CN114049765A - Urban road network traffic flow OD estimation method based on automatic vehicle number plate identification data - Google Patents

Urban road network traffic flow OD estimation method based on automatic vehicle number plate identification data Download PDF

Info

Publication number
CN114049765A
CN114049765A CN202111296123.0A CN202111296123A CN114049765A CN 114049765 A CN114049765 A CN 114049765A CN 202111296123 A CN202111296123 A CN 202111296123A CN 114049765 A CN114049765 A CN 114049765A
Authority
CN
China
Prior art keywords
vehicle
track
road network
number plate
data
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202111296123.0A
Other languages
Chinese (zh)
Other versions
CN114049765B (en
Inventor
欧吉顺
聂庆慧
周志刚
邓社军
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.)
Yangzhou University
Original Assignee
Yangzhou 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 Yangzhou University filed Critical Yangzhou University
Priority to CN202111296123.0A priority Critical patent/CN114049765B/en
Publication of CN114049765A publication Critical patent/CN114049765A/en
Application granted granted Critical
Publication of CN114049765B publication Critical patent/CN114049765B/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
    • G08G1/0129—Traffic data processing for creating historical data or processing based on historical data
    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G1/00—Traffic control systems for road vehicles
    • G08G1/01—Detecting movement of traffic to be counted or controlled
    • G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
    • G08G1/0137—Measuring and analyzing of parameters relative to traffic conditions for specific applications
    • 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
    • G08G1/017—Detecting movement of traffic to be counted or controlled identifying vehicles
    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G1/00—Traffic control systems for road vehicles
    • G08G1/065—Traffic control systems for road vehicles by counting the vehicles in a section of the road or in a parking area, i.e. comparing incoming count with outgoing count

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Chemical & Material Sciences (AREA)
  • Analytical Chemistry (AREA)
  • Traffic Control Systems (AREA)

Abstract

The invention discloses an urban road network traffic flow OD estimation method based on automatic vehicle number plate identification data, which is used for researching and analyzing vehicle number plate monitoring data in a selected road network area, carrying out precision processing on the vehicle number plate data, carrying out travel chain division on the vehicle number plate data according to time and space intervals, reconstructing all travel tracks with track loss and acquiring OD flow of the reconstructed travel tracks in the road network area. Compared with the traditional method based on manual investigation, the method extracts complete vehicle travel track information in an automatic mode, estimates the dynamic OD flow of the road network on the basis, obviously reduces the manual workload, and effectively improves the authenticity and reliability of the OD estimation result. Compared with a classical method for reversely deducing OD flow based on the road section flow, the method disclosed by the invention can greatly reduce the complexity of modeling and solving and has better implementability.

Description

Urban road network traffic flow OD estimation method based on automatic vehicle number plate identification data
Technical Field
The invention belongs to the technical field of road traffic information monitoring, and particularly relates to an urban road network traffic flow OD estimation method based on automatic vehicle number plate identification data.
Background
Along with the increase of social economy and the progress of science and technology, the urbanization process of all countries in the world is accelerated continuously, and the scale of urban population is increased continuously. Meanwhile, the problem of traffic jam caused by unbalanced supply and demand of traffic is gradually highlighted, and further development of cities is restricted in various aspects.
Vehicle Origin-Destination (OD) traffic is one of the basic data indexes of an urban transportation system, and describes travel distribution conditions between each pair of ODs in the whole transportation network. In an urban traffic system, OD flow directly reflects the space-time distribution of traffic flow in an urban traffic network, and has important significance for management and control, development planning, perspective prediction and the like of urban road traffic. An OD matrix composed of OD flow as elements is often used as basic input data in a traffic simulation technology to effectively simulate the running condition of urban traffic. Therefore, obtaining accurate and complete OD demand information is a key and hot research content in the traffic field.
Early OD flux was mainly obtained by manual investigation. The investigation methods have the defects of huge consumption of manpower, financial resources and material resources, strong manual dependence, low precision, slow data updating and the like. In addition, for a large-scale traffic network, manual investigation for obtaining OD traffic is more difficult to implement effectively. Thereafter, researchers have proposed indirect estimation methods that reverse the road network OD traffic through partial section traffic. However, since such methods involve the linkage optimization modeling of both dynamic OD estimation and dynamic traffic distribution, the modeling and solving processes are full of great complexity and challenges. In recent years, with the development of traffic detection technology, automatic vehicle identification technology represented by an automatic number plate identification system has been widely used. By recording the time and the identity information of the vehicle entering the road intersection, the data of the automatic number plate recognition system can be used for restoring the travel track of the vehicle in a given road network area. Furthermore, by extracting and aggregating the track information of the vehicles in the road network, the dynamic OD demand information of the road network can be effectively estimated.
Currently, research on OD estimation can be divided into two main categories, static OD estimation and dynamic OD estimation. As cities develop and motor vehicle reserves grow, traffic demand patterns also become more complex. Because the real-time change rule of the traffic demand cannot be reflected, the static OD cannot meet the real-time and accurate requirements required by the modern urban traffic management and control. The existing dynamic OD estimation method is generally based on road section flow to reversely deduce OD flow, complex modeling and solving are needed, and the operation efficiency is low.
Disclosure of Invention
In order to solve the problems, the invention discloses an urban road network traffic flow OD estimation method based on automatic vehicle number plate identification data. According to the method, the vehicle number plate monitoring data in the selected road network area are researched and analyzed, the vehicle number plate data are subjected to precision processing, the vehicle number plate data are subjected to trip chain division according to time and space intervals, all trip tracks with track loss are reconstructed, and the OD flow of the reconstructed trip tracks in the road network area is obtained.
In order to achieve the purpose, the technical scheme of the invention is as follows:
the urban road network traffic flow OD estimation method based on automatic vehicle number plate identification data comprises the following steps:
s1: setting a vehicle number plate detector at a key intersection of a selected road network area, monitoring and processing traffic basic information through the vehicle number plate detector, wherein the traffic basic information comprises vehicle number plate data, and constructing a complete travel track of vehicles in the selected road network area according to the vehicle number plate data and a topological structure of a road network to obtain a travel chain divided by vehicles in the road network area on a single day;
s2: collecting travel chain track information of the vehicle, acquiring space-time track information of the vehicle in a selected road network area, counting the quantity of track information with the same path in the road network area, superposing the quantity of the same travel chain track, and analyzing a flow estimation result of each path in the road network area;
s3: counting the estimation results of all path flows in the road network area, carrying out sample expansion processing on the calculated path flows, determining a specific sample expansion coefficient, and analyzing the total flow of each path in the road network;
s4: collecting the traffic flow OD pairs with the same starting point and the same end point, aiming at each traffic flow OD pair, obtaining a path set contained in each traffic flow OD pair, analyzing the OD flow of all the traffic flow OD pairs with the same starting point and the same end point according to each path in the path set, and analyzing the average OD flow in a road network area according to the OD flow of all the traffic flow OD pairs.
Further, in step S1, the process of monitoring and processing the traffic basic information by the vehicle license plate detector includes the following sub-steps:
s10: arranging a vehicle number plate detector on an entrance lane of a key intersection in a selected road network area, collecting data of each vehicle number plate in the selected road network area, and acquiring an intersection adjacent relation in the road network area to form a selected road network topological structure;
s11: preprocessing acquired vehicle license plate data, screening unrecognizable vehicle license plate data and repeated vehicle license plate data, removing the unrecognizable vehicle license plate data, respectively comparing the acquisition date, the vehicle license plate number, the bayonet number, the equipment number and the lane number of the repeated vehicle license plate data, detecting the acquisition time of the repeated vehicle license plate data when all characteristic values of the data are the same, and keeping the first detected data and removing other repeated data when the acquired time interval is smaller than a set threshold value;
s12: and performing association matching on the preprocessed vehicle number plate data and the intersections of the selected road network area, extracting the time information of a plurality of number plate data of each vehicle, sequencing according to the time sequence, and acquiring the travel track of each vehicle on the current day.
Further, in step S1, the process of performing topology construction on the vehicle travel track in the selected road network area according to the vehicle number plate data to obtain a travel chain divided by the vehicles in the road network area on a single day includes the following steps:
S10-A: dividing the single-day trip chains of all vehicles according to the time and space information in the road network area, screening the trip chains divided by all vehicles in the road network area in a single day manner, judging whether the trip chain track divided by each vehicle is continuous or not, collecting the trip chain with track information loss, carrying out path analysis on the path with track information loss, complementing the missing track information and perfecting the trip chain divided by vehicles in the road network area in a single day manner.
Further, the step S10-a specifically includes the following steps:
s100: counting the number plate data of each vehicle in the selected road network area, and grouping all the vehicle track data to be processed in the selected road network area according to the number plate ID of each vehicle;
s101: obtaining track data of each grouped vehicle in the selected road network area, and sequencing the track data of each group of vehicles according to the time sequence;
s102: preprocessing each group of vehicle track data, and removing repeated records in the vehicle track data grouping and the grouping with only a single track point;
s103: when the track points in the group of vehicle track data are complete, dividing a trip chain of each group of preprocessed vehicle track data according to a timestamp difference value between two adjacent vehicle track point data in the group to obtain a complete track data set of each group of vehicles, and when part of track points in the group of vehicle track data are missing, marking the track to be reconstructed and summarizing the track to be reconstructed into a track set to be reconstructed;
s104: aiming at the complete vehicle license plate data track set, splitting each vehicle license plate data track, taking head and tail track points of a given time period as an input sequence, taking the rest middle track points as an output sequence, and constructing a Seq2Seq training data set;
s105: constructing a track sequence prediction model by utilizing the constructed training data set and a recurrent neural network GRU algorithm;
s106: and aiming at the track set to be reconstructed, splitting adjacent track point pairs of each track, and reconstructing the track of each track point pair by combining a track sequence prediction model.
Further, the step 106 specifically includes the following steps:
s106-1: extracting a group of adjacent track point pairs, and judging whether the track point pairs are adjacent in space;
s106-2: when the set of track points are adjacent in space, reconstruction processing is not needed;
s106-3: when the set of track points are not adjacent in space, judging whether the set of track points have a history mode;
s106-4: when the set of track points has a historical mode, predicting and completing the track-lacking points by using the constructed track sequence prediction model;
s106-5: when the set of track points does not have a historical mode, completing the missing track points by using a shortest path method based on travel time;
s106-6: counting the supplemented missing track points, and screening whether all the track points of the vehicles in the research area are processed;
s106-7: if all track points of the vehicle are not completed, repeating the steps S106-1 to S106-6, and repeatedly extracting a group of adjacent track points to complete the track.
Further, in the step S10-a, the process of dividing the single day travel chains of all vehicles according to the time and space information in the road network area includes the following steps:
s10-a: counting the number plate data of the vehicles in the selected road network area, extracting all number plate data of a certain vehicle in one day, and counting the number plate data as N;
s10-b: arranging N license plate data according to the time sequence according to the acquisition time of each license plate data of the vehicle, and enabling N to be 1;
s10-c: in the arranged vehicle number plate data, the acquisition time of the nth data and the acquisition time of the (n + 1) th data are differed to obtain the actual running time of the vehicle between adjacent number plate detectors, and the actual running time is set as Ti；
S10-d: calculating the shortest path between the two vehicle license plate detectors corresponding to the two vehicle license plate data, and setting the length of the shortest path to be LminSetting the road flow velocity between two vehicle number plate detectors to VFSetting the shortest travel time between two vehicle number plate detectors to TminAccording to the formula: t ismin＝Lmin/VFSetting the driving budget time between two vehicle number plate detectors as
Figure BDA0003336639600000041
Wherein, the value of delta is a constant larger than 1;
s10-e: comparing the actual running time between two vehicle number plate detectors with the shortest running time and the estimated time between two vehicle number plate detectors respectively, and when T isi<TminIf the data is wrong, the data of the vehicle number plate detected by the corresponding vehicle number plate detector is deleted, and if the data is wrong, the data of the vehicle number plate is deleted
Figure BDA0003336639600000042
Explaining the process of the vehicle staying at the corresponding position of one vehicle detector, the trip chain should be disconnected, and the nth data and the (n + 1) th data should belong to the end point and the starting point of two different trip chains;
s10-f: and when N is less than N, making N equal to N +1, and repeating the steps S100-3 to S100-6 until N is equal to N, thereby completing the division of the trip chains of different vehicles in the road network area.
Further, in the step S10-a, screening the trip chains divided by one day for all vehicles in the road network area, determining whether the trip chain trajectory divided by each vehicle is continuous, collecting the trip chain with missing trajectory information, performing path analysis on the path with missing trajectory information, complementing the missing trajectory information, and completing the trip chain divided by one day for the vehicles in the road network area, includes the following steps:
s10-1: screening a trip chain with missing track information in a routing network area, and extracting an alternative path between two vehicle number plate detectors corresponding to the trip chain;
s10-2: when the distance between two vehicle number plate detectors corresponding to the trip chain is close, the alternative path distances between the two vehicle number plate detectors are sequenced, and the length of the ith alternative path between some two vehicle number plate detectors is set to be LiSetting a summary of the i-th alternative route between two vehicle number plate detectorsA rate of Pi1，Pi1＝1/Li；
S10-3: when the distance between two vehicle number plate detectors corresponding to a trip chain is long, the lengths of high-grade roads in alternative paths between the two vehicle number plate detectors are sequenced, wherein the high-grade road is a main road, the total number of the alternative paths between the two vehicle number plate detectors is set to be n, and the length of the main road in the ith alternative path between the two vehicle number plate detectors is set to be n
Figure BDA0003336639600000051
Setting the probability that the ith alternative path between two vehicle license plate detectors is selected as Pi2According to the formula:
Figure BDA0003336639600000052
calculating the probability of selecting the ith alternative path between two current vehicle license plate detectors;
s10-4: detecting whether each alternative path is contained by number plate data detected by a vehicle number plate detector in an original trip chain, comparing the number of the vehicle number plate detectors contained by the original trip chain in a certain alternative path i with the total number of the vehicle number plate detectors installed on the original trip chain, and setting the number of the vehicle number plate detectors on the alternative path i and the original trip chain as MiThe number of the vehicle number plate detectors installed on the original trip chain is NoriginSetting the probability that the ith alternative path between two vehicle number plate detectors is selected as Pi3，Pi3＝Mi/Norigin；
S10-5: according to the steps, all alternative path selection probabilities between two vehicle license plate detectors corresponding to the trip chain with track information missing in the road network area are compared, one path with the highest probability is extracted to serve as a track reconstruction path, and the trip chain divided by vehicles in the road network area on a single day is perfected.
Further, the step S2 includes the following sub-steps:
s21: counting the trip chains with the complement missing track information to obtain the real trip chain of each trip of each vehicle in the selected road network area;
s22: dividing travel chains of all vehicles in a selected road network area by taking hours as units, and screening all tracks of which the track end time of the vehicle travel chains is positioned in a certain period;
s23: counting the number of vehicles with the same track, setting the number of the intersection at the starting point as an intersection 1 and the number of the intersection at the finishing point as an intersection 2 by taking the track of the intersection at a certain starting point as an object, screening the number of the vehicles with the same track from the intersection 1 to the intersection 2, and overlapping, wherein the quantity of screening results is the path flow from the intersection 1 to the intersection 2 in the time period;
s24: and step S23 is repeated, and every two intersections in the selected road network area are processed until the OD flow estimation of all paths in the selected road network area is completed.
Further, the step S3 includes the following sub-steps:
s31: analyzing the vehicle number plate detection permeability of each intersection in the selected road network area, setting the vehicle number plate detection permeability of a certain intersection as gamma, and setting the number of the effective vehicle number plate records collected by the intersection as NiSetting the number of all vehicle number plate records collected at the intersection as
Figure BDA0003336639600000061
According to the formula:
Figure BDA0003336639600000062
calculating the vehicle number plate detection permeability of the current intersection, and calculating the vehicle number plate detection permeability of different intersections in the selected road network area according to the formula;
s32: analyzing the sampling rate of the vehicle license plates at different moments in the selected road network area, acquiring the number of the vehicle license plates collected by the vehicle license plate detector at a certain moment in the road network area, extracting the total number of the vehicles passing through the road network area at the moment, wherein the sampling rate at different moments is the ratio of the number of the vehicle license plates collected by the vehicle license plate detector in the road network area at different moments to the total number of the vehicles in the road network area, and determining the scaling factor of all the paths in the selected road network at the corresponding moment according to the sampling rate;
s33: determining a sample expansion coefficient of a selected road network area according to the detection permeability and the sampling rate of the vehicle license plates at different intersections, wherein the sample expansion coefficient is the product of the permeability and the sampling rate of the corresponding intersection;
s34: the method comprises the steps of obtaining vehicle license plate detection permeability and scaling coefficients of all paths corresponding to different intersections and different time periods in a road network area, setting total flow of each path to be Q, setting OD flow acquired by each path to be Q', setting the scaling coefficient of a certain path corresponding to a time period to be lambda, and according to a formula:
Q＝Q'/(γ*λ)
and calculating to obtain the total flow of each path in the currently selected road network area.
Further, the step S4 includes the following sub-steps:
s41: aiming at a certain traffic flow OD pair, each path in a path set contained in the traffic flow OD pair is obtained, and the path flow of each path is obtained;
s42: stacking the path flow of all paths in the path set to obtain the OD flow of the OD pair of the traffic flow;
s43: analyzing the path flow inside each traffic flow OD pair in the road network area, and counting the OD flow of each traffic flow OD pair;
s44: analyzing OD flow of the same OD pair of the traffic flow in all fixed time intervals in one day, and calculating average OD flow, wherein the average OD flow is the OD flow of the same OD pair in all fixed time intervals in one day divided by the number of the fixed time intervals in one day;
s45: and repeating the steps S41-S44, and analyzing the average OD flow of all the traffic flow OD pairs in the road network area.
The invention has the beneficial effects that:
1. the method can acquire the same path flow of the starting point or the end point of each path in a fixed time period, analyze the travel generation amount and the travel attraction amount of each intersection in the road network area, analyze the average traffic flow OD of the road network in different time periods according to the travel generation amount and the travel attraction amount of each intersection, and perform visual analysis on the OD estimation result according to the average traffic flow OD.
2. The invention analyzes and processes the vehicle number plate data in an automatic mode, extracts complete vehicle travel track information from the vehicle number plate data, estimates the dynamic OD flow of the road network on the basis of aggregation and sample expansion, obviously reduces the manual workload, and effectively improves the authenticity and reliability of the OD estimation result. Compared with a classical method for reversely deducing OD flow based on the road section flow, the method disclosed by the invention can greatly reduce the complexity of modeling and solving, has better implementability, and provides reliable decision support for making scientific and reasonable road traffic control strategies.
Drawings
Fig. 1 is an overall flowchart of an urban road network traffic flow OD estimation method based on automatic vehicle number plate identification data according to the present invention.
Fig. 2 is a partial flowchart of step S1.
Fig. 3 is a partial flowchart of step S1.
Fig. 4 is a flowchart of step S2.
Fig. 5 is a flowchart of step S3.
Fig. 6 is a flowchart of step S4.
FIG. 7 is a flowchart of step S10-A in example 1.
Fig. 8 is a flowchart of step S106 in embodiment 1.
FIG. 9 is a partial flowchart of step S10-A in example 2.
FIG. 10 is a partial flowchart of step S10-A in example 2.
Detailed Description
The present invention will be further illustrated with reference to the accompanying drawings and specific embodiments, which are to be understood as merely illustrative of the invention and not as limiting the scope of the invention.
Example 1:
the flow of the method for estimating the traffic flow OD of the urban road network based on the automatic vehicle number plate identification data, which is provided by the embodiment, is shown in fig. 1, and comprises the following steps:
s1: setting a vehicle number plate detector at a key intersection of a selected road network area, monitoring and processing traffic basic information through the vehicle number plate detector, wherein the traffic basic information comprises vehicle number plate data, and constructing a complete travel track of vehicles in the selected road network area according to the vehicle number plate data and a topological structure of a road network to obtain a travel chain divided by vehicles in the road network area on a single day;
in step S1, the process of monitoring and processing the traffic basic information by the vehicle number plate detector is shown in fig. 2, and specifically includes the following steps:
s10: arranging a vehicle number plate detector on an entrance lane of a key intersection in a selected road network area, collecting data of each vehicle number plate in the selected road network area, and acquiring an intersection adjacent relation in the road network area to form a selected road network topological structure;
s11: preprocessing acquired vehicle license plate data, screening unrecognizable vehicle license plate data and repeated vehicle license plate data, removing the unrecognizable vehicle license plate data, respectively comparing the acquisition date, the vehicle license plate number, the bayonet number, the equipment number and the lane number of the repeated vehicle license plate data, detecting the acquisition time of the repeated vehicle license plate data when all characteristic values of the data are the same, and keeping the first detected data and removing other repeated data when the acquired time interval is smaller than a set threshold value;
s12: and performing association matching on the preprocessed vehicle number plate data and the intersections of the selected road network area, extracting the time information of a plurality of number plate data of each vehicle, sequencing according to the time sequence, and acquiring the travel track of each vehicle on the current day.
Specifically referring to fig. 3, in step S1, the process of performing topology construction on the vehicle travel track in the selected road network area according to the vehicle license plate data specifically includes the following steps:
S10-A: dividing the single-day trip chains of all vehicles according to the time and space information in the road network area, screening the trip chains divided by all vehicles in the road network area in a single day manner, judging whether the trip chain track divided by each vehicle is continuous or not, collecting the trip chain with track information loss, carrying out path analysis on the path with track information loss, complementing the missing track information and perfecting the trip chain divided by vehicles in the road network area in a single day manner.
The specific flow of step S10-a is shown in fig. 3, and includes the following steps:
s100: counting the number plate data of each vehicle in the selected road network area, and grouping all the vehicle track data to be processed in the selected road network area according to the number plate ID of each vehicle;
s101: obtaining track data of each grouped vehicle in the selected road network area, and sequencing the track data of each group of vehicles according to the time sequence;
s102: preprocessing each group of vehicle track data, and removing repeated records in the vehicle track data grouping and the grouping with only a single track point;
s103: when the track points in the group of vehicle track data are complete, dividing a trip chain of each group of preprocessed vehicle track data according to a timestamp difference value between two adjacent vehicle track point data in the group to obtain a complete track data set of each group of vehicles, and when part of track points in the group of vehicle track data are missing, marking the track data as a track to be reconstructed, and summarizing the track data into a track set to be reconstructed;
s104: aiming at the complete vehicle license plate data track set, splitting each vehicle license plate data track, taking head and tail track points of a given time period as an input sequence, taking the rest middle track points as an output sequence, and constructing a Seq2Seq training data set;
s105: constructing a track sequence prediction model by utilizing the constructed training data set and a recurrent neural network GRU algorithm;
s106: and aiming at the track set to be reconstructed, splitting adjacent track point pairs of each track, and reconstructing the track of each track point pair by combining a track sequence prediction model.
Referring specifically to fig. 8, step S106 includes the steps of:
s106-1: extracting a group of adjacent track point pairs, and judging whether the track point pairs are adjacent in space;
s106-2: when the set of track points are adjacent in space, reconstruction processing is not needed;
s106-3: when the set of track points are not adjacent in space, judging whether the set of track points have a history mode;
s106-4: when the set of track points has a historical mode, predicting and completing the track-lacking points by using the constructed track sequence prediction model;
s106-5: when the set of track points does not have a historical mode, completing the missing track points by using a shortest path method based on travel time;
s106-6: counting the supplemented missing track points, and screening whether all the track points of the vehicles in the research area are processed;
s106-7: if all track points of the vehicle are not completed, repeating the steps S106-1 to S106-6, and repeatedly extracting a group of adjacent track points to complete the track.
S2: collecting travel chain track information of the vehicle, acquiring the time-space information of the vehicle in the selected road network area, counting the quantity of track information with the same path in the road network area, superposing the quantity of the same travel chain track, and analyzing the flow estimation result of each path in the road network area;
referring specifically to fig. 4, step S2 includes the following steps:
s21: counting the trip chains with the complement missing track information to obtain the real trip chain of each trip of each vehicle in the selected road network area;
s22: dividing travel chains of all vehicles in a selected road network area by taking hours as units, and screening all tracks of which the track end time of the vehicle travel chains is positioned in a certain period;
s23: counting the number of vehicles with the same track, setting the number of the intersection at the starting point as an intersection 1 and the number of the intersection at the finishing point as an intersection 2 by taking the track of the intersection at a certain starting point as an object, screening the number of the vehicles with the same track from the intersection 1 to the intersection 2, and overlapping, wherein the quantity of screening results is the path flow from the intersection 1 to the intersection 2 in the time period;
s24: and step S23 is repeated, and every two intersections in the selected road network area are processed until the OD flow estimation of all paths in the selected road network area is completed.
S3: counting the estimation results of all path flows in the road network area, carrying out sample expansion processing on the calculated path flows, determining a specific sample expansion coefficient, and analyzing the total flow of each path in the road network;
referring specifically to fig. 5, step S3 includes the steps of:
s31: analyzing the vehicle number plate detection permeability of each intersection in the selected road network area, setting the vehicle number plate detection permeability of a certain intersection as gamma, and setting the number of the effective vehicle number plate records collected by the intersection as NiSetting the number of all vehicle number plate records collected at the intersection as
Figure BDA0003336639600000101
According to the formula:
Figure BDA0003336639600000102
calculating the vehicle number plate detection permeability of the current intersection, and calculating the vehicle number plate detection permeability of different intersections in the selected road network area according to the formula;
s32: analyzing the sampling rate of the vehicle license plates at different moments in the selected road network area, acquiring the number of the vehicle license plates collected by the vehicle license plate detector at a certain moment in the road network area, extracting the total number of the vehicles passing through the road network area at the moment, wherein the sampling rate at different moments is the ratio of the number of the vehicle license plates collected by the vehicle license plate detector in the road network area at different moments to the total number of the vehicles in the road network area, and determining the scaling factor of all the paths in the selected road network at the corresponding moment according to the sampling rate;
s33: determining a sample expansion coefficient of a selected road network area according to the detection permeability and the sampling rate of the vehicle license plates at different intersections, wherein the sample expansion coefficient is the product of the permeability and the sampling rate of the corresponding intersection;
s34: the method comprises the steps of obtaining vehicle license plate detection permeability and scaling coefficients of all paths corresponding to different intersections and different time periods in a road network area, setting total flow of each path to be Q, setting OD flow acquired by each path to be Q', setting the scaling coefficient of a certain path corresponding to a time period to be lambda, and according to a formula:
Q＝Q′/(γ*λ)
and calculating to obtain the total flow of each path in the currently selected road network area.
S4: collecting the traffic flow OD pairs with the same starting point and the same end point, aiming at each traffic flow OD pair, obtaining a path set contained in each traffic flow OD pair, analyzing the OD flow of all the traffic flow OD pairs with the same starting point and the same end point according to each path in the path set, and analyzing the average OD flow in a road network area according to the OD flow of all the traffic flow OD pairs.
Referring specifically to fig. 6, step S4 includes the following steps:
s41: aiming at a certain traffic flow OD pair, each path in a path set contained in the traffic flow OD pair is obtained, and the path flow of each path is obtained;
s42: stacking the path flow of all paths in the path set to obtain the OD flow of the OD pair of the traffic flow;
s43: analyzing the path flow inside each traffic flow OD pair in the road network area, and counting the OD flow of each traffic flow OD pair;
s44: analyzing OD flow of the same OD pair of the traffic flow in all fixed time intervals in one day, and calculating average OD flow, wherein the average OD flow is the OD flow of the same OD pair in all fixed time intervals in one day divided by the number of the fixed time intervals in one day;
s45: and repeating the steps S41-S44, and analyzing the average OD flow of all the traffic flow OD pairs in the road network area.
Example 2:
this example provides another implementation manner of step S10-a, and the flow of the remaining steps is the same as that in embodiment 1:
preferably, in step S10-a, the process of dividing the single-day travel chains of all vehicles according to the time and space information in the road network area further includes the following steps as shown in fig. 9:
s10-a: counting the number plate data of the vehicles in the selected road network area, extracting all number plate data of a certain vehicle in one day, and counting the number plate data as N;
s10-b: arranging N license plate data according to the time sequence according to the acquisition time of each license plate data of the vehicle, and enabling N to be 1;
s10-c: in the arranged vehicle number plate data, the acquisition time of the nth data and the acquisition time of the (n + 1) th data are differed to obtain the actual running time of the vehicle between adjacent number plate detectors, and the actual running time is set as Ti；
S10-d: calculating the shortest path between the two vehicle license plate detectors corresponding to the two vehicle license plate data, and setting the length of the shortest path to be LminSetting the road flow velocity between two vehicle number plate detectors to VFSetting the shortest travel time between two vehicle number plate detectors to TminAccording to the formula: t ismin＝Lmin/VFSetting the driving budget time between two vehicle number plate detectors as
Figure BDA0003336639600000111
Wherein, the value of delta is a constant larger than 1;
s10-e: comparing the actual running time between two vehicle number plate detectors with the shortest running time and the estimated time between two vehicle number plate detectors respectively, and when T isi<TminIf the data is wrong, the data of the vehicle number plate detected by the corresponding vehicle number plate detector is deleted, and if the data is wrong, the data of the vehicle number plate is deleted
Figure BDA0003336639600000112
Explaining the process of the vehicle staying at the corresponding position of one vehicle detector, the trip chain should be disconnected, and the nth data and the (n + 1) th data should belong to the end point and the starting point of two different trip chains;
s10-f: and when N is less than N, making N equal to N +1, and repeating the steps S100-3 to S100-6 until N is equal to N, thereby completing the division of the trip chains of different vehicles in the road network area.
Referring specifically to fig. 10, step S10-a further includes the following specific steps:
s10-1: screening a trip chain with missing track information in a routing network area, and extracting an alternative path between two vehicle number plate detectors corresponding to the trip chain;
s10-2: when the distance between two vehicle number plate detectors corresponding to the trip chain is close, the alternative path distances between the two vehicle number plate detectors are sequenced, and the length of the ith alternative path between some two vehicle number plate detectors is set to be LiSetting the probability that the ith alternative path between two vehicle number plate detectors is selected as Pi1，Pi1＝1/Li；
S10-3: when the distance between two vehicle number plate detectors corresponding to a trip chain is long, the lengths of high-grade roads in alternative paths between the two vehicle number plate detectors are sequenced, wherein the high-grade road is a main road, the total number of the alternative paths between the two vehicle number plate detectors is set to be n, and the length of the main road in the ith alternative path between the two vehicle number plate detectors is set to be n
Figure BDA0003336639600000121
Setting the probability that the ith alternative path between two vehicle license plate detectors is selected as Pi2According to the formula:
Figure BDA0003336639600000122
calculating the probability of selecting the ith alternative path between two current vehicle license plate detectors;
s10-4: detecting whether each alternative path is contained by the number plate data detected by the vehicle number plate detector in the original trip chain, and detecting the vehicle number plate contained by the original trip chain in a certain alternative path iThe number of the vehicle number plate detectors is compared with the total number of the vehicle number plate detectors arranged on the original trip chain, and the number of the vehicle number plate detectors which are simultaneously positioned on the alternative path i and the original trip chain is set to be MiThe number of the vehicle number plate detectors installed on the original trip chain is NoriginSetting the probability that the ith alternative path between two vehicle number plate detectors is selected as Pi3，Pi3＝Mi/Norigin；
S10-5: according to the steps, all alternative path selection probabilities between two vehicle license plate detectors corresponding to the trip chain with track information missing in the road network area are compared, one path with the highest probability is extracted to serve as a track reconstruction path, and the trip chain divided by vehicles in the road network area on a single day is perfected.
It should be specifically stated that the selection probabilities of the alternative paths are simultaneously constrained in steps S10-2, S10-3, and S10-4.
It should be noted that the above-mentioned contents only illustrate the technical idea of the present invention, and the protection scope of the present invention is not limited thereby, and it is obvious to those skilled in the art that several modifications and decorations can be made without departing from the principle of the present invention, and these modifications and decorations fall within the protection scope of the claims of the present invention.

Claims (8)

1. The urban road network traffic flow OD estimation method based on automatic vehicle number plate identification data is characterized by comprising the following steps of:
s1: setting a vehicle number plate detector at a key intersection of a selected road network area, monitoring and processing traffic basic information through the vehicle number plate detector, wherein the traffic basic information comprises vehicle number plate data, and constructing a complete travel track of vehicles in the selected road network area according to the vehicle number plate data and a topological structure of a road network to obtain a travel chain divided by vehicles in the road network area on a single day;
the process of monitoring and processing the traffic basic information through the vehicle number plate detector comprises the following substeps:
s10: arranging a vehicle number plate detector on an entrance lane of a key intersection in a selected road network area, collecting data of each vehicle number plate in the selected road network area, and acquiring an intersection adjacent relation in the road network area to form a selected road network topological structure;
s11: preprocessing acquired vehicle license plate data, screening unrecognizable vehicle license plate data and repeated vehicle license plate data, removing the unrecognizable vehicle license plate data, respectively comparing the acquisition date, the vehicle license plate number, the bayonet number, the equipment number and the lane number of the repeated vehicle license plate data, detecting the acquisition time of the repeated vehicle license plate data when all characteristic values of the data are the same, and keeping the first detected data and removing other repeated data when the acquired time interval is smaller than a set threshold value;
s12: the preprocessed vehicle number plate data is associated and matched with intersections of the selected road network area, time information of a plurality of number plate data of each vehicle is extracted, the number plate data are sequenced according to the time sequence, and the travel track of each vehicle on the day is obtained;
the process of topologically constructing the vehicle travel track in the selected road network area according to the vehicle number plate data to obtain the travel chain divided by the vehicles in the road network area in one day comprises the following steps:
S10-A: dividing the single-day trip chains of all vehicles according to time and space information in a road network area, screening the trip chains divided by all vehicles in the road network area in a single day manner, judging whether the trip chain track divided by each vehicle is continuous or not, collecting the trip chain with track information loss, carrying out path analysis on the path with the track information loss, complementing the missing track information, and perfecting the trip chain divided by the vehicles in the road network area in a single day manner;
s2: collecting travel chain track information of the vehicle, acquiring space-time track information of the vehicle in a selected road network area, counting the quantity of track information with the same path in the road network area, superposing the quantity of the same travel chain track, and analyzing a flow estimation result of each path in the road network area;
s3: counting the estimation results of all path flows in the road network area, carrying out sample expansion processing on the calculated path flows, determining a specific sample expansion coefficient, and analyzing the total flow of each path in the road network;
s4: collecting the traffic flow OD pairs with the same starting point and the same end point, aiming at each traffic flow OD pair, obtaining a path set contained in each traffic flow OD pair, analyzing the OD flow of all the traffic flow OD pairs with the same starting point and the same end point according to each path in the path set, and analyzing the average OD flow in a road network area according to the OD flow of all the traffic flow OD pairs.
2. The method for estimating the OD of urban road network traffic flow based on the automatic vehicle number plate identification data of claim 1, wherein the step S10-a specifically comprises the following steps:
s100: counting the number plate data of each vehicle in the selected road network area, and grouping all the vehicle track data to be processed in the selected road network area according to the number plate ID of each vehicle;
s101: obtaining track data of each grouped vehicle in the selected road network area, and sequencing the track data of each group of vehicles according to the time sequence;
s102: preprocessing each group of vehicle track data, and removing repeated records in the vehicle track data grouping and the grouping with only a single track point;
s103: when the track points in the group of vehicle track data are complete, dividing a trip chain of each group of preprocessed vehicle track data according to a timestamp difference value between two adjacent vehicle track point data in the group to obtain a complete track data set of each group of vehicles, and when part of track points in the group of vehicle track data are missing, marking the track to be reconstructed and summarizing the track to be reconstructed into a track set to be reconstructed;
s104: aiming at the complete vehicle license plate data track set, splitting each vehicle license plate data track, taking head and tail track points of a given time period as an input sequence, taking the rest middle track points as an output sequence, and constructing a Seq2Seq training data set;
s105: constructing a track sequence prediction model by utilizing the constructed training data set and a recurrent neural network GRU algorithm;
s106: and aiming at the track set to be reconstructed, splitting adjacent track point pairs of each track, and reconstructing the track of each track point pair by combining a track sequence prediction model.
3. The method for estimating the OD of a traffic flow in a city road network based on the automatic vehicle number plate identification data as claimed in claim 2, wherein said step 106 comprises the following steps:
s106-1: extracting a group of adjacent track point pairs, and judging whether the track point pairs are adjacent in space;
s106-2: when the set of track points are adjacent in space, reconstruction processing is not needed;
s106-3: when the set of track points are not adjacent in space, judging whether the set of track points have a history mode;
s106-4: when the set of track points has a historical mode, predicting and completing the track-lacking points by using the constructed track sequence prediction model;
s106-5: when the set of track points does not have a historical mode, completing the missing track points by using a shortest path method based on travel time;
s106-6: counting the supplemented missing track points, and screening whether all the track points of the vehicles in the research area are processed;
s106-7: if all track points of the vehicle are not completed, repeating the steps S106-1 to S106-6, and repeatedly extracting a group of adjacent track points to complete the track.
4. The method for estimating the OD of urban road network traffic flow based on the automatic vehicle number plate identification data of claim 1, wherein in step S10-a, the process of dividing the single day trip chain of all vehicles according to the time and space information in the road network area comprises the following steps:
s10-a: counting the number plate data of the vehicles in the selected road network area, extracting all number plate data of a certain vehicle in one day, and counting the number plate data as N;
s10-b: arranging N license plate data according to the time sequence according to the acquisition time of each license plate data of the vehicle, and enabling N to be 1;
s10-c: in the arranged vehicle number plate data, the acquisition time of the nth data and the acquisition time of the (n + 1) th data are differed to obtain the actual running time of the vehicle between adjacent number plate detectors, and the actual running time is set as Ti(ii) a S10-d: calculating the shortest path between the two vehicle license plate detectors corresponding to the two vehicle license plate data, and setting the length of the shortest path to be LminSetting the road flow velocity between two vehicle number plate detectors to VFSetting the shortest travel time between two vehicle number plate detectors to TminAccording to the formula: t ismin＝Lmin/VFSetting the driving budget time between two vehicle number plate detectors as
Figure FDA0003336639590000031
Figure FDA0003336639590000032
Wherein, the value of delta is a constant larger than 1;
s10-e: comparing the actual running time between two vehicle number plate detectors with the shortest running time and the estimated time between two vehicle number plate detectors respectively, and when T isi<TminIf the data is wrong, the data of the vehicle number plate detected by the corresponding vehicle number plate detector is deleted, and if the data is wrong, the data of the vehicle number plate is deleted
Figure FDA0003336639590000033
Explaining the process of the vehicle staying at the corresponding position of one vehicle detector, the trip chain should be disconnected, and the nth data and the (n + 1) th data should belong to the end point and the starting point of two different trip chains;
s10-f: and when N is less than N, making N equal to N +1, and repeating the steps S100-3 to S100-6 until N is equal to N, thereby completing the division of the trip chains of different vehicles in the road network area.
5. The method for estimating the OD of urban road network traffic flow based on automatic vehicle number plate identification data according to claim 1, wherein in step S10-a, screening the travel chains divided by all vehicles in the road network area on a single day, determining whether the travel chain track of each vehicle is continuous, collecting the travel chain with missing track information, performing path analysis on the path with missing track information, complementing the missing track information, and completing the travel chain divided by vehicles in the road network area on a single day, comprises the following steps:
s10-1: screening a trip chain with missing track information in a routing network area, and extracting an alternative path between two vehicle number plate detectors corresponding to the trip chain;
s10-2: when the distance between two vehicle number plate detectors corresponding to the trip chain is close, the alternative path distances between the two vehicle number plate detectors are sequenced, and the length of the ith alternative path between some two vehicle number plate detectors is set to be LiSetting the probability that the ith alternative path between two vehicle number plate detectors is selected as Pi1，Pi1＝1/Li；
S10-3: when the distance between two vehicle number plate detectors corresponding to a trip chain is long, the lengths of high-grade roads in alternative paths between the two vehicle number plate detectors are sequenced, wherein the high-grade road is a main road, the total number of the alternative paths between the two vehicle number plate detectors is set to be n, and the length of the main road in the ith alternative path between the two vehicle number plate detectors is set to be n
Figure FDA0003336639590000042
Setting the probability that the ith alternative path between two vehicle license plate detectors is selected as Pi2According to the formula:
Figure FDA0003336639590000041
calculating the probability of selecting the ith alternative path between two current vehicle license plate detectors;
s10-4: detect eachWhether a standby route is contained by number plate data detected by a vehicle number plate detector in an original trip chain or not is judged, the number of the vehicle number plate detectors contained in a certain standby route i by the original trip chain is compared with the total number of the vehicle number plate detectors installed on the original trip chain, and the number of the vehicle number plate detectors on the standby route i and the number of the vehicle number plate detectors on the original trip chain are set to be MiThe number of the vehicle number plate detectors installed on the original trip chain is NoriginSetting the probability that the ith alternative path between two vehicle number plate detectors is selected as Pi3，Pi3＝Mi/Norigin；
S10-5: according to the steps, all alternative path selection probabilities between two vehicle license plate detectors corresponding to the trip chain with track information missing in the road network area are compared, one path with the highest probability is extracted to serve as a track reconstruction path, and the trip chain divided by vehicles in the road network area on a single day is perfected.
6. The method for estimating the OD of urban road network traffic flow based on the automatic vehicle number plate identification data of claim 1, wherein said step S2 comprises the following substeps:
s21: counting the trip chains with the complement missing track information to obtain the real trip chain of each trip of each vehicle in the selected road network area;
s22: dividing travel chains of all vehicles in a selected road network area by taking hours as units, and screening all tracks of which the track end time of the vehicle travel chains is positioned in a certain period;
s23: counting the number of vehicles with the same track, setting the number of the intersection at the starting point as an intersection 1 and the number of the intersection at the finishing point as an intersection 2 by taking the track of the intersection at a certain starting point as an object, screening the number of the vehicles with the same track from the intersection 1 to the intersection 2, and overlapping, wherein the quantity of screening results is the path flow from the intersection 1 to the intersection 2 in the time period;
s24: and step S23 is repeated, and every two intersections in the selected road network area are processed until the OD flow estimation of all paths in the selected road network area is completed.
7. The method for estimating the OD of urban road network traffic flow based on the automatic vehicle number plate identification data of claim 1, wherein said step S3 comprises the following substeps:
s31: analyzing the vehicle number plate detection permeability of each intersection in the selected road network area, setting the vehicle number plate detection permeability of a certain intersection as gamma, and setting the number of the effective vehicle number plate records collected by the intersection as NiSetting the number of all vehicle number plate records collected at the intersection as
Figure FDA0003336639590000051
According to the formula:
Figure FDA0003336639590000052
calculating the vehicle number plate detection permeability of the current intersection, and calculating the vehicle number plate detection permeability of different intersections in the selected road network area according to the formula;
s32: analyzing the sampling rate of the vehicle license plates at different moments in the selected road network area, acquiring the number of the vehicle license plates collected by the vehicle license plate detector at a certain moment in the road network area, extracting the total number of the vehicles passing through the road network area at the moment, wherein the sampling rate at different moments is the ratio of the number of the vehicle license plates collected by the vehicle license plate detector in the road network area at different moments to the total number of the vehicles in the road network area, and determining the scaling factor of all the paths in the selected road network at the corresponding moment according to the sampling rate;
s33: determining a sample expansion coefficient of a selected road network area according to the detection permeability and the sampling rate of the vehicle license plates at different intersections, wherein the sample expansion coefficient is the product of the permeability and the sampling rate of the corresponding intersection;
s34: the method comprises the steps of obtaining vehicle license plate detection permeability and scaling coefficients of all paths corresponding to different intersections and different time periods in a road network area, setting total flow of each path to be Q, setting OD flow acquired by each path to be Q', setting the scaling coefficient of a certain path corresponding to a time period to be lambda, and according to a formula:
Q＝Q'/(γ*λ)
and calculating to obtain the total flow of each path in the currently selected road network area.
8. The method for estimating the OD of urban road network traffic flow based on the automatic vehicle number plate identification data of claim 1, wherein said step S4 comprises the following substeps:
s41: aiming at a certain traffic flow OD pair, each path in a path set contained in the traffic flow OD pair is obtained, and the path flow of each path is obtained;
s42: stacking the path flow of all paths in the path set to obtain the OD flow of the OD pair of the traffic flow;
s43: analyzing the path flow inside each traffic flow OD pair in the road network area, and counting the OD flow of each traffic flow OD pair;
s44: analyzing OD flow of the same OD pair of the traffic flow in all fixed time intervals in one day, and calculating average OD flow, wherein the average OD flow is the OD flow of the same OD pair in all fixed time intervals in one day divided by the number of the fixed time intervals in one day;
s45: and repeating the steps S41-S44, and analyzing the average OD flow of all the traffic flow OD pairs in the road network area.
CN202111296123.0A 2021-11-03 2021-11-03 Urban road network traffic flow OD estimation method based on automatic vehicle number plate identification data Active CN114049765B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111296123.0A CN114049765B (en) 2021-11-03 2021-11-03 Urban road network traffic flow OD estimation method based on automatic vehicle number plate identification data

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111296123.0A CN114049765B (en) 2021-11-03 2021-11-03 Urban road network traffic flow OD estimation method based on automatic vehicle number plate identification data

Publications (2)

Publication Number Publication Date
CN114049765A true CN114049765A (en) 2022-02-15
CN114049765B CN114049765B (en) 2022-07-19

Family

ID=80207167

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111296123.0A Active CN114049765B (en) 2021-11-03 2021-11-03 Urban road network traffic flow OD estimation method based on automatic vehicle number plate identification data

Country Status (1)

Country Link
CN (1) CN114049765B (en)

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113656979A (en) * 2021-08-26 2021-11-16 浙江数智交院科技股份有限公司 Road network data generation method and device, electronic equipment and storage medium
CN114937366A (en) * 2022-07-22 2022-08-23 深圳市城市交通规划设计研究中心股份有限公司 Traffic flow calculation method based on multi-scale traffic demand and supply conversion
CN114999162A (en) * 2022-08-02 2022-09-02 北京交研智慧科技有限公司 Road traffic flow obtaining method and device
CN115331433A (en) * 2022-07-28 2022-11-11 东南大学 Urban main road multi-vehicle track reconstruction method based on multi-source data
CN116257797A (en) * 2022-12-08 2023-06-13 江苏中路交通发展有限公司 Single trip track identification method of motor vehicle based on Gaussian mixture model
CN116307318A (en) * 2023-03-06 2023-06-23 深圳市城市交通规划设计研究中心股份有限公司 Road traffic tracing system and method based on bayonet license plate identification data
CN116665439A (en) * 2023-04-28 2023-08-29 东南大学 Urban road network vehicle flow OD uncertainty estimation method based on license plate identification data
CN117711174A (en) * 2023-12-07 2024-03-15 山东高速集团有限公司 Data processing method and system for vehicle passing information

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20200042799A1 (en) * 2018-07-31 2020-02-06 Didi Research America, Llc System and method for point-to-point traffic prediction
CN111243277A (en) * 2020-03-09 2020-06-05 山东大学 Commuting vehicle space-time trajectory reconstruction method and system based on license plate recognition data
CN112365711A (en) * 2020-10-21 2021-02-12 东南大学 Vehicle track reconstruction method based on license plate recognition data
CN113593220A (en) * 2021-07-02 2021-11-02 南京泛析交通科技有限公司 Road network bearing capacity estimation method based on macroscopic basic graph

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20200042799A1 (en) * 2018-07-31 2020-02-06 Didi Research America, Llc System and method for point-to-point traffic prediction
CN111243277A (en) * 2020-03-09 2020-06-05 山东大学 Commuting vehicle space-time trajectory reconstruction method and system based on license plate recognition data
CN112365711A (en) * 2020-10-21 2021-02-12 东南大学 Vehicle track reconstruction method based on license plate recognition data
CN113593220A (en) * 2021-07-02 2021-11-02 南京泛析交通科技有限公司 Road network bearing capacity estimation method based on macroscopic basic graph

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
马金麟 等: "基于车牌识别数据的车辆OD矩阵获取研究", 《重庆理工大学学报（自然科学）》 *

Cited By (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113656979A (en) * 2021-08-26 2021-11-16 浙江数智交院科技股份有限公司 Road network data generation method and device, electronic equipment and storage medium
CN113656979B (en) * 2021-08-26 2024-05-14 浙江数智交院科技股份有限公司 Road network data generation method and device, electronic equipment and storage medium
CN114937366A (en) * 2022-07-22 2022-08-23 深圳市城市交通规划设计研究中心股份有限公司 Traffic flow calculation method based on multi-scale traffic demand and supply conversion
CN114937366B (en) * 2022-07-22 2022-11-25 深圳市城市交通规划设计研究中心股份有限公司 Traffic flow calculation method based on multi-scale traffic demand and supply conversion
CN115331433B (en) * 2022-07-28 2023-09-29 东南大学 Urban arterial road multi-vehicle track reconstruction method based on multi-source data
CN115331433A (en) * 2022-07-28 2022-11-11 东南大学 Urban main road multi-vehicle track reconstruction method based on multi-source data
CN114999162A (en) * 2022-08-02 2022-09-02 北京交研智慧科技有限公司 Road traffic flow obtaining method and device
CN116257797A (en) * 2022-12-08 2023-06-13 江苏中路交通发展有限公司 Single trip track identification method of motor vehicle based on Gaussian mixture model
CN116307318A (en) * 2023-03-06 2023-06-23 深圳市城市交通规划设计研究中心股份有限公司 Road traffic tracing system and method based on bayonet license plate identification data
CN116307318B (en) * 2023-03-06 2023-10-20 深圳市城市交通规划设计研究中心股份有限公司 Road traffic tracing system and method based on bayonet license plate identification data
CN116665439B (en) * 2023-04-28 2024-05-07 东南大学 Urban road network vehicle flow OD uncertainty estimation method based on license plate identification data
CN116665439A (en) * 2023-04-28 2023-08-29 东南大学 Urban road network vehicle flow OD uncertainty estimation method based on license plate identification data
CN117711174A (en) * 2023-12-07 2024-03-15 山东高速集团有限公司 Data processing method and system for vehicle passing information

Also Published As

Publication number Publication date
CN114049765B (en) 2022-07-19

Similar Documents

Publication Publication Date Title
CN114049765B (en) Urban road network traffic flow OD estimation method based on automatic vehicle number plate identification data
CN106997669B (en) Method for judging traffic jam cause based on feature importance
CN108629978B (en) Traffic track prediction method based on high-dimensional road network and recurrent neural network
CN108345666B (en) Vehicle abnormal track detection method based on time-space isolated points
CN103871246B (en) Based on the Short-time Traffic Flow Forecasting Methods of road network spatial relation constraint Lasso
CN107330217B (en) Mesoscopic oil consumption prediction method based on RBFNN
CN111210612B (en) Method for extracting bus route track based on bus GPS data and station information
CN109658695A (en) A kind of multifactor Short-time Traffic Flow Forecasting Methods
CN107563566B (en) Inter-bus-station operation time interval prediction method based on support vector machine
CN110176139A (en) A kind of congestion in road identification method for visualizing based on DBSCAN+
CN110299011A (en) A kind of traffic flow forecasting method of the highway arbitrary cross-section based on charge data
CN105702029A (en) Express way traffic state prediction method taking spatial-temporal correlation into account at different times
CN103984994B (en) Method for predicting urban rail transit passenger flow peak duration
CN110836675B (en) Decision tree-based automatic driving search decision method
CN102081846A (en) Expressway charge data track matching based traffic state recognition method
CN101783074A (en) Method and system for real-time distinguishing traffic flow state of urban road
CN105405293A (en) Short-term prediction method of road travel time and system
CN112489426A (en) Urban traffic flow space-time prediction scheme based on graph convolution neural network
CN110738855A (en) Road traffic flow condition prediction method for data sparse time periods
CN108648445A (en) Dynamic traffic Tendency Prediction method based on traffic big data
CN102890862A (en) Traffic condition analyzing device and method based on vector mode
CN112613225A (en) Intersection traffic state prediction method based on neural network cell transmission model
CN112884014A (en) Traffic speed short-time prediction method based on road section topological structure classification
CN107945510A (en) A kind of section detecting method for considering transport need and road network operational efficiency
CN114882069A (en) Taxi track abnormity detection method based on LSTM network and attention mechanism

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