CN111047860A - Vehicle running track extraction method - Google Patents

Abstract

The vehicle running track extraction method can solve the technical problem that the running path of the driving vehicle in the urban area is difficult to determine. The method comprises the following steps: acquiring vehicle passing data of all electronic policemen and checkpoints on a certain day; determining that adjacent detection point positions of a certain vehicle can be directly connected to form an effective path; extracting all possible effective paths between the connected detection points under the condition that the adjacent detection points cannot be directly connected; determining the consumed time of different paths based on the extracted effective path and the operation data information of different road sections of the internet at each time interval; comparing the time consumption of the possible path with the time difference value of the adjacent detection points to determine an optimal vehicle driving path set; and determining the optimal driving path by adopting a fuzzy comprehensive evaluation method. The method can provide key data support for road traffic distribution, traffic travel guidance and road network planning of the urban road traffic network at any time period, can reduce traffic cost, improves travel efficiency, and has wide application prospect.

Description

Vehicle running track extraction method

The invention relates to the technical field of traffic planning construction and vehicle path extraction, in particular to a vehicle running track extraction method.

Background

Traffic is a fundamental industry and a service industry for national economy and social development. With economic development, and progress of computers and information technology, intelligent transportation is under the background of big data, and the acquisition of urban traffic data and the extraction of effective traffic information are fundamental stones of the development of future intelligent transportation applications.

Under the requirement of a scientific and technological strong police, a large number of electronic police and bayonet devices are distributed in the road network of various cities in China, and the devices in the whole city can generate massive data information every day. Besides inquiring illegal vehicles and some law enforcement requirements, the data information is not applied as traffic data for improving traffic operation and traffic efficiency. This is mainly due to the lack of methods for deep mining of these data. Under the rapid development of an internet map, the acquired real-time road section operation data information and the massive traffic information of deep mining electronic policemen and bayonet equipment are a big data basis for realizing intelligent travel, safe travel, efficient travel and green travel, the vehicle path information acquired by data mining can be used for carrying out scientific traffic management and planning on the intelligent, safe, efficient and green travel demands of travelers, and in addition, based on the massive vehicle path information of urban areas, the traffic managers can deeply know the root cause of traffic jam and help the traffic managers to make more scientific and various management measures to relieve the traffic pressure of the urban areas.

The urban whole-time vehicle running track extraction is the basis of traffic path planning, is the key for determining the induced path and making scientific, efficient and reasonable management and control measures, and is directly related to the traffic distribution precision. At present, a great deal of research is carried out on methods for extracting vehicle running tracks based on electronic police, and a lot of theoretical and practical achievements are obtained. Wherein, the running path information of a single vehicle in the whole time period is extracted based on the electronic police vehicle passing data. However, the situations such as incomplete arrangement in urban areas of electronic police, loss and errors of detection data and the like are not considered, and a relevant research method for a possible driving path existing between adjacent detection points is lacked. Therefore, under various possible conditions in the actual application process, it is very important to accurately determine the information of the urban all-time vehicle driving path, and in addition, the traffic data generated by the vehicle running track extraction method can provide key data support for the formulation of urban road management and control measures, traffic travel guidance and road network planning. The method has very wide application prospect in order to reduce social cost and improve social travel efficiency.

Disclosure of Invention

The vehicle running track extraction method can solve the technical problem that the running path of the driving vehicle in the urban area is difficult to determine.

In order to achieve the purpose, the invention adopts the following technical scheme:

a vehicle running track extraction method comprises the following steps:

the vehicle running track extraction method is characterized in that vehicle passing data information captured by different point location devices is obtained through an electric alarm and a bayonet device, and running data information of vehicles on different road sections and time lengths on an internet platform is obtained, and the method comprises the following steps:

the method comprises the following steps:

s100, obtaining vehicle passing data of all electronic police and checkpoints on a certain day, and classifying and processing according to license plates.

S200, determining that adjacent detection points of a certain vehicle can be directly connected to form an effective path.

S300, extracting all possible effective paths between the connected detection points under the condition that the adjacent detection points of a certain vehicle cannot be directly connected.

S400, determining the consumed time of different paths based on the effective path extracted in S300 and the operation data information of different road sections of the Internet platform at each time period.

And S500, comparing the consumed time of different paths with the time difference value of adjacent detection points based on the S400, and determining an optimal vehicle driving path set.

And S600, analyzing influence factors of the traffic path based on the driving path set generated in the S500, and constructing a traffic path reliability comprehensive function to determine an optimal driving path.

Further, the step S100 is to acquire passing data of all electronic police officers and checkpoints in a certain day, and to classify and process the passing data according to license plates. The method specifically comprises the following steps:

s101, vehicle passing data of all electronic police and checkpoints in a certain day in an urban area are obtained, information such as detection equipment IDs, license plate numbers, detection time, road section numbers and lengths of inspection areas and downstream intersections of all the vehicle passing data are extracted, data strips are generated, and all the data strips are classified according to the license plate numbers.

And S102, generating a new data table according to the sequence of the detection time for each type of data, wherein a column of direct connection is additionally arranged in the new data table, the data filled in the column is only 0 and 1, and the last detection point is not required to be filled in. The hurdle data is determined by step S200.

Further, the step S200 determines that adjacent detection points of a certain vehicle can be directly connected to form an effective path; the method specifically comprises the following steps:

s201, acquiring all urban road network information databases, constructing a road network G representing original geographic information, and abstracting road network nodes into nodes p in the road network G_iAnd the road sections between adjacent nodes of the road network can be abstracted into the edge l in the road network G_ijG ═ { V, L, T }, where V (G) } { p₁,p₂,…,p_nThe node set is used as the node set, and n represents the number of nodes; l (p)_i,p_j)＝{l_ij,l_ji,d_ijDenotes the set of vectors for road segments between the nodes of the road network, where l_ijRepresenting a slave node v_iTo node v_jRoad section of d_ijRepresenting a slave node v_iTo node v_jA road segment length value of; t (l)_ij)＝{T_pj，t_s,t_l,t_r,t_uMeans road section l in road network_ijThe downstream crossing signal period and the vector matrix set of the steering information thereof, wherein T_pjFor intersection p_jSignal period of (d), port t_sIs 1_ijIs connected with the road section l of the straight going of the downstream crossing_jxOr null, that is to say, the straight going is prohibited; t is t_lIs 1_ijLeft-turn connected road section l of downstream crossing_jxOr null, i.e. left turn is prohibited; t is t_rIs 1_ijDownstream crossing right turn connected road section l_jxOr null, i.e., right turn is prohibited; t is t_uIs 1_ijSection of road l connected with downstream intersection in turn_jiOr null, i.e., no turning around is prohibited.

S202, sequentially extracting two adjacent passing data information according to the sequence in the classification data table, taking the first data and the second data as a group, taking the second data and the third data as a group, and so on.

S203, through all road network information databases in urban areas, the road section numbers in all the grouped first vehicle passing data can be used for knowing downstream junctions of the positions of electronic police and bayonets in a snapshot mode, and the clear junction information can be used for knowing a downstream road section number set through the steering information of the road network information database.

And S204, comparing the road section number of the second vehicle passing data in the group with the road section number in the road section number set at the downstream of the S203 one by one, wherein if consistent road section numbers exist, adjacent detection point positions can be directly connected to form an effective path, and a column of the data table is directly connected with 1. If not, the column of the data table "direct connection" is filled with 0, and the process proceeds to step S300.

Further, in the step S300, under the condition that adjacent detection points of a certain vehicle cannot be directly connected, all possible effective paths between the connected detection points are extracted; the method specifically comprises the following steps:

s301, determining and calculating two adjacent detection time t_iAnd t_i+1The difference of (a):

△t＝t_i+1-t_i,i＝1,2,…,n-1 (13)

s302, determining all possible driving paths between adjacent detection points by adopting bidirectional tracing, namely, taking the first vehicle passing detection point of the group as a starting point, and gradually extending downstream according to road network information to construct a path network with a tree structure; and taking the second grouped vehicle passing detection point as a terminal point, and gradually extending upstream according to the road network information to construct a path network of a tree structure. The number of road stiffness network layers for constructing the bidirectional tree structure is preset to be Q.

S303, through all road network information databases in urban areas, the road section number in all the grouped first vehicle passing data can be used for knowing a downstream intersection at the snapshot position of the electronic police, and the clear intersection information can be used for knowing a downstream road section number set D through the steering information of the road network information database₁. Determining the downstream road junction according to the downstream road segment number set, and obtaining the downstream road segment number set D according to the steering information of the road network information database₂(ii) a The calculation is continuously carried out to obtain the Q generation branch road part number information set D_q. And based on Q generation operation, sequentially generating information tables of different traffic paths by collecting road section numbers in different stages

S304, determining the upstream road junction of the road segment number of the second passing data in each group according to the steering information of all the road network information databases in the urban area, and knowing the upstream road segment number set U according to the specific road junction information and the steering information of the road network information databases₁. Determining the road junction of the upstream road section according to the road section number set of the upstream, and knowing the road section number set U of the upstream through the steering information of the road network information database₂(ii) a The calculation is continuously carried out to obtain the Q generation branch road part number information set U_q. And based on Q generation operation, sequentially generating information tables of different traffic paths by collecting road section numbers in different stages

S305, information table

The nodes in certain path information extract node information according to the node sequence from upstream crossing to downstream crossing, and the node information is stored in an information table

Which searches for traffic paths containing the same nodes. Extracting information tables

Extracting information table from upstream traffic path information having common node

The upstream traffic path information with common node is merged into a complete traffic path information table

Further, the step S400 determines time consumption values of different paths based on the effective paths extracted in steps S300 and S200 and the operation data information of different road sections of the interconnection network platform at each time interval; the method specifically comprises the following steps:

s401, sequentially acquiring all traffic paths in the step S305, and extracting length information of a plurality of road sections in the traffic paths; acquiring two adjacent detection time intervals t of each road section_i，t_i+1]And obtaining the consumed time of different paths in the first stage by the average running speed data of the internal Internet platform, namely:

t_s1time-consuming first-stage traffic path, unit: second;

d_isegment length of segment i, in units: rice;

-the length of the road section i detection area from the downstream intersection,unit: rice;

o-transit time in a crossing, different area values are different, and an average value needs to be obtained through investigation, wherein the unit is as follows: second;

v′_i-time interval [ t ]_i，t_i+1]Average running speed of inner road section i, unit: m/s.

S402, calculating to obtain that the time when the vehicle enters the road section i is (t)_i+△t′_i) The time of entering the section i +1 is (t)_i+△t′_i+1) Of which △ t'_iComprises the following steps:

△t′_i-the difference between the time of entering the road section i and the first detection time in the second stage, unit: second;

d_isegment length of segment i, in units: rice;

v′_i-average speed of travel of section i, in units: m/s;

o-transit time in a crossing, different area values are different, and an average value needs to be obtained through investigation, wherein the unit is as follows: second;

△t_itwo adjacent detection times t_iAnd t_i+1The difference of (a).

S403, acquiring the time interval t of the Internet platform_i+△t′_i，t_i+△t′_i+1]And obtaining the time consumption of different paths in the second stage by the average driving speed data on the road section i, namely:

t_s2time consumption of the second phase traffic path, unit: second;

d_isegment length of segment i, in units: rice;

-the length of the road section i detection area from the downstream intersection, unit: rice;

o-transit time in a crossing, different area values are different, and an average value needs to be obtained through investigation, wherein the unit is as follows: second;

v″_i-time interval [ t ]_i+△t'_i，t_i+△t'_i+1]Average running speed of inner road section i, unit: m/s;

s404, calculating to obtain that the time when the vehicle enters the road section i is (t)_i+△t″_i) The time of entering the section i +1 is (t)_i+△t″_i+1) Namely:

△t″_i-the third phase difference, unit, between the time of entering the section i and the first detection time: second;

d_isegment length of segment i, in units: rice;

v″_i-time interval [ t ]_i+△t″_i，t_i+△t″_i+1]Average travel speed of link i, unit: m/s;

o-transit time in a crossing, different area values are different, and an average value needs to be obtained through investigation, wherein the unit is as follows: second;

△t_itwo adjacent detection times t_iAnd t_i+1A difference of (d);

s405, obtaining the time interval t of the Internet platform_i+△t″_i，t_i+△t″_i+1]And obtaining the time consumption of different paths in the third stage by the average running speed data on the road section i, namely:

t_s3-time-consuming, unit of third-stage traffic route: second;

d_isegment length of segment i, in units: rice;

-the length of the road section i detection area from the downstream intersection, unit: rice;

o-transit time in a crossing, different area values are different, and an average value needs to be obtained through investigation, wherein the unit is as follows: second;

v″′_i-time interval [ t ]_i+△t″′_i，t_i+△t″′_i+1]Average running speed of inner road section i, unit: m/s;

s406, presetting the time consumption for calculating different paths and needing to pass through a K stage; according to the processing from step S404 to step S405, the time interval of the Internet platform is finally obtained

And (3) obtaining the time consumption of different paths in the K stage according to the average running speed data on the road section i, namely:

t_sKtime-consuming, unit of phase k traffic path: second;

d_isegment length of segment i, in units: rice;

-the length of the road section i detection area from the downstream intersection, unit: rice;

o-transit time in a crossing, different area values are different, and an average value needs to be obtained through investigation, wherein the unit is as follows: second;

-time interval

Average running speed of inner road section i, unit: m/s.

Further, in the step S500, based on S400, the consumed time of different paths is compared with the time difference between adjacent detection points, so as to determine an optimal vehicle driving path set. The method specifically comprises the following steps:

s501, obtaining the absolute value of the time difference value of the adjacent detection points subtracted by the time consumed by different paths in the K stage based on the operation data of the Internet platform for judgment, namely

|t_sK-△t_i|≤△t_i/10 (20)

t_sK-time cost value of phase k traffic path, in units: s;

△t_itwo adjacent detection times t_iAnd t_i+1The difference of (a).

S502, judging all possible vehicle running paths between adjacent detection points, storing the paths meeting the formula (20) in a vehicle running path set, and reducing the absolute values of the time difference values of the adjacent detection points from small to large according to the time consumed by different paths.

Further, in the step S600, based on the driving path set generated in the step S500, influence factors of the traffic path are analyzed, and a traffic path reliability comprehensive function is constructed to determine the optimal driving path. The method specifically comprises the following steps:

s601, difference value of time consumed by a path and detection time, number of turns required by the path and long and short node period are important factors influencing the path selection of a driver, and a traffic path reliability comprehensive function F (x) containing the three factors is constructed₁,x₂,x₃). The influencing factors include the difference value between the time consumed by the path and the detection time, the number of turns required by the path, and the length of the node cycle, and the set of factors U can be defined as: u ═ difference between the time consumed by the path and the detection time, the number of turns required by the path, and the node cycle length, expressed as: u ═ U₁,u₂,u₃}。

S602, factor weight; the fuzzy subset A of the factors in the factor set U is marked as fuzzy vector A ═ a₁,a₂,a₃) Wherein a is_iRepresents a single factor u_iMeasure of the magnitude of the contribution in the overall evaluation factor, 0 ≦ a_iLess than or equal to 1; different areas a_iThe values are different, and the fuzzy vector is obtained through questionnaire survey to obtain the average value A which is (0.54,0.29, 0.17).

And S603, normalizing the influencing factors. The fuzzy subset B of the factors on the factor set U is marked as a fuzzy vector B ═ B₁,b₂,b₃)^TWherein b is_jRepresents a single factor u_iThe normalized value.

(1) Normalization value of difference value between path time consumption and detection time

b₁-normalizing the difference between the time taken for the path and the detection time;

t_sk-time consumption for phase K traffic route, unit: s;

△t_itwo adjacent detection times t_iAnd t_i+1A difference of (d);

(2) normalized value of number of turns required for path

b₂-a normalized value of the number of turns required for the path;

g₁the left turn times and the unit of the different paths between the adjacent detection points need to be counted: secondly;

g₂-different paths between adjacent detection points need to count the right turn times in units: secondly;

g₃-different paths between adjacent detection points need to count the straight row total times, unit: secondly;

m₁-left turn offCoefficient, different area values are different, and the average value can be 0.5;

m₂the values of the right turn relation coefficients in different areas are different, and the average value can be 0.8;

m₃the straight-going relation coefficient has different values in different areas, and the average value can be 1.0;

(3) normalized value of node period length

Four intervals are divided according to the signal period length of the existing signal intersection, and the intersection period length of each interval is investigated to obtain the intersection signal period coefficient.

b₃-a normalization value of the node period length;

h₁the number of intersections with different paths passing through signal periods less than or equal to 80 seconds, unit: a plurality of;

h₂the number of intersections with the signal period of more than 80 seconds and less than or equal to 110 seconds for different paths is as follows: a plurality of;

h₃the number of intersections with the signal period of more than 110 seconds and less than or equal to 140 seconds for different paths is as follows: a plurality of;

h₄the number of intersections with different paths passing through a signal period greater than 140 seconds, unit: a plurality of;

w₁-the different paths pass through a coefficient with a signal period of less than or equal to 80 seconds;

w₂-the different paths pass through coefficients with a signal period greater than 80 seconds and less than or equal to 110 seconds;

w₃-the different paths pass through coefficients with a signal period greater than 110 seconds and less than or equal to 140 seconds;

w₄-the different paths pass through coefficients with a signal period greater than 140 seconds;

obtaining factor fuzzy subset B ═ (B) of different vehicle driving paths₁,b₂,b₃)^T；

S604, constructing a comprehensive function F (x)₁,x₂,x₃)。F(x₁,x₂,x₃) The larger the function value is, the higher the reliability of the extracted path is, namely the optimal path is;

F(x₁,x₂,x₃)＝AB＝(a₁,a₂,a₃)(b₁,b₂,b₃)^T(24)

and (3) calculating by using an equation (24) to obtain a traffic route with the highest reliability comprehensive function value of different vehicle running routes, wherein the traffic route is the vehicle running route in the adjacent detection time.

The method can provide key data support for road traffic distribution, traffic travel guidance and road network planning of the urban road traffic network at any time period. The method has very wide application prospect in order to reduce social cost and improve social travel efficiency.

Drawings

FIG. 1 is a schematic flow diagram of the process of the present invention;

fig. 2 is a schematic diagram of a network of urban vehicle travel paths.

Detailed Description

In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention.

The vehicle running track extraction method disclosed by the embodiment of the invention is used for displaying the working flow of the method by acquiring vehicle passing data, road network information and vehicle running data of an interconnection platform, which are acquired by an urban electronic police and a gate device, as shown in the attached drawings 1 and 2:

s100, obtaining vehicle passing data of all electronic police and checkpoints on a certain day, and classifying and processing according to license plates.

S200, determining that adjacent detection points of a certain vehicle can be directly connected to form an effective path.

S300, extracting all possible effective paths between the connected detection points under the condition that the adjacent detection points of a certain vehicle cannot be directly connected.

S400, determining the consumed time of different paths based on the effective path extracted in S300 and the operation data information of different road sections of the Internet platform at each time period.

And S500, comparing the consumed time of different paths with the time difference value of adjacent detection points based on the S400, and determining an optimal vehicle driving path set.

And S600, analyzing influence factors of the traffic path based on the driving path set generated in the S500, and constructing a traffic path reliability comprehensive function to determine an optimal driving path.

The above steps can be further explained as:

s100, obtaining vehicle passing data of all electronic police and checkpoints at a certain day, and classifying and processing according to license plates. The method specifically comprises the following steps: the method comprises the steps of obtaining vehicle passing data of all electronic policemen and checkpoints in a certain day in an urban area, extracting information such as detection equipment IDs, license plate numbers, detection time, road section numbers and lengths of inspection areas and downstream intersections of the vehicle passing data, generating data strips, and classifying all the data strips according to the license plate numbers. And each kind of data generates a new data table according to the sequence of the detection time, a column of direct connection is additionally arranged on the new data table, the data filled in the column are only 0 and 1, and the last detection point is not required to be filled. See Table 1-1.

TABLE 1-1 passing data information Table

S200, determining that adjacent detection points of a certain vehicle can be directly connected to form an effective path. Specifically, the method comprises the following steps:

s201, acquiring all urban road network information databases, constructing a road network G representing original geographic information, and abstracting road network nodes into nodes p in the road network G_iAnd the road sections between adjacent nodes of the road network can be abstracted into the edge l in the road network G_ijG ═ { V, L, T }, where V (G) } { p₁,p₂,…,p_nThe node set is used as the node set, and n represents the number of nodes; l (p)_i,p_j)＝{l_ij,l_ji,d_ijDenotes the set of vectors for road segments between the nodes of the road network, where l_ijRepresenting a slave node v_iTo node v_jRoad section of d_ijRepresenting a slave node v_iTo node v_jA road segment length value of;

indicating a road segment l in a road network_ijThe downstream crossing signal cycle and the vector matrix set of the steering information thereof, wherein

For intersection p_jSignal period of (d), port t_sIs 1_ijIs connected with the road section l of the straight going of the downstream crossing_jxOr null, that is to say, the straight going is prohibited; t is t_lIs 1_ijLeft-turn connected road section l of downstream crossing_jxOr null, i.e. left turn is prohibited; t is t_rIs 1_ijDownstream crossing right turn connected road section l_jxOr null, i.e., right turn is prohibited; t is t_uIs 1_ijSection of road l connected with downstream intersection in turn_jiOr null, i.e., no turning around is prohibited. See tables 1-2.

TABLE 1-2 urban traffic network and associated Attribute information Table

S202, sequentially extracting two adjacent passing data information according to the sequence in the classification data table, taking the first data and the second data as a group, taking the second data and the third data as a group, and so on.

S203, through all road network information databases in urban areas, the road section numbers in all the grouped first vehicle passing data can be used for knowing downstream junctions of the positions of electronic police and bayonets in a snapshot mode, and the clear junction information can be used for knowing a downstream road section number set through the steering information of the road network information database.

And S204, comparing the road section number of the second vehicle passing data in the group with the road section number in the road section number set at the downstream of the S203 one by one, wherein if consistent road section numbers exist, adjacent detection point positions can be directly connected to form an effective path, and a column of the data table is directly connected with 1. If not, the column of the data table "direct connection" is filled with 0, and the process proceeds to step S300. See Table 1-1.

S300, extracting all possible effective paths between the connected detection points under the condition that the adjacent detection points of a certain vehicle cannot be directly connected; the method specifically comprises the following steps:

s301, determining the difference between two adjacent detection times, △ t being 98S (25)

S302, determining all possible driving paths between adjacent detection points by adopting bidirectional tracing, namely, taking the first vehicle passing detection point of the group as a starting point, and gradually extending downstream according to road network information to construct a path network with a tree structure; and taking the second grouped vehicle passing detection point as a terminal point, and gradually extending upstream according to the road network information to construct a path network of a tree structure. And setting the road force network layer number Q of the bidirectional tree structure to be 3.

S303, through all road network information databases in urban areas, the road section number in all the grouped first vehicle passing data can be used for knowing a downstream intersection at the snapshot position of the electronic police, and the clear intersection information can be used for knowing a downstream road section number set D through the steering information of the road network information database₁. Determining the downstream road junction according to the downstream road segment number set, and obtaining the downstream road segment number set D according to the steering information of the road network information database₂(ii) a The calculation is continuously carried out to obtain the 3 generation branch segment number information set D₃. And based on 3 generation operation, sequentially generating information tables of different traffic paths by collecting road section numbers in different stages

See tables 1-3.

Tables 1-3 information tables

Serial number	Starting road segment numbering	D1	D2	D3
					1	l5，6	l6,10	l10,9	l9,5
2	l5,6	l6,10	l10，9	l9,13
					3	l5,6	l6,10	l10,14	l14,13
4	l5,6	l6,10	l10,14	l14,15
					5	l5,6	l6,10	l10,11	l11,15
6	l5,6	l6,10	l10,11	l11,12
					7	l5,6	l6,10	l10,11	l11,7
8	l5,6	l6,7	l7,3	l3,2
					9	l5,6	l6,7	l7,3	l3,4
10	l5,6	l6,7	l7,8	l8,4
					11	l5,6	l6,7	l7,8	l8,12
12	l5,6	l6,7	l7,11	l11,10
					13	l5,6	l6,7	l7,11	l11,15
14	l5,6	l6,7	l7,11	l11,12
					15	l5,6	l6,2	l2,3	l3,7
16	l5,6	l6,2	l2,3	l3,4
					17	l5,6	l6,2	l2,1	l1,5

S304, determining the upstream road junction of the road segment number of the second passing data in each group according to the steering information of all the road network information databases in the urban area, and knowing the upstream road segment number set U according to the specific road junction information and the steering information of the road network information databases₁. Determining the road junction of the upstream road section according to the road section number set of the upstream, and knowing the road section number set U of the upstream through the steering information of the road network information database₂(ii) a The calculation is continuously carried out to obtain the 3 generation branch road part number information set U_q. And based on 3 generation operation, sequentially generating information tables of different traffic paths by collecting road section numbers in different stages

See tables 1-4.

Tables 1-4 information tables

Serial number	Terminal road segment numbering	u1	u2	u3
					1	l11,15	l12,11	l16,12	l15,16
2	l11,15	l12,11	l8,12	l7,8
					3	l11,15	l12,11	l8,12	l4,8
4	l11,15	l7,11	l6,7	l5,6
					5	l11,15	l7,11	l6,7	l10,6
6	l11,15	l7,11	l6,7	l2,6
					7	l11,15	l7,11	l3,7	l2,3
8	l11,15	l7,11	l3,7	l4,3
					9	l11,15	l7,11	l8,7	l12,8
10	l11,15	l7,11	l8,7	l4,8
					11	l11,15	l10,11	l9,10	l13,9
12	l11,15	l10,11	l9,10	l5,9
					13	l11,15	l10,11	l6,10	l2,6
14	l11,15	l10,11	l6,10	l5,6
					15	l11,15	l10,11	l6,10	l7,6
16	l11,15	l10,11	l14,10	l13,14
					17	l11,15	l10,11	l14,10	l15,14

S305, information table

The nodes in certain path information extract node information according to the node sequence from upstream crossing to downstream crossing, and the node information is stored in an information table

Which searches for traffic paths containing the same nodes. Extracting information tables

Downstream traffic path information having common nodes, extracting information table

The upstream traffic path information with common node is merged into a complete traffic path information table

See tables 1-5.

Tables 1-5 information tables

S400, determining time consumption values of different paths based on the effective paths extracted in S300 and S200 and the operation data information of different road sections of the Internet platform at each time period; the method specifically comprises the following steps:

s401, sequentially acquiring all traffic paths in the step S305, and extracting length information of a plurality of road sections in the traffic paths; acquiring two adjacent detection time intervals t of each road section_i，t_i+1]And obtaining the consumed time of different paths in the first stage by the average running speed data of the internal Internet platform, namely:

t_s1time-consuming first-stage traffic path, unit: second;

d_isegment length of segment i, in units: rice;

-the length of the road section i detection area from the downstream intersection, unit: rice;

o-transit time in the intersection, difference in the values of different regions, and the average value of 5, unit: second;

v′_i-time interval [ t ]_i，t_i+1]Average running speed of inner road section i, unit: m/s.

S402, calculating to obtain that the time when the vehicle enters the road section i is (t)_i+△t′_i) The time of entering the section i +1 is (t)_i+△t′_i+1) Of which △ t'_iComprises the following steps:

△t′_i-the difference between the time of entering the road section i and the first detection time in the second stage, unit: second;

d_isegment length of segment i, in units: rice;

v′_i-average speed of travel of section i, in units: m/s;

o-transit time in a crossing, different area values are different, and an average value needs to be obtained through investigation, wherein the unit is as follows: second;

△t_itwo adjacent detection times t_iAnd t_i+1The difference of (a).

S403, acquiring the time interval t of the Internet platform_i+△t′_i，t_i+△t′_i+1]And obtaining the time consumption of different paths in the second stage by the average driving speed data on the road section i, namely:

t_s2time consumption of the second phase traffic path, unit: second;

d_isegment length of segment i, in units: rice;

-the length of the road section i detection area from the downstream intersection, unit: rice;

o-transit time in the intersection, difference in the values of different regions, and the average value of 5, unit: second;

v″_i-time interval [ t ]_i+△t′_i，t_i+△t′_i+1]Average running speed of inner road section i, unit: m/s;

s404, calculating to obtain that the time when the vehicle enters the road section i is (t)_i+△t″_i) Entering road sectionThe time of i +1 is (t)_i+△t″_i+1) Namely:

△t″_i-the third phase difference, unit, between the time of entering the section i and the first detection time: second;

d_isegment length of segment i, in units: rice;

v″_i-time interval [ t ]_i+△t″_i，t_i+△t″_i+1]Average travel speed of link i, unit: m/s;

o-transit time in the intersection, difference in the values of different regions, and the average value of 5, unit: second;

△t_itwo adjacent detection times t_iAnd t_i+1A difference of (d);

s405, obtaining the time interval t of the Internet platform_i+△t″_i，t_i+△t″_i+1]And obtaining the time consumption of different paths in the third stage by the average running speed data on the road section i, namely:

t_s3-time-consuming, unit of third-stage traffic route: second;

d_isegment length of segment i, in units: rice;

-the length of the road section i detection area from the downstream intersection, unit: rice;

o-transit time in the intersection, difference in the values of different regions, and the average value of 5, unit: second;

v″′_i-time interval [ t ]_i+△t″_i，t_i+△t″_i+1]Average running speed of inner road section i, unit: m/s;

s406, setting time consumption for calculating different paths to pass through 5 stages; according to the processing from step S404 to step S405, the time interval of the Internet platform is finally obtained

And (3) obtaining the time consumption of different paths in the 5 th stage by using the average traveling speed data on the section i, namely:

t_s5-time consumption of phase 5 traffic routes, unit: second;

d_isegment length of segment i, in units: rice;

-the length of the road section i detection area from the downstream intersection, unit: rice;

o-transit time in the intersection, difference in the values of different regions, and the average value of 5, unit: second;

-time interval

Average running speed of inner road section i, unit: m/s.

Tables 1-6 are derived from steps S401-S406.

TABLE 1-6 TIME-TAKING TABLE FOR THE DIFFERENT PATH IN STAGE 5

Serial number	Starting road segment numbering	Intermediate link numbering	Terminal road segment numbering	ts5(second)
					1	l5,6	l6,10--l10,11	l11,15	332.1
2	l5,6	l6,7--l7,11	l11,15	322.5
					3	l5,6	l6,2--l2,3--l3,7--l7,11	l11,15	502.3
4	l5,6	l6,7--l7,8--l8,12--l12,11	l11,15	427.6
					5	l5,6	l6,10--l10,9--l9,13--l13,14--l14,10--l10,11	l11,15	661.5
6	l5,6	l6,10--l10,14--l14,15--l15,16--l16,12--l12,11	l11,15	508.7
					7	l5,6	l6,10--l10,11--l11,12--l12,8--l8,7--l7,11	l11,15	617.8
8	l5,6	l6,7--l7,3--l3,2--l2,6--l6,10--l10,11	l11,15	808.9
					9	l5,6	l6,7--l7,3--l3,4--l4,8--l8,12--l12,11	l11,15	582.3
10	l5,6	l6,7--l7,3--l3,4--l4,8--l8,7--l7,11	l11,15	609.4
					11	l5,6	l6,7--l7,11--l11,12--l12,8--l8,7--l7,11	l11,15	607.2
12	l5,6	l6,7--l7,11--l11,10--l10,6--l6,7--l7,11	l11,15	808.0
					13	l5,6	l6,2--l2,3--l3,7--l7,8--l8,12--l12,11	l11,15	608.0
14	l5,6	l6,2--l2,3--l3,4--l4,8--l8,7--l7,11	l11,15	603.8
					15	l5,6	l6,2--l2,3--l3,4--l4,8--l8,12--l12,11	l11,15	576.8

And S500, comparing the consumed time of different paths with the time difference value of adjacent detection points based on the S400, and determining an optimal vehicle driving path set. The method specifically comprises the following steps:

s501, obtaining the absolute value of the time difference value of the adjacent detection points subtracted from the time consumption of different paths in the 5 th stage based on the operation data of the Internet platform for judgment, namely judging

|t_s5-△t_i|≤△t_i/20 (32)

t_s5-time cost value of the fifth phase traffic path, in units: s;

△t_itwo adjacent detection times t_iAnd t_i+1The difference of (a).

S502, judging all possible vehicle running paths between adjacent detection points, storing the paths meeting the formula (32) in a vehicle running path set, and sequencing the absolute values of the time difference values of the adjacent detection points from small to large according to the time consumed by different paths. See tables 1-7.

TABLE 1-7 optimal vehicle travel Path Table

Serial number	Starting road segment numbering	Intermediate link numbering	Terminal road segment numbering	ts5(second)
					1	l5,6	l6,10--l10,11	l11,15	332.1
2	l5,6	l6,7--l7,11	l11,15	322.5

And S600, analyzing influence factors of the traffic paths based on the driving path set generated in the step S500, and constructing a traffic path reliability comprehensive function. The method specifically comprises the following steps:

s601, difference value of time consumed by the path and detection time, number of turns required by the path and long and short node period are important factors influencing the path selection of the driver, and intersection containing the three factors is constructedPath reliability synthesis function F (x)₁,x₂,x₃). The influencing factors include the difference value between the time consumed by the path and the detection time, the number of turns required by the path, and the length of the node cycle, and the set of factors U can be defined as: u ═ difference between the time consumed by the path and the detection time, the number of turns required by the path, and the node cycle length, expressed as: u ═ U₁,u₂,u₃}。

And S602, factor weighting. The fuzzy subset A of the factors in the factor set U is marked as fuzzy vector A ═ a₁,a₂,a₃) Wherein a is_iRepresents a single factor u_iMeasure of the magnitude of the contribution in the overall evaluation factor, 0 ≦ a_iLess than or equal to 1. Different areas a_iThe values are different, and the fuzzy vector A is obtained by questionnaire survey as (0.54,0.29, 0.17).

And S603, normalizing the influencing factors. The fuzzy subset B of the factors on the factor set U is marked as a fuzzy vector B ═ B₁,b₂,b₃)^TWherein b is_jRepresents a single factor u_iThe normalized value.

(1) Normalization value of difference between path consumed time and detection time:

b₁-the difference between the time taken for the path and the detection time is normalized;

t_s5-time cost value of the fifth phase traffic path, in units: s;

△t_itwo adjacent detection times t_iAnd t_i+1The difference of (a).

(2) Normalized value of number of turns required for a path:

b₂-the number of turns required for the path is normalized;

g₁the left turn times and the unit of the different paths between the adjacent detection points need to be counted: next time；

g₂-different paths between adjacent detection points need to count the right turn times in units: secondly;

g₃-different paths between adjacent detection points need to count the straight row total times, unit: secondly;

m₁the left turn relation coefficient has different values in different areas, and the value is 0.5;

m₂the value of the right turn relation coefficient is different in different areas, and the value is 0.8;

m₃the value of the straight-going relation coefficient is different in different areas and is 1.0 at this time.

(3) Normalized value of node period length

Four intervals are divided according to the signal period length of the existing signal intersection, and the intersection period length of each interval is investigated to obtain an intersection signal period coefficient table 1-8.

b₃-node cycle length is short to unity;

h₁the number of intersections with different paths passing through signal periods less than or equal to 80 seconds, unit: a plurality of;

h₂the number of intersections with the signal period of more than 80 seconds and less than or equal to 110 seconds for different paths is as follows: a plurality of;

h₃the number of intersections with the signal period of more than 110 seconds and less than or equal to 140 seconds for different paths is as follows: a plurality of;

h₄the number of intersections with different paths passing through a signal period greater than 140 seconds, unit: a plurality of;

w₁-the different paths pass through a coefficient with a signal period of less than or equal to 80 seconds;

w₂-the different paths pass through coefficients with a signal period greater than 80 seconds and less than or equal to 110 seconds;

w₃-system of different path passing signal period greater than 110 seconds and less than or equal to 140 secondsCounting;

w₄-the different paths pass through coefficients with a signal period greater than 140 seconds;

TABLE 1-8 crossing signal period coefficient table

Serial number	Signal period interval	Coefficient of signal period wi
			1	T≤80s	0.9
2	80s＜T≤110s	0.8
			3	110s＜T≤140s	0.7
4	T＞140s	0.6

In summary, (B) is the factor fuzzy subset B of different vehicle driving paths₁,b₂,b₃)^T. See tables 1-9.

TABLE 1-9 fuzzy subsets of factors for different vehicle travel paths

Serial number	Starting road segment numbering	Intermediate link numbering	Terminal road segment numbering	Factor ambiguity subset B
					1	l5,6	l6,10--l10,11	l11,15	(0.975,0.700,0.767)T
2	l5，6	l6,7--l7,11	l11,15	(0.995,0.833,0.733)T

S604, constructing a comprehensive function F (x)₁,x₂,x₃)。F(x₁,x₂,x₃) The larger the function value is, the higher the reliability of the extracted path is, that is, the optimal path is.

F(x₁,x₂,x₃)＝AB＝(a₁,a₂,a₃)(b₁,b₂,b₃)^T(36)

The calculation of the formula (36) can obtain the running paths of different vehiclesAnd (4) integrating the function values dependently. See tables 1-10 for the traffic route l of the adjacent monitoring points_5,6--l_6,7--l_7,11--l_11,15The comprehensive function value is the highest, and the traffic path of the vehicle in the whole time period is obtained as l_1,5--l_5,6--l_6,7--l_7,11--l_11,15--l_15,16。

TABLE 1-10 fuzzy subsets of factors for different vehicle travel paths

Serial number	Starting road segment numbering	Intermediate link numbering	Terminal road segment numbering	Synthesis function F (x)1,x2,x3)
					1	l5,6	l6,10--l10,11	l11,15	0.86
2	l5,6	l6,7--l7,11	l11,15	0.90

The embodiment of the invention is designed aiming at a vehicle running track extraction method, and is characterized in that data contents acquired by the existing electronic police and a bayonet device are considered, an algorithm flow for extracting a vehicle path is constructed by combining internet data, a road network information graph adaptive to a path extraction algorithm needs to be perfected in the process, and the deep mining treatment of the data information is a basis for constructing scientific and effective path extraction; and then analyzing influence factors influencing the driving path of the driver, setting the weight and normalization processing of the influence factors, and finally establishing a path reliability comprehensive function model. The research result can provide key data support for formulation of urban road management and control measures, traffic travel guidance and road network planning. The method has very wide application prospect in order to reduce social cost and improve social travel efficiency.

The above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments can be modified, or some technical features can be equivalently replaced; and the modifications or the substitutions do not make the essence of the corresponding technical solution depart from the spirit and scope of the technical solution of the embodiments of the present invention.

Claims (7)

1. The utility model provides a vehicle movement track extraction method, obtains the data information of crossing that the different positions on single vehicle time length were taken a candid photograph through electric police, bayonet socket equipment, through the vehicle on the internet platform in the different highway sections and the operation data information on time length, its characterized in that:

the method comprises the following steps:

s100, obtaining vehicle passing data of all electronic policemen and checkpoints on a certain day, and classifying and processing the vehicle passing data according to license plates;

s200, determining that adjacent detection point positions of a certain vehicle can be directly connected to form an effective path;

s300, extracting all possible effective paths between the connected detection points under the condition that the adjacent detection points of a certain vehicle cannot be directly connected;

s400, determining the consumed time of different paths based on the effective path extracted in S300 and the operation data information of different road sections of the Internet platform at each time interval;

s500, comparing the consumed time of different paths with the time difference value of adjacent detection points based on S400, and determining an optimal vehicle running path set;

and S600, analyzing influence factors of the traffic path based on the driving path set generated in the S500, and constructing a traffic path reliability comprehensive function to determine an optimal driving path.

2. The vehicle running track extraction method according to claim 1, characterized in that: the S100 acquires the passing data of all electronic policemen and checkpoints at a certain day, and classifies and processes the data according to license plates; the method specifically comprises the following steps:

s101, obtaining vehicle passing data of all electronic policemen and checkpoints in a certain day in an urban area, extracting information such as detection equipment IDs (identity) of all vehicle passing data, license plate numbers, detection time, road section numbers, lengths of inspection areas from downstream intersections and the like, generating data strips, and classifying all the data strips according to the license plate numbers;

and S102, generating a new data table by each type of data according to the sequence of the detection time, wherein a column of direct connection is additionally arranged in the new data table, the column only has 0 and 1 filled with data, and the last detection point is not required to be filled.

3. The vehicle running track extraction method according to claim 2, characterized in that: step S200, determining that adjacent detection point positions of a certain vehicle can be directly connected to form an effective path; the method specifically comprises the following steps:

s201, acquiring all urban road network information databases, constructing a road network G representing original geographic information, and abstracting road network nodes into nodes p in the road network G_iThe road sections between adjacent nodes of the road network can be abstracted into the edge l in the road network G_ijG ═ { V, L, T }, where V (G) } { p₁,p₂,…,p_nThe node set is used as the node set, and n represents the number of nodes; l (p)_i,p_j)＝{l_ij,l_ji,d_ijRepresents a vector set of road sections among road network nodes, wherein l_ijRepresenting a slave node v_iTo node v_jRoad section of d_ijRepresenting a slave node v_iTo node v_jA road segment length value of;

indicating a road segment l in a road network_ijThe downstream crossing signal cycle and the vector matrix set of the steering information thereof, wherein

For intersection p_jSignal period of (d), t_sIs 1_ijIs connected with the road section l of the straight going of the downstream crossing_jxOr null, i.e. no straight going is prohibited; t is t_lIs 1_ijLeft-turn connected road section l of downstream crossing_jxOr null, i.e. left turn is prohibited; t is t_rIs 1_ijDownstream crossing right turn connected road section l_jxOr null, i.e., right turn is prohibited; t is t_uIs 1_ijSection of road l connected with downstream intersection in turn_jiOr null, namely, the turning around is forbidden;

s202, sequentially extracting two adjacent passing data information according to the sequence in the classification data table, taking the first data and the second data as a group, taking the second data and the third data as a group, and repeating the steps;

s203, through all road network information databases in urban areas, the road section numbers in all grouped first passing vehicle data can be used for knowing downstream junctions of positions of electronic police and bayonets in a snapshot manner, and the clear junction information can be used for knowing a downstream road section number set through steering information of the road network information database;

s204, comparing the road section number of the second vehicle passing data in the group with the road section number in the road section number set at the downstream of the S203 one by one, if consistent road section numbers exist, adjacent detection point positions can be directly connected to form an effective path, and a data table is filled with 1 in a 'direct connection' column; if not, the column of the data table "direct connection" is filled with 0, and the process proceeds to step S300.

4. The vehicle running track extraction method according to claim 3, characterized in that: in the step S300, under the condition that adjacent detection points of a certain vehicle cannot be directly connected, all possible effective paths between the connected detection points are extracted; the method specifically comprises the following steps:

s301, determining and calculating two adjacent detection time t_iAnd t_i+1The difference of (a):

△t＝t_i+1-t_i,i＝1,2,…,n-1 (1)

s302, determining all possible driving paths between adjacent detection points by adopting bidirectional tracing, namely, taking the first vehicle passing detection point of the group as a starting point, and gradually extending downstream according to road network information to construct a path network with a tree structure; gradually extending the second grouped vehicle passing detection point to the upstream to construct a tree-structured path network according to the road network information by taking the second grouped vehicle passing detection point as a terminal point; the number of road stiffness network layers for constructing the bidirectional tree structure is preset to be Q.

S303, through all road network information databases in urban areas, the road section number in all the grouped first vehicle passing data can be used for knowing the downstream intersection of the electronic police snapshot position, and the clear intersection information can be used for knowing a downstream road section number set D through the steering information of the road network information database₁(ii) a Determining the downstream road junction according to the downstream road segment number set, and obtaining the downstream road segment number set D according to the steering information of the road network information database₂(ii) a The calculation is continuously carried out to obtain the Q generation branch road part number information set D_q(ii) a And based on Q generation operation, sequentially generating information tables of different traffic paths by collecting road section numbers in different stages

S304, determining the upstream intersection of the road section number of the second passing data in each group according to the steering information of all road network information databases in the urban area, and determiningThe road junction information can know the upstream road section number set U through the steering information of the road network information database₁(ii) a Determining the road junction of the upstream road section according to the upstream road section number set, and knowing the upstream road section number set U through the steering information of the road network information database₂(ii) a The calculation is continuously carried out to obtain the Q generation branch road part number information set U_q(ii) a And based on Q generation operation, sequentially generating information tables of different traffic paths by collecting road section numbers in different stages

S305, information table

The nodes in certain path information extract node information according to the node sequence from upstream crossing to downstream crossing, and the node information is stored in an information table

Searching for a traffic path containing the same node; extracting information tables

Extracting information table from upstream traffic path information having common node

The upstream traffic path information with common node is merged into a complete traffic path information table

5. The vehicle running track extraction method according to claim 4, characterized in that: the step S400 is to determine time consumption values of different paths based on the effective path extracted in the step S300 and the operation data information of different road sections of the Internet platform at each time period; the method specifically comprises the following steps:

s401, sequentially acquiring all traffic paths in the step S305, and extracting length information of a plurality of road sections in the traffic paths; acquiring two adjacent detection time intervals t of each road section_i，t_i+1]And obtaining the consumed time of different paths in the first stage by the average running speed data of the internal Internet platform, namely:

t_s1time-consuming first-stage traffic path, unit: second;

d_isegment length of segment i, in units: rice;

-the length of the road section i detection area from the downstream intersection, unit: rice;

o-transit time in a crossing, different area values are different, and an average value needs to be obtained through investigation, wherein the unit is as follows: second;

v′_i-time interval [ t ]_i，t_i+1]Average running speed of inner road section i, unit: m/s;

s402, calculating that the time p for the vehicle to enter the road section is (t)_i+△t′_i) The time of entering the section i +1 is (t)_i+△t′_i+1) Of which △ t'_iComprises the following steps:

△t′_i-the difference between the time of entering the road section i and the first detection time in the second stage, unit: second;

d_isegment length of segment i, in units: rice;

v′_i-average speed of travel of section i, in units: m/s;

o-transit time in a crossing, different area values are different, and an average value needs to be obtained through investigation, wherein the unit is as follows: second;

△t_itwo adjacent detection times t_iAnd t_i+1The difference of (a).

S403, acquiring the time interval t of the Internet platform_i+△t'_i，t_i+△t'_i+1]And obtaining the time consumption of different paths in the second stage by the average driving speed data on the road section i, namely:

t_s2time consumption of the second phase traffic path, unit: second;

d_isegment length of segment i, in units: rice;

-the length of the road section i detection area from the downstream intersection, unit: rice;

o-transit time in a crossing, different area values are different, and an average value needs to be obtained through investigation, wherein the unit is as follows: second;

v″_i-time interval [ t ]_i+△t'_i，t_i+△t'_i+1]Average running speed of inner road section i, unit: m/s;

s404, calculating to obtain that the time when the vehicle enters the road section i is (t)_i+△t″_i) The time of entering the section i +1 is (t)_i+△t″_i+1) Namely:

△t″_i-the third phase difference, unit, between the time of entering the section i and the first detection time: second;

d_isegment length of segment i, in units: rice;

v″_i-time interval [ t ]_i+△t″_i，t_i+△t″_i+1]Average travel speed of link i, unit: m/s;

o-transit time in a crossing, different area values are different, and an average value needs to be obtained through investigation, wherein the unit is as follows: second;

△t_itwo adjacent detection times t_iAnd t_i+1A difference of (d);

s405, obtaining the time interval t of the Internet platform_i+△t″_i，t_i+△t″_i+1]And obtaining the time consumption of different paths in the third stage by the average running speed data on the road section i, namely:

t_s3-time-consuming, unit of third-stage traffic route: second;

d_isegment length of segment i, in units: rice;

-the length of the road section i detection area from the downstream intersection, unit: rice;

o-transit time in a crossing, different area values are different, and an average value needs to be obtained through investigation, wherein the unit is as follows: second;

v″′_i-time interval [ t ]_i+△t″_i，t_i+△t″_i+1]Average running speed of inner road section i, unit: m/s;

s406, presetting the time consumption for calculating different paths and needing to pass through a K stage; according to the processing from step S404 to step S405, the time interval of the Internet platform is finally obtained

And (3) obtaining the time consumption of different paths in the K stage according to the average running speed data on the road section i, namely:

t_sKtime-consuming, unit of phase k traffic path: second;

d_isegment length of segment i, in units: rice;

-the length of the road section i detection area from the downstream intersection, unit: rice;

o-transit time in a crossing, different area values are different, and an average value needs to be obtained through investigation, wherein the unit is as follows: second;

-time interval

Average running speed of inner road section i, unit: m/s.

6. The vehicle running track extraction method according to claim 5, characterized in that: the step S500 is that based on the step S400, the consumed time of different paths is compared with the time difference value of adjacent detection points, and an optimal vehicle driving path set is determined; the method specifically comprises the following steps:

s501, obtaining the absolute value of the time difference value of the adjacent detection points subtracted by the time consumed by different paths in the K stage based on the operation data of the Internet platform for judgment, namely

|t_sK-△t_i|≤△t_i/10 (8)

t_sK-time cost value of phase k traffic path, in units: s;

△t_itwo adjacent detection times t_iAnd t_i+1A difference of (d);

s502, judging all possible vehicle running paths between adjacent detection points, storing the paths meeting the formula (8) in a vehicle running path set, and sequencing from small to large according to the absolute value of the time difference value of subtracting the adjacent detection points from the time consumed by different paths.

7. The vehicle running track extraction method according to claim 6, characterized in that: the step S600 is to analyze the influence factors of the traffic path and construct a traffic path reliability comprehensive function based on the driving path set generated in the step S500; the method specifically comprises the following steps:

s601, constructing a traffic path reliability comprehensive function F (x) containing the three factors, wherein the difference value of the path time consumption and the detection time, the number of turns required by the path and the period length of the nodes are important factors influencing the path selection of the driver₁,x₂,x₃) (ii) a The influencing factors include the difference value between the time consumed by the path and the detection time, the number of turns required by the path, and the length of the node cycle, and the set of factors U can be defined as: u ═ difference between time consumed by the path and detection time, number of turns required by the path, and node cycle length, expressed as: u ═ U₁,u₂,u₃}。

S602, factor weight; the fuzzy subset A of the factors in the factor set U is marked as fuzzy vector A ═ a₁,a₂,a₃) Wherein a is_iRepresents a single factor u_iMeasure of the magnitude of the contribution in the overall evaluation factor, 0 ≦ a_iLess than or equal to 1; different areas a_iValues are different, and fuzzy vectors are obtained through questionnaire to obtain the average value A which is (0.54,0.29, 0.17).

And S603, normalizing the influencing factors. The fuzzy subset B of the factors on the factor set U is marked as a fuzzy vector B ═ B₁,b₂,b₃)^TWherein b is_jRepresents a single factor u_iThe normalized value.

(1) Normalization value of difference value between path time consumption and detection time

b₁-normalizing the difference between the time taken for the path and the detection time;

t_sk-time consumption for phase K traffic route, unit: s;

△t_itwo adjacent detection times t_iAnd t_i+1A difference of (d);

(2) normalized value of number of turns required for path

b₂-a normalized value of the number of turns required for the path;

g₁the left turn times and the unit of the different paths between the adjacent detection points need to be counted: secondly;

g₂-different paths between adjacent detection points need to count the right turn times in units: secondly;

g₃-different paths between adjacent detection points need to count the straight row total times, unit: secondly;

m₁the left turn relation coefficient has different values in different areas, and the average value can be 0.5;

m₂the values of the right turn relation coefficients in different areas are different, and the average value can be 0.8;

m₃the straight-going relation coefficient has different values in different areas, and the average value can be 1.0;

(3) normalized value of node period length

Four intervals are divided according to the signal period length of the existing signal intersection, and the intersection period length of each interval is investigated to obtain the intersection signal period coefficient.

b₃-a normalization value of the node period length;

h₁the number of intersections with different paths passing through signal periods less than or equal to 80 seconds, unit: a plurality of;

h₂-different paths pass through paths with signal period greater than 80 seconds and less than or equal to 110 secondsNumber of mouths, unit: a plurality of;

h₃the number of intersections with the signal period of more than 110 seconds and less than or equal to 140 seconds for different paths is as follows: a plurality of;

h₄the number of intersections with different paths passing through a signal period greater than 140 seconds, unit: a plurality of;

w₁-the different paths pass through a coefficient with a signal period of less than or equal to 80 seconds;

w₂-the different paths pass through coefficients with a signal period greater than 80 seconds and less than or equal to 110 seconds;

w₃-the different paths pass through coefficients with a signal period greater than 110 seconds and less than or equal to 140 seconds;

w₄-the different paths pass through coefficients with a signal period greater than 140 seconds;

obtaining factor fuzzy subset B ═ (B) of different vehicle driving paths₁,b₂,b₃)^T；

S604, constructing a comprehensive function F (x)₁,x₂,x₃)。F(x₁,x₂,x₃) The larger the function value is, the higher the reliability of the extracted path is, namely the optimal path is;

F(x₁,x₂,x₃)＝AB＝(a₁,a₂,a₃)(b₁,b₂,b₃)^T(12)

and (3) calculating by the formula (12) to obtain a traffic route with the highest reliability comprehensive function value of different vehicle running routes, wherein the traffic route is the vehicle running route in the adjacent detection time.

Images