CN108417039A - A Method for Estimating Traffic Demand at Signalized Intersections Influenced by Traffic Flow Composition - Google Patents

Abstract

The invention belongs to the technical field of intelligent traffic control, and relates to a method for estimating traffic demand at signalized intersections influenced by traffic flow composition, which is suitable for arterial and regional road systems. Considering various types of vehicles and discrete characteristics of platoons, vehicle detectors are set at the stop line of the intersection entrance and its upstream section respectively, and multiple traffic flow data sets are obtained from this, and a platoon dispersion coefficient calibration method is proposed based on these data sets. Then a traffic demand estimation model is established. In addition, the technical application process of the above models and methods is given. Using the computer programming software MATLAB and the traffic simulation software VISSIM, use cases illustrate the technology application process. The results show that the new method can accurately estimate the traffic demand of lane groups at signalized intersections for the traffic flow composed of multiple types of vehicles.

Description

A Method for Estimating Traffic Demand at Signalized Intersections Influenced by Traffic Flow Composition

technical field

The invention belongs to the technical field of intelligent traffic control, and relates to a method for estimating traffic demand at signalized intersections influenced by traffic flow composition.

Background technique

With the rapid development of social economy, the increasing number of motor vehicles has caused the problem of urban traffic congestion to become increasingly serious. Signal timing optimization at intersections is an effective way to alleviate urban traffic congestion. As the basis of signal timing optimization, the accuracy of traffic demand data determines whether the signal timing scheme is reliable.

In recent years, research results on traffic demand estimation at home and abroad can be summarized as follows: (1) From a macro perspective, estimate the annual traffic demand or daily traffic demand. The four-stage model requires a large-scale traffic survey to obtain the travel data of the origin and destination, which often consumes a lot of manpower, material resources and financial resources; compared with the travel data of the origin and destination, the road section traffic flow data in the OD back-estimation method is easier to obtain. (2) From the micro level, some scholars combined the positioning data and ArcGIS software to obtain the hourly traffic demand of each road section in the urban road network, which can show the differences in traffic demand between different regions in the city; some scholars used road detector data to estimate The traffic demand of lane groups and road sections in the urban road network is calculated, and the time scale of the estimated results can be less than one hour, which can reflect the time-varying characteristics of traffic demand in more detail. Most of these studies ignore the effect of signal timing on road impedance.

Usually, traffic demand has strong time-varying characteristics. As far as signal control is concerned, it is particularly important to accurately grasp its time-varying law in order to effectively meet the traffic demand during peak hours. For arterial or regional road systems, the traffic demand on the downstream approach depends on the traffic demand on the upstream approach. Here, the traffic demand of the entrance refers to the arrival traffic flow, which can be obtained by setting the detection section. If the detection section is set at a position far upstream from the stop line, because the platoon dispersion phenomenon will occur when the fleet drives from the upstream to the downstream, the obtained change curve of the arriving traffic flow does not coincide with the change curve of the actual traffic demand; if the detection section is set at Near the stop line, the queue length of vehicles formed by signal control often exceeds the detection section. At this time, the measured traffic flow is essentially the flow of leaving traffic, so it cannot reflect the real traffic demand. When designing the signal timing scheme, in practice, the measured traffic flow at the entrance is often taken as its traffic demand, but this approach is not scientific. In addition, traffic flow is usually composed of multiple types of vehicles, and the discrete characteristics of different types of vehicles are very different due to differences in driving characteristics. In view of this, for arterial and regional road systems, the present invention proposes a method for estimating traffic demand at signalized intersections influenced by traffic flow composition.

Contents of the invention

The invention provides a method for estimating traffic demand of lane groups at signalized intersections for arterial and regional road systems, and the estimated traffic demand data can be used for signal timing optimization.

Technical scheme of the present invention:

A method for estimating traffic demand at signalized intersections influenced by traffic flow composition. Firstly, vehicle detectors are installed at the stop line of the entrance road of the signalized intersection and its upstream section to obtain the data of each traffic flow, and the fleet dispersion is calibrated based on the obtained data. coefficient, and then estimate the traffic demand of the intersection lane group, the specific steps are as follows:

(1) Data set definition

A is the relationship matrix of any two intersections i and i′ in the road network; Α=(a _ii′ ) _I×I , i, i′∈{1,2,…,I}, I is the number of intersections; when When the intersection i' is adjacent to i, a _ii' = 1, otherwise, a _ii' = 0;

B _ii' is the relationship matrix between the entrance road j of intersection i and the exit road k' of intersection i'; j∈{1,2,…,J _i }, k′∈{1,2,…,K _i′ }, J _i is the number of entrance lanes at intersection i, and K _i′ is the number of exit lanes at intersection i’; When the entrance road j of the intersection i and the exit road k' of the intersection i′ belong to the same section, b _jk′ = 1, otherwise, b _jk′ = 0;

C _i is the relationship matrix between the exit road k and the traffic flow m in the intersection i; k∈{1,2,…,K _i }, m∈{1,2,…,M _i }, K _i , M _i are the number of exit lanes and traffic flow of intersection i respectively; When k has the right of way, c _km =1, otherwise, c _km =0;

is the distance between the stop line of intersection i′ entrance j′ and the upstream section of intersection i entrance j′, j′∈{1,2,…,J _i′ }, J _i′ is the intersection i′ entrance number;

is the distance between the stop line of intersection i' entrance road j' and the stop line of intersection i entrance road j';

is the expected speed of the class w vehicle when it travels from intersection i′ to i;

is the expected travel time of class w vehicles from the stop line of intersection i′entrance j′ to the upstream section of intersection i′entrance j′,

is the expected travel time of the class w vehicle from the intersection i' entrance road j' stop line to the intersection i entrance road j stop line,

a is an integer variable, w∈{1,2,…,W}, W is the number of vehicle types, ceil is rounded up, and max is the maximum value;

T is the length of the analysis period;

For the time period (t, t+T), the collection of the actual number of vehicles passing the stop line in all sampling intervals of the wth class vehicle of the traffic flow m at the intersection i;

For the time period (ta, t), the set of the actual number of vehicles passing the stop line in all sampling intervals of the wth class vehicle of the traffic flow m at the intersection i;

For the time period (ta, t+T), the collection of the actual number of vehicles passing the stop line in all sampling intervals of the wth class vehicle of the traffic flow m at the intersection i;

For the time period (ta,t+T), the intersection The collection of the actual number of vehicles that pass the stop line in all sampling intervals of class w vehicles of each traffic flow; for intersection i and entrance road j, when a _ii′ =1 and b _jk′ =1, record At this time the intersection is the upstream intersection of entry road j;

For the time period (t, t+T), the collection of the actual number of vehicles passing the upstream section in all sampling intervals of the wth class vehicle at the intersection i entrance road j;

For the time period (ta, t), the set of the actual number of vehicles passing through the upstream section in all sampling intervals of the w class vehicle at the intersection i entrance road j;

For the time period (ta, t+T), the collection of the actual number of vehicles passing through the upstream section in all sampling intervals of the wth class vehicle at the intersection i entrance road j;

(2) Detector layout

Arrange vehicle detectors at each intersection stop line and its upstream section, and the distance between the stop line and its upstream section is greater than the maximum queuing length of the entrance road at the peak period; according to and calculated Then obtain the value of a; use the vehicle detector at the stop line to obtain the number of vehicles leaving in all sampling intervals of each type of vehicle in the section, that is, the data set Use the upstream section detector of the entrance road to obtain the number of arriving vehicles in all sampling intervals of various vehicles in the section, that is, the data set

In the formula: S _ij is the distance between the upstream section of intersection i entrance j and its stop line; β _ii′ is the distance coefficient, β _ii′ ∈ (0,1); D _i′,i is the Entrance road j, distance between intersection i′ and the entrance road stop line of i;

(3) Data processing

(3.1) Detector data at the stop line

Use formula (2) to pair the data set processed to get the data set and Using formula (3) to the data set processed, get Using formula (4) to pair the data set processed to get the data set

In the formula: is the actual number of vehicles in the traffic flow m of the intersection i passing the stop line in the time period z; is the data sampling interval; Estimate intervals for traffic flows, for Integer multiples of ; M _i' is the traffic flow number of intersection i'; is the set of the actual number of vehicles passing the stop line in all the sampling intervals of vehicles of class w at the intersection i′ traffic flow m′ in the time period (ta, t+T); c _k′m′ is the exit road k′ inside the intersection i′ The identifier of the relationship with the traffic flow m', when the traffic flow m' has the right of way at the exit road k', c _k'm' = 1, otherwise, c _k'm' = 0;

(3.2) Upstream section detector data

Use formula (5) to pair the data set processed to get the data set and Using formula (6) to the data set processed, get

In the formula: is the actual number of vehicles of class w passing through the upstream section in the period z at the entrance j of intersection i;

(4) Discrete coefficient calibration

In order to estimate the traffic demand, it is necessary to calibrate the correction coefficient α ^w in the traffic demand estimation model. The specific calibration method is as follows:

Use formula (7) to estimate the number of vehicles of class w passing through the upstream section in the period z With 0.01 as the interval, α ^w is taken from 0 to 1, seeking to make and The absolute value of the average relative error _Δij is the smallest correction factor Then, use this coefficient and formula (9) to get the correction coefficient

In the formula: is the intersection i′traffic flow m′ in the time period The actual number of vehicles in category w passing through the stop line; Respectively, the estimated number of vehicles of class w vehicles passing through the upstream section in the period z and z-1 of the entrance road j of the intersection i; is the dispersion coefficient of the wth class vehicle from the stop line of each entrance road at the intersection i′ to the upstream section of the intersection i entrance road j; α ^w is the correction coefficient of the wth class vehicle, α ^w ∈ [0,1]; is the correction coefficient of the travel time of class w vehicles from the stop line of each entrance road at intersection i′ to the upstream section of intersection i entrance road j, is the correction coefficient of the travel time of class w vehicles from the stop line of each entrance road at intersection i′ to the stop line of intersection i entrance road j, Z is the time period; is the adjustment factor,

(5) Traffic demand estimation model

Use correction factor Estimate the traffic demand of each traffic flow at intersection i, entrance road j, and formula (10)

In the formula: is the estimated number of vehicles passing the stop line in the time period z of the traffic flow m at the intersection i entrance road j; Estimated number of vehicles of type w vehicles passing the stop line at the intersection i entrance road j traffic flow m in the time period z-1; is the intersection i′traffic flow m′ in the time period The actual number of vehicles of class w in the vehicle passing the stop line; _χijm is the proportion of traffic flow m in all traffic flow at intersection i entrance road j; is the dispersion coefficient of the wth class vehicle from the stop line of each entrance road at the intersection i′ to the stop line of the entrance road j of the intersection i.

Beneficial effects of the present invention: In arterial and regional road systems, the present invention can accurately estimate the traffic demand of lane groups at signalized intersections for the traffic flow composed of multiple types of vehicles, and the estimated traffic demand is used for signal timing optimization Good results can be obtained.

Description of drawings

Figure 1 is a schematic diagram of the arterial intersection group and its lane setting.

Figure 2 is a schematic diagram of the regional intersection group and its lane setting, where each intersection can be a three-way, four-way or five-way intersection, and each entrance road is channelized with one short left-turn lane and one dedicated left-turn lane lanes, 2 straight lanes, and 1 straight right mixed lane. For each entrance road, there may be no or more than one left-turn short lane or left-turn special lane, there may be no or one or more through lanes, or there may be no straight and right mixed lanes, or there may be One or more right-turn lanes. In addition, D _i-1,i is the distance between the stop line of the entrance road at intersection i-1 and i in the upward direction, D _i,i+1 is the distance between the stop line of the entrance road at intersection i and i+1 in the upward direction, D _{i +1,i+2} is the distance between the stop line of the entrance road at the intersection i+1 and i+2 in the upward direction, D _i+2,i-1 is the stop line of the entrance road at the intersection i+2 and i-1 in the upward direction spacing. Here, the direction from west to east is the upward direction of the arterial intersection group, and the counterclockwise direction is the upward direction of the regional intersection group.

Figure 3 is a schematic diagram of the intersection detector setting. Taking intersection i as an example, starting from the west entrance, the left-turn lane groups are numbered clockwise as M1, M3, M5, and M7, and the straight right lane groups that conflict with it They are numbered M2, M4, M6, M8 in sequence. The detector number in the figure consists of four digits, the first digit is the entrance lane number, and the west, north, east, and south entrance lane numbers are 1, 2, 3, and 4 respectively; the second digit is the detector position number, when When it is set at the stop line, the number is 0, and when it is set at the upstream section, it is numbered 1; the third digit is the traffic flow number, and the numbers of left turn, straight ahead, straight right, and right turn flow are respectively 1, 2, 3, and 4 , the number is 0 regardless of the flow direction; the fourth digit is a sequential number, numbered from the inside to the outside. For example, 1011 is the number of the detector installed at the stop line of the short left-turn lane at the west entrance. S _ij is the distance between the upstream section of intersection i, entrance j and its stop line, starting from the west entrance, numbered 1, 2, 3, 4 in the clockwise direction j. For three-way intersections, if the number of entrances is less than 4, the number of detectors will be reduced; for five-way intersections, if the number of entrances is more than 4, the number of detectors will be increased.

Fig. 4 is a schematic diagram of an intersection signal phase scheme, which consists of eight phases, including two rings and two barriers.

Figure 5 is a schematic diagram of the case intersection and its lane setting. Each entrance road is channelized with 2 special left-turn lanes, 1 through lane, and 1 straight-right mixed lane. Here, U and D are the upper and lower intersections, respectively. D _{U, D} is the distance between the stop line of the west entrance of the upstream and downstream intersections, and S _D1 is the distance between the upstream section of the west entrance of the downstream intersection and its stop line.

Figure 6 is a schematic diagram of the signal phase scheme at the intersection of the case, which is adopted at both intersections U and D.

Detailed ways

The specific implementation manners of the present invention will be further described below in conjunction with the accompanying drawings and technical solutions.

1. Acquisition of traffic flow data

Construct a two-intersection system. The channelization scheme and signal phase scheme of each intersection are shown in Figure 5 and Figure 6 respectively. The length of the entrance road is 50m, and the road gradient is 0. Among them, the distance between the stop line of the west entrance road at the upper and lower intersections is It is 600m. For intersections U and D, the west, north, east and south entrances are numbered 1, 2, 3 and 4 respectively, and the west, north, east and south exits are numbered 1, 2, 3 and 4 respectively; , east, and south entrances, the left-turn, straight-going, and right-turn traffic flow numbers are 11, 12, 13, 21, 22, 23, 31, 32, 33, 41, 42, and 43 respectively. Assume that the saturated flow rates of the left-turn lane, the through lane, and the right-turn lane are 1810, 1850, and 1810 pcu/h, respectively. Table 1 shows the preset traffic flow of each intersection entrance in the system, assuming that its composition is 80% small cars, 10% medium cars, and 10% large cars. Table 2 shows the proportion of each traffic flow at each intersection entrance lane. Taking the estimation of the traffic demand of each type of vehicle in the traffic flows of the west entrance of the intersection D as an example, the specific implementation manner of the present invention is introduced.

Table 1 Preset traffic flow of each intersection entrance road

Table 2 The proportion of each traffic flow at each intersection entrance lane

2. Signal timing scheme

Table 3 shows the public cycle duration, phase display green light time, and intersection phase difference of the signal timing scheme. The full red time and yellow light time of each phase are 2s and 3s, respectively.

Table 3 Signal Timing Scheme

3. Calibration of discrete coefficient

Set the analysis period to T=3600s, and w takes the values 1, 2, and 3 in turn, representing small cars, medium-sized cars, and large cars, respectively. For the traffic flow data shown in Tables 1 and 2 and the signal timing scheme shown in Table 3, the microscopic traffic simulation software VISSIM is used to simulate the traffic flow operation status. In the simulation model, the expected speed of small cars, medium cars, and large cars is 50km/h, and D _{U, D} is 600m. In order to make S _D1 greater than the maximum queuing length at the west entrance of intersection D during peak hours, β _UD is taken as 2/3, and S _D1 = 400m. According to the specific conditions of the road section, in the case Take 0.05. Assuming that the simulation time is 4500s, the data sampling interval Take 5s, traffic flow estimation interval Take 60s, The data acquisition period is 840-4500 s. Table 4 is the relationship matrix of intersection U and D. For the west entrance of intersection D, use the data collection tool to obtain the traffic flow data set, and then obtain the correction coefficient according to the parameter calibration process

Table 4 Relationship matrix of intersection U and D

4. Traffic demand estimation

For the west entrance of intersection D, based on the above correction coefficients, the traffic demand of various vehicles in each traffic flow within the time period of 900-4500s can be estimated according to formula (10). Here, the arrival numbers of small cars, medium-sized cars, and large cars that turn left, go straight, and turn right at the stop line are estimated at intervals of 1 min. The results are shown in Table 5.

Table 5 Traffic Demand of West Entrance Road at Intersection D

Claims (1)

1. a kind of signalized intersections transport need evaluation method that traffic flow composition influences, first stops in signalized intersections entrance driveway The wagon detector for obtaining each stock traffic flow data is respectively set at line and its upstream section, and demarcates fleet based on data acquired Then coefficient of dispersion estimates the group transport need of intersection track, which is characterized in that be as follows again：

(1) data set definition

A is the relational matrix of any two intersection i and i ' in road network；Α=(a_ii′)_I×I, i, i ' ∈ { 1,2 ..., I }, I are to hand over Prong number；When intersection i ' is adjacent with i, a_ii′=1, otherwise, a_ii′=0；

B_ii′For the relational matrix of intersection i entrance driveway j and intersection i ' exit ramps k '；j∈{1,2,…, J_i, k ' ∈ { 1,2 ..., K_i′, J_iFor intersection i entrance driveway numbers, K_i′For intersection i ' exit ramp numbers；As intersection i entrance driveway j When belonging to same a road section with intersection i ' exit ramps k ', b_jk′=1, otherwise, b_jk′=0；

C_iFor the relational matrix of intersection i inner outlets road k and wagon flow m；k∈{1,2,…,K_i, m ∈ 1, 2,…,M_i, K_i、M_iRespectively the exit ramp number of intersection i, wagon flow number；When wagon flow m is when exit ramp k has right-of-way, c_km=1, Otherwise, c_km=0；

For the distance between intersection i ' entrance driveway j ' stop lines and the intersection upstreams i entrance driveway j section, j ' ∈ 1, 2,…,J_i′, J_i′For intersection i ' entrance driveway numbers；

For the distance between intersection i ' entrance driveway j ' stop lines and intersection i entrance driveway j stop lines；

Desired speed when i is driven towards by intersection i ' for w classes vehicle；

To be travelled from intersection i ' entrance driveway j ' stop lines to the expectation of the w class vehicles of the intersection upstreams i entrance driveway j section Time,

For from intersection i ' entrance driveway j ' stop lines to the expectation of the w class vehicles of intersection i entrance driveway j stop lines when driving Between,

A is integer variable,W ∈ { 1,2 ..., W }, W are type of vehicle number, ceil To round up, max is to be maximized；

T is analysis phase duration；

For period (t, t+T), to pass through stop line in intersection i wagon flow m w class vehicle all sampling intervals Actual vehicle number set；

For period (t-a, t), to pass through stop line in intersection i wagon flow m w class vehicle all sampling intervals Actual vehicle number set；

For for period (t-a, t+T), intersection i wagon flow m w class vehicle all sampling intervals are interior by stopping The only set of the actual vehicle number of line；

For for period (t-a, t+T), intersectionLead in each stock wagon flow w class vehicle all sampling intervals Cross the set of the actual vehicle number of stop line；For intersection i entrance driveway j, work as a_ii′=1 and b_jk′When=1, noteAt this time IntersectionFor the upstream intersection of entrance driveway j；

For period (t, t+T), to pass through upstream in intersection i entrance driveway j w class vehicle all sampling intervals The set of the actual vehicle number of section；

For period (t-a, t), to pass through upstream in intersection i entrance driveway j w class vehicle all sampling intervals The set of the actual vehicle number of section；

For period (t-a, t+T), to pass through in intersection i entrance driveway j w class vehicle all sampling intervals The set of the actual vehicle number of upstream section；

(2) Loop detector layout

Lay wagon detector in the entrance driveway stop line of each intersection every and its upstream section, stop line and its upstream section it Between distance be more than peak period where entrance driveway maximum queue length；According toWithIt is calculatedAnd then it obtains A values；It is obtained in the per share wagon flow of the section all kinds of vehicles all sampling intervals using wagon detector at stop line and sails out of vehicle number, That is data setIt is obtained between all samplings of section various types of vehicles using entrance driveway upstream section detector Every interior arrival vehicle number, i.e. data set

In formula：S_ijFor the distance between intersection i entrance driveway j upstream sections and its stop line；β_ii′For distance coefficient, β_ii′∈ (0,1)；D_i′,iTo be directed to the entrance driveway stop line spacing of intersection i entrance driveway j, intersection i ' and i；

(3) data processing

(3.1) detector data at stop line

Using formula (2) to data setIt is handled, obtains data setWithUsing formula (3) to data setIt is handled, is obtainedUtilize formula (4) logarithm According to collectionIt is handled, obtains data set

In formula：For intersection i wagon flows m, w class vehicles are logical in period z Cross the actual vehicle number of stop line；For data sampling interval；For magnitude of traffic flow estimated intervals,ForIntegral multiple；M_i′For The wagon flow number of intersection i '；It is adopted for period (t-a, t+T) interior intersection i ' wagon flows m ' w class vehicles are all Pass through the set of the actual vehicle number of stop line in sample interval；c_k′m′For the relationship mark of intersection i ' inner outlets road k ' and wagon flow m ' Know symbol, when wagon flow m ' is when exit ramp k ' has right-of-way, c_k′m′=1, otherwise, c_k′m′=0；

(3.2) upstream section detector data

Using formula (5) to data setIt is handled, obtains data setWithUsing formula (6) to data setIt is handled, is obtained

In formula：For the intersection i entrance driveway j actual vehicle numbers that w classes vehicle passes through upstream section in period z；

(4) coefficient of dispersion is demarcated

To estimate transport need, need to demarcate the correction factor α in transport need appraising model^w, specific scaling method is as follows：

The vehicle number for passing through upstream section using w classes vehicle in formula (7) estimation period zIt is interval, α with 0.01^wIt is taken from 0 To 1, seek to makeWithAverage relative error absolute value delta_ijMinimum correction factorThen, it utilizes The coefficient and formula (9) obtain correction factor

In formula：For intersection i ' wagon flows M ' is in the periodThe actual vehicle number that interior w classes vehicle passes through stop line；Respectively intersection i The entrance driveway j estimation vehicle numbers that w classes vehicle passes through upstream section in period z, z-1；To stop from each entrance driveway of intersection i ' Only line to the w class vehicles of the intersection upstreams i entrance driveway j section coefficient of dispersion；α^wFor the correction factor of w class vehicles, α^w∈[0, 1]；For repairing from each entrance driveway stop lines of intersection i ' to the w class vehicle running times of the intersection upstreams i entrance driveway j section Positive coefficient,For from each entrance driveway stop lines of intersection i ' to the w class vehicle running times of intersection i entrance driveway j stop lines Correction factor,Hop count when Z is；For regulation coefficient,

(5) transport need appraising model

Utilize correction factorThe transport need of the per share wagon flows of intersection i entrance driveway j is estimated with formula (10)

In formula：For intersection i entrance driveway j wagon flows m Pass through the estimation vehicle number of stop line in period z；Intersection i entrance driveway j wagon flows m w classes in period z-1 The estimation vehicle number that vehicle passes through stop line；It is intersection i ' wagon flow m ' in the periodInterior w class vehicles pass through The actual vehicle number of stop line；χ_ijmFor wagon flow m ratios shared in all wagon flows of intersection i entrance driveway j；For from intersection Mouthful each entrance driveway stop lines of i ' to the w class vehicles of intersection i entrance driveway j stop lines coefficient of dispersion.