CN101707000B - Urban road traffic multiobjective optimization control method - Google Patents

Abstract

The invention belongs to the technical field of urban road traffic control, particularly to an urban road traffic multiobjective optimization control method. The method comprises the following technical schemes of: establishing a multiobjective optimization control model, optimizing the cycle and the split green ratio by adopting a cyclic discrete variable positive and negative stepwise bidirectional adjustment method, adding constraint conditions in the multiobjective optimization control model and solving the multiobjective optimization control model by adopting a genetic algorithm. The invention realizes traffic control, not only satisfies requirement for traffic stream rapid evacuation, but also meets requirement for traffic stream equilibrium distribution, and improves the service level of road intersections.

Description

Urban road traffic multiobjective optimization control method

Technical field

The invention belongs to the urban road transportation control technical field, relate in particular to a kind of urban road traffic multiobjective optimization control method.

Background technology

Along with the development in city and the increase of vehicle, carry out effective traffic control to guarantee the unimpeded of traffic, become the major issue that vehicle supervision department faces day by day.Traffic flow control plays an important role to the traffic capacity, the mitigation urban traffic congestion of improving urban road, its important content is unimpeded, orderly, the effective control measure of safety employing to urban road traffic flow, improves the service efficiency of road equipment to greatest extent.

Along with transport need continue to increase and to smoothness and balanced requirement, traditional traffic control target such as intersection delay, queue length and parking rate etc. can not satisfy the needs of urban road transportation control, in order to improve the traffic circulation situation, improve smoothness and the harmony at traffic networking, must set up urban road traffic multiobjective optimization control method, and in the multiple-objection optimization process, according to the real road traffic environment, give smoothness and the different weight of harmonious distribution.

Smoothness is that quantitative description traffic trip person is in participating in traffic, because a kind of index of the prolongation hourage that road and environmental baseline, traffic interference and factors such as traffic administration and control cause.Being contemplated to be of this index can be as far as possible give one of traffic participant unimpeded, traffic trip environment efficiently, satisfy the requirement of traffic flow rapid evacuation, can selectively make certain direction traffic flow more unimpeded simultaneously.For cross junction, the smoothness index is embodied by the weighting expectation value of the track mean delay of all directions.

Harmony be the quantitative description traffic flow in the index of the balanced intensity of all directions, its meaning is to make traffic control method satisfy the requirement of traffic flow equiblibrium mass distribution, this has embodied the fair management of traffic control scheme to crossing all directions traffic flow.Harmonious index is embodied by the standard variance of the average queue length of all directions.

Mostly traditional crossing control target is to incur loss through delay with total vehicle of crossing all directions is that target is calculated, and uses this target to have following problem:

1, the traffic control scheme of certain optimization has realized the target that total vehicle delay reduces, and this may be to be incured loss through delay significantly to reduce by the vehicle of certain minor direction of crossing to cause, but the vehicle of other one or more crossings directions delay has simultaneously increased;

2, each phase place queue length of crossing is very unbalanced, and the less importer of queue length is to distributing too much green time, and the long importer of queue length is to lacking effective green time.

At the problems referred to above, realize a kind of urban road traffic multiobjective optimization control method, make traffic control method satisfy traffic flow rapid evacuation and traffic flow equiblibrium mass distribution, become the key of the service level that improves the crossing.

Summary of the invention

The objective of the invention is to, a kind of urban road traffic multiobjective optimization control method is provided, the control target is smoothness and balanced weighted sum of crossing, in order to overcome the problem that tradition control target exists in traffic control and management.

Technical scheme of the present invention is that a kind of urban road traffic multiobjective optimization control method is characterized in that described method comprises the following steps:

Step 1: set up the multiobjective optimal control model

$\mathrm{Obj}:Z=k1\mathrm{CT}+k2\mathrm{JH}=k1\underset{i=1}{\overset{M}{\mathrm{\Σ}}}{\mathrm{\α}}_i{\stackrel{\‾}{T}}_i+k2\sqrt{\frac{{(\frac{\underset{i=1}{\overset{M}{\mathrm{\Σ}}}{\stackrel{\‾}{P}}_i}{M}-{\stackrel{\‾}{P}}_i)}^2}{M}};$ Wherein, k1 represents that weight, the k2 of smoothness represent balanced weight, α _iPriority weight, the CT of expression phase place i represent that smoothness desired value, JH represent harmonious desired value, T _iThe delay time at stop of expression phase place i, P _iWhat represent is the queue length of phase place i, and M represents the phase place sum of crossing.

Step 2: adopt the method for the positive and negative progressively two-way adjustment of period discrete variable that cycle and split are optimized;

Step 3: in the multiobjective optimal control model, add constraint condition;

Step 4: adopt genetic algorithm to the multiobjective optimal control model solution.

It is that the cycle forward is adjusted from C that the method for the positive and negative progressively two-way adjustment of described employing period discrete variable is optimized specifically cycle and split _MinBeginning progressively is incremented to C _Max, calculate minimum total delay time T _Cmin1(x); Periodic reverse is adjusted from C _MaxBeginning progressively is decremented to C _Min, calculate minimum total delay time T _Cmin2(x); Total delay time T with described two minimums _Cmin1(x) and T _Cmin2(x) compare, cycle and the split at the total delay time place that numerical value is little are optimal period and split.

Described constraint condition is:

x _i-a _i≥0

$\underset{i=1}{\overset{n}{\mathrm{\Σ}}}(x_i+\mathrm{\Δ}{t}_i)=C$

$0.75\≤\frac{y_{i}C}{x_i}\≤1.2$

C _min≤C≤C _maxf

v＝75km/h

1≤i≤N；

Wherein, a _iIt is the minimum effective green time of i phase place; Δ t _iIt is the i phase loss time.

Described step 4 specifically comprises the following steps:

Step 41: to the model initialization, set harmonious index and smoothness index weight, population number, chromosome length, the total algebraically of iteration, copy, hybridize the variation probability;

Step 42: adopt binary coding, in feasible zone, produce the chromosome of population number size at random;

Step 43: calculate the population adaptive value, and by the ordering of adaptive value size;

Step 44: copy and have adaptive value high dyeing body to of future generation by copying probability;

Step 45: except duplicated chromosome, satisfy population number purpose chromosome to a new generation by crossover probability, the generation of variation probability;

Step 46: judged whether the total algebraically of iteration, if do not have, then gone to step 43;

Step 47: calculate each phase place timing by the optimal-adaptive value;

Step 48: the weight of adjusting harmonious index and smoothness index; Go to step 41.

Described crossover probability is:

$P_c=\left\{\begin{array}{cc}{P}_{c1}-\frac{(P_{c1}-P_{c2})(f^{\&prime;}-f_{\mathrm{avg}})}{f_{\max }-f_{\mathrm{avg}}},& f\&GreaterEqual;f_{\mathrm{avg}}\\ P_{c1},& f<f_{\mathrm{avg}}\end{array}\right.;$

Wherein, f _MaxBe fitness maximum in the colony; f _AvgAverage fitness value for per generation colony; F ' is bigger fitness value in two individualities that will intersect; F is the individual fitness value that will make a variation; P _C1=0.9, P _C2=0.6.

Described variation probability is:

$P_m=\left\{\begin{array}{cc}{P}_{m1}-\frac{(P_{m1}-P_{m2})(f_{\max }-f_{\mathrm{avg}})}{f_{\max }-f_{\mathrm{avg}}},& f\&GreaterEqual;f_{\mathrm{avg}}\\ P_{m1},& f<f_{\mathrm{avg}}\end{array}\right.;$

Wherein, f _MaxBe fitness maximum in the colony; f _AvgAverage fitness value for per generation colony; F ' is bigger fitness value in two individualities that will intersect; F is the individual fitness value that will make a variation; P _M1=0.1, P _M2=0.001.

The present invention has realized that traffic control both satisfied the requirement of traffic flow rapid evacuation, satisfies the requirement of traffic flow equiblibrium mass distribution again, has improved the service level of intersection.

Description of drawings

Fig. 1 is the urban road traffic multiobjective optimization control method process flow diagram;

Fig. 2 is multiobjective optimal control model solution process flow diagram;

Fig. 3 be the embodiment of the invention provide with SCOOT schemes synthesis index synoptic diagram relatively.

Embodiment

Below in conjunction with accompanying drawing, preferred embodiment is elaborated.Should be emphasized that following explanation only is exemplary, rather than in order to limit the scope of the invention and to use.

Fig. 1 is the urban road traffic multiobjective optimization control method process flow diagram.Among Fig. 1, implementation process of the present invention is:

Step 1: set up the multiobjective optimal control model.

Multiobjective optimal control model of the present invention is:

$\mathrm{Obj}:Z=k1\mathrm{CT}+k2\mathrm{JH}=k1\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{\mathrm{\&alpha;}}_i{\stackrel{\&OverBar;}{T}}_i+k2\sqrt{\frac{{(\frac{\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{\stackrel{\&OverBar;}{P}}_i}{M}-{\stackrel{\&OverBar;}{P}}_i)}^2}{M}}$

Wherein, k1 represents that weight, the k2 of smoothness represent balanced weight, α _iPriority weight, the CT of expression phase place i represent that smoothness desired value, JH represent harmonious desired value, T _iThe delay time at stop of expression phase place i, P _iWhat represent is the queue length of phase place i, and M represents the phase place sum of crossing.Power can be manually composed in the weight configuration, also can carry out self study, configurable multiple goal in the optimized process of the value of pursuing one's goal.

Step 2: adopt the method for the positive and negative progressively two-way adjustment of period discrete variable that cycle and split are optimized.

After the selected model, need to seek signal period duration C and the effective green time g of an optimum _eWhat generally adopt at present is earlier by the approximate C that tries to achieve of the computing formula weber optimal period duration calculation formula of optimal period duration C, increase cycle duration during again according to C≤60N, can improve the principle of the traffic capacity, adjust C, be that independent variable is optimized it at last with the effective green time.Obviously the method can not guarantee to obtain making the C of target function value minimum, and the applicability of this method is limited, and when saturation degree is increased to 1 gradually, the computing formula of the total delay of vehicle will be no longer suitable, cause result of calculation incorrect.

In order to overcome above drawback, the present invention adopts the method to the positive and negative progressively two-way adjustment of period discrete variable simultaneously cycle and split to be optimized.The cycle forward is adjusted from X _MinBeginning progressively is incremented to C _Max, calculate minimum total delay time T _Cmin1(x); Oppositely method of adjustment is identical with the forward method of adjustment, from C _MaxBeginning progressively is decremented to C _Min, calculate minimum total delay T _Cmin2(x).Two minimum total delay values are compared the optimal period and the split that are model than cycle and the split at fractional value place, i.e. best timing scheme.

Step 3: in the multiobjective optimal control model, add constraint condition.

Consider the physical constraint condition of signal controlling, constraint condition be expressed as follows:

x _i-a _i≥0

$\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}(x_i+\mathrm{\&Delta;}{t}_i)=C$

$0.75\&le;\frac{y_{i}C}{x_i}\&le;1.2$

C _min≤C≤C _maxf

v＝75km/h

1≤i≤N

In the formula: a _iIt is the minimum effective green time of i phase place; Δ t _iIt is the i phase loss time; Constraint condition 3 is the saturation degree (ratios of flow and the entrance driveway traffic capacity to the crossing, the traffic capacity equals saturation volume and this, and to flow to the split of place phase place long-pending) restriction, 0.75 be used for avoiding too small in saturation degree, be that the traffic capacity is during much larger than transport need, meaningless increase vehicle is incured loss through delay and stop frequency, 1.2 be used for avoiding saturation degree excessive and cause crowdedly, its value is variable.Constraint condition 4 is the constraints to the signal period, and the formula that generally adopts is C at present _Min=20N, C _Max=60N, N are the phase place sum.

Step 4: adopt genetic algorithm to the multiobjective optimal control model solution.

Because intersection timing problem is a stochastic process, adopts traditional algorithm often to be difficult to obtain satisfied optimum solution, the present invention adopts genetic algorithm to carry out model solution.Fig. 2 is multiobjective optimal control model solution process flow diagram.Among Fig. 2, multiobjective optimal control model solution process is:

Step 41: to the model initialization, set harmonious index and smoothness index weight, population number, chromosome length, the total algebraically of iteration, copy, hybridize the variation probability.

Step 42: adopt binary coding, in feasible zone, produce the chromosome of population number size at random.

In the embodiments of the invention, the phase place of selecting for use is 4.With binary code numeric string [d _N/4..., d ₁, c _N/3... c ₁, b _N/3..., b ₁, a _N/3..., a ₁] chromosome of expression, n is chromosome length.Binary code numeric string [d _N/4..., d ₁], [c _N/3... c ₁], [b _N/3..., b ₁], [a _N/3..., a ₁] the timing g of corresponding phase 1, phase place 2, phase place 3, phase place 4 respectively ₁, g ₂, g ₃, g ₄, and satisfy relational expression:

$g_1=e+(C-4e)\underset{i=1}{\overset{n/4}{\mathrm{\&Sigma;}}}\frac{a_i{2}^i}{2^{n/4+1}-1}$

$g_2=e+(C-4e)\underset{i=1}{\overset{n/4}{\mathrm{\&Sigma;}}}\frac{b_i{2}^i}{2^{n/4+1}-1}$

$g_3=e+(C-4e)\underset{i=1}{\overset{n/4}{\mathrm{\&Sigma;}}}\frac{c_i{2}^i}{2^{n/4+1}-1}$

$g_4=e+(C-4e)\underset{i=1}{\overset{n/4}{\mathrm{\&Sigma;}}}\frac{d_i{2}^i}{2^{n/4+1}-1}$

E is phase transition lost time in the formula.

Step 43: calculate the population adaptive value, and by the ordering of adaptive value size.

Step 44: copy and have adaptive value high dyeing body to of future generation by copying probability.

Step 45: except duplicated chromosome, satisfy population number purpose chromosome to a new generation by crossover probability, the generation of variation probability.

Generate initial population and through the part kind group time that intersects, the mutation operator generation is new, must consider the constraint condition of model.The circulation interlace operation of selecting for use Smith to equal to propose in 1987 intersects, and the execution that circulation intersects recombinates each progeny population with the feature of father population as a reference under constraint condition, and the variation strategy selects for use equity to vary one's tactics.

Crossover probability P _cWith the variation probability P _mBe the key point that influences genetic algorithm behavior and performance, directly influence convergence.Adopt self-adapted genetic algorithm to determine that formula is as follows:

$P_c=\left\{\begin{array}{cc}{P}_{c1}-\frac{(P_{c1}-P_{c2})(f^{\&prime;}-f_{\mathrm{avg}})}{f_{\max }-f_{\mathrm{avg}}},& f\&GreaterEqual;f_{\mathrm{avg}}\\ P_{c1},& f<f_{\mathrm{avg}}\end{array}\right.P_m=\left\{\begin{array}{cc}{P}_{m1}-\frac{(P_{m1}-P_{m2})(f_{\max }-f_{\mathrm{avg}})}{f_{\max }-f_{\mathrm{avg}}},& f\&GreaterEqual;f_{\mathrm{avg}}\\ P_{m1},& f<f_{\mathrm{avg}}\end{array}\right.$

In the formula, f _MaxBe fitness maximum in the colony; f _AvgAverage fitness value for per generation colony; F ' is bigger fitness value in two individualities that will intersect; F is the individual fitness value that will make a variation; P _C1=0.9, P _C2=0.6, P _M1=0.1, P _M2=0.001.

Step 46: judged whether the total algebraically of iteration, if do not have, then gone to step 43;

Step 47: calculate each phase place timing by the optimal-adaptive value;

Step 48: the weight of adjusting harmonious index and smoothness index; Go to step 41.

Fig. 3 be the embodiment of the invention provide with SCOOT schemes synthesis index synoptic diagram relatively.Among Fig. 3, embodiment provided by the invention is that object calculates with crossroad, Beijing.The harmonious index in this crossing and the weight of smoothness index and the priority weight value of each phase place are:

k1＝0.7，k2＝0.3

α ₁＝0.1，α ₂＝0.2，α ₃＝0.2，α ₄＝0.5

In the formula, k1 represents harmonious index, and k2 represents the smoothness index.α ₁α ₂α ₃α ₄Represent eastern import respectively, southing mouth, western import, northing mouth phase place.

Bring model into, through calculating the timing result be: g ₁=42, g ₂=51, g ₃=49, g ₄=48, the s of unit (second).

According to shown in Figure 3, control method smoothness in this paper and harmony have had certain improvement.Smoothness numerical value is more low, and the traffic flow delay time at stop of expression crossing is more few; Harmonious index value is more little, and the traffic flow of expression crossing all directions more is tending towards balanced.Overall target is reflection crossing smoothness and balanced unified index, at equal weight k ₁, k ₂Under the condition, the more low expression of comprehensive index value crossing smoothness and harmony are more good.

The present invention has realized that traffic control both satisfied the requirement of traffic flow rapid evacuation, satisfies the requirement of traffic flow equiblibrium mass distribution again, has improved the service level of intersection.

The above; only for the preferable embodiment of the present invention, but protection scope of the present invention is not limited thereto, and anyly is familiar with those skilled in the art in the technical scope that the present invention discloses; the variation that can expect easily or replacement all should be encompassed within protection scope of the present invention.Therefore, protection scope of the present invention should be as the criterion with the protection domain of claim.

Claims (1)

1. a urban road traffic multiobjective optimization control method is characterized in that described method comprises the following steps:

Step 1: set up the multiobjective optimal control model

Obj： $Z=k1\mathrm{CT}+k2\mathrm{JH}=k1\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{\mathrm{\&alpha;}}_i{\stackrel{\&OverBar;}{T}}_i+k2\sqrt{\frac{{(\frac{\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{\stackrel{\&OverBar;}{P}}_i}{M}-{\stackrel{\&OverBar;}{P}}_i)}^2}{M}};$ Wherein, k1 represents that weight, the k2 of smoothness represent balanced weight, α _iPriority weight, the CT of expression phase place i represent that smoothness desired value, JH represent harmonious desired value, T _iThe delay time at stop of expression phase place i, P _iWhat represent is the queue length of phase place i, and M represents the phase place sum of crossing;

Step 2: adopt the method for the positive and negative progressively two-way adjustment of period discrete variable that cycle and split are optimized; Specifically be that the cycle forward is adjusted from C _MinBeginning progressively is incremented to C _Max, calculate minimum total delay time T _Cmin1(x); Periodic reverse is adjusted from C _MaxBeginning progressively is decremented to C _Min, calculate minimum total delay time T _Cmin2(x); Total delay time T with described two minimums _Cmin1(x) and T _Cmin2(x) compare, cycle and the split at the total delay time place that numerical value is little are optimal period and split;

Step 3: in the multiobjective optimal control model, add constraint condition;

Step 4: adopt genetic algorithm to the multiobjective optimal control model solution; Comprise the following steps:

Step 41: to the model initialization, set harmonious index and smoothness index weight, population number, chromosome length, the total algebraically of iteration, copy probability and hybridization variation probability;

Step 42: adopt binary coding, in feasible zone, produce the chromosome of population number size at random;

Step 43: calculate the population adaptive value, and by the ordering of adaptive value size;

Step 44: copy and have adaptive value high dyeing body to of future generation by copying probability;

Step 45: except duplicated chromosome, satisfy population number purpose chromosome to a new generation by crossover probability and the generation of hybridization variation probability;

Step 46: judged whether the total algebraically of iteration, if do not have, then gone to step 43;

Step 47: calculate each phase place timing by the optimal-adaptive value;

Step 48: the weight of adjusting harmonious index and smoothness index; Go to step 41.

Images