CN101604479B - Method for evaluating service level of plane signal intersection under mixed traffic environment - Google Patents

Abstract

The invention discloses a method for evaluating service level of a plane signal intersection under a mixed traffic environment in the field of urban road traffic planning and management. The method comprises the following steps: determining evaluation indexes of the service level of the plane signal intersection under the mixed traffic environment; calculating the weight of the plane signal intersection under the mixed traffic environment of each index by using an analytic hierarchy process; calculating central points and standard differences of five service levels of the plane signal intersection under the mixed traffic environment by using a fuzzy clustering method, and solving a membership function of each evaluation index; acquiring index data of each index under the mixed traffic environment; calculating and controlling the membership degree of each level of each index to form a single-factor evaluation matrix; and using the single-factor evaluation matrix and the index weight to carry out integrated evaluation for each index under each level, and determining the service level according to the maximum membership rule. The method integrally considers the influences of various traffic factors comprising non-motor vehicle and pedestrian on the plane signal intersection, and evaluates the service level of the plane signal intersection more accurately.

Description

The evaluation method of service level of plane signal intersection under mixed traffic environment

Technical field

The invention belongs to urban highway traffic planning and management domain, relate in particular to a kind of evaluation method of service level of plane signal intersection under mixed traffic environment.

Background technology

Since eighties of last century the fifties; The traffic engineering researchist of more external developed countries just is devoted to tolerance of seeking and method and the model of estimating signalized intersections service quality always, and the effort of passing through decades has formed relatively independent theory and method.The most representative with U.S.'s " HCM " (HighwayCapacity Manual is abbreviated as HCM), influence is also maximum, and some countries and regions are carried out service level research based on this.The signalized intersections service level that new edition HCM2000 proposed is on average controlled to incur loss through delay with per car and is index and it is divided into six grades of A-F; And confirmed that control incurs loss through delay computing method.

With abroad compare, China is starting late aspect the research of signalized intersections service level, but has also obtained certain achievement.The signalized intersections service level evaluation index of advising in " traffic engineering " book of appointing Feitian etc. to write comprises traffic loading coefficient, efficiency factor, crossing be obstructed vehicle, delay time at stop and queue length, and is five grades with this with level of service division; The suggestion of Beijing municipal administration designing institute is I-IV level through on average hold up traffic length and traffic user's the index that is felt as of crossing mean delay time, red light with level of service division with load factor of intersection, vehicle; Xiao Qianjin etc. estimate employing saturation degree, efficiency index, all delays of car, safety coefficient and indexs such as traffic administration and environmental baseline to the crossing in Chongqing City's urban road reorganization and expansion technological project; Money cold wind etc. has also been accomplished research in this respect, with the per car road traffic capacity, saturation degree, delay and twice parking rate as the crossing evaluation index and provided the grade scale of A-E Pyatyi.

Above achievement in research has been enriched the theory and the method for crossing service level evaluation; But not enough in conjunction with the research of Chinese mixed traffic characteristic; The crossing service level evaluation index and the index calculating method that are proposed, all with single motor-driven wagon flow as considering object.Therefore, existing service level evaluation method is not suitable for the traffic flow situation of China's high mixed.The service level of being estimated out can't correctly reflect objective reality.Influence shows the following aspects to the crossing service level in mixed traffic:

1, pedestrian, bicycle are to the interference of motor vehicle operation

For the planed signal crossing, in the same green light signals cycle, there are the pedestrian, bicycle of different directions and the mutual interference mutually of motor vehicle, cause the service condition of motor vehicle, user's impression to change, and then service level is exerted an influence.And domestic and international crossing service level evaluation system does not all reflect the index of this respect content, and this phenomenon can't obtain embodying in service level;

2, to the influence of evaluation index

In the crossing service level evaluation index of domestic proposition, the more indexs such as saturation degree, efficiency factor that have appear.Owing to there be the phase mutual interference of mixed traffic in the crossing, the traffic capacity loss with causing the crossing reduces the passage rate of motor vehicle in the crossing, and then can influence saturation degree, efficiency factor; And current calculating to saturation degree, efficiency factor, being mostly of employing is external based on the computing method under the single automobile traffic environment, thereby can't calculate the desired value under the mixed traffic environment effectively.The computing method of these indexs need to study again.

Summary of the invention

The objective of the invention is to; In the evaluation method to existing service level of plane signal intersection under mixed traffic environment; Fail the deficiency of non-disturbance regime of consideration machine and domestic mixed traffic environment; Thereby cause the irrational problem of evaluation result, propose a kind of method of service level of plane signal intersection under mixed traffic environment.

Technical scheme of the present invention is, a kind of evaluation method of service level of plane signal intersection under mixed traffic environment is characterized in that, said evaluation method comprises the following steps:

Step 1: confirm the service level of plane signal intersection under mixed traffic environment evaluation index, comprising: control delay, jamming rate, saturation degree, efficiency factor, queue length;

Step 2: utilize analytical hierarchy process to calculate control delay, jamming rate, saturation degree, the efficiency factor of plane signal intersection under mixed traffic environment, the weights separately of 5 evaluation indexes of queue length;

Step 3: use fuzzy clustering method, obtain the central point and the standard deviation of crossing service level each item at different levels evaluation index under the mixed traffic environment, try to achieve the subordinate function of evaluation indexes at different levels;

Step 4: utilize motor vehicle and bicycle in the INSSIM simulation software to produce model, pass through decision model with speed model, pedestrian's generation model, queuing model and right-hand rotation motor vehicle motor vehicle, bicycle and the pedestrian's of hybrid cross mouth operation conditions is carried out emulation, thereby obtain the desired value of mixed traffic environment control delay down, jamming rate, saturation degree, efficiency factor, 5 evaluation indexes of queue length;

Step 5: use the subordinate function of each evaluation index of service level at different levels, the degree of membership of calculation control delay, jamming rate, saturation degree, efficiency factor, 5 each grades of evaluation index of queue length forms single factor and passes judgment on matrix respectively;

Step 6: utilize single factor to pass judgment on matrix and index weight value, following 5 evaluation indexes of each rank are carried out multifactorial evaluation, confirm the service level rank by maximum subjection principle.

In the said step 2, utilize analytical hierarchy process to calculate control delay, jamming rate, saturation degree, the efficiency factor of plane signal intersection under mixed traffic environment, the weights separately of 5 evaluation indexes of queue length, comprise the following steps:

Step 21: set up the hierarchical structure of estimating;

Step 22: structure judgment matrix;

Step 23: carry out single preface of level and consistency check;

Step 24: carry out total ordering of level and consistency check;

Step 25: calculate each index weight value.

Said right-hand rotation motor vehicle passes through decision model and sets up based on the BP neural network, comprising:

Step 41: set up three layers of BP network structure respectively, confirm input layer, output layer neuron number and hidden neuron number range;

Step 42: train the interior neural network of number range of hidden neuron, confirm the number of hidden neuron.

In the said step 5, the degree of membership of the delay of difference calculation control, jamming rate, saturation degree, efficiency factor, 5 each grades of evaluation index of queue length, specifically:

Wherein, c _jBe the achievement data of control delay, jamming rate, saturation degree, efficiency factor, 5 indexs of queue length, x _IjBe the central point of service levels at different levels, y _IjThe standard deviation of service levels at different levels.

Said single factor is passed judgment on matrix: $R=\left\{\begin{array}{cccc}{r}_{11}& r_{12}& ...& r_{15}\\ r_{21}& r_{22}& ...& r_{25}\\ ...& ...& ...& ...\\ r_{51}& r_{52}& ...& r_{55}\end{array}\right\}.$

In the said step 6, following 5 indexs of each rank are carried out multifactorial evaluation specifically:

For each index weights W that calculates by step 1 ^T=(W ₁, W ₂... W ₅), with it and single factor pass judgment on matrix R carry out model M (∧ ∨) calculates, multifactorial evaluation

In the said step 6, confirm the service level rank, specifically be, get max (b by maximum subjection principle ₁, b ₂... b ₅)=b _k, then service level is confirmed as the k level.

Its effect of method that the present invention proposes is; Take all factors into consideration the influence of the various traffic factors that comprise bicycle and pedestrian and index to planed signal crossing service level contribution degree, more accurate to the evaluation of service level of plane signal intersection under mixed traffic environment.

Description of drawings

Fig. 1 is the evaluation method process flow diagram of the service level of plane signal intersection under mixed traffic environment that proposes of the present invention;

Fig. 2 is the hierarchical chart that utilizes each item index of the plane signal intersection under mixed traffic environment that analytical hierarchy process sets up.

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 scope of the present invention and application thereof.

The present invention utilizes magnitude of traffic flow sensor, traffic signal light condition sensor and traffic simulation server to realize the evaluation of service level of plane signal intersection under mixed traffic environment.Wherein, Magnitude of traffic flow sensor links to each other with server through data transmission link respectively with the traffic signal light condition sensor; The traffic signal light condition sensor is arranged on intersection; Magnitude of traffic flow sensor is arranged on car lane, cycle track and the walkway of crossing adjacency; Server is accepted from the data-signal of magnitude of traffic flow sensor and traffic signal light condition sensor and is stored in the database, and the traffic simulation server comprises a cover and is equipped with the computer system of network interface and runs on the planed signal crossing service level evaluation software-INSSIM simulation software on this computer system.Traffic simulation software is divided into traffic simulation program and basic database two parts.

Fig. 1 is the evaluation method process flow diagram of the service level of plane signal intersection under mixed traffic environment that proposes of the present invention.Among Fig. 1, the implementation procedure of the method that the present invention proposes is:

Step 1: confirm the service level of plane signal intersection under mixed traffic environment evaluation index, comprising: control delay, jamming rate, saturation degree, efficiency factor, queue length.

Appoint in " traffic engineering " book that the Feitian professor writes, service level is defined as: an index weighing traffic flow service condition and the service quality that driver and passenger experienced.Therefore, in conjunction with China's mixed traffic environment, the service level of crossing should be defined as: the service condition of signalized intersections mixed traffic flow and user (comprising automobile driver or cycling person) are to the satisfaction of service that the crossing provides.From key concept, the present invention confirms that the service level evaluation index of hybrid cross mouth is control delay, queue length, saturation degree, efficiency factor and jamming rate.

Control is incured loss through delay: control is incured loss through delay and is comprised stop delay and the acceleration and deceleration lost time of vehicle in the scope of crossing.

Queue length: queue length is meant that stopped vehicle takies the space length of road.

Saturation degree (traffic loading coefficient): this index is meant under mixed traffic environment, and entrance driveway actual traffic amount is (v) with the ratio of the entrance driveway traffic capacity (c).

Efficiency factor: this index is meant under mixed traffic environment, the ratio of vehicle Vehicle Speed on the road speed of crossing and highway section.

Saturation degree (loading coefficient), efficiency factor index in calculating of the present invention, have all been considered the mutual disturbing factor of mixed traffic.

Jamming rate: be that the present invention combines the characteristics of China's mixed traffic to propose.Its implication is meant in the planed signal crossing, and motor vehicle is owing to the loss of time that the non-interference of machine causes, and accounts under the noiseless situation the normally ratio of current required time of motor vehicle.Jamming rate can be used for describing in the crossing bicycle and pedestrian to the interference of motor vehicle.

Step 2: utilize analytical hierarchy process to calculate control delay, jamming rate, saturation degree, the efficiency factor of plane signal intersection under mixed traffic environment, the weights separately of 5 evaluation indexes of queue length.Detailed process is:

Step 21: set up the hierarchical structure of estimating.Fig. 2 is a hierarchical structure of utilizing each item index of the plane signal intersection under mixed traffic environment that analytical hierarchy process sets up.Among Fig. 2, setting the service evaluation level is destination layer A; Type formation B comprises the outer evaluation index B of stop line ₁With fare inner evaluation index B ₂5 indexs are arranged in type formation C, belong to the outer evaluation index B of stop line ₁With fare inner evaluation index B ₂, wherein, the outer evaluation index B of stop line ₁Under comprise: C is incured loss through delay in control ₁, saturation degree C ₂, queue length C ₃, fare inner evaluation index B ₂Comprise: jamming rate C ₄, efficiency factor C ₅

Step 22: structure judgment matrix.From the 2nd layer (type formation B) beginning of the hierarchy Model set up, for the same one deck factors that are subordinated to each factor of last layer, with 1～9 relatively yardstick construct judgment matrix through expert's scoring, up to orlop.Promptly, compare yardstick to the relative importance degree assignment between each element by 1～9 to the important in twos degree of the index in type formation B and the type formation C.What following table (table 1) was represented is the implication of 1～9 comparison yardstick.

Table 1: compare yardstick implication table

For example, draw judgment matrix according to expertise wherein: A-B, B ₁-C, B ₂-C:

Matrix $A-B=\left[\begin{array}{cc}1& 1\\ 1& 1\end{array}\right]$

Matrix $B_1-C=\left[\begin{array}{ccc}1& 4& 3\\ 1/4& 1& 1/2\\ 1/3& 2& 1\end{array}\right]$

Matrix $B_2-C=\left[\begin{array}{cc}1& 3\\ 1/3& 1\end{array}\right]$

Step 23: carry out single preface of level and consistency check.

The single preface of level is exactly to confirm the influence degree of each factor of this layer to the upper strata factor, calculates the maximum characteristic root and the characteristic of correspondence vector of judgment matrix, and the most frequently used method is and amasss method and root method.Method is selected and is amassed in this calculating for use, and its concrete steps are following:

(1) each column element of judgment matrix A being done normalization handles;

(2) judgment matrix of each row after normalization is handled pressed the row addition, get vectorial

(3) to vector

Carry out normalization and handle resulting W=(W ₁, W ₂... W _n) ^T

Be proper vector;

(4) the maximum characteristic root of calculating judgment matrix:

${\mathrm{\λ}}_{\max }=\underset{i=1}{\overset{n}{\mathrm{\Σ}}}\frac{{\left(\mathrm{AW}\right)}_i}{{\mathrm{nW}}_i}---\left(1\right)$

(5) consistency check:

Reasonable in order to guarantee the conclusion that analytical hierarchy process obtains, must will carry out consistency check the deviation limits of judgment matrix within the specific limits.Generally judge with CI and two indexs of CR.

Coincident indicator: $\mathrm{CI}=\frac{{\mathrm{\λ}}_{\mathrm{Max}}-n}{n-1}---\left(2\right)$

Consistency Ratio at random: $\mathrm{CR}=\frac{\mathrm{CI}}{\mathrm{RI}}---\left(3\right)$

RI wherein, i.e. average homogeneity property index is repeatedly to repeat to get arithmetical mean after the calculating of judgment matrix eigenwert at random to obtain.

When CR＜0.1 item thinks that the consistance of judgment matrix is an acceptable, otherwise, think that inconsistency is too serious, need construct judgment matrix again or do necessary adjustment.

Step 24: carry out total ordering of level and consistency check.Calculate the sequencing weight of all factors of same level, claim that level always sorts whole general objective relative importance.It is with the combination weights of the weights of next each factor of level and last layer time factor, obtains the relative importance of orlop factor with respect to whole general objective.

Calculate the A-C weight, computing formula is following:

W _A-C＝W _A-B×W _B-C (4)

Wherein, W _A-BThe weight of expression B layer index, W _B-CExpression C layer index is with respect to the weight of the B layer index of affiliated associated layers.

Total ordering also need be done consistency check to level, and check is still successively carried out to low layer by high level as level always sorts.Each level has all passed through the consistency check of the single preface of level though this is, each compares judgment matrix in pairs and has all had comparatively satisfied consistance.But when integrated survey, the nonuniformity of each level still might accumulate, and causes the nonuniformity that the final analysis result is more serious.

If in the C layer with B _jThe paired relatively judgment matrix of relevant factor in single preface through consistency check; Trying to achieve single preface coincident indicator is CI (j); (j=1 ..., m); Corresponding mean random coincident indicator is RI (j) (CI (j), RI (j) try to achieve when the single preface of level), and then the C layer always sorts at random that the consistance ratio does

$\mathrm{CR}=\frac{\underset{j=1}{\overset{m}{\mathrm{\Σ}}}\mathrm{CI}\left(j\right)b_j}{\underset{j=1}{\overset{m}{\mathrm{\Σ}}}\mathrm{RI}\left(j\right)b_j}---\left(5\right)$

Wherein, b _jBe factor B ₁..., B _nAbout A _jThe single preface weight of level.When CR＜0.10, think that the total ranking results of level has satisfied consistance and accepts this analysis result.

Step 25: calculate each index weight value.

After a plurality of expert's marking, obtain the total sorting calculation result of level separately.It is comprehensive, obtain the final weights W=(W of each index of assessment indicator system ₁, W ₂... W _n) ^TWherein, W _jBe the comprehensive weights of j index.

${\mathrm{\ω}}_j=\frac{\underset{i=1}{\overset{s}{\mathrm{\Σ}}}{\mathrm{\λ}}_i{W}_{\mathrm{ij}}}{\underset{i=1}{\overset{s}{\mathrm{\Σ}}}{\mathrm{\λ}}_i},(j=\mathrm{1,2},...n)---\left(6\right)$

S is expert's number in the formula; λ _iBe i position expert's weights; W _IjBe the weights of i position expert to j index.

Step 3: the utilization fuzzy clustering method obtains the central point and the standard deviation of crossing service level each item at different levels evaluation index under the mixed traffic environment, and tries to achieve the subordinate function of evaluation indexes at different levels.

The fuzzy c means clustering algorithm (FCM) of based target function is grouped into the nonlinear programming problem of a belt restraining with cluster analysis, finds the solution through optimization that the optimum that obtains data set is fuzzy to be divided and cluster.

FCM fuzzy clustering problem can be expressed as the objective function below optimizing on mathematics, obtain to make it reach the degree of membership matrix U={ u of minimum value _Ik} _{C * n}With cluster centre P={p ₁, p ₂... p _c,

$\∀i\∈\left\{i\right|1\≤i\≤c\}:$

$J_m(U,P)=\underset{k=1}{\overset{n}{\mathrm{\Σ}}}\underset{i=1}{\overset{c}{\mathrm{\Σ}}}{\left(u_{\mathrm{ik}}\right)}^m{d}^2(x_k,p_i)---\left(7\right)$

Constraint condition is:

$\left\{\begin{array}{c}\underset{i=1}{\overset{c}{\mathrm{\Σ}}}{u}_{\mathrm{ik}}=\mathrm{1,1}\≤k\≤n\\ u_{\mathrm{ik}}\∈\left[\mathrm{0,1}\right],1\≤i\≤c,1\≤k\≤n\end{array}\right.---\left(8\right)$

In the formula; M is a weighted index, is called level and smooth index or fuzzy parameter again

5 evaluation index control delays, saturation degree, queue length, efficiency factor, jamming rate that the front is selected are designated as x respectively _i(i=1,2,3,4,5) are as the index domain: X=(x ₁, x ₂..., x ₅; Obtain 360 data samples under the different traffic environments through emulation, as sample domain: U={x ₁, x ₂..., x ₃₆₀, thereby obtain one 360 * 5 matrix.

Use the Fuzzy C clustering method of objective function, calculate the central point of service levels at different levels, step is following:

Step (1) initialization, delivery is stuck with paste Weighting exponent m=2, and the classification of cluster is counted c=5, the number n of data sample point=360, iteration stopping threshold epsilon=0.001.Represent iterations with l.

Step (2) adopts random approach to select the initial cluster center point Put as initial cluster center for c before the picked at random, and calculate initial degree of membership matrix U ⁽⁰⁾

Step (3) algorithm begins iteration.According to formula (7), (8), bring in constant renewal in cluster centre p _i(l) and the degree of membership matrix U ^(l), up to || p ^(l+1)-p ^(l)||＜ε, (U P) converges to minimal value to objective function J at this moment.

This paper utilizes the programming of MATLAB simulation software according to the aforementioned calculation step, and the cluster centre that obtains (central points of service levels at different levels) sees the following form (table 2).

Table 2: crossing service level central points at different levels under the mixed traffic environment

Under calculating mixed traffic environment, in the central point of the 5 grades of service levels in crossing, can also obtain the standard deviation of the 5 grades of service levels in crossing under the mixed traffic environment.Following table (table 3) is the standard deviation of service levels at different levels.

Table 3: crossing LOS criterias at different levels are poor under the mixed traffic environment

Step 4: utilize motor vehicle and bicycle to produce model, pass through decision model with speed model, pedestrian's generation model, queuing model and right-hand rotation motor vehicle motor vehicle, bicycle and the pedestrian's of hybrid cross mouth operation conditions is carried out emulation, thereby obtain the desired value of mixed traffic environment control delay down, jamming rate, saturation degree, efficiency factor, 5 indexs of queue length.

The present invention has used for reference the existing model of forefathers when the parameter data, and to craspedodrome bicycle and the pedestrian influence to the generation of right-hand rotation motor vehicle, has creatively proposed the right-hand rotation motor vehicle and passed through decision model.Right-hand rotation is passed through decision model and is meant: the motor vehicle of no right turn signal control, because of not receiving the influence of signal lamp, vehicle all can get into the crossing at any time, so in equidirectional straightgoing vehicle green light phase time, possibly produce with the same-phase straight-going bicycle and disturb.When go in right-hand rotation motor vehicle entering crossing, in interference range,, then need to pass through decision model and judge whether to obtain right-of-way, if can then pass through according to the motor vehicle that trains if run into synchronous straight-going bicycle; Otherwise, ramp to stop is waited for.

The present invention's motor vehicle of turning right passes through decision model and sets up based on the BP neural network, comprises the following steps,

Step 41: set up three layers of BP network structure respectively, confirm input layer, output layer neuron number and latent number of layers scope.

This paper adopts three layers of BP neural network structure of typical case: input layer, the latent layer of individual layer and output layer.Input layer is accepted outside input data, and the input parameter that this paper confirms is respectively: quantity, the bicycle of the interior bicycle of interference range provide a motor vehicle crossing gap and a back bicycle speed when speed before the right-hand rotation motor vehicle passes through, motor vehicle entering interference range.At random the sample that collects is divided into two parts, is respectively training set and test set.

Output layer node number gets 2, when input sample decision-making when passing through, desirable output result be (1,0), when the decision-making of input sample for not passing through, it is (0,1) that ideal is exported the result.To the output result of output layer is 1 neuron, if the output result, thinks then that the output result is 1 more than or equal to 0.9 o'clock; To the output result of output layer is 0 neuron, if the output result, thinks then that the output result is 0 smaller or equal to 0.1 o'clock; As output result between 0.1 and 0.9 time, think that then this output valve can not pass through or not pass through and make corresponding judgment the right-hand rotation motor vehicle.Hidden neuron is counted n and is calculated with experimental formula,

(n _iBe the input neuron number; n ₀The output neuron number; A is the constant between 1～10).

Step 42: train the interior neural network of number range of hidden neuron, confirm the number of hidden neuron.

When confirming the number of hidden neuron, the number of training hidden neuron respectively is from 10 neural networks such as 3～12 grades, and every type all keeps error and satisfies the convergence target (the convergence target is for smaller or equal to 1 * 10 among the present invention ^-3) 20 neural networks carry out emulation and statistical study; Add up the total degree of all kinds of neural metwork trainings in the time of training.20 network utilisation checking sample sets that satisfy performance index are carried out the emulation statistical study; Calculate the average correct recognition rata of every type of network simulation; From 20 training results of every type, pick out the best simulation result of best simulation result simultaneously as disparate networks.

Table 4 and table 5 are respectively the training and the tables of simulation results of right-hand rotation motor vehicle crossing gap experiment and the training and the tables of simulation results of passing through the time-delay experiment.In the table 4; The hidden neuron number is that the average correct recognition rata of emulation of 5 and 7 neural network is 99.21%; And the hidden neuron number is 7 neural metwork training total degree is 58 times; Less than the hidden neuron number is 5 neural network, and therefore, it is 7 the neural network simulation result as the best that right-hand rotation motor vehicle crossing gap model is selected the hidden neuron number.

Table 4: the training and the tables of simulation results of the experiment of right-hand rotation motor vehicle crossing gap

And in the table 5, the hidden neuron number is that 3 the average correct recognition rata of neural network emulation is the highest, is 99.04%, passes through decision model so to select the hidden neuron number be 3 neural network as the right-hand rotation motor vehicle of accepting based on time-delay of the best.

Table 5: the right-hand rotation motor vehicle passes through the training and the tables of simulation results of time-delay experiment

Step 5: use the subordinate function of the evaluation index of service levels at different levels, calculate control delay, jamming rate, saturation degree, efficiency factor, every grade of other degree of membership of 5 indexs of queue length respectively, form single factor and pass judgment on matrix.

With reference to the central point and the standard deviation of the service level that obtains in the step 3, the subordinate function of each evaluation index of service level at different levels

Wherein, c _jBe the achievement data that calculates in the step 4, x _IjBe the central point of service level, y _IjBe standard deviation.

Single factor is passed judgment on matrix R:

$R=\left\{\begin{array}{cccc}{r}_{11}& r_{12}& ...& r_{15}\\ r_{21}& r_{22}& ...& r_{25}\\ ...& ...& ...& ...\\ r_{51}& r_{52}& ...& r_{55}\end{array}\right\}$

Step 6: utilize single factor to pass judgment on matrix and index weight, following 5 indexs of each rank are carried out multifactorial evaluation, confirm the service level rank by maximum subjection principle.

By the definite weights W of step 1 ^T=(W ₁, W ₂... W ₅) and single factor of confirming of step 5 pass judgment on matrix R, with model M (∧, ∨) calculating can get multifactorial evaluation:

B promptly is the evaluation result to service level, presses maximum subjection principle, gets max (b ₁, b ₂... b ₅)=b _k, then service level is confirmed as the k level.

The present invention has overcome can't reflect in the existing service level evaluation technology that the non-disturbance regime of mixed traffic machine in service, some evaluation indexes fail to consider the deficiency of mixed traffic environment, and the evaluation that makes service level of plane signal intersection under mixed traffic environment more rationally and science.

The above; Be merely the preferable embodiment of the present invention, but protection scope of the present invention is not limited thereto, any technician who is familiar with the present technique field is 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. the evaluation method of a service level of plane signal intersection under mixed traffic environment is characterized in that, said evaluation method comprises the following steps:

Step 1: confirm the service level of plane signal intersection under mixed traffic environment evaluation index, comprising: control delay, jamming rate, saturation degree, efficiency factor, queue length;

Said control is incured loss through delay and is comprised stop delay and the acceleration and deceleration lost time of vehicle in the scope of crossing;

Said jamming rate is meant in the planed signal crossing, and motor vehicle is owing to the loss of time that the non-interference of machine causes, and accounts under the noiseless situation the normally ratio of current required time of motor vehicle;

Said saturation degree is meant under mixed traffic environment, the ratio of entrance driveway actual traffic amount and the entrance driveway traffic capacity;

Said efficiency factor is meant under mixed traffic environment, the ratio of vehicle Vehicle Speed on the road speed of crossing and highway section;

Step 2: utilize analytical hierarchy process to calculate control delay, jamming rate, saturation degree, the efficiency factor of plane signal intersection under mixed traffic environment, the weights separately of 5 evaluation indexes of queue length; Comprise the following steps:

Step 21: set up the hierarchical structure of estimating;

Step 22: structure judgment matrix;

Step 23: carry out single preface of level and consistency check;

Step 24: carry out total ordering of level and consistency check;

Step 25: calculate each index weight value;

Step 3: use fuzzy clustering method, obtain the central point and the standard deviation of crossing service level each item at different levels evaluation index under the mixed traffic environment, try to achieve the subordinate function of evaluation indexes at different levels;

Step 4: utilize motor vehicle and bicycle in the INSSIM simulation software to produce model, pass through decision model with speed model, pedestrian's generation model, queuing model and right-hand rotation motor vehicle motor vehicle, bicycle and the pedestrian's of hybrid cross mouth operation conditions is carried out emulation, thereby obtain the desired value of mixed traffic environment control delay down, jamming rate, saturation degree, efficiency factor, 5 indexs of queue length; Said right-hand rotation motor vehicle passes through decision model and sets up based on the BP neural network, comprising: set up three layers of BP network structure respectively earlier, confirm input layer, output layer neuron number and hidden neuron number range; Neural network in the number range of retraining hidden neuron is confirmed the number of hidden neuron;

Step 5: use the subordinate function of each evaluation index of service level at different levels, respectively the degree of membership of calculation control delay, jamming rate, saturation degree, efficiency factor, 5 each grades of evaluation index of queue length; Specifically:

J=1,2,3,4,5 wherein, c _jBe the achievement data of control delay, jamming rate, saturation degree, efficiency factor, 5 indexs of queue length, x _IjBe the central point of service levels at different levels, y _IjStandard deviation for service levels at different levels; Form single factor and pass judgment on matrix, said single factor is passed judgment on matrix and is:

$R=\left\{\begin{array}{cccc}{r}_{11}& r_{12}& ...& r_{15}\\ r_{21}& r_{22}& ...& r_{25}\\ ...& ...& ...& ...\\ r_{51}& r_{52}& ...& r_{55}\end{array}\right\},$

Step 6: utilize single factor to pass judgment on matrix and index weight value, following 5 evaluation indexes of each rank are carried out multifactorial evaluation, specifically: for each index weight value W that calculates by step 2 ^T=(W ₁, W ₂W ₅), with it and single factor pass judgment on matrix R carry out model M (∧ ∨) calculates, multifactorial evaluation

Confirm the service level rank by maximum subjection principle, specifically: get max (b ₁, b ₂... b ₅)=b _k, then service level is confirmed as the k level.

Images