd (4) zero clearing, and counting the next analysis interval.
Through steps S131 to S134, the flow turning ratio is calculated, and step S140 in fig. 1 is explained.
S140, determining Bayesian networks in different flow directions according to the vehicle travel time model, the number of arriving vehicles and the traffic flow steering ratio;
specifically, according to a vehicle travel time model, the number of arriving vehicles and the traffic flow steering ratio, a Bayesian network is established for different flow directions. The Bayesian network is of a three-layer structure, and the number n of arriving vehicles at a road section entrance in an analysis interval is set in the embodiment of the applicationi,zObeying the Poisson distribution, the first layer parameter of the Bayesian network is a Poisson distribution parameter lambda, and the second layer parameter comprises a target mean value
Figure GDA0003505065980000121
And target standard deviation
Figure GDA0003505065980000122
Number of arriving vehicles ni,zSum traffic steering ratio betazThe third layer parameter includes the vehicle track STi,z。
Referring to fig. 8, fig. 8 is a flowchart illustrating steps of constructing a bayesian network according to an embodiment of the present application, where the method includes, but is not limited to, steps S141 to S143:
s141, determining a first probability distribution model between a first layer parameter and a second layer parameter of the Bayesian network according to the Poisson distribution parameters and the number of arriving vehicles;
specifically, in an analysis interval, if the number of vehicles arriving at a road section entrance obeys Poisson distribution, a first-layer parameter lambda and a second-layer parameter n in the Bayesian network are establishedi,zFirst probability distribution model Q (n) betweeni,z(ii) a λ), the first probability distribution model is represented as follows:
Figure GDA0003505065980000123
wherein:
Δts,i,z＝ts,i+1,z-ts,i,z
λ is the poisson distribution parameter over an analysis interval.
S142, determining a second probability distribution model between the second-layer parameter and the third-layer parameter of the Bayesian network according to the speed distribution characteristics, the number of arriving vehicles, the traffic flow steering ratio and the vehicle track;
in particular, velocity profile characteristics are incorporated
Figure GDA0003505065980000124
And
Figure GDA0003505065980000125
number of arriving vehicles ni,zTraffic steering ratio betazAnd vehicle track STi,zEstablishing second layer parameters in a Bayesian network
Figure GDA0003505065980000126
And third layer parameter STi,zSecond probability distribution module in betweenModel (III)
Figure GDA0003505065980000127
Wherein the content of the first and second substances,
Figure GDA0003505065980000128
is composed of
Figure GDA0003505065980000129
Has already been mentioned in the above step S110, and in summary, Ti,zAlso obey parameter
Figure GDA00035050659800001210
The determined normal distribution is expressed as follows:
Figure GDA00035050659800001211
and S143, determining a joint probability distribution model according to the first probability distribution model and the second probability distribution model.
Specifically, according to a first probability distribution model and a second probability distribution model, n in the ith signal period of the flow direction z is establishedi,zAnd all of
Figure GDA0003505065980000131
The joint probability distribution model of (2), the expression of which is as follows:
Figure GDA0003505065980000132
wherein, Tz＝{T1,z,T2,z,...,TIAnd the travel time set of the networked vehicles in an analysis interval is represented by the flow direction z.
For the joint probability distribution model described above, ni,zCan be determined from the data of the fixed-point detector,
Figure GDA0003505065980000133
and betazCan be determined by the above-mentioned steps, and therefore, only the poisson distribution parameter λ, the velocity distribution characteristic, and the sum of the unknown parameters in the bayesian network are those of the shaded portion
Figure GDA0003505065980000134
And
Figure GDA0003505065980000135
through steps S141-S143, a bayesian network is obtained, and step S150 in fig. 1 is set forth below.
And S150, determining the maximum queuing lengths and the average vehicle travel time in different flow directions according to the traffic wave theory and the solution of the Bayesian network.
Specifically, as can be seen from the bayesian network constructed in step S140, the bayesian network in this application has three unknown parameters to be solved, which are: poisson distribution parameter λ, velocity distribution characteristic
Figure GDA0003505065980000136
And
Figure GDA0003505065980000137
therefore, the solving process of the Bayesian network is mainly the solving process of the three unknown parameters, and the maximum queuing lengths and the vehicle average travel time in different flow directions can be determined according to the solution of the Bayesian network. The solution process for the bayesian network is first set forth below.
Referring to fig. 9, fig. 9 is a flowchart illustrating steps of solving a bayesian network according to an embodiment of the present application, where the method includes, but is not limited to, steps S151 to S153:
s151, converting the joint probability distribution model into a log-likelihood function;
specifically, the joint probability distribution model obtained in step S143 is converted into a log-likelihood function, and the specific mathematical expression is as follows:
Figure GDA0003505065980000138
s152, according to a log-likelihood function, respectively carrying out derivation on the target mean value, the target standard deviation and the Poisson distribution parameter, and determining a solution of the Bayesian network;
specifically, according to the above-mentioned log-likelihood function, for the poisson distribution parameter λ,
Figure GDA0003505065980000139
and
Figure GDA00035050659800001310
and (4) carrying out derivation, wherein the derivation result is zero, and the estimation values of the three variables can be obtained, and the estimation values of the three variables are the solutions of the Bayesian network. The specific derivation formula is as follows:
Figure GDA0003505065980000141
Figure GDA0003505065980000142
Figure GDA0003505065980000143
wherein the content of the first and second substances,
Figure GDA0003505065980000144
is an estimate of the target mean,
Figure GDA0003505065980000145
Is an estimate of the target standard deviation and,
Figure GDA0003505065980000146
for convenience of description, the estimate of the Poisson distribution parameter is hereinafter denoted by the superscript ^ for the estimate of the variable.
The solution to the bayesian network is completed through steps S151-S152, and the following sets forth the steps of determining the maximum queue length and the vehicle mean travel time for different flow directions according to the traffic wave theory and the solution to the bayesian network.
Referring to fig. 10, fig. 10 is a flowchart illustrating steps of traffic state parameter estimation provided in the embodiment of the present application, where the method includes, but is not limited to, steps S153-S156:
s153, acquiring an arrival flow estimation value of a road section inlet and an initial queuing length estimation value in the current flow direction;
specifically, for
Figure GDA0003505065980000147
To pair
Figure GDA0003505065980000148
And the data of the road section inlet arrival flow collected by the fixed point detector is collected and measured to obtain
Figure GDA0003505065980000149
And
Figure GDA00035050659800001410
number of vehicles arriving at the entrance of the road section
Figure GDA00035050659800001411
Wherein the content of the first and second substances,
Figure GDA00035050659800001412
and
Figure GDA00035050659800001413
are each ts,i,zAnd tc,i,zCan be calculated by the formula in step S101, except that p isf,zIs replaced by
Figure GDA00035050659800001414
Qi,0,zIs replaced by
Figure GDA00035050659800001415
The specific expression is as follows:
Figure GDA00035050659800001416
Figure GDA00035050659800001417
Figure GDA00035050659800001418
can be composed of
Figure GDA00035050659800001419
The calculation results in that,
Figure GDA00035050659800001420
specifically, the formula in step S103 can be referred to, and it should be noted that in this embodiment, it is assumed that the initial queue length is known at the beginning of an analysis interval, that is, the initial queue length is known, that is, the analysis interval is a queue length of a queue of a
Figure GDA00035050659800001421
S154, determining the estimated value of the maximum queuing length in the current flow direction according to the estimated value of the arrival flow and the estimated value of the initial queuing length in the current signal period;
referring to fig. 4, according to the traffic wave theory, the maximum queue length in the ith signal period of the flow direction z can be expressed as: estimate of initial queue length flowing into z-th signal period
Figure GDA00035050659800001422
And
Figure GDA00035050659800001423
and
Figure GDA00035050659800001424
number of vehicles arriving at the entrance of the road section
Figure GDA00035050659800001425
Sum, calculateThe formula is as follows:
Figure GDA00035050659800001426
wherein the content of the first and second substances,
Figure GDA0003505065980000151
is the maximum queue length estimate in the ith signal period flowing to z.
S155, determining a total delay estimated value according to the estimated value of the maximum queuing length;
specifically, referring to FIG. 5, the vehicle delay to the z-th signal cycle consists of two parts, one is the delay resulting from the queue not being completely cleared in the current signal cycle and the initial queue length being established in the next signal cycle, which can be expressed as
Figure GDA0003505065980000152
The other part is the delay generated by the red light in the current signal period, the vehicle delay of the part is in a linear decreasing trend along with the increase of the arrival time of the vehicle, and the average value of the vehicle delay is TR,i,zAnd/2, the total delay of this part can be expressed as: mean value of vehicle delay multiplied by
Figure GDA0003505065980000153
And
Figure GDA0003505065980000154
number of vehicles arriving at the entrance of road between
Figure GDA0003505065980000155
Therefore, the total delay in the ith signal period flowing to z is calculated as follows:
Figure GDA0003505065980000156
wherein the content of the first and second substances,
Figure GDA0003505065980000157
as an estimate of the total vehicle delay in the ith signal period flowing to z, TR,i,zThe duration of the red light is the flow z ith signal period.
And S156, determining an estimated value of the average travel time of the vehicle according to the total delay estimated value.
Specifically, the vehicle average travel time of the ith signal cycle flowing to z can be calculated according to the determination in the above steps, and the calculation formula is as follows:
Figure GDA0003505065980000158
wherein the content of the first and second substances,
Figure GDA0003505065980000159
is an estimate of the average travel time of the vehicle during the ith signal period flowing to z,
Figure GDA00035050659800001510
is composed of
Figure GDA00035050659800001511
And
Figure GDA00035050659800001512
the road section between the two reaches the flow.
The estimation of the maximum queue length and the vehicle mean travel time for different flows is accomplished through steps S153-S156.
Through the steps S100-S150, a vehicle delay model is built according to a traffic wave theory, and a flow-direction-level vehicle travel time model is built according to the vehicle delay model and the speed distribution characteristics of different flow-direction vehicles at the intersection; detecting the flow obtained to the road section entrance by a fixed point detector to determine the number of vehicles at the road section entrance; according to the vehicle track obtained by the movement detector, a virtual travel route is set, and the traffic flow steering ratio of the road section is determined according to the condition that the vehicle passes through the virtual travel route; determining Bayesian networks in different flow directions according to the vehicle travel time model, the number of arriving vehicles and the traffic flow steering ratio; and determining the maximum queuing lengths and the average vehicle travel of the intersection in different flow directions by solving the Bayesian network. The embodiment of the application combines the vehicle track data and the fixed point detector data, and can still provide a more reliable traffic state parameter estimation result under the condition of lower vehicle track data sampling rate, so that the method provided by the application has better robustness; secondly, the method for estimating the parameters by fusing the traffic wave theory and the statistical analysis is established, a vehicle delay model is established based on the traffic wave theory, the statistical relationship between observation data and an analysis model is established through a Bayesian network, the random volatility of actual traffic conditions is considered, the physical characteristics of a traffic system are combined, and the accuracy of the estimation of the traffic state parameters is effectively improved, so that the method provided by the application has strong interpretability; in addition, the method and the device can realize joint estimation of the intersection free flow speed and the traffic state variable, and the estimation result can be widely applied to signal-controlled intersection optimization, so that the urban intersection traffic running state monitoring level is further improved, and information support is provided for an intelligent traffic system more efficiently.
Referring to fig. 11, fig. 11 is a schematic diagram of an intersection traffic state parameter estimation system based on multi-source data according to an embodiment of the present application, where the system 1100 includes: a delay calculation module 1110, a travel time calculation module 1120, a flow calculation module 1130, a turn ratio calculation module 1140, a statistical analysis module 1150, and a traffic state parameter estimation module 1160. The delay calculation module is used for establishing a vehicle delay model according to a traffic wave theory; the travel time calculation module is used for determining a vehicle travel time model of a flow direction level according to the vehicle delay model and the speed distribution characteristics of vehicles with different flow directions; the flow calculation module is used for determining the number of arriving vehicles at the road section inlet according to the road section inlet flow acquired by the fixed point detector; the steering ratio calculation module is used for determining the traffic flow steering ratio of a road section by setting a virtual travel line according to the vehicle track acquired by the movement detector; the statistical analysis module is used for determining Bayesian networks in different flow directions according to the vehicle travel time model, the number of arriving vehicles and the traffic flow steering ratio; and the traffic state parameter estimation module is used for determining the maximum queuing lengths and the vehicle average travel time in different flow directions according to the traffic wave theory and the solution of the Bayesian network.
It will be understood by those of ordinary skill in the art that all or some of the steps, systems, and methods disclosed above may be implemented as software, firmware, hardware, or suitable combinations thereof. Some or all of the physical components may be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application specific integrated circuit. Such software may be distributed on computer readable media, which may include computer storage media (or non-transitory media) and communication media (or transitory media). The term computer storage media includes volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data, as is well known to those of ordinary skill in the art. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, Digital Versatile Disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can accessed by a computer. In addition, communication media typically embodies computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism and includes any information delivery media as is well known to those skilled in the art.
While the preferred embodiments of the present invention have been described, the present invention is not limited to the above embodiments, and those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of the present invention, and such equivalent modifications or substitutions are included in the scope of the present invention defined by the claims.

Claims (10)

1. An intersection traffic state parameter estimation method based on multi-source data is characterized by comprising the following steps:
establishing a vehicle delay model according to a traffic wave theory;
determining a vehicle travel time model of a flow direction level according to the vehicle delay model and the speed distribution characteristics of vehicles in different flow directions;
determining the number of arriving vehicles at the road section entrance according to the road section entrance flow acquired by the fixed point detector;
determining the traffic flow steering ratio of a road section by setting a virtual travel line according to the vehicle track acquired by the movement detector;
determining Bayesian networks in different flow directions according to the vehicle travel time model, the number of arriving vehicles and the traffic steering ratio;
and determining the estimated values of the maximum queuing lengths in different flow directions according to the traffic wave theory and the solution of the Bayesian network, and determining the estimated value of the average travel time of the vehicle.
2. The multi-source data-based intersection traffic state parameter estimation method according to claim 1, wherein the vehicle delay model establishing step specifically comprises:
determining a first moment when a current flowing vehicle reaches a road section inlet in a current signal period;
if the vehicle in the current flow direction is not saturated in the current signal period, determining a second moment when the vehicle reaches the vehicle entrance after the red light delay, and determining the vehicle delay model according to the first moment and the second moment;
and if the current flowing vehicle is oversaturated in the current signal period, determining a third moment when the vehicle reaches the vehicle entrance after the red light delay, and determining the vehicle delay model according to the first moment and the third moment.
3. The multi-source data-based intersection traffic state parameter estimation method according to claim 2, characterized in that:
the current flowing direction vehicle is not saturated: in the current signal period, all vehicles flowing to the current intersection pass through the intersection, and the initial queuing length does not exist at the current intersection in the next signal period;
the current direction of vehicle over-saturation means: in the current signal period, the part of the vehicle which flows to the current intersection passes through the intersection, and the initial queuing length exists at the current intersection in the next signal period.
4. The multi-source data-based intersection traffic state parameter estimation method according to claim 1, wherein the speed distribution characteristics of the vehicles flowing in different directions are specifically: the target reciprocal obeys normal distribution, and the speed distribution characteristics comprise a target mean value and a target standard deviation;
wherein the target reciprocal is the reciprocal of the free flow speed of the vehicle in the current flow direction;
wherein the target mean is a mean of the target reciprocal, and the target standard deviation is a standard deviation of the target reciprocal.
5. The method for estimating the intersection traffic state parameters based on the multi-source data according to claim 1, wherein the step of determining the traffic flow turning ratio of the road section by setting a virtual travel route according to the vehicle track acquired by the movement detector comprises the following steps:
acquiring the vehicle tracks in different flow directions in different signal periods, and acquiring track data of a specified section in the vehicle tracks according to a set virtual travel line;
when the vehicle passes through the virtual travel line, acquiring the current moment, the travel line ID and the vehicle ID;
matching the downstream vehicle ID acquired by the downstream road section virtual travel line with the upstream vehicle ID acquired by the upstream road section virtual travel line;
when the downstream vehicle ID is matched with the upstream vehicle ID, the counter of the downstream road section counts;
when the current analysis interval is finished, determining the traffic flow steering ratio of the road section according to the counter value of the downstream road section corresponding to the current flow direction and the counter values of all the downstream road sections;
wherein the analysis interval comprises a number of signal periods.
6. The multi-source data based intersection traffic state parameter estimation method according to claim 4, wherein the Bayesian network is a three-layer structure, a first-layer parameter of the Bayesian network is a Poisson distribution parameter, a second-layer parameter includes the target mean, the target standard deviation, the number of arriving vehicles, and the traffic-flow steering ratio, and a third-layer parameter includes the vehicle trajectory.
7. The multi-source data based intersection traffic state parameter estimation method according to claim 6, wherein the Bayesian network construction step comprises:
determining a first probability distribution model between a first layer parameter and a second layer parameter of the Bayesian network according to the Poisson distribution parameters and the number of arriving vehicles;
determining a second probability distribution model between the second layer parameter and the third layer parameter of the Bayesian network according to the speed distribution characteristics, the number of arriving vehicles, the traffic flow steering ratio and the vehicle track;
determining a joint probability distribution model from the first probability distribution model and the second probability distribution model.
8. The method for estimating intersection traffic state parameters based on multi-source data according to claim 7, wherein the Bayesian network is solved by using maximum likelihood estimation, and the concrete steps of solving the Bayesian network comprise:
converting the joint probability distribution model to a log-likelihood function;
according to a log-likelihood function, respectively carrying out derivation on the target mean value, the target standard deviation and the Poisson distribution parameters, and determining a solution of the Bayesian network;
wherein the solution to the Bayesian network comprises an estimate of the target mean, an estimate of the target standard deviation, and an estimate of the Poisson distribution parameter.
9. The method for estimating intersection traffic state parameters based on multi-source data according to claim 1, wherein determining estimated values of maximum queuing lengths in different flow directions and determining estimated values of vehicle mean travel time according to a traffic wave theory and a solution of the Bayesian network comprises:
acquiring an estimated value of arrival flow of a road section inlet and an estimated value of initial queuing length in the current flow direction;
in the current signal period, determining the estimated value of the maximum queuing length in the current flow direction according to the estimated value of the arrival flow and the estimated value of the initial queuing length;
determining a total delay estimation value according to the estimation value of the maximum queuing length;
and determining the estimated value of the average travel time of the vehicle according to the total delay estimated value.
10. An intersection traffic state parameter estimation system based on multi-source data is characterized by comprising:
the delay calculation module is used for establishing a vehicle delay model according to a traffic wave theory;
the travel time calculation module is used for determining a vehicle travel time model of a flow direction level according to the vehicle delay model and the speed distribution characteristics of vehicles with different flow directions;
the flow calculation module is used for determining the number of arriving vehicles at the road section entrance according to the road section entrance flow acquired by the fixed point detector;
the steering ratio calculation module is used for determining the traffic flow steering ratio of the road section by setting a virtual travel line according to the vehicle track acquired by the movement detector;
the statistical analysis module is used for determining Bayesian networks with different flow directions according to the vehicle travel time model, the number of arriving vehicles and the traffic flow steering ratio;
and the traffic state parameter estimation module is used for determining the estimated values of the maximum queuing lengths in different flow directions and the estimated value of the vehicle average travel time according to the traffic wave theory and the solution of the Bayesian network.
CN202110598416.8A 2021-05-31 2021-05-31 Intersection traffic state parameter estimation method and system based on multi-source data Active CN113380027B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110598416.8A CN113380027B (en) 2021-05-31 2021-05-31 Intersection traffic state parameter estimation method and system based on multi-source data

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110598416.8A CN113380027B (en) 2021-05-31 2021-05-31 Intersection traffic state parameter estimation method and system based on multi-source data

Publications (2)

Publication Number Publication Date
CN113380027A CN113380027A (en) 2021-09-10
CN113380027B true CN113380027B (en) 2022-07-05

Family

ID=77575020

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110598416.8A Active CN113380027B (en) 2021-05-31 2021-05-31 Intersection traffic state parameter estimation method and system based on multi-source data

Country Status (1)

Country Link
CN (1) CN113380027B (en)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114155724B (en) * 2021-12-23 2022-11-22 连云港杰瑞电子有限公司 Intersection traffic signal control method in Internet of vehicles environment
CN114822055B (en) * 2022-06-06 2023-06-09 深圳英博达智能科技有限公司 Intelligent traffic road cooperation system based on machine vision detection
CN115019507B (en) * 2022-06-06 2023-12-01 上海旷途科技有限公司 Urban road network travel time reliability real-time estimation method
CN115083162B (en) * 2022-06-15 2023-11-17 北京三快在线科技有限公司 Road condition prediction method, device, equipment and storage medium
CN115691105B (en) * 2022-09-06 2023-09-15 昆明理工大学 Maximum-voltage signal control method considering pedestrian crossing of discrete road section of motorcade

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20130103300A1 (en) * 2011-10-25 2013-04-25 Nokia Corporation Method and apparatus for predicting a travel time and destination before traveling
US10706720B2 (en) * 2017-09-16 2020-07-07 Tata Consultancy Services Limited Predicting vehicle travel times by modeling heterogeneous influences between arterial roads
CN111862602A (en) * 2020-06-24 2020-10-30 南京理工大学 Single-intersection signal timing optimization method based on track data

Also Published As

Publication number Publication date
CN113380027A (en) 2021-09-10

Similar Documents

Publication Publication Date Title
CN113380027B (en) Intersection traffic state parameter estimation method and system based on multi-source data
CN108053645B (en) Signal intersection periodic flow estimation method based on track data
CN113327416B (en) Urban area traffic signal control method based on short-term traffic flow prediction
JP2002008183A (en) Method of determining traffic situation based on reported vehicle data for traffic network including traffic-controlled network nodes
JP6647402B2 (en) Method and device for obtaining traffic light duration data
US8798897B2 (en) Real-time traffic analysis through integration of road traffic prediction and traffic microsimulation models
CN106571030B (en) Queuing length prediction method under multi-source traffic information environment
WO2018064931A1 (en) Method for estimating travel time distribution of taxi on urban roads when operating states of taxis are considered
CN109064741B (en) Method for reconstructing running track of main road vehicle based on multi-source data fusion
CN108597235B (en) Intersection signal parameter optimization and effect evaluation method based on traffic video data
CN109166313B (en) Overflow early warning method according to vehicle passing data
CN101727746A (en) Method for dynamically estimating travel time of motor vehicle on urban road under control of signal lamps
CN107590999A (en) A kind of traffic state judging method based on bayonet socket data
CN115311854A (en) Vehicle space-time trajectory reconstruction method based on data fusion
CN110610118A (en) Traffic parameter acquisition method and device
CN111613051B (en) Method and device for estimating saturated headway
Sheng et al. Real-time queue length estimation with trajectory reconstruction using surveillance data
Bagheri et al. Real-time estimation of saturation flow rates for dynamic traffic signal control using connected-vehicle data
JP2004118700A (en) Prediction device of driving time required
CN113538902A (en) Intersection vehicle track data restoration method based on traffic state
CN111640303A (en) City commuting path identification method and equipment
CN116153084B (en) Vehicle flow direction prediction method, prediction system and urban traffic signal control method
CN110428612A (en) Vehicle management-control method, device, computer equipment and storage medium
JP2002216287A (en) Method of calculating right turn, left turn and straight travel rate at intersection
Wang et al. Estimation of Queue Length at Signalized Intersections based on Electronic Police Data

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant