CN103927890B - A kind of Trunk Road Coordination signal control method based on dynamic O-D Matrix Estimation - Google Patents

Abstract

The invention discloses a kind of Trunk Road Coordination signal control method based on dynamic O-D Matrix Estimation, the method utilizes the link flow that on main line, each crossing turnover stomatodeum detects, Kalman filtering and reverse transmittance nerve network algorithm is adopted to estimate the dynamic O-D matrix at crossing respectively, and design the precision and stability that Bayes's combined method improves estimated result, set up single intersection Multiple Target Signals Controlling model on this basis, and using each crossing signals cycle maximal value of calculating as main line common period.Further design with main line vehicle not the rate of being obstructed be the Trunk Road Coordination signal control method of objective function to the maximum, solve the phase differential between split and adjacent intersection obtaining main line direction, each crossing, thus form Trunk Road Coordination signal timing plan.This method, under the prerequisite ensureing main line vehicle priority pass, takes into account the traffic efficiency of each single intersection, solves prior art cannot adjust control program in real time problem according to volume of traffic change, has the advantage such as high precision, application on site.

Description

A kind of Trunk Road Coordination signal control method based on dynamic O-D Matrix Estimation

Technical field

The invention belongs to technical field of control over intelligent traffic, be specifically related to a kind of Trunk Road Coordination signal control method based on dynamic O-D Matrix Estimation.

Background technology

Real-time whistle control system, as the core of advanced traveler information systems, plays vital effect for alleviation urban traffic blocking.Meanwhile, because the control program of real-time signal control system constantly changes along with the time, the basic data using real-time traffic flow data as signal timing plan is therefore needed.As the elementary cell that signal controls, the real-time signal control at crossing needs dynamic crossing import and export flow and steering flow as input data, and under the condition of existing flow quantity detecting system, crossing is easy to obtain at the link flow at each turnover stomatodeum place, but in real time steering flow then cannot detect and obtains.Crossing dynamic O-D Matrix Estimation model can import and export the time series of flow according to the crossing detected, be back-calculated to obtain the dynamic O-D matrix in crossing, i.e. dynamic steering flow.Along with the development of intelligent transport technology, this model is subject to extensive concern, propose recurrence estimation algorithm (1987), Bell fleet diffusion method (1991), genetic algorithm (2005), Kalman filtering algorithm (2006), backpropagation (Backpropagation, be called for short BP) the dynamic O-D matrix estimation method in crossing such as neural network algorithm (2007), these methods can control provide good shoring of foundation for the signal of single intersection.

In addition; in urban road network; interval between crossing is general all little; especially for the multiple independent crossing on certain main line; when implementing separate signal timing plan; often there will be the situation that fleet does not dissipate completely or fleet always stops because of red light between crossing, cause vehicle queue's phenomenon frequently occur and accumulate, thus the traffic congestion causing main line serious.In order to make the major part on arterial highway even rolling stock pass through smoothly at green time, avoid to the next signal cycle arrive vehicle pass-through impact, need to set up a kind of Arterial Coordination Control method that fleet dissipates of considering.In existing Arterial Coordination Control method, the major parameter of cooperation control has common period, split and phase differential.In order to obtain optimum controling parameters, scholars propose different control methods, as maximum green wave band method with based on incuring loss through delay minimum offset optimization method etc.

The dynamic O-D matrix estimation method in existing crossing and Trunk Road Coordination signal control method also have the following disadvantages:

In the dynamic O-D matrix estimation method in crossing, recurrence estimation algorithm, Bell fleet diffusion method are all derive with linear model and estimate dynamic O-D matrix, be applicable to the estimation of long period through inflow-rate of water turbine smoothing processing, be difficult to the dynamic O-D matrix estimating real time nonlinear change, be unsuitable for application on site; Genetic algorithm is used to solve the Optimized model of the Error Absolute Value sum minimizing observed reading and estimated value in the dynamic O-D in crossing estimates, result is evolved to comprise or close to the state of optimum solution, efficiency is relatively low through iteration; Kalman filtering algorithm recursion nature of that its efficiency is higher but precision is relatively not good enough; BP neural network algorithm carries out training and learning according to historical data, and compare with real data and obtain error, the weights and threshold of network is constantly adjusted by reverse propagated error, make the error sum of squares of network minimum, under stable weights and threshold condition, realize the estimation to current data, but there is training speed be absorbed in the weak points such as local optimum slowly, easily.

In Trunk Road Coordination signal control method, maximum green wave band method and being all widely applied based on methods such as incuring loss through delay minimum offset optimization method, but these methods all have stricter restriction to the arrival rule etc. of the distance between the geometric condition at crossing each on main line, adjacent intersection, each crossing place vehicle, and the signal timing plan obtained is fixing within a certain period of time, real real-time adjustment can not be carried out according to the change of the magnitude of traffic flow.

Summary of the invention

The technical problem to be solved in the present invention is in prior art, and the distinct methods of crossing dynamic O-D Matrix Estimation has different shortcoming, as large in estimated bias, poor stability or efficiency low etc.; Trunk Road Coordination signal control method requires high to road conditions, and the signal timing plan obtained can not carry out real real-time adjustment according to the change of traffic conditions; And then a kind of Trunk Road Coordination signal control method based on dynamic O-D Matrix Estimation is provided, based on the link flow detected value of each crossing turnover stomatodeum, calculate the parameters such as main line common period, each crossing split and adjacent intersection phase differential, the real time coordination signal realizing main line controls, and improves the traffic capacity.

For solving the problems of the technologies described above, the invention provides a kind of Trunk Road Coordination signal control method based on dynamic O-D Matrix Estimation:

This control method comprises based on the main line common period computing method of dynamic O-D Matrix Estimation and not to be obstructed the maximum main line split of rate and phase difference calculating method based on vehicle, two kinds of methods form the Trunk Road Coordination signal control method based on dynamic O-D Matrix Estimation jointly, road segment segment flow is imported and exported according to each crossing that flow detector obtains, the best common period at each crossing of main line, the phase differential between split and adjacent intersection can be calculated, realize Arterial Coordination Control, its key step is as follows:

(1) crossing turnover stomatodeum flow detection: each crossing turnover stomatodeum runs link flow detecting device on main line, detects the road section traffic volume flow obtaining crossing turnover of each period stomatodeum;

(2) estimation of the dynamic O-D matrix in crossing: using the dynamic steering ratio at crossing as independent variable, stomatodeum flow is passed in and out for known quantity to detect the crossing obtained, the crossing dynamic steering ratio estimate model program based on Bayes's weighting is run in far-end computer, solve the dynamic steering ratio at each crossing, obtain the dynamic O-D matrix at crossing further;

(3) determination of main line common period: the algorithm routine running Multiple Target Signals Controlling model in far-end computer, with the dynamic O-D matrix in crossing for known quantity, solve with the Multiple Target Signals Controlling model that vehicle is incured loss through delay and average queue length is minimum, the effective Maximum Traffic Capacity of road is target, obtain the signal timing dial cycle of each crossing optimum, select maximum periodic quantity as main line common period;

(4) calculating of vehicle resolution time: judge whether often pair of next crossing fleet of contiguous crossing dissipated before a upper crossing fleet arrives by detecting device, and be divided into the two kinds of situations that do not dissipate and dissipated, calculate the vehicle resolution time at each crossing respectively;

(5) determination of Arterial Coordination Control scheme: main line common period, each crossing vehicle resolution time are input to and are not obstructed in the maximum main line split of rate and phase difference calculating model based on vehicle, objective function is to the maximum with the not rate of being obstructed of main line, solve split and the phase differential at each crossing, common period, split and phase differential three groups of parameters form Trunk Road Coordination signal timing plan jointly;

(6) controling parameters obtained is transferred to teleseme to implement, realizes Trunk Road Coordination signal and control.

Calculate on the basis of each independent crossing optimal period on main line adopting the single intersection Multiple Target Signals Controlling model based on dynamic O-D Matrix Estimation, select maximum periodic quantity as main line common period C, and set up and not to be obstructed the maximum main line split of rate and phase difference calculating method based on vehicle, solve split and the phase differential of Arterial Coordination Control, ensure that main line direction vehicle passes through smoothly;

Vehicle travels required gap periods number between adjacent intersection:

In formula, l is the distance on main line between adjacent intersection; V is the average velocity that vehicle travels between adjacent intersection; C is main line common period; INT () is bracket function.

The up vehicle sailed in main line section is when arriving crossing, and there will be two kinds of situations: the first situation is belisha beacon is green light, and vehicle directly can pass through crossing, namely not interruptedly directly passes through; The second situation is belisha beacon is amber light or red light, vehicle need waiting signal lamp to become after green light just by, namely vehicle is by being obstructed.

1. the first situation: travelled to time of crossing n by crossing n-1 and be less than the time that the vehicle of crossing n under upper cycle red light accumulation dissipate, vehicle travels and passes through to queuing up during the n of crossing, that is:

$t_n^{\′}\≤\frac{{\∫}_{t_n+{\mathrm{\λ}}_{n}C+(N-1)C}^{t_n+NC}{k}_n{q}_n\left(t\right)dt}{S_n}$

T ' in formula _nfor vehicle travels to time of crossing n by crossing n-1, t _nfor crossing n is relative to the phase differential of crossing n-1, λ _nfor the split of n main line direction, crossing green light phase place, k _nfor the flow regulation coefficient of crossing n, q _nt () is the arrival rate function of crossing n vehicle, S _nfor the traffic capacity in n main line direction, crossing.

Now the resolution time of crossing n vehicle is:

$T_n=\frac{{\∫}_{t_n^{\′}}^{t_n+C}{k}_n{q}_n\left(t\right)dt+{\∫}_{t_n+{\mathrm{\λ}}_{n}C+(N-1)C}^{t_n+NC}{k}_n{q}_n\left(t\right)dt}{S_n}$

What above formula Middle molecule represented is the vehicle number altogether dissipated in the main line direction green light phase place of crossing n, and this part vehicle comprises two parts, and Part I is the queuing vehicle that crossing n of upper cycle does not dissipate, and Part II is the vehicle driving to crossing n from crossing, upstream.

2. the second situation: vehicle travels to time of crossing n from crossing n-1 and is greater than the time that the vehicle of crossing n under upper cycle red light accumulation dissipate, vehicle travel to during the n of crossing without the need to queuing up, directly pass through, that is:

$t_n^{\′}\≤\frac{{\∫}_{t_n+{\mathrm{\λ}}_{n}C}^{t_n+C}{k}_n{q}_n\left(t\right)dt}{S_n}$

Now the resolution time of crossing n vehicle is:

T _n＝0

Definition main line direction through vehicles in the not rate of being obstructed of crossing n is:

$M_n=\frac{Q_n^{\′}}{Q_n}=\frac{{\∫}_{t_n+T_n}^{t_n+{\mathrm{\λ}}_{n}C}{q}_n\left(t\right)dt}{Q_n}$

Q ' in formula _nfor the flow not producing delay because of lamp control, directly pass through in the n-th crossing cycle in main line, Q _nfor the total flow within the n-th crossing cycle.

On main line, rate of the not being obstructed sum at all crossings is maximum as objective function, sets up Trunk Road Coordination signal Controlling model: $\max f(t_n,{\mathrm{\λ}}_n)=\underset{n}{\Σ}{M}_n$

$s.t.\left\{\begin{array}{c}{t}_n\≥0\\ 0\≤{\mathrm{\λ}}_n\≤{\mathrm{\λ}}_{n,max}\end{array}\right.$

λ in formula _{n, max}represent the maximum split of crossing n on main line.

Solve Trunk Road Coordination signal Controlling model, obtain the split λ in main line direction, each crossing _n, and the phase differential t of adjacent intersection _n, in conjunction with main line common period C, Trunk Road Coordination signal timing plan can be obtained.

Technical scheme of the present invention has following beneficial effect relative to prior art:

1. the link flow that obtains according to crossing turnover stomatodeum detecting device of the present invention, Kalman filtering and BP neural network algorithm is utilized to carry out the dynamic O-D Matrix Estimation in crossing, and then through Bayes's weighting algorithm, correction is weighted to two kinds of estimated values, obtain overall more excellent dynamic O-D matrix, avoid estimated result local error excessive, improve the precision and stability of the dynamic O-D Matrix Estimation in crossing.

2. the present invention is directed to single intersection signal to control, with the dynamic O-D Matrix Estimation value at each crossing for initial conditions, devise with the multi-objective nonlinear optimization model that each single intersection is incured loss through delay and queue length is minimum, the effective Maximum Traffic Capacity of road is target, solve the signal period obtaining each crossing, and using maximal value as main line common period, while realizing Arterial Coordination Control, take into account the traffic efficiency of each single intersection.

3. the present invention devise consider that fleet dissipates, be not obstructed with vehicle the minimum Trunk Road Coordination signal Controlling model for target of rate, solve the impact that vehicle queue controls main signal, make wagon flow with green waveshape by each crossing, the traffic efficiency of main line can be provided.

4. the present invention obtain main line common period, each crossing the main signal controling parameters such as split and adjacent intersection phase differential along with the change of traffic conditions real-time change, really achieve real-time Trunk Road Coordination signal to control, and counting yield is high, signal timing plan can be generated in real time, meet precision and the efficiency requirements of application on site.

Accompanying drawing explanation

Fig. 1 is Trunk Road Coordination signal control method (queuing up and do not dissipate in the crossing) structural drawing based on dynamic O-D Matrix Estimation

Fig. 2 is Trunk Road Coordination signal control method (queuing up and dissipate in the crossing) structural drawing based on dynamic O-D Matrix Estimation

Fig. 3 is the Trunk Road Coordination signal control method process flow diagram based on dynamic O-D Matrix Estimation

Embodiment

Each detailed problem involved in technical solution of the present invention is described in detail below in conjunction with accompanying drawing.Be to be noted that described embodiment is only intended to be convenient to the understanding of the present invention, and any restriction effect is not play to it.

When judging that vehicle arrives crossing by the link flow detecting device at each crossing, whether crossing queuing vehicle dissipates, and Trunk Road Coordination signal control method is divided into two kinds of situations: crossing queue up do not dissipate, crossing queues up and dissipates.

When queuing and do not dissipate in crossing, based on dynamic O-D Matrix Estimation Trunk Road Coordination signal control method structural drawing as shown in Figure 1, now the queuing vehicle at crossing place needs the regular hour to dissipate and by crossing, the vehicle of arrival needs to wait in line equally after green light starts.Fig. 1 the first half illustrates relation on main line between each crossing, by being embedded in the link flow detecting device below each crossing turnover stomatodeum, obtaining the import and export road segment segment flow at crossing, as given data, and being transferred to far-end computer.Fig. 1 the latter half illustrates the principle of the Trunk Road Coordination signal control method based on dynamic O-D Matrix Estimation, road segment segment flow is imported and exported at each crossing according to detecting, Bayes's combined method is adopted to estimate the dynamic O-D matrix at each crossing, and be entered in single intersection Multiple Target Signals Controlling model, solve the signal period of each crossing optimum, using maximal value as main line common period; Simultaneously when crossing queuing vehicle does not dissipate, calculate the resolution time of queuing vehicle, it is input to together with main line common period and minimizes in the arterial control model of rate of not being obstructed, solve the split at each crossing and the phase differential of adjacent intersection; Main line common period, main line direction, each crossing split, adjacent intersection phase differential form real-time Trunk Road Coordination signal timing plan jointly.

Queue up when dissipating in crossing, based on dynamic O-D Matrix Estimation Trunk Road Coordination signal control method structural drawing as shown in Figure 2, now the queuing vehicle at crossing place dissipates when upstream vehicle arrives, the vehicle of arrival without the need to wait namely by.The structure of Fig. 2 and Fig. 1 is substantially identical, and unique difference is that crossing queuing vehicle resolution time should be 0.

Based on dynamic O-D Matrix Estimation Trunk Road Coordination signal control method process flow diagram as shown in Figure 3.Whole flow process is made up of following 6 steps: (1) crossing turnover stomatodeum flow detection, the estimation of the dynamic O-D matrix in (2) crossing, the determination of (3) main line common period, the calculating of (4) vehicle resolution time, the determination of (5) Arterial Coordination Control scheme, (6) transfer to the application of whistle control system.Concrete steps comprise:

Step 1: crossing turnover stomatodeum flow detection

Utilize and be arranged on each track flow detector that road segment segment place is imported and exported at crossing, detect the turnover stomatodeum magnitude of traffic flow obtained in time interval k, i.e. Q _i(k), i=1,2 ..., r represents that period k flows into the flow at crossing from entrance driveway i, Y _j(k), j=1,2 ..., s represents that period k flows out the flow at crossing from exit ramp j, judges whether crossing has queuing vehicle simultaneously, and is transferred to far-end computer and processes.

Step 2: the estimation of the dynamic O-D matrix in crossing

According to the time series detecting the crossing import and export road segment segment flow obtained, with the dynamic steering ratio B of period k _ijk (), as state variable, carries out the estimation of crossing dynamic steering ratio, comprise the estimation of historical period and the estimation of present period.

Clearly, dynamic steering ratio in crossing should meet following constraint condition:

①B _ij(k)≥0，i＝1，2，…，r；j＝1，2，…，s

② $\underset{j=1}{\overset{s}{\Σ}}{B}_{ij}\left(k\right)=1,i=1,2,...,r$

First, use Kalman filtering algorithm to estimate dynamic steering ratio, set up state-space model as follows:

State equation: B (k)=B (k-1)+W (k)

Observation equation: Y (k)=Q (k) * B (k)+e (k)

In formula, B (k), Q (k), Y (k) are respectively B _ij(k), Q _i(k), Y _jthe matrix of (k) or vector form, W (k) to be average be 0 white Gaussian noise vector, e (k) to be average be 0 observation white Gaussian noise vector.

Adopt existing order Kalman filtering algorithm to solve, and in correction algorithm flow process, the initial value of dynamic steering ratio is as follows:

L in formula _ijbe the track quantity realizing being turned to by i entrance driveway j exit ramp, for all-purpose road, respectively turn to average value.

To the dynamic steering ratio that order Kalman filtering algorithm calculates, carry out cutting and standardized process, make it meet the intrinsic constraint condition of ratio of turning.

On the basis detecting flow, with the M Programming with Pascal Language of Matlab software, realize Kalman filtering algorithm, export the real-time estimated value of dynamic steering ratio, comprise history estimated value with current estimated value

Then, use BP neural network algorithm to estimate dynamic steering ratio, algorithm flow is as follows:

Design the BP neural network of three layers, comprise input layer, hidden layer and output layer:

Input layer: 3 neurons, the respectively inlet flow rate in each track, corresponding entrance driveway upstream, when track, entrance driveway upstream quantity is different, neuronal quantity does respective change;

Hidden layer: 15 neurons, transport function adopts logarithm S type function, and its output valve, in the interval range of [0,1], is coincide with ratio of turning scope;

Output layer: adopt linear transfer function, have 3 neurons, the ratio of turning in corresponding left-hand rotation, craspedodrome, 3 directions of turning right, has 3 output valves.

For making each neuronic output valve after initial weighting close to zero, ensureing that each neuronic weights can both change maximum part at their S type activation function and regulate, getting initial weight for the random number between (-1,1).

Adopt momentum-adjusting learning rate adjustment algorithm, carry out the weights and threshold in round-off error back-propagation process, make BP neural network algorithm both can find globally optimal solution, can the training time be shortened again.

Utilize the M Programming with Pascal Language of Matlab, realize solving of BP neural network, export the real-time estimated value of dynamic steering ratio, comprising history estimated value with current estimated value

The history estimated value of Kalman filtering and BP neural network is made comparisons with corresponding dynamic steering ratio history actual value, obtain mean absolute percentage error, and then obtain by following formula the probability P r (H selecting Kalman filtering algorithm according to history estimated bias ^kF), select the probability P r (H of BP neural network algorithm ⁿ):

$\mathrm{Pr}\left(H^{KF}\right)=\left\{\begin{array}{cc}1-{\mathrm{EH}}^{KF},& ({\mathrm{EH}}^{KF}<1)\\ 0,& ({\mathrm{EH}}^{KF}\&GreaterEqual;1)\end{array}\right.$

$\mathrm{Pr}\left(H^N\right)=\left\{\begin{array}{cc}1-{\mathrm{EH}}^N,& ({\mathrm{EH}}^N<1)\\ 0,& ({\mathrm{EH}}^N\&GreaterEqual;1)\end{array}\right.$

Wherein, EH ^kF, EH ⁿbe respectively Kalman filtering algorithm, the history estimated value of BP neural network algorithm and the mean absolute percentage error of corresponding history actual value.

The computing method of mean absolute percentage error: wherein for estimated value, B _ijk () is actual value.

Further consider the estimated bias on the same day, in order to improve precision, dynamically update the weight of combined method simultaneously, front 5 period estimated values of employing Kalman filtering and BP neural network algorithm current estimation period and the deviation of corresponding period combined method estimated value are as current estimated bias, thus obtain under the prerequisite of history estimated bias, the probability P r (D|H of Kalman filtering algorithm is selected according to current estimated bias ^kF), select the probability P r (D|H of BP neural network algorithm ⁿ):

$\mathrm{Pr}\left(D\right|H^{KF})=\left\{\begin{array}{cc}1-E^{KF},& (E^{KF}<1)\\ 0,& (E^{KF}\&GreaterEqual;1)\end{array}\right.$

$\mathrm{Pr}\left(D\right|H^N)=\left\{\begin{array}{cc}1-E^N,& (E^N<1)\\ 0,& (E^N\&GreaterEqual;1)\end{array}\right.$

Wherein, E ^kF, E ⁿbe respectively Kalman filtering algorithm, front 5 period estimated values of BP neural network algorithm and the mean absolute percentage error of corresponding Bayes's combinational estimation value.

Solve Bayes's weights W of Kalman filtering algorithm, BP neural network algorithm ^kFand W ⁿ:

Pr(D)＝Pr(D|H ^KF)Pr(H ^KF)+Pr(D|H ^N)Pr(H ^N)

$W^{KF}=\frac{\mathrm{Pr}\left(D\right|H^{KF})\mathrm{Pr}\left(H^{KF}\right)}{\mathrm{Pr}\left(D\right)}$

$W^N=\frac{\mathrm{Pr}\left(D\right|H^N)\mathrm{Pr}\left(H^N\right)}{\mathrm{Pr}\left(D\right)}$

Utilize the dynamic steering ratio that Kalman filtering and BP neural network estimate separately, according to Bayes's weighted formula, the dynamic steering ratio estimate value that present period is final can be obtained:

${\tilde{B}}_{ij}\left(k\right)=W_{ij}^{KF}\left(k\right)\&times;B_{ij}^{KF}\left(k\right)+W_{ij}^N\left(k\right)\&times;B_{ij}^N\left(k\right)$

Using the deviation of Bayes's weighting modified value of present period Kalman filtering and BP neural network estimated result and present period as present period deviation stored in current deviation database, use it for and calculate the next current estimated bias estimating the period, and more new estimation period.

According to the dynamic steering ratio that Bayes's combined method is estimated and the link flow Q of each crossing entrance driveway _ik (), can obtain the dynamic O-D Matrix Estimation value at each crossing.

Step 3: the determination of main line common period

Set up with the signal control cycle at crossing for independent variable, using vehicle incur loss through delay and queue length minimum with the effective Maximum Traffic Capacity of road as the Non-linear Optimal Model of objective function.

The weight coefficient K of definition delay, queue length, the effective traffic capacity of road three evaluation indexes _x ¹, K _x ², K _x ³:

K _x ¹＝2s _xp _x(1-P)；K _x ²＝s _xp _x(1-P)T；K _x ³＝2(3600/T)P；

S in formula _xfor the saturation volume of an xth phase place, p _xfor an xth phase place magnitude of traffic flow and the ratio of saturation volume, P be each phase place magnitude of traffic flow with the ratio of saturation volume and, T is the signal period at crossing.

At weight coefficient K _x ¹, K _x ², K _x ³calculating in, p _xall need crossing dynamic O-D Matrix Estimation value as given data with the acquisition of P.

Single intersection Multiple Target Signals Controlling model is as follows:

$\min f\left(T\right)=\underset{x=1}{\overset{n}{\&Sigma;}}\&lsqb;{K_x}^1{d}_x+{K_x}^2{L}_x-{K_x}^3{Q}_x\&rsqb;$

$s.t.\left\{\begin{array}{c}\underset{x=1}{\overset{n}{\&Sigma;}}(G_x+A_x+R_x)=T\\ 0.9p_x\&le;\frac{G_x}{T}\&le;1.1p_x,1\&le;x\&le;m\\ gree{n}_{x,\min }\&le;G_x\&le;gree{n}_{x,max},1\&le;x\&le;m\\ \underset{x=1}{\overset{n}{\&Sigma;}}{G}_x+L\&le;J,1\&le;x\&le;m\\ G_x\&GreaterEqual;0,1\&le;x\&le;m\end{array}\right.$

In formula:

D _x: the mean delay time that an xth phase place vehicle arrives;

L _x: the average queue length of vehicle of an xth phase place;

Q _x: the effective traffic capacity of road of an xth phase place;

G _x: the effective green time of an xth phase place;

A _x: the yellow time of an xth phase place, be taken as 3 seconds;

R _x: the complete red time of an xth phase place, be taken as 3 seconds;

M: signal phase number;

Green _{x, min}, green _{x, max}: the minimum effective green time of a crossing xth phase place, maximum effective green time, be taken as 15 seconds and 60 seconds respectively;

L: total losses time signal period, be taken as 16 seconds;

L _x: the vehicle launch lost time of an xth phase place, be taken as 3 seconds;

I _x: the copper sulfate basic of an xth phase place, I _x=A _x+ R _x;

J: maximum cycle time, is taken as 180 seconds;

Delay, queue length, the effective traffic capacity of road three Performance Evaluating Indexes are defined as follows:

The vehicles average delay time: vehicle is obstructed the difference of walking required time and without hindrance walking required time at the entrance place of leading the way, crossing, the vehicles average delay time of an xth phase place:

$d_x=\frac{T{(1-G_x/T)}^2}{2(1-p_x)}+\frac{{(1-\frac{\underset{x=1}{\overset{n}{\&Sigma;}}{l}_x}{T})}^2}{2\frac{\underset{x=1}{\overset{n}{\&Sigma;}}{l}_x}{T}}$

The average queue length of vehicle: within a signal period, the mean value of maximum queue length when each bar track green light phase place is initial, the average queue length of vehicle of an xth phase place:

L _x＝2q _xR _x

Q in formula _xrepresent that the vehicle of an xth phase place arrives flow rate, the forms such as Poisson distribution can be taken as according to actual conditions.

The effective traffic capacity of road: within a certain period of time by certain crossing all entrance driveway stop lines vehicle number sum, for signal junction, the effective traffic capacity of road of xth phase place:

Q _x＝λ _xs _x

λ in formula _xrepresent the split of an xth phase place.

The crossing that step 2 is obtained dynamic O-D Input matrix single intersection Multiple Target Signals Controlling model, and adopt Lingo Program, obtain signal timing dial parameter and the evaluation index of single intersection, and using the signal period maximum for each crossing as main line common period C.

Step 4: the calculating of vehicle resolution time

Calculate vehicle between adjacent intersection, travel required gap periods number:

In formula, l is the distance on main line between adjacent intersection; V is the average velocity that vehicle travels between adjacent intersection, can obtain according to floating vehicle system or other existing Vehicle Speed Forecasting Methodologies; INT () is bracket function;

The up vehicle sailed in main line section is when arriving crossing, and there will be two kinds of situations: the first situation is belisha beacon is green light, and vehicle directly can pass through crossing, namely not interruptedly directly passes through; The second situation is belisha beacon is amber light or red light, vehicle need waiting signal lamp to become after green light just by, namely vehicle is by being obstructed.

1. the first situation: travelled to time of crossing n by crossing n-1 and be less than the time that the vehicle of crossing n under upper cycle red light accumulation dissipate, vehicle travels and passes through to queuing up during the n of crossing, that is:

$t_n^{\&prime;}\&le;\frac{{\&Integral;}_{t_n+{\mathrm{\&lambda;}}_{n}C+(N-1)C}^{t_n+NC}{k}_n{q}_n\left(t\right)dt}{S_n}$

T ' in formula _nfor vehicle travels to time of crossing n by crossing n-1, t _nfor crossing n is relative to the phase differential of crossing n-1, λ _nfor the split of n main line direction, crossing green light phase place, k _nfor the flow regulation coefficient of crossing n, S _nfor the traffic capacity in n main line direction, crossing, q _nt () is the arrival rate function of crossing n vehicle, can be taken as the forms such as Poisson distribution according to actual conditions.

Now the resolution time of crossing n vehicle is:

$T_n=\frac{{\&Integral;}_{t_n^{\&prime;}}^{t_n+C}{k}_n{q}_n\left(t\right)dt+{\&Integral;}_{t_n+{\mathrm{\&lambda;}}_{n}C+(N-1)C}^{t_n+NC}{k}_n{q}_n\left(t\right)dt}{S_n}$

What above formula Middle molecule represented is the vehicle number altogether dissipated in the main line direction green light phase place of crossing n, and this part vehicle comprises two parts, and Part I is the queuing vehicle that crossing n of upper cycle does not dissipate, and Part II is the vehicle driving to crossing n from crossing, upstream.

2. the second situation: vehicle travels to time of crossing n from crossing n-1 and is greater than the time that the vehicle of crossing n under upper cycle red light accumulation dissipate, vehicle travel to during the n of crossing without the need to queuing up, directly pass through, that is:

$t_n^{\&prime;}\&le;\frac{{\&Integral;}_{t_n+{\mathrm{\&lambda;}}_{n}C}^{t_n+C}{k}_n{q}_n\left(t\right)dt}{S_n}$

Now the resolution time of crossing n vehicle is:

T _n＝0

Step 5: the determination of Arterial Coordination Control scheme

The vehicle resolution time that main line common period step 3 obtained and step 4 obtain is input to Trunk Road Coordination signal Controlling model:

$\max f(t_n,{\mathrm{\&lambda;}}_n)=\underset{n}{\&Sigma;}{M}_n$

$s.t.\left\{\begin{array}{c}{t}_n\&GreaterEqual;0\\ 0\&le;{\mathrm{\&lambda;}}_n\&le;{\mathrm{\&lambda;}}_{n,max}\end{array}\right.$

λ in formula _{n, max}represent the maximum split of crossing n on main line, the present invention is taken as 0.75.

Adopt the M Programming with Pascal Language of Matlab, solve Trunk Road Coordination signal Controlling model, for crossing n, obtain the split λ of n main line direction, crossing green light phase place _n, crossing n is relative to the phase differential t of crossing n-1 _n.

The phase differential of main line common period, main line direction, each crossing split, adjacent intersection constitutes arterial control scheme jointly.Period is upgraded, real-time Trunk Road Coordination signal timing plan can be obtained.

Step 6: the application transferring to whistle control system

The phase differential of real-time main line common period, each crossing main line direction split and adjacent intersection is transferred to whistle control system, real-time Trunk Road Coordination signal can be realized and control.

The present invention passes through traffic study, by the result of the Trunk Road Coordination signal control method gained in concrete main line case based on dynamic O-D Matrix Estimation, contrast with conventional maximum green wave band method, based on dynamic O-D Matrix Estimation Trunk Road Coordination signal control method can according to time become the magnitude of traffic flow, real-time arterial control scheme is provided, all obvious due to maximum green wave band method in delay, queue length and stop frequency three indexs, there is good effect, under the prerequisite meeting precision and efficiency requirements, the traffic efficiency of main line can be improved.

Specifically describe embodiment of the present invention above, should be appreciated that the people of the common skill one to the art, do not departing from any modification or partial replacement of the scope of the invention, all belong to the scope of claims of the present invention protection.

Claims (1)

1., based on a Trunk Road Coordination signal control method for dynamic O-D Matrix Estimation, it is characterized in that:

This control method comprises based on the main line common period computing method of dynamic O-D Matrix Estimation and not to be obstructed the maximum main line split of rate and phase difference calculating method based on vehicle, road segment segment flow is imported and exported according to each crossing that flow detector obtains, calculate the best common period at each crossing of main line, the phase differential between split and adjacent intersection, realize Arterial Coordination Control, its key step is as follows:

(1) crossing turnover stomatodeum flow detection: each crossing turnover stomatodeum runs link flow detecting device on main line, detects the road section traffic volume flow obtaining crossing turnover of each period stomatodeum;

(2) estimation of the dynamic O-D matrix in crossing: using the dynamic steering ratio at crossing as independent variable, stomatodeum flow is passed in and out for known quantity to detect the crossing obtained, the crossing dynamic steering ratio estimate model program based on Bayes's weighting is run in far-end computer, solve the dynamic steering ratio at each crossing, obtain the dynamic O-D matrix at crossing further;

(3) determination of main line common period: the algorithm routine running Multiple Target Signals Controlling model in far-end computer, with the dynamic O-D matrix in crossing for known quantity, solve with the Multiple Target Signals Controlling model that vehicle is incured loss through delay and average queue length is minimum, the effective Maximum Traffic Capacity of road is target, obtain the signal timing dial cycle of each crossing optimum, select maximum periodic quantity as main line common period;

(4) calculating of vehicle resolution time: judge whether often pair of next crossing fleet of contiguous crossing dissipated before a upper crossing fleet arrives by detecting device, and be divided into the two kinds of situations that do not dissipate and dissipated, calculate the vehicle resolution time at each crossing respectively;

(5) determination of Arterial Coordination Control scheme: main line common period, each crossing vehicle resolution time are input to and are not obstructed in the maximum main line split of rate and phase difference calculating model based on vehicle, objective function is to the maximum with the not rate of being obstructed of main line, solve split and the phase differential at each crossing, common period, split and phase differential three groups of parameters form Trunk Road Coordination signal timing plan jointly;

(6) controling parameters obtained is transferred to teleseme to implement, realizes Trunk Road Coordination signal and control.