CN105405294A - Early warning method of traffic congestion roads - Google Patents

Abstract

The invention relates to an early warning method of traffic congestion roads. The method includes the following steps that: a road congestion index function lambda i with road section speed vi adopted as an independent variable is constructed according to roads of different levels, wherein i represents the i-th road section; the quotient of the number of vehicles Ni which can be accommodated by each road and the level k of each road is adopted as the weight of the corresponding road; the congestion index lambda <-> of a road network is obtained through utilizing the product of the weight and the congestion index of each road; and finally, the congestion situations of each road and the road network are determined, and early warning is carried out. The early warning method has the advantages of simple operation and easiness in realization with a large quantity of data provided.

Description

The method for early warning in congestion in road section

Technical field

The present invention relates to traffic congestion assessment technique field, particularly a kind of method for early warning of congestion in road section.

Background technology

Traffic congestion index is according to road situation, and the unimpeded or conceptual exponential quantity of blocking up of the concentrated expression road net that some cities are arranged, it is equivalent to jam situation digitizing.

Traffic congestion is that the population distribution that causes of economic development imbalance is uneven, urban transportation supply is limited and city layout and economic development such as not to mate at the concentrated reflection of various social contradications, is a global problem.In order to deeply understand the essence of traffic congestion comprehensively, for the every aspect work such as traffic administration, planning, policy appearance provide support and correct guidance Public Traveling, present situation that what practical alleviation was increasingly serious block up, need a set of assessment indicator system that conscientiously can reflect road congestion conditions, wherein just comprise traffic congestion index.

Traffic index carries out deep processing process to the dynamic vehicle positional information being distributed in streets and lanes, city (abbreviation floating car data) to obtain, and is to data processing centre (DPC) in Beijing by the vehicle GPS passback dynamic data on more than 30,000, whole city taxi.First to vehicle position data process, obtain the travelling speed of difference in functionality grade road, then and data on flows different according to function path calculates this road shared weight in the whole network, judges, provide the index desired value being converted to 0-10 finally by people to the perception of blocking up.

But returning dynamic data to data processing centre (DPC) by the vehicle GPS of the taxi more than 30,000 is that cost is too large, and very high to the accuracy requirement of the GPS information of each car and the data of passback, inconvenient operation.

Summary of the invention

Technical matters to be solved by this invention be to provide a kind of for assessment of roadway congestion situation, when data volume is large the method for early warning in the congestion in road section of realization simple and easy to operate.

It is as follows that the present invention solves the problems of the technologies described above adopted technical scheme:

The method for early warning in congestion in road section, comprises the following steps:

S1. at each section layout data harvester, for calculating each section vehicle flowrate at a time and average velocity;

S2. the function of section congestion index is constructed:

λ _i＝α _kln(C _k-v _i)+s(i＝1,2,…,n)

Wherein, subscript k represents that category of roads is numbered, a total m grade; Subscript i represents i-th section; λ _irepresent the congestion index in i-th section, be quantified as the integer of 0 to 10; C _kfor the threshold values that blocks up; α and S is normaliztion constant;

S3. the threshold values C that blocks up in each grade section is obtained _k: by recording the different threshold values C that blocks up in the section of different brackets _k;

S4. the average velocity v in every bar section is calculated:

Computing formula is as follows:

$v=\frac{1}{p}\underset{l=1}{\overset{p}{\&Sigma;}}{\&part;}_1(a-{\left(b\&times;S_l-c\right)}^d)$

Wherein, p represents the quantity of sensor, and by the flow in certain section in the Sl representation unit time, a, b, c, d are the passage situation observing a large amount of real roads, and adopt least square method approximate evaluation to obtain, v is the average speed value in section, for weight, by manually arranging by experience or adopting True Data automatic Fitting method of estimation to obtain;

S5. the congestion index of system-wide net is calculated:

The function that blocked up in the average velocity substitution section, real-time road surface in each section obtains the congestion index λ in this section _i, N _ifor the vehicle number that each section holds, weighting obtains the congestion index of road network:

$\stackrel{\&OverBar;}{\mathrm{\&lambda;}}=\underset{i=1}{\overset{n}{\&Sigma;}}\frac{N_i}{k\left(i\right)}{\mathrm{\&lambda;}}_i$

Wherein, for the traffic congestion index of system-wide net, N _ibe the road accommodation vehicle number in i-th section, k (i) represents the grade of i-th road, λ _irepresent the congestion index of i-th road;

S6. pass through λ _inumerical value different, determine the congestion of each section and system-wide net, and make early warning.

The present invention first according to different stage road construction corresponding with section speed v _ithe section congestion index function lambda that (subscript i represents i-th section) is independent variable _i, then with the vehicle number N that each section can hold _iwith the business of section grade k as the weights in this section, utilize the product of the congestion index in weights and each section, try to achieve the congestion index of this road network finally determine the congestion of each section and system-wide net, and make early warning.Have when data volume is large, the advantage of realization simple and easy to operate.

As preferably, in step S1, data collector is sensor, for collection vehicle speed and vehicle flowrate, arranges 2-5 sensor in each section respectively.Its advantage is, the passage rate of diverse location can be calculated by the sensor of relevant position and obtain, and therefore ensemble average passage rate in section is the result that each segmentation is integrated.Generally, the segmentation that sensor can be observed is thinner, and the data of acquisition are more complete, and model is more accurate, and the degree of accuracy of estimation is higher.

As preferably, in step S4, the citation form of computing formula is: V=a-(b × S-c) ^d; In the average velocity computing formula of l observation station be: V _l=a-(b × S _l-c) ^d, wherein l represents the distance at this point and crossing;

Average velocity for system-wide section is calculated as follows: wherein, L represents the length in this section;

By V _l=a-(b × S _l-c) ^dbring above formula into obtain: $V=\left({\&Integral;}_0^L\right(a-{\left(b\&times;S_l-c\right)}^d)dl)/L;$

Can obtain after above formula is simplified: wherein, p represents this section number of sensors, is 2-5.

Its advantage is, because in sectional type speed-flow traffic flow model, formula 5 exists integral and calculating process, namely needs the road traffic condition knowing each segmentation.But, in actual applications, the speed flowrate situation of change of each position on section can not be obtained, therefore can simplify it, thus make whole process simplification.

As preferably, in step S2, the category of roads criteria for classifying is as following table:

Its advantage is, after specifying the category of roads criteria for classifying, can construct the function of section congestion index more easily.

The present invention compared with the existing technology has the following advantages and effect:

1, because the present invention adopts the mode of sensor image data, cost is low, returns dynamic data without the need to a large amount of vehicle loading GPS, also not high to the equipment requirement of sensor.

2, because the present invention is simple to operate when data volume is large, reduce cost on the one hand, also improve the objectivity of result and the efficiency of whole process on the other hand.

3, due to the model easy to understand that the present invention uses, exploitation is convenient to.

4, the degree of modularity due to the model of the present invention's use is higher, when current demand occurs only need revise one of them part when changing, is easy to expansion multiplexing.

Accompanying drawing explanation

In order to be illustrated more clearly in the embodiment of the present invention or technical scheme of the prior art, be briefly described to the accompanying drawing used required in embodiment or description of the prior art below, apparently, accompanying drawing in the following describes is only some embodiments of the present invention, for those of ordinary skill in the art, under the prerequisite not paying creative work, other accompanying drawing can also be obtained according to these accompanying drawings.

Fig. 1 is calculation process schematic diagram of the present invention.

Embodiment

Below in conjunction with embodiment, the present invention is described in further detail, and following examples are explanation of the invention and the present invention is not limited to following examples.

Embodiment 1: the present embodiment generally speaking, first according to different stage road construction corresponding with section speed v _ithe section congestion index function lambda that (subscript i represents i-th section) is independent variable _i, then with the vehicle number N that each section can hold _iwith the business of section grade k as the weights in this section, utilize the product of the congestion index in weights and each section, try to achieve the congestion index of this road network concrete steps are as follows:

1, the congestion index in each section is calculated

(1) in each section placement sensor, for calculating each section vehicle flowrate at a time and average velocity, 3 sensors are set at least respectively in each section.

(2) functional form of section congestion index is first constructed.

Here consider that traffic threshold values, this section vehicle average velocity at that time in section jam situation and this type of section itself are relevant.So constructed fuction is as follows:

λ _i＝α _kln(C _k-v _i)+s(i＝1,2,…,n)

Wherein, subscript k represents category of roads, a total m grade; Subscript i represents i-th section; λ _irepresent the congestion index in i-th section; C _kfor the threshold values that blocks up (which grade this section belongs to, and just gets the threshold values that blocks up that this grade section is corresponding); α and S is normaliztion constant, λ _ibe quantified as 0 to 10, during by v=C, when λ=0 and v=0, the value of normaliztion constant α and S is determined in λ=10.

(3) the threshold values C that blocks up in each grade section is obtained _k.

There is the different threshold values C that blocks up in the section of different brackets _k, this is relevant with section itself, describes here by speed---when road surface becomes the Road average-speed before blocking up from unimpeded.

2, the average velocity v in every bar section is calculated _i

(1) average velocity in the flow rate calculation section detected according to a section diverse location, final formula is as follows:

$v=\frac{1}{p}\underset{l=1}{\overset{p}{\&Sigma;}}{\&part;}_1(a-{(b\&times;S_l-c)}^d)$

Wherein, represent the weight (section having p sensor) of each sensor on section, obtain by manually arranging by experience or carry out estimation by historical data by least square method.By the flow in certain section in the S representation unit time; A, b, c, d need by observing a large amount of real roads passage situation Least Square Method out.

In detail:

Sectional type speed flow models is mainly used to the relation of simulate day travel phase speed-flow.Multinomial model is often used to describe the stable speed-flow situation of change flowing to the blocked flow stage, and its citation form is:

V＝a-(b×S-c) ^d

Wherein, by the flow in certain section in the s representation unit time, v represents the average overall travel speed of this period.These four parameters of a, b, c, d need the passage situation observing a large amount of real roads, and adopt optimization method approximate evaluation to obtain, common method is least square method.Section, Hangzhou a=66.1, b=-4.5E-2, c=1.6 and d=0.2.

We suppose to appeal the restriction that formula is subject to observation position, and namely different flows, at the diverse location in section, can observe different speed, in the average velocity computing formula of l observation station is so:

V _i＝a-(b×S _i-c) ^d

Wherein, l represents that speed and flow are all the results observed when distance crossing l.For under the average velocity calculating of system-wide section:

$V=\left({\&Integral;}_0^L{V}_{l}dl\right)/L$

Wherein, L represents road section length.The average passage rate that this formula can be understood as section is determined by the passage rate of diverse location in this section (or different segmentation).The passage rate of diverse location can be obtained by the flow rate calculation of relevant position, and therefore ensemble average passage rate in section is the result that each segmentation is integrated.Generally, the segmentation that sensor can be observed is thinner, and the data of acquisition are more complete, and model is more accurate, and the degree of accuracy of estimation is higher.In order to represent the corresponding relation of local flow and road average velocity, by V _i=a-(b × S _i-c) ^dsubstitute into above formula, we can obtain:

$V=\left({\&Integral;}_0^L\right(a-{\left(b\&times;S_l-c\right)}^d)dl)/L$

Above-mentioned formula describes the Road average-speed of the magnitude of traffic flow calculating observed according to diverse location in section.In the model, the flow that Road average-speed obtains according to different observation station, because we are its called after sectional type speed-flow traffic flow model, is used for calculating the average velocity in section.

Because in sectional type speed-flow traffic flow model, appeal last publicity and there is integral and calculating process, namely need the road traffic condition knowing each segmentation.But, in actual applications, the speed flowrate situation of change of each position on section can not be obtained.Therefore, this model can be reduced to:

$v=\left({\mathrm{\&Sigma;}}_{l=1}^{1=p}\frac{{\&part;}_l(a-{\left(b\&times;S_l-c\right)}^d)}{p}\right)$

Have the section of 3 sensors for one, this model can be reduced to:

$V=\left({\mathrm{\&Sigma;}}_{l=1}^{l=3}{\&part;}_l\right(a-{\left(b\&times;S_l-c\right)}^d)/3)$

Simplify and need to consider that each sensors observe point is to the importance of last estimated result, that is need to consider weight manually can arrange by experience, True Data automatic Fitting method of estimation also can be adopted to obtain.

3, the congestion index of system-wide net is calculated

The function that blocked up in the average velocity substitution section, real-time road surface in each section obtains the congestion index λ in this section _i, then according to the vehicle number N that each section holds _iwith the grade of this road, weighting obtains the congestion index of road network:

$\stackrel{\&OverBar;}{\mathrm{\&lambda;}}=\underset{i=1}{\overset{n}{\&Sigma;}}\frac{N_i}{k\left(i\right)}{\mathrm{\&lambda;}}_i$

Wherein, for the traffic congestion index of system-wide net, N _ibe the road accommodation vehicle number in i-th section, k (i) represents the grade of i-th road, λ _irepresent the congestion index of i-th road.

4, pass through λ _inumerical value different, determine the congestion of each section and system-wide net, and make early warning.

The criteria for classifying of above-mentioned category of roads k is as following table:

For example:

(1) the vehicle average velocity of each section in a certain period first, is calculated.For speed computing formula

$V=\left({\mathrm{\&Sigma;}}_{l=1}^{l=3}{\&part;}_l\right(a-{\left(b\&times;S_l-c\right)}^d)/3)$

, first want historical data to estimate parameter a wherein, b, c, d.The section of 3 sensors is installed for one, near its outlet, near near middle and entrance, respectively sensor is housed, need the relation manually marking this observation station history vehicle flow and speed, Least Square Method is utilized to go out parameter a, the value of b, c, d.After specify that the value of unknown parameter, just obtain the formula that calculates Road average-speed, unique unknown quantity is exactly the flow of this section at three sensor places.

The data utilizing sensor to obtain obtain the data on flows in this section, substitute into above-mentioned computing formula and obtain the average velocity of section in this period.

(2) threshold values that blocks up in different brackets section is obtained.The section threshold values that blocks up can Negotiation speed, flow and road surface occupation rate be weighed, here speed is adopted to weigh, draw speed change curves, clearly can observe the catastrophe point (being blocked up from smooth and easy becoming) of speed, this speed is taken as the threshold values that blocks up of this grade road.

(3) section congestion index is calculated.To block up in section function lambda _i=α _kln (C _k-v _i) in+s, λ _ibe quantified as 0 to 10, during by v=C, when λ=0 and v=0, the value of normaliztion constant α and S is determined in λ=10, substitutes into Road average-speed v, calculates the congestion index in this section.The congestion index λ of every bar road is calculated by the method _i(i=1,2 ..., n)

(4) congestion index of system-wide net is calculated.The congestion index in each section is substituted into

$\stackrel{\&OverBar;}{\mathrm{\&lambda;}}=\underset{i=1}{\overset{n}{\&Sigma;}}\frac{N_i}{k\left(i\right)}{\mathrm{\&lambda;}}_i$

In, obtain system-wide net congestion index.

(5) pass through λ _inumerical value different, determine the congestion of each section and system-wide net, and make early warning.

In addition, it should be noted that, the specific embodiment described in this instructions, the shape, institute's title of being named etc. of its parts and components can be different.All equivalences of doing according to structure, feature and the principle described in inventional idea of the present invention or simple change, be included in the protection domain of patent of the present invention.Those skilled in the art can make various amendment or supplement or adopt similar mode to substitute to described specific embodiment; only otherwise depart from structure of the present invention or surmount this scope as defined in the claims, protection scope of the present invention all should be belonged to.

Claims (4)

1. the method for early warning in congestion in road section, is characterized in that, comprises the following steps:

S1. at each section layout data harvester, for calculating each section vehicle flowrate at a time and average velocity;

S2. the function of section congestion index is constructed:

λ _i＝α _kln(C _k-v _i)+s(i＝1,2,…,n)

Wherein, subscript k represents that category of roads is numbered, a total m grade; Subscript i represents i-th section; λ _irepresent the congestion index in i-th section, be quantified as the integer of 0 to 10; C _kfor the threshold values that blocks up; α and S is normaliztion constant;

S3. the threshold values C that blocks up in each grade section is obtained _k: by recording the different threshold values C that blocks up in the section of different brackets _k;

S4. the average velocity v in every bar section is calculated:

Computing formula is as follows:

$v=\frac{1}{p}\underset{l=1}{\overset{p}{\&Sigma;}}{\&part;}_1(a-{\left(b\&times;S_l-c\right)}^d)$

Wherein, p represents the quantity of sensor, S _lby the flow in certain section in the representation unit time, a, b, c, d are the passage situation observing a large amount of real roads, and adopt least square method approximate evaluation to obtain, v is the average speed value in section, for weight;

S5. the congestion index of system-wide net is calculated:

The function that blocked up in the average velocity substitution section, real-time road surface in each section obtains the congestion index λ in this section _i, N _ifor the vehicle number that each section holds, weighting obtains the congestion index of road network:

$\stackrel{\&OverBar;}{\mathrm{\&lambda;}}=\underset{i=1}{\overset{n}{\&Sigma;}}\frac{N_i}{k\left(i\right)}{\mathrm{\&lambda;}}_i$

Wherein, for the traffic congestion index of system-wide net, N _ibe the road accommodation vehicle number in i-th section, k (i) represents the grade of i-th road, λ _irepresent the congestion index of i-th road;

S6. pass through numerical value different, determine the congestion of each section and system-wide net, and make early warning.

2. the method for early warning in congestion in road section according to claim 1, is characterized in that: in described step S1, data collector is sensor, for collection vehicle speed and vehicle flowrate, arranges 2-5 sensor in each section respectively.

3. the method for early warning in congestion in road section according to claim 1, is characterized in that: in described step S4, and the citation form of computing formula is: V=a-(b × S-c) ^d; In the average velocity computing formula of l observation station be: V _l=a-(b × S _l-c) ^d, wherein l represents the distance at this point and crossing;

Average velocity for system-wide section is calculated as follows: wherein, L represents the length in this section;

By V _l=a-(b × S _l-c) ^dbring above formula into obtain:

Can obtain after above formula is simplified: wherein, p represents this section number of sensors, is 2-5.

4. the method for early warning in congestion in road section according to claim 1, it is characterized in that: in described step S2, the category of roads criteria for classifying is as following table: