CN105427394A - Congestion charging optimal toll rate determining method based on trial-and-error method and motor vehicle flow - Google Patents

Abstract

The invention discloses a congestion charging optimal toll rate determining method based on a trial-and-error method and motor vehicle flow. The optimal toll rate can be obtained through adjustment on the basis of vehicle flow data of each entrance road section in a charging region. For the optimal charging scheme, the congestion charging optimal toll rate determining method adopts a set of strict mathematical proof to establish the ''trial-and-error method'' for toll rate adjustment, so as to ensure that the method can be used for converging the optimal toll rate. The congestion charging optimal toll rate determining method comprises the steps of: (1) evaluating each available congestion charging pattern by utilizing a stochastic user equilibrium theory based on an asymmetric road section travel time equation; (2) establishing a monotonous and continuous variable inequality model for forecasting network equilibrium flow; (3) and determining step length of adjustment and a charging rate value of each step by utilizing a projection algorithm for solving the variable inequality model.

Description

Based on the congestion-pricing optimum toll rate defining method of trial and error and motor vehicle flow

Technical field

The present invention relates to a kind of according to trial and error and area of charge entrance section motor vehicle flow data, determine the method for congestion-pricing Optimal Toll Rate, belong to urban traffic control and control field.

Background technology

Traffic congestion be exactly all the time urban transportation running subject matter, it not only make a lot of driver to urban transportation dejected and also pollute urban environment, increase social cost.Traffic congestion increases hourage, increases the uncertainty of hourage, creates air and noise pollution and reduces social productive forces.Congestion-pricing is considered to urban traffic control person and regulates the traffic demand, an important tool of transfer bus passenger number.Much existing research is absorbed in how to determine optimum congestion-pricing rate.First (not limiting for toll collection location) and second (only to subnetwork charge) best charge is studied as urban congestion charging policy widely as two economic concepts.

The practical application of urban traffic blocking charge is significant concern of the present invention point, the scheme that three famous (Singapore, London, Stockholm) congestion-pricing embodiments existed all adopt warning line to charge.London uses the charge of warning line license, and two other city uses the charge method in warning line entrance section.As everyone knows, the charge method in warning line entrance section is more effective and fair.Because it has two practical character.First, when warning line congestion-pricing comes into effect time, transportation network supvr is more concerned about the traffic in area of charge; Such as, when congestion-pricing within 1975, set up in Singapore by first time time, the target of land transportation administrative authority of Singapore this scheme of original adoption is total magnitude of traffic flow that minimizing 25% to 35% enters toll zone.That is, enter the total magnitude of traffic flow of a certain area of charge and should be limited to a boundary value determined in advance.Secondly, each entrance charging value entering a certain area of charge is identical, so that driver's identification and authorities' management.

Generally speaking, although existing research obtains good achievement in theory, but wherein a lot of determining method for congestion-pricing rate be not suitable for practical application, it is main because they need the precise information of a lot of transportation network attribute, comprise trip requirements equation, the time value of road trip time equation and transportation network user.For whole transportation network, these data are difficult to obtain.

Summary of the invention

Technical matters: the invention provides a kind ofly does not need the congestion-pricing optimum toll rate defining method based on trial and error and motor vehicle flow of mass data for warning line congestion-pricing.

Technical scheme: the congestion-pricing optimum toll rate defining method based on trial and error and motor vehicle flow of the present invention, comprises the steps:

Step one: according to the point of urban traffic network, line and partition data, set up transportation network topological diagram;

Step 2: based on the transportation network topological diagram set up in step one, determine all entrances in area of charge and this region, and each entrance is defined as toll collection location;

Step 3: the total vehicle flowrate boundary value of target entering each area of charge is set, and implement the initial charge price plan of a setting at toll collection location;

Step 4: current tariff price plan implement during this period of time in, observe and record the vehicle flowrate entering each charging section after vehicle flowrate is stablized;

Step 5: the difference between " the total vehicle flowrate boundary value of target " that the vehicle flowrate of each charging section observed by step 4 and step 3 are determined, according to adaptive prediction algorithm for correction, calculates new paying price scheme;

Step 6: calculate the difference between new paying price scheme and old paying price scheme, judges whether this difference is less than decision threshold, in this way, then the new paying price scheme finally obtained is exported as optimal result, otherwise returns step 4.

Further, in the inventive method, transportation network topological diagram comprises the point of urban traffic network, line and partition data, the total vehicle flowrate boundary value of described target be block up according to reducing, the relative ratio of optimizing the environment determines, described initial charge price plan is arbitrary price plan, and described adaptive prediction algorithm for correction comprises prediction and corrects two processes.

Congestion-pricing measure is usually pointed, charge for central business district, city, through street and the city such as Section of Outer Ring Line and the downtown region that easily gets congestion, and these cities have higher private motor vehicles trip proportion usually, have perfect loop or congested area to be easy to divide, the public transport that there is still unsaturated road network or prosperity outside traffic congestion area of charge can bear the huge volume of traffic implementing transfer after congestion-pricing.

In preferred version of the present invention, to suppose in network total I area of charge, determine to enter each area of charge i, i=1,2 according to urban traffic network ..., I, all entrance sections, and to be charged in these entrance sections and toll rate is identical, use τ _irepresent.τ=(τ _i, i=1,2 ... I) ^trepresent the charge in all regions, subscript " T " represents the transposition of vector.H _irepresentative enters flow (immigration total flow) boundary value of predetermined area of charge i.

A represents the section set in network. represent region i, i=1,2 ..., I, the set in all entrance sections. represent the set in all charge entrance sections.If then τ _arepresent the charge on a of section, if section a is not the entrance section of area of charge, then τ _a=0.Different toll project τ can affect the routing of the network user and cause different balanced traffic stream.V _a(τ) the balance road traffic delay of section a ∈ A is represented, T _a(v, τ) represents broad sense road trip time function:

T _a(v，τ)＝t _a(v)+τ _a/α，a∈A(1)

Wherein, α represents the time value of the network user.T _av () represents the asymmetric road trip time function of section a ∈ A, it be a non-negative about link flow vector v, monotone increasing and the function can led continuously, v represents v _a(τ) set.

In the step 3 of preferred version of the present invention, block up to reduce, optimize the environment as target, the total vehicle flowrate boundary value of target entering each area of charge is set, such as reduce region, urban district block up requirements of plan target carriage flow reduce half, then total for target vehicle flowrate boundary value can be set to the half of currency, and implement an arbitrary paying price scheme at toll collection location, and as five yuan, i.e. initial charge price vector

The step 4 of preferred version of the present invention: to current tariff price plan implement week age, current tariff price plan implement during this period of time in, observe and record the vehicle flowrate entering each charging section after vehicle flowrate is stablized.

Acquisition methods and the Congestion Toll mode of motor vehicle flow are closely related.Such as by being laid in the video detector statistics motor vehicle flow of each charge entrance, or the electronic charging label on each motor vehicle of process can be read based on trackside short-range wireless communication technologies, and then statistics motor vehicle flow.

The step 5 of preferred version of the present invention: by the difference between " motor vehicle flow that observation obtains " and " motor vehicle flow boundary value ", utilize following mathematical formulae to calculate, obtain new paying price scheme.

The computation process of new paying price is as follows:

A, structure model

Φ(τ ^*)(τ-τ ^*)≥0，τ∈Ω(2)

Wherein Ω=and τ | τ _i>=0, i=1,2 ... I} is the feasible set of τ, and subscript " * " represents optimum solution, and variable inequality function phi (τ) is by such as giving a definition:

Wherein represent the real number set be made up of I element.

B, calculating variable inequation Φ (τ ⁽ⁿ⁾) difference namely between " observation obtain motor vehicle flow " and " motor vehicle flow boundary value ".

C, to be found auxiliary toll project vector by projection operation, auxiliary toll project is loaded on network, then observes the magnitude of traffic flow in corresponding entrance section, calculate corresponding variable inequation value with the flow value observed subsequently.

D, by further calculating ratio r ⁽ⁿ⁾with step value π ⁽ⁿ⁾, obtain new toll rate scheme.

Concrete solution steps (adaptive prediction algorithm for correction):

Step 1 initialization;

Three constant κ are set ₁, κ ₂, γ, wherein 0 < κ ₂< κ ₁< 1, γ ∈ (0,2), arranges initial step length η ⁽¹⁾> 0.Arranging iteration sequence number is n=1.

Step 2 forecasting process;

Step 2.1: toll project τ is set in a network ⁽ⁿ⁾, then observe the magnitude of traffic flow in each area of charge entrance section, by i=1,2 ... .I represent, then calculate variable inequation

$\mathrm{\&Phi;}\left({\mathrm{\&tau;}}^{\left(n\right)}\right)={(H_i-\underset{a\&Element;{\stackrel{\&OverBar;}{A}}_i}{\&Sigma;}{v}_a\left({\mathrm{\&tau;}}^{\left(n\right)}\right),i=1,2,\mathrm{...},I)}^T---\left(4\right)$

Step 2.2: pass through projection operation

${\stackrel{\&OverBar;}{\mathrm{\&tau;}}}^{\left(n\right)}=P_{\mathrm{\&Omega;}}\&lsqb;{\mathrm{\&tau;}}^{\left(n\right)}-{\mathrm{\&eta;}}^{\left(n\right)}\mathrm{\&Phi;}\left({\mathrm{\&tau;}}^{\left(n\right)}\right)\&rsqb;---\left(5\right)$

Find auxiliary toll project vector wherein P _Ωvectorial τ ' is projected to the projection operation on toll project feasible set Ω by [τ '] expression, and the following formula of its value represents:

P _Ω[τ′]＝(maX(0，τ′ _i)，i＝1，2，...，I) ^T(6)

Step 2.3: will toll project be assisted be loaded on network, then observe the magnitude of traffic flow in corresponding entrance section i=1,2 ... I, calculates corresponding variable inequation value with the flow value of observation subsequently

$\mathrm{\&Phi;}\left({\stackrel{\&OverBar;}{\mathrm{\&tau;}}}^{\left(n\right)}\right)={(H_i-\underset{a\&Element;{\stackrel{\&OverBar;}{A}}_i}{\&Sigma;}{v}_a\left({\stackrel{\&OverBar;}{\mathrm{\&tau;}}}^{\left(n\right)}\right),i=1,2,\mathrm{...},I)}^T---\left(7\right)$

Step 2.4: calculate ratio r by equation below ⁽ⁿ⁾,

$r^{\left(n\right)}={\mathrm{\&eta;}}^{\left(n\right)}\left|\right|\mathrm{\&Phi;}\left({\mathrm{\&tau;}}^{\left(n\right)}\right)-\mathrm{\&Phi;}\left({\stackrel{\&OverBar;}{\mathrm{\&tau;}}}^{\left(n\right)}\right)|{|}_2/||{\mathrm{\&tau;}}^{\left(n\right)}-{\stackrel{\&OverBar;}{\mathrm{\&tau;}}}^{\left(n\right)}|{|}_2---\left(8\right)$

If r ⁽ⁿ⁾≤ κ ₁, carry out step 3, otherwise according to

${\mathrm{\&eta;}}^{\left(n\right)}=\frac{2}{3}{\mathrm{\&eta;}}^{\left(n\right)}\min \{1,\frac{1}{r^{\left(n\right)}}\}---\left(9\right)$

Reduce step value, then carry out step 2.2.

Step 3 correction process;

Based on τ ⁽ⁿ⁾, η ⁽ⁿ⁾, for correction process calculates a step value π ⁽ⁿ⁾, then obtain the toll project vector τ upgraded ⁽ⁿ⁺¹⁾.

Step 3.1: according to the step value π of the correction process of equation calculating below ⁽ⁿ⁾:

${\mathrm{\&pi;}}^{\left(n\right)}=\mathrm{\&gamma;}\&times;{\mathrm{\&eta;}}^{\left(n\right)}\&times;\frac{\underset{i=1}{\overset{I}{\&Sigma;}}\&lsqb;({\mathrm{\&tau;}}_i^{\left(n\right)}-{\stackrel{\&OverBar;}{\mathrm{\&tau;}}}_i^{\left(n\right)})\&times;h_i^{\left(n\right)}\&rsqb;}{\underset{i=1}{\overset{I}{\&Sigma;}}{\left(h_i^{\left(n\right)}\right)}^2}---\left(10\right)$

Wherein

$h_i^{\left(n\right)}=({\mathrm{\&tau;}}_i^{\left(n\right)}-{\stackrel{\&OverBar;}{\mathrm{\&tau;}}}_i^{\left(n\right)})-{\mathrm{\&eta;}}^{\left(n\right)}\left({\mathrm{\&Phi;}}_i\right({\mathrm{\&tau;}}^{\left(n\right)})-{\mathrm{\&Phi;}}_i({\stackrel{\&OverBar;}{\mathrm{\&tau;}}}^{\left(n\right)}\left)\right),i=1,2,...,I---\left(11\right)$

Step 3.2: upgrade toll project vector τ by following projection operation ⁽ⁿ⁺¹⁾:

${\mathrm{\&tau;}}^{(n+1)}=P_{\mathrm{\&Omega;}}\&lsqb;{\mathrm{\&tau;}}^{\left(n\right)}-{\mathrm{\&pi;}}^{\left(n\right)}\mathrm{\&Phi;}\left({\stackrel{\&OverBar;}{\mathrm{\&tau;}}}^{\left(n\right)}\right)\&rsqb;---\left(12\right)$

Step 3.3: judge whether following condition is set up:

${\mathrm{\&eta;}}^{\left(n\right)}\left|\right|\mathrm{\&Phi;}\left({\mathrm{\&tau;}}^{\left(n\right)}\right)-\mathrm{\&Phi;}\left({\stackrel{\&OverBar;}{\mathrm{\&tau;}}}^{\left(n\right)}\right)|{|}_2/||{\mathrm{\&tau;}}^{\left(n\right)}-{\stackrel{\&OverBar;}{\mathrm{\&tau;}}}^{\left(n\right)}|{|}_2\&le;{\mathrm{\&kappa;}}_2---\left(13\right)$

If above formula is set up, then according to the step value of scheme increase below η ⁽ⁿ⁾, then carry out step 6:

${\mathrm{\&eta;}}^{(n+1)}=\frac{3}{2}{\mathrm{\&eta;}}^{\left(n\right)}---\left(14\right)$

Otherwise, directly carry out step 6.

Step 6: calculate the difference between new paying price scheme and old paying price scheme, judges whether this difference is less than decision threshold, in this way, then the new paying price scheme finally obtained is exported as optimal result, otherwise returns step 4.

If then stop, ε is a positive limit value set in advance.Otherwise, make n=n+1, carry out step 4.

Beneficial effect: the present invention compared with prior art, has the following advantages:

Although existing research obtains good achievement in theory, but wherein a lot of determining method for congestion-pricing rate be not suitable for practical application, it is main because they need the precise information of a lot of transportation network attribute, comprise trip requirements equation, the time value of road trip time equation and transportation network user.For whole transportation network, these data are difficult to obtain.Therefore, the present invention proposes a kind ofly to avoid the decision-making technique of the toll rate using the aforementioned data that those not easily obtain for warning line congestion-pricing, consider warning line congestion-pricing two practical character simultaneously, namely it is identical for entering each entrance charging value that the total magnitude of traffic flow of a certain area of charge should be limited to a boundary value determined in advance and enter a certain area of charge, so that driver's identification and authorities' management.From proposed trial and error, only need the motor vehicle flow data on charging section, and these data are easy to obtain from charge station or vehicle detection coil.

The inventive method is workable, and the vehicle flowrate data that only need record Congestion Toll area entry section automatically can regulate and obtain the Optimal Toll Rate meeting and reduce targets such as blocking up, optimize the environment.

Accompanying drawing explanation

Fig. 1 is the topology diagram of road network.

Fig. 2 is optimum toll rate defining method process flow diagram.

Embodiment

Below in conjunction with example and Figure of description, the present invention is further illustrated.

This part content illustrates the present invention in conjunction with the embodiments further, these embodiments should be understood only be not used in for illustration of the present invention and limit the scope of the invention, after having read the present invention, the amendment that those skilled in the art tackle the various equivalent form of value of the present invention has all fallen within the application's claims limited range.

Step one: input target cities transportation network related data (point, line and subregion) thus obtain transportation network topological diagram.

Fig. 1 is the network structure of embodiment, the terminus that table 1 gives present networks to and trip requirements.This example comprises 7 summits, 11 sections and an area of charge.Hourage, equation used BPR type equation,

$t_a\left(v_a\right)=t_a^0(1.0+0.15{\left(\frac{v_a}{C_a}\right)}^4,a\&Element;A)---\left(15\right)$

The transport need of table 1 origin and destination

Origin and destination pair	Transport need (vehicle/hour)
		1→7	6000
2→7	5000
		3→7	5000
6→7	4000

Wherein represent and freely flow hourage, C _arepresent section tolerance limit.Consider section effect of confluxing, such as, the flow on section 1 and 7 is flowed on section 3, and the magnitude of traffic flow therefore on section 1 can affect the hourage in section 7 simultaneously.Two are had to section of confluxing: section 1 and 7, section 2 and 6 in this example.Therefore, for these two pairs of sections, their equation hourage adopts as Types Below:

$t_a\left(v\right)=t_a^0(1+0.15\&times;{\left(\frac{v_a+0.5{\hat{v}}_{\hat{a}}}{1.5C_a}\right)}^4),a\&Element;A---\left(16\right)$

represent the flow on section a ∈ A in paired section.Such as, for section 1, represent the flow on section 7.Such equation result in asymmetrical road trip time function.Provide in table 2 and C _aoccurrence.

Table 2 road trip time function parameter

Step 2: based on all entrances in the transportation network figure determination area of charge in step one and this region, and each entrance is being defined as toll collection location.

This area of charge is by its summit Isosorbide-5-Nitrae, and 5,7 define.Charge in three area entry sections 5,6,7.Optimum charging value on all area entry sections is identical.

Step 3: setting enters target total motor vehicle traffic limit value of each area of charge (to reduce targets such as blocking up, optimize the environment, determine total motor vehicle traffic limit value), the present embodiment adopts three kinds of boundary values as a comparison, be respectively 6000 vehicles/hour, 5000 vehicles/hour and 4000 vehicles/hour, and implement an initial charge price 5 yuan at toll collection location, i.e. initial charge price vector

Step 4: to current tariff price plan implement week age, current tariff price plan implement during this period of time in, observe and record the vehicle flowrate entering each charging section after vehicle flowrate is stablized.

A, for initial charge scheme, need to observe the link flow in each entrance section corresponding to it.In the present embodiment, we separate a Stochastic User Equilibrium problem based on probit, and use the balance link flow in gained entrance section to estimate these motor vehicle flow data.Here, we suppose that the user equilibrium immediately that the trip behavior of the network user is deferred to based on probit is theoretical, and the user equilibrium immediately that the methodology simultaneously proposed also is applicable to other types is theoretical.We suppose that the time value of the network user is 0.6 yuan/minute simultaneously.In practical application, the time value of user is unwanted in trial and error.

B, we use straight average method to separate this Stochastic User Equilibrium problem based on probit in an embodiment.Wherein, use Monte Carlo simulation to estimate random network loading procedure, namely the first step carries out initialization, arranges iteration count l=1; Second step samples, from T _a~ N (t _a, β t _a) in be each section a's sampling; 3rd step has entirely completely without distribution, namely based on obtaining distribute { q _wto connecting on the right shortest path in each origin and destination, q _wrepresent the travelling demand of origin and destination to w ∈ W.This step creates road traffic delay duration set it is average that 4th step carries out flow, makes 5th step carries out end condition detection, makes ${\mathrm{\&sigma;}}_a^{\left(l\right)}=\sqrt{\frac{1}{l(l-1)}\underset{m=1}{\overset{l}{\&Sigma;}}{\&lsqb;V_a^{\left(m\right)}-v_a^{\left(l\right)}\&rsqb;}^2},\&ForAll;a,$ If ${\max }_a\{{\mathrm{\&sigma;}}_a^{\left(l\right)}/v_a^{\left(l\right)}\}\&le;\mathrm{\&kappa;},$ Then stop, solution is otherwise, make l=l+1, carry out second step.

C, in order to ensure accuracy, we use 1000 simulation processes, in each Monte Carlo simulation process, have three tasks: the first, sample by the cognitive error term of pseudo random number to normal distribution; The second, search each origin and destination between shortest path; Finally, by all origin and destination demand assignment to search shortest path on.Because the cognition mistake for hourage is normally based on path definition, in order to avoid enumerating path, we are by cognitive error definition on each section, and namely the road trip time of the broad sense that user is cognitive equals:

wherein T _arepresent the road trip time function of broad sense, and suppose the cognitive error ξ of road trip time _abe average be 0, variance is the normal distribution of constant.

Step 5: by the difference between " motor vehicle flow that observation obtains " and " motor vehicle flow boundary value ", according to adaptive prediction algorithm for correction, obtain new paying price scheme.

Step 1 initialization;

κ ₁＝0.9，κ ₂＝0.1，γ＝1.8，η ⁽⁰⁾＝1.0。And to arrange iteration sequence number be n=1.

Step 2 forecasting process;

Step 2.1: toll project τ is set in a network ⁽ⁿ⁾, then observe the magnitude of traffic flow in each area of charge entrance section, by v _a(τ ⁽ⁿ⁾), i=1,2 ... I represents, then calculates variable inequation $\mathrm{\&Phi;}\left({\mathrm{\&tau;}}^{\left(n\right)}\right)={(H_i-\underset{a\&Element;{\stackrel{\&OverBar;}{A}}_i}{\&Sigma;}{v}_a\left({\mathrm{\&tau;}}^{\left(n\right)}\right),i=1,2,\mathrm{...},I)}^T,,$ Namely the difference between " motor vehicle flow that observation obtains " and " motor vehicle flow boundary value ".

Step 2.2: pass through projection operation find auxiliary toll project vector wherein projection operation P _Ω[τ ']=(maX (0, τ ' _i), i=1,2 ..., I) ^t, represent and vectorial τ ' projected on the feasible set Ω of toll project.

Step 2.3: will toll project be assisted be loaded on network, then observe the magnitude of traffic flow in corresponding entrance section i=1,2 ... I, calculates corresponding variable inequation value with the flow value of observation subsequently $\mathrm{\&Phi;}\left({\stackrel{\&OverBar;}{\mathrm{\&tau;}}}^{\left(n\right)}\right)={(H_i-\underset{a\&Element;{\stackrel{\&OverBar;}{A}}_i}{\&Sigma;}{v}_a\left({\stackrel{\&OverBar;}{\mathrm{\&tau;}}}^{\left(n\right)}\right),i=1,2,\mathrm{...},I)}^T.$

Step 2.4: calculate ratio r by equation below ⁽ⁿ⁾,

$r^{\left(n\right)}={\mathrm{\&eta;}}^{\left(n\right)}\left|\right|\mathrm{\&Phi;}\left({\mathrm{\&tau;}}^{\left(n\right)}\right)-\mathrm{\&Phi;}\left({\stackrel{\&OverBar;}{\mathrm{\&tau;}}}^{\left(n\right)}\right)|{|}_2/||{\mathrm{\&tau;}}^{\left(n\right)}-{\stackrel{\&OverBar;}{\mathrm{\&tau;}}}^{\left(n\right)}|{|}_2$

If r ⁽ⁿ⁾≤ κ ₁, carry out step 3, otherwise according to reduce step value, then carry out step 2.2.

Step 3 correction process;

Based on τ ⁽ⁿ⁾, η ⁽ⁿ⁾, for correction process calculates a step value π ⁽ⁿ⁾, then obtain the toll project vector τ upgraded ⁽ⁿ⁺¹⁾.

Step 3.1: according to the another one of equation calculating below step value π ⁽ⁿ⁾

${\mathrm{\&pi;}}^{\left(n\right)}=\mathrm{\&gamma;}\&times;{\mathrm{\&eta;}}^{\left(n\right)}\&times;\frac{\underset{i=1}{\overset{I}{\&Sigma;}}\&lsqb;({\mathrm{\&tau;}}_i^{\left(n\right)}-{\stackrel{\&OverBar;}{\mathrm{\&tau;}}}_i^{\left(n\right)})\&times;h_i^{\left(n\right)}\&rsqb;}{\underset{i=1}{\overset{I}{\&Sigma;}}{\left(h_i^{\left(n\right)}\right)}^2}$

Wherein $h_i^{\left(n\right)}=({\mathrm{\&tau;}}_i^{\left(n\right)}-{\stackrel{\&OverBar;}{\mathrm{\&tau;}}}_i^{\left(n\right)})-{\mathrm{\&eta;}}^{\left(n\right)}\left({\mathrm{\&Phi;}}_i\right({\mathrm{\&tau;}}^{\left(n\right)})-{\mathrm{\&Phi;}}_i({\stackrel{\&OverBar;}{\mathrm{\&tau;}}}^{\left(n\right)}\left)\right),$ i＝1，2，...，I

Step 3.2: upgrade toll project vector τ by following projection operation ⁽ⁿ⁺¹⁾

${\mathrm{\&tau;}}^{(n+1)}=P_{\mathrm{\&Omega;}}\&lsqb;{\mathrm{\&tau;}}^{\left(n\right)}-{\mathrm{\&pi;}}^{\left(n\right)}\mathrm{\&Phi;}\left({\stackrel{\&OverBar;}{\mathrm{\&tau;}}}^{\left(n\right)}\right)\&rsqb;$

Step 3.3: Rule of judgment ${\mathrm{\&eta;}}^{\left(n\right)}\left|\right|\mathrm{\&Phi;}\left({\mathrm{\&tau;}}^{\left(n\right)}\right)-\mathrm{\&Phi;}\left({\stackrel{\&OverBar;}{\mathrm{\&tau;}}}^{\left(n\right)}\right)|{|}_2/||{\mathrm{\&tau;}}^{\left(n\right)}-{\stackrel{\&OverBar;}{\mathrm{\&tau;}}}^{\left(n\right)}|{|}_2\&le;{\mathrm{\&kappa;}}_2$ Whether set up, if set up, then carry out step 6 again; Otherwise, directly carry out step 6.

Step 6: by the difference between new paying price scheme and old paying price scheme, judge whether termination algorithm, and export optimal result.If do not stopped, then returning step 4, carrying out loop iteration.

If namely then stop, ε=1 × 10 ^-4.Otherwise, make n=n+1, carry out step 4.

Solution of the present invention is based on the trial and error benefit evaluation of the congestion-pricing Optimal Toll Rate of motor vehicle flow data:

Be keep the traffic in area of charge based on the object of the congestion-pricing of motor vehicle flow data, and this target is not more than certain boundary value set in advance by the inbound traffic flow that restriction enters area of charge and realize.The optimum toll project under three kinds of different situations is provided in table 3.Wherein end condition adopts 1 × 10 ^-4, parameter value uses κ ₁=0.9, κ ₂=0.1, γ=1.8, η ⁽⁰⁾=1.0.The third line in table 3 provides the optimum payment collector case value of output, fourth line shows, total inbound traffic stream of area of charge equals predetermined threshold value, implied that its solution meets aforesaid three mathematic condition, namely obtained toll project successfully can limit total inbound traffic stream and be not more than predetermined boundary value.From situation 1 to 3, along with the predetermined threshold value that setting is less, that is require better traffic in area of charge, the charging value of area of charge becomes increasing.This also illustrates that the network user needs to pay and morely reaches and keep better traffic.

Three kinds of schemes of table 3 boundary value and optimum charging value

Above-described embodiment is only the preferred embodiment of the present invention; be noted that for those skilled in the art; under the premise without departing from the principles of the invention; some improvement and equivalent replacement can also be made; these improve the claims in the present invention and are equal to the technical scheme after replacing, and all fall into protection scope of the present invention.

Claims (2)

1., based on a congestion-pricing optimum toll rate defining method for trial and error and motor vehicle flow, it is characterized in that, the method comprises the following steps:

Step one: according to the point of urban traffic network, line and partition data, set up transportation network topological diagram;

Step 2: based on the transportation network topological diagram set up in step one, determine all entrances in area of charge and this region, and each entrance is defined as toll collection location;

Step 3: the total vehicle flowrate boundary value of target entering each area of charge is set, and implement the initial charge price plan of a setting at toll collection location;

Step 4: current tariff price plan implement during this period of time in, observe and record the vehicle flowrate entering each charging section after vehicle flowrate is stablized;

Step 5: the difference between " the total vehicle flowrate boundary value of target " that the vehicle flowrate of each charging section observed by step 4 and step 3 are determined, according to adaptive prediction algorithm for correction, calculates new paying price scheme;

Step 6: calculate the difference between new paying price scheme and old paying price scheme, judges whether this difference is less than decision threshold, in this way, then the new paying price scheme finally obtained is exported as optimal result, otherwise returns step 4.

2. the congestion-pricing optimum toll rate defining method based on trial and error and motor vehicle flow according to claim 1, it is characterized in that, described transportation network topological diagram comprises the point of urban traffic network, line and partition data, the total vehicle flowrate boundary value of described target be block up according to reducing, the relative ratio of optimizing the environment determines, described initial charge price plan is arbitrary price plan, and described adaptive prediction algorithm for correction comprises prediction and corrects two processes.