CN102938203B - Basic traffic flow parameter based automatic identification method for traffic congestion states - Google Patents

Abstract

The invention discloses a basic traffic flow parameter based automatic identification method for traffic congestion states. The method comprises the steps of determining traffic state characteristic threshold values and judging real-time traffic states, wherein firstly, characteristic threshold values of free-flowing, congestion and blockage are determined through normal fitting of the traffic flow and the occupancy phase angle tangent value, and self correction of threshold values can be conducted according to changes of traffic states; and secondly, real-time traffic states are judged through calculation of the distribution of fitted values of real-time traffic flow phase angle characteristic values according to continuous twice real-time traffic volumes. The method has the advantages that three important basic parameters of the traffic flow, such as the flow, the speed and the occupancy, serve as evidences for judgment of traffic states, the problem that traffic states of city roads at rush hours are difficult to judge correctly, and traffic conditions comprise the free-flowing state, the congestion state and the blockage state. By the aid of the method, the reaction speed for emergencies can be accelerated, and suggestions on selection of traveling routes are provided for travelers.

Description

Road congestion state automatic distinguishing method based on basic traffic flow parameter

Technical field

The present invention relates to a kind of road congestion state automatic distinguishing method based on basic traffic flow parameter, for real time discriminating road traffic state.

Background technology

Correct understanding, objective evaluation urban traffic blocking, define the actual state of traffic congestion, is to analyse in depth traffic congestion, solve the element task of blocking up, and has great importance.

External vehicle supervision department and research institution have carried out a large amount of traffic congestions and have evaluated correlative study, mainly concentrate on USA and Europe Deng state.Wherein, the U.S. has carried out research and the practice of the traffic congestion assessment indicator system of system the earliest, has set up the fairly perfect assessment indicator system of blocking up; Europe Japan waits other developed countries and area aspect the assessment indicator system of blocking up, also to carry out relevant research.

The relevant item of the traffic congestion judgement that recent years is carried out is mainly the evaluation for the overall traffic circulation quality in city, as security department in 2008 and the Ministry of Construction combine, formulated < < urban traffic management assessment indicator system > >, relevant traffic circulation assessment item has also been put into effect in the city such as Beijing-Shanghai voluntarily.But domesticly there is no at present traffic congestion evaluation index research special, system, also do not set up the index system of traffic congestion evaluation.

The domestic research for traffic congestion judgement, general all foundations of judgement using this main traffic behavior parameter of flow as traffic behavior, these class methods have certain effect under the comparatively unobstructed state of road traffic flow.But significantly decline for the traffic behavior judging nicety rate under traffic peak state.In urban road network during in rush-hour, major urban arterial highway flow may the long period in state of saturation, even if obvious variation has occurred traffic behavior, as there is emergency case, cause traffic flow to be stopped up, the variation of flow is also also not obvious.So adopt flow can not reflect exactly the truth of road traffic state as single discriminant parameter.

Summary of the invention

The object of the invention is to overcome the deficiencies in the prior art, a kind of road congestion state automatic distinguishing method based on basic traffic flow parameter is provided, road traffic state judgement more is accurately provided.

According to technical scheme provided by the invention, the described road congestion state automatic distinguishing method based on basic traffic flow parameter comprises the following steps:

One, determining of threshold value:

1.1, in arithmetic for real-time traffic flow historical data base, calculate phase angle [alpha] tangent value size, l is flow, and O is time occupancy, and sample point is arranged according to phase angle [alpha] tangent value ascending order, and the horizontal ordinate of sample point is time occupancy O, and ordinate is flow L; Select whole sample points to carry out normal approach, obtain matching statistic Q ₁; Then, since second phasing degree tangent minimum value, select all sample points, carry out normal approach, obtain matching statistic Q ₂; Obtain by that analogy matching statistic sequence Q _k, the value of matching statistic hour is best traffic behavior separation;

1.2, the phasing degree sequence calculating is divided three classes according to the best traffic behavior separation obtaining in step 1.1: in the sequence of phasing degree, adjacent 2 are all unimpeded state in separation left side, adjacent 2 is all blocked state on separation right side, and adjacent two to put another point in separation left side in order be congestion state on separation right side;

1.3, the minimum value L of flow in the sample point of congestion state in obtaining step 1.2 _min;

Two, the judgement of real-time traffic states:

Obtain real-time traffic flow data, if flow is less than L _min, adopt and only with flow, as the method for parameter, carry out the judgement of traffic behavior, according to uninterrupted, directly differentiate traffic behavior; If flow is not less than L _min, according to the speed of this traffic flow data, flow and time occupancy, calculate its phasing degree size, according to the separation of gained in step 1, judge between this location, phasing degree, judge real-time traffic behavior.

Described in step 1, the update method of threshold value is: according to up-to-date arithmetic for real-time traffic flow historical data, repeating step one, calculates new best traffic behavior separation and L _minthe threshold value that replaces previous step.

In step 1.1, the computing formula of matching statistic is as follows

$Q_k=\frac{n_k\mathrm{\&Sigma;}{(x_i-\stackrel{\&OverBar;}{x_k})}^3}{(n_k-1)(n_k-2)s^3},$

Q wherein _kthe matching statistic that represents k sample sequence, n _kbe the sample size of k sample sequence, the standard deviation that s is sample, be k sample sequence mean value.

Advantage of the present invention is: this method is by three important basic parameters of traffic flow, it is the foundation that flow, speed, occupation rate are judged as traffic behavior, solved urban road peak period traffic behavior and be difficult to the accurately problem of judgement, and traffic behavior has been divided into unimpeded, crowded and three kinds of states of obstruction.Can improve the reaction velocity for accident, and can be traveler provide suggestion on traffic path is selected.

Accompanying drawing explanation

Fig. 1 is the relation curve of flow and occupation rate.

Fig. 2 is congestion state method of discrimination process flow diagram of the present invention.

Embodiment

Below in conjunction with drawings and Examples, the invention will be further described.

At urban highway traffic net, during in rush-hour, the variation of the magnitude of traffic flow is not remarkable.For flow supersaturation, and the traffic hazard of burst, this single judgment mode is difficult to traffic behavior to provide judgement accurately.This method is by three important basic parameters of traffic flow, it is the foundation that flow, speed, occupation rate are judged as traffic behavior, solved urban road peak period traffic behavior and be difficult to the accurately problem of judgement, and traffic behavior has been divided into unimpeded, crowded and three kinds of states of obstruction.Described flow refers in certain hour section by the vehicle fleet of certain section, described speed refers in certain hour section by the average velocity of all vehicles of certain section, described occupation rate refers to that the time occupancy in certain hour section (exists on type wagon detector, in one section of sense cycle, on wagon detector, exist the temporal summation of vehicle and the ratio of sense cycle to be called the time occupancy of this sense cycle, be called for short occupation rate).

The relation of flow L and occupation rate O is as Fig. 1, and if A point and B point are all in flow higher position, flow can not reflect the sudden change of traffic behavior exactly.

According to Fig. 1, can obtain the computing formula of speed:

$V=c\frac{L}{O}$

Wherein, V is wagon flow average velocity, and L is flow, and O is occupation rate, and c is coefficient.According to above formula, can obtain the phasing degree formula in (O, L) plane:

$\tan \mathrm{\&alpha;}=\frac{L}{O}$

Can calculate phase angle [alpha] tangent value according to this formula, and utilize this to be worth to carry out the division of traffic behavior.Phasing degree can approach and reach 90 degree in theory, and now occupation rate is close to 0, and flow is also very little, but speed is very large, is now free traffic flow speed, and traffic is in definitely unobstructed.Phasing degree also can equal 0, and now occupation rate is very large, be tending towards saturated, and flow is close to 0, and traffic is in absolute blocked state.

Variation along with occupation rate, traffic behavior changes between also from unimpeded to crowded, the distribution of traffic flow speed, by the relation of flow and occupation rate, has been embodied in phasing degree, and this distribution has comprised all traffic behavior points (sample point that namely will analyze) below.The velocity distribution of normal traffic states answers Normal Distribution or logarithm to distribute, and the velocity distribution of Traffic Congestion also has this feature.

As shown in Figure 2, the road congestion state automatic distinguishing method that the present invention proposes comprises the following steps.

1, threshold value determines.

(1) in arithmetic for real-time traffic flow historical data base, calculate phase angle [alpha] size, and sample point is arranged according to phase angle [alpha] ascending order, the horizontal ordinate of sample point is occupation rate O, ordinate is flow L; Select whole sample points to carry out normal approach, obtain matching statistic (calculating of statistic is that the degree of bias of sample is calculated in fact); Then, since second phasing degree minimum value, select all sample points, carry out normal approach, obtain matching statistic; Obtain by that analogy matching statistic sequence Q _k, the value of matching statistic hour is best traffic behavior separation.

The computing formula of matching statistic is as follows

$Q_k=\frac{n_k\mathrm{\&Sigma;}{(x_i-\stackrel{\&OverBar;}{x_k})}^3}{(n_k-1)(n_k-2)s^3},$

Q wherein _kthe matching statistic that represents k sample sequence, n _kbe the sample size of k sample sequence, the standard deviation that s is sample, be k sample sequence mean value.

(2) these phasing degree sequences are divided three classes according to calculating the best traffic behavior separation obtaining in step (1), in the sequence of phasing degree, adjacent 2 are all unimpeded state in separation left side, adjacent 2 is all blocked state on separation right side, and adjacent two to put another point in separation left side in order be congestion state on separation right side.

(3) the minimum value L of flow in the sample point of the middle congestion state point of obtaining step (2) _min.

2, the judgement of real-time traffic states.

(1) obtain real-time traffic flow data, if flow is less than L _min, adopt conventional employing flow as the method for parameter, to carry out the judgement of congestion status.According to uninterrupted, directly differentiate traffic behavior.

(2) if flow is not less than L _min, according to the speed of this traffic flow data, flow and occupation rate, calculate its phasing degree size, according to the separation of gained in step 1, judge between this location, phasing degree, judge real-time traffic behavior.

3, the renewal of threshold value.

Because data on flows is being upgraded always, road conditions are also in continuous variation, and according to up-to-date historical data, repeating step 1, calculates new best traffic behavior separation and L _minthe threshold value that replaces previous step.

Claims (3)

1. the road congestion state automatic distinguishing method based on basic traffic flow parameter, is characterized in that, comprises the following steps:

One, determining of threshold value:

1.1, in arithmetic for real-time traffic flow historical data base, calculate phase angle [alpha] tangent value size, l is flow, and O is time occupancy, and sample point is arranged according to phase angle [alpha] tangent value ascending order, and the horizontal ordinate of sample point is time occupancy O, and ordinate is flow L; Select whole sample points to carry out normal approach, obtain matching statistic Q ₁; Then, since second phasing degree tangent minimum value, select all sample points, carry out normal approach, obtain matching statistic Q ₂; Obtain by that analogy matching statistic sequence Q _k, the value of matching statistic hour is best traffic behavior separation;

1.2, the phasing degree sequence calculating is divided three classes according to the best traffic behavior separation obtaining in step 1.1: in the sequence of phasing degree, adjacent 2 are all unimpeded state in separation left side, adjacent 2 is all blocked state on separation right side, and adjacent two to put another point in separation left side in order be congestion state on separation right side;

1.3, the minimum value L of flow in the sample point of congestion state in obtaining step 1.2 _min;

Two, the judgement of real-time traffic states:

Obtain real-time traffic flow data, if flow is less than L _min, adopt and only with flow, as the method for parameter, carry out the judgement of traffic behavior, according to uninterrupted, directly differentiate traffic behavior; If flow is not less than L _min, according to the speed of this traffic flow data, flow and time occupancy, calculate its phasing degree size, according to the separation of gained in step 1, judge between this location, phasing degree, judge real-time traffic behavior.

2. the road congestion state automatic distinguishing method based on basic traffic flow parameter as claimed in claim 1, it is characterized in that, the update method of described threshold value is: according to up-to-date arithmetic for real-time traffic flow historical data, repeating step one, calculates new best traffic behavior separation and L _minthe threshold value that replaces previous step.

3. the road congestion state automatic distinguishing method based on basic traffic flow parameter as claimed in claim 1, is characterized in that, the computing formula of matching statistic is as follows

$Q_k=\frac{n_k\mathrm{\&Sigma;}{(x_i-\stackrel{\&OverBar;}{x_k})}^3}{(n_k-1)(n_k-2)s^3},$

Q wherein _kthe matching statistic that represents k sample sequence, n _kbe the sample size of k sample sequence, the standard deviation that s is sample, be k sample sequence mean value.