CN103050016A - Hybrid recommendation-based traffic signal control scheme real-time selection method - Google Patents
Hybrid recommendation-based traffic signal control scheme real-time selection method Download PDFInfo
- Publication number
- CN103050016A CN103050016A CN2012105684179A CN201210568417A CN103050016A CN 103050016 A CN103050016 A CN 103050016A CN 2012105684179 A CN2012105684179 A CN 2012105684179A CN 201210568417 A CN201210568417 A CN 201210568417A CN 103050016 A CN103050016 A CN 103050016A
- Authority
- CN
- China
- Prior art keywords
- traffic
- control program
- crossing
- signals control
- scene
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000010187 selection method Methods 0.000 title abstract 2
- 238000012545 processing Methods 0.000 claims abstract description 57
- 238000000034 method Methods 0.000 claims abstract description 37
- 238000004088 simulation Methods 0.000 claims abstract description 22
- 238000001914 filtration Methods 0.000 claims abstract description 7
- 238000003066 decision tree Methods 0.000 claims description 30
- 238000011156 evaluation Methods 0.000 claims description 12
- 230000008569 process Effects 0.000 claims description 10
- 238000004891 communication Methods 0.000 claims description 8
- 238000010276 construction Methods 0.000 claims description 5
- 230000005540 biological transmission Effects 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 claims description 3
- 238000004422 calculation algorithm Methods 0.000 abstract description 7
- 230000000694 effects Effects 0.000 abstract description 4
- 238000005516 engineering process Methods 0.000 abstract description 3
- 230000004044 response Effects 0.000 abstract description 2
- 230000006870 function Effects 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 3
- 230000008901 benefit Effects 0.000 description 3
- 238000005457 optimization Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000011217 control strategy Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000006698 induction Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
Images
Abstract
The invention discloses a hybrid recommendation-based traffic signal control scheme real-time selection method, which comprises the following steps of: processing intersection characteristics and traffic signal control scheme characteristics to obtain intersection categories and scheme categories; and constructing a standard traffic scene on a parallel simulation platform for each intersection category to test a control scheme so as to obtain performance of the intersection category under the simulation platform as system starting knowledge. When the actual intersection control performance cannot meet the requirements, the control scheme suitable for an intersection is found through a hybrid recommendation technology, including based on content recommendation and collaborative filtering recommendation, and is transmitted to an actual intersection to improve the traffic control effect of the intersection. Meanwhile, a system knowledge base is further perfected by recording and processing actual intersection traffic information to improve the accuracy of a recommendation algorithm. According to the method, dynamic response can be made to short-term traffic state fluctuation variation and an emergency situation on an urban traffic road network, so that the method is particularly suitable for being used in an urban traffic control system.
Description
Technical field
The present invention relates to traffic signals control technology field, particularly a kind of based on mixing the real-time choosing method of traffic signals control program of recommending.
Background technology
The traffic system of good operation and city are developed in a healthy way closely bound up.Traffic jam issue has become the bottleneck problem of large-and-medium size cities development, and the period difference that the dynamic perfromance that traffic system itself has produces and the difference between the crossing self, so that single a kind of crossing control program can't satisfy crossing traffic demands for control whole on the road network.Therefore according to the traffic characteristics of different periods at different crossings, carry out and have the operational efficiency that otherness crossing control program will be optimized whole traffic network, the whole control performance of raising traffic control system.
In the middle of existing traffic intersection control method for designing, multi-period crossing control method adopts various control scheme wheel to bring optimization of road joints traffic control effect according to crossing state period characteristic (namely there is its different characteristics the different periods).Wherein the prior art division (being that the time point rotated of control program is selected) and the suitable control program of every single control time that comprise control time chosen two steps.The former mainly adopts by the mode of crossing traffic history data being identified to find out control in one day and rotates time point, and the latter designs appropriate traffic control scheme according to the traffic behavior historical data.This type of multi-period crossing control method depends on the traffic historical data, is a kind of off-line method of reply traffic system dynamic perfromance.When using, these class methods require the dynamic perfromance of traffic system to have repeatable preferably at space-time.In addition, can't be to fine granularity, short time traffic conditions fluctuation changes makes a response, and such as the traffic emergency situations etc., and this is unavoidable in the actual traffic system.
Summary of the invention
The technical matters that (one) will solve
The objective of the invention is to overcome off-line and divide the deficiency that control time and the control program of Off-line control period are chosen, in network range, take existing traffic control scheme as the basis, difference and traffic intersection real time status according to the traffic control scheme, adopt Dynamic Matching, optimization of road joints traffic control performance is invented a kind of based on mixing the real-time choosing method of traffic signals control program of recommending for this reason to greatest extent.
(2) technical scheme
For addressing the above problem, the present invention proposes a kind of real-time choosing method of traffic signals control program based on mixing recommendation, the method comprises:
Off-line knowledge acquisition layer module construction standard crossing model and standard traffic scene, and obtain the traffic behavior parameter of each suitable traffic signals control program under the standard traffic scene, and described traffic behavior parameter and corresponding standard traffic scene data are sent to transport information store and the processing layer module;
Transport information storage and processing layer module are according to the traffic scene that calculates each traffic signals control program from the resulting described traffic behavior parameter of off-line knowledge acquisition layer module and standard traffic scene data-scheme score, also continue to fill and revise traffic scene-scheme call according to the real crossing traffic information that obtains from on-the-spot execution level module, whether the traffic signals control program of judging simultaneously real crossing needs to switch, and sends new traffic signals control program when needing to switch to on-the-spot execution level module;
On-the-spot execution level module is obtained real crossing traffic information, and the real crossing traffic information that obtains is sent to transport information storage and processing layer module; When receiving the renewal traffic signals control program message of transport information storage and the transmission of processing layer module, existing signal timing plan is switched to new traffic signals control program, and according to the physical communication agreement crossing control signal is carried out lamp control signal and encode, carry out described new traffic signals control program.
The invention allows for a kind of real-time selecting device of traffic signals control program based on mixing recommendation, it comprises:
Off-line knowledge acquisition layer module, it comprises that the traffic simulation scene makes up module, scene simulation module and artificial transport information acquisition module, wherein the traffic simulation scene makes up module construction standard crossing model and standard traffic scene, each traffic signals control program that is fit under the scene simulation modular simulation various criterion traffic scene, artificial transport information acquisition module obtains the traffic behavior parameter in the simulation process, and described traffic behavior parameter and corresponding standard traffic scene data are sent to transport information storage and processing layer module;
Transport information storage and processing layer module, it comprises transport information assessment and processing module and traffic signals control program administration module; The assessment of wherein said transport information and processing module are according to the traffic scene that calculates each traffic signals control program from the resulting described traffic behavior parameter of off-line knowledge acquisition layer module and standard traffic scene data-scheme score, described transport information assessment also continues to fill and revise traffic scene-scheme call according to the real crossing traffic information that obtains from on-the-spot execution level module with processing module, whether the traffic signals control program of judging simultaneously real crossing needs to switch, if need to switch, then traffic signals control program administration module sends new traffic signals control program to on-the-spot execution level module;
On-the-spot execution level module, it comprises real crossing traffic information acquisition module, signal controlling execution module and signal timing plan handover module; Wherein said real crossing traffic information acquisition module obtains real crossing traffic information, and the real crossing traffic information that obtains is sent to transport information storage and processing layer module; Described signal timing plan handover module switches to new traffic signals control program with existing traffic signals control program when receiving the renewal traffic signals control program message of transport information storage and the transmission of processing layer module; Described signal controlling execution module carries out the lamp control signal coding according to the physical communication agreement to the crossing control signal, carries out the traffic signals control program of upgrading.
(3) beneficial effect
Beneficial effect of the present invention: after employing is chosen in real time based on the traffic signals control program of mixing recommendation, realized that traffic intersection control can change the traffic control scheme of dynamically adjusting according to real-time traffic state at road cross, improved by a relatively large margin control system and dealt with the ability that short time traffic conditions fluctuates, particularly to the accident processing power.Simultaneously can take full advantage of the scheme base existing program, optimization of road joints traffic control effect is brought preferably comprehensive benefit.
Description of drawings
Fig. 1 is networked layering traffic signal control system structural drawing in the embodiment of the invention;
Fig. 2 is the example of a large class decision tree of traffic intersection in the embodiment of the invention;
Fig. 3 is off-line knowledge acquisition process workflow journey figure in the embodiment of the invention;
Fig. 4 is that the transport information storage is chosen process workflow journey figure in real time with the processing layer control program in the embodiment of the invention;
Fig. 5 is on-the-spot execution level workflow diagram in the embodiment of the invention.
Embodiment
For making the purpose, technical solutions and advantages of the present invention clearer, below in conjunction with specific embodiment, and with reference to accompanying drawing, the present invention is described in more detail.
As shown in Figure 1, the real-time choosing method of traffic signals control program based on mixing recommendation of the present invention is realized by the various level computer distribution type, is divided into off-line knowledge acquisition layer A1, transport information storage and processing layer A2 and on-the-spot execution level A3 according to function of the present invention in the various level computing machine.Communication can be adopted existing signal controller network communication medium and method between each layer.Used Pearson's correlation calculations in the recommended flowsheet, based on this coefficient of paddy method for measuring similarity, that Slope One algorithm can be selected is common, for example mahout increases income, and the machine learning platform is realized.Knowledge and scheme database adopt common, for example MySQL database.Common, for example Paramics traffic simulation platform that traffic simulation platform adopts.
Contain among the off-line knowledge acquisition layer A1: scene simulation modules A 11, traffic simulation scene make up modules A 12, road net data processing module A13, artificial transport information acquisition module A14, wherein:
Recorded the basic parameter of actual traffic road network in the off-line knowledge acquisition layer among the road net data processing module A13, such as the crossing characteristic.Described road net data processing module A13 constructs the large class decision tree of the most shallow traffic intersection and writes traffic knowledge base A16 by the ID3 algorithm based on the basic parameter of described actual traffic road network.Obtain the large class realization in some crossings to the classification at all controlled crossings in the traffic network by the large class decision tree of traffic intersection.As shown in Figure 2, the root node of the large class decision tree of described traffic intersection is crossing type (or other crossing characteristic attributes), and leaf node is the large class in crossing, and the large class in each crossing comprises several special intersection features.Based on the large class in crossing, the traffic simulation scene make up modules A 12 construct corresponding standard crossing model (such as standard crossroad model, standard T-shaped road junction model etc.) and as described in the standard traffic scene (such as the trip peak traffic scene of standard, the trip rareness traffic scene of standard etc.) of standard crossing model, the implication of its Plays refers to that the crossing makes up and being structured on the special intersection feature of traffic scene got default value, is 24 etc. such as large crossing default channel number.Then, scene simulation modules A 11 is at the traffic signals control program parallel artificial that is fit under to the various criterion traffic scene on the distributed platform, the road network crossing operational factor target function that artificial transport information acquisition module A14 calls emulation platform obtains transport information, described transport information comprises traffic state at road cross parameter (such as car speed etc.) and traffic scene data (such as existing crossing traffic wagon flow state etc.), artificial transport information acquisition module A14 is also with the traffic state at road cross parameter read-in traffic knowledge base A16 that obtains, simultaneously the traffic scene data are write the traffic scene property data base, to be committed to transport information storage and processing layer A2; The traffic signals control program that is fit under the wherein said standard scene is provided by traffic signals control program administration module A15.
Contain among transport information storage and the processing layer A2: traffic signals control program administration module A15, traffic knowledge base A16, transport information assessment and processing module A17, content-based recommending module A18 and based on collaborative filtering recommending modules A 19, wherein:
Comprise altogether the large class decision tree of traffic intersection, the large class decision tree of traffic signals control program, traffic plan-scene call, traffic scene data four class traffic knowledge among the traffic knowledge base A16 of transport information storage and processing layer A2.The large class decision tree of traffic intersection is produced by road net data processing module A13, the large class decision tree of traffic signals control program is utilized the basic parameter of traffic signals control program to produce and write traffic knowledge base A16 by the ID3 algorithm by traffic signals control and management modules A 15, realization is to the classification of all signal timing plans in the system, its root node is the applicable crossing type (or other traffic signals control program characteristic attributes) of traffic signals control program, leaf node is the large class of traffic signals control program, and the large class of each traffic signals control program comprises several special traffic signals control program features.What traffic signals control program administration module A15 stored all optional traffic signals control programs and control program relates to feature (the crossing feature that is suitable for and traffic scene feature).Described traffic plan-scene call and traffic scene data are then provided with processing module A17 by the transport information assessment, and the basic parameter of wherein said traffic signals control program comprises applicable traffic intersection feature and applicable traffic scene feature.Transport information assessment and processing module A17 are responsible for assessing and the transport information that provides from off-line knowledge acquisition layer A1 and on-the-spot execution level A3 are provided.After receiving the transport information that produces from the artificial of off-line knowledge acquisition layer A1, described transport information assessment can be to traffic knowledge base A16 record standard traffic scene data and standard traffic scene-scheme score with processing module A17.After receiving the real crossing traffic information from on-the-spot execution level A3, whether this module can be assessed real intersection signal control scheme needs to switch, and also can record real traffic scene data and real traffic scene-scheme score to traffic knowledge base A16 simultaneously.When the transport information assessment judges that with processing module A17 the signal timing plan of real traffic intersection need to switch, can trigger recommending module A18 and A19, wherein the content-based recommending module A18 large class of traffic signals control program that can will be fit to be switched real crossing under the large class decision tree of traffic control scheme is supported is found out and is submitted to based on collaborative filtering recommending modules A 19, based on collaborative filtering recommending modules A 19 with the traffic scene score in traffic knowledge base A16, find out the traffic signals control program that is fit to be switched real crossing under the support of traffic scene data, and the submission scheme is number to traffic signals control program administration module A15, this module extract from the traffic signals control program storehouse of self and the transmitted signal control program to on-the-spot execution level A3.
Contain among the on-the-spot execution level A3: signal timing plan handover module A20, signal controlling execution module A21, real crossing traffic information acquisition module A22, wherein:
The default timing signal control program of on-the-spot execution level A3 operation, when transport information storage provided the switching signal control program with processing layer A2 to on-the-spot execution level A3, signal timing plan handover module A20 was responsible for receiving and finishing new arrival traffic signals control program and the former traffic signals control program switch operating that carries into execution a plan.Run signal control execution module A21 after to be switched work is finished, this module is responsible for the lamp control signal coding and is supported physics communication protocol, finishes traffic signals control function.Reality crossing traffic information acquisition module A22 finishes traffic information collection and the transport information that collects is committed to transport information storage and processing layer A2.
Off-line knowledge acquisition layer A1 obtains off-line knowledge, according to the traffic behavior under the physical characteristics of different traffic intersections and the different traffic scenes, on distributed traffic simulating platform, construct simultaneously a plurality of standard traffic scenes, the traffic signals control program that is fit under the described standard traffic scene is tested, obtain the traffic state at road cross parameter of described traffic signals control program under described standard traffic scene, such as the crossing vehicle mean delay time, the average passage rate of crossing vehicle etc.Standard traffic scene data will be submitted to transport information storage and processing layer A2 as off-line knowledge with the traffic state at road cross parameter of corresponding traffic signals control program, solve whole commending system cold start-up problem.
Off-line knowledge acquisition process workflow journey figure as shown in Figure 3, wherein:
Step B11: the performance data of traffic signals control program comprises the crossing type that each traffic signals control program is fit to and the traffic scene data that are fit to by traffic signals control program storehouse among the solution designers typing A15;
Step B12: with all characteristic typings that relates to the crossing, the crossing characteristic comprises the crossing shape, geographic position, detection means information, phase path ratio, crossing scale; Wherein detection means information comprises the crossing traffic information obtaining means, and as based on ground induction coil, infrared detector is made a video recording first-class; Phase place refers to obtain simultaneously one group of road combination of channels of right-of-way within a signal period, a phase place both can represent that the right-of-way of motor vehicle also can represent pedestrian's right-of-way, and these two right-of-ways are consistent; Described phase path compares the number of active lanes ratio for the phase place of the phase place that comprises the number of active lanes maximum in all phase places and number of active lanes minimum.
Step B13: adopt the ID3 decision trees to generate the large class decision tree of traffic signals control program and the large class decision tree of traffic intersection according to the performance data of described traffic signals control program and the characteristic at described crossing;
Step B14: make up standard crossing model: according to the physical characteristics that the large class in different crossings in the large class decision tree of described traffic intersection has, build different standard crossing models at distributed traffic simulating platform;
Step B15: make up the standard traffic scene: the traffic behavior parameter of each the standard crossing model by changing distributed traffic simulating platform makes up different standard traffic scenes, and wherein the traffic behavior parameter comprises vehicle queue length on traffic intersection vehicle congestion situation such as each passage, the outside wagon flow input status in crossing such as each passage vehicle arrival rate, road speed etc.;
Step B16: under the described standard traffic scene each traffic signals control program that is fit to is being carried out simulation work: under the support in the large class decision tree of traffic signals control program and traffic signals control program storehouse, obtain the large class of traffic scene that each existing traffic signals control program is fit to, and corresponding traffic signals control program sent in the standard traffic scene corresponding on the emulation platform finish intersection signal control work;
Step B17: at the control performance record of each traffic signals control program that is fit under the standard traffic scene: after a standard traffic scene emulation finishes, the traffic state at road cross parameter will go on record in its scene, and and corresponding traffic scene data send to together the transport information storage and be used for generating the control performance score of described traffic signals control program under this traffic scene with processing layer A2, with as traffic scene-scheme call in traffic knowledge base A16.
That the control effect of traffic signals control program in the traffic scene that the certain hour span is arranged (comprising emulation and reality) marked about traffic plan scoring step among transport information assessment among transport information storage and the processing layer A2 and the processing module A17, the traffic state data of each road that relates to according to this time period crossing when traffic scene finishes calculates, and record simultaneously correspondingly traffic scene data, make up the scheme of described traffic signals control program-scene score knowledge, wherein said traffic behavior parameter comprises the crossing traffic Control performance standard.If a traffic scene time span does not finish, the traffic signals control program just occurs to be switched, the traffic signals control program that then replaces is marked related traffic scene time span take switching instant as the concluding time.
The concrete evaluate formula of the traffic scene of traffic signals control program-scheme score is as follows:
Score in the following formula
iBe the traffic scene score of traffic signals control program i, whole traffic scene time span is t crossing traffic evaluation time point; γ is discount factor, span (0,1); T is crossing traffic evaluation time point in service, and span is [0, K]; The traffic behavior assessment comprises n crossing traffic Control performance standard, such as the crossing vehicle mean delay time, and the average passage rate of crossing vehicle etc., wherein
For described traffic signals control program i t crossing traffic evaluation time the score o'clock on d item traffic control performance index, w
dBe the weight of d item traffic control performance index, d is the index numbering, and span is [1, n].
The traffic scene record then can record the relevant road in crossing at the scene vehicle queue length of the zero hour, and the vehicle mean arrival rate of this time span.
That the transport information storage is chosen process workflow journey figure in real time with processing layer A2 control program in the embodiment of the invention as shown in Figure 4, wherein:
Step C11: intercept the control crossing traffic information from on-the-spot execution level A3;
Step C12: whether the control performance fluctuating range of judging existing traffic signals control program according to the traffic behavior at target crossing in the described control crossing traffic information surpasses certain threshold value, if do not surpass then execution in step C11, if surpass then execution in step C13, to start crossing scheme selection process;
Step C13: content-based recommendation, utilize the large class decision tree of the data based described traffic signals control program of the present traffic scene in described target crossing, find the large class of traffic signals control program of the present traffic scene that is fit to described target crossing;
Step C14: whether traffic signals control program related in the large class of traffic signals control program that determining step C13 finds out is unique, if unique then directly with this traffic signals control program as recommending the traffic signals control program, and execution in step C21, if it is not unique, execution in step C15 then is to start the collaborative filtering recommending process;
Step C15: implement scene matching according to the traffic scene data that record among the present traffic scene data in described target crossing and the described traffic knowledge base A6, obtain the highest similar scene of similarity according to the similarity of traffic scene feature, the degree of correlation is utilized Pearson's correlation calculations;
Step C16: the call scheme that Scene-the scheme score is the highest of in the traffic scene of traffic knowledge base A16-scheme call, finding out all signal timing plans under the similar scene;
Step C17: judge whether the scene of described top score scheme-scheme score surpasses threshold value, choosing according to two parameters of vehicle arrival rate and vehicle queue length of described threshold value determine, its concrete value is utilized the fuzzy logic realization.If surpass, then with this top score scheme as recommending the traffic signals control program, and execution in step C21 does not then carry out neighborhood search step C18 if surpass;
Step C18: whether the neighborhood range size of judging similar scene can increase, if the neighborhood scope has been extended to the result that candidate's traffic signals control program in traffic signals control program corresponding to all traffic plans of recording among the traffic knowledge base A6-scene call or the neighborhood has covered content-based recommendation, then the neighborhood scope can not increase.If can increase, execution in step C19 then, if can not increase, then with described top score scheme as recommending the traffic signals control program, and execution in step C21;
Step C19: seek the neighborhood scene set that the traffic scene that makes new advances is filled similar scene.Namely gather (under the initial situation in the neighborhood scene, the set of this neighborhood scene is for empty) outside the existing traffic scene, according under the traffic scene to the traffic scene of different traffic signals control programs-scheme score, in traffic knowledge base A6, use and search out other traffic scenes the highest to similar scene similarity based on this coefficient of paddy method for measuring similarity, and it is inserted among neighborhood scene set of similar scene;
Step C20: reappraise scene under described similar scene of each emerging traffic signals control program of relating in the set of described neighborhood scene-scheme score based on Slope One algorithm, and find out the top score scheme, then execution in step C17; Wherein, described emerging traffic signals control program is not for there being the scheme of scene-scheme call under described similar scene, described Slope One algorithm is according to the described scene that has the traffic signals control program of scene-scheme call under described similar scene that relates to of described neighborhood scene-scheme score, calculates the scene of described emerging traffic signals control program under described similar scene-scheme call;
Step C21: with the existing traffic signals control program comparison of carrying out on described recommendation traffic signals control program number and the to be switched crossing, if consistent, illustrate that then scheme base without better scheme, will not replace; If inconsistent, send to upgrade the message of crossing traffic signal timing plan to traffic control program storehouse, comprise to be switched road slogan and described recommendation traffic signals control program number;
Transport information assessment and processing module arrange by the fluctuation to the control performance score of the current traffic signals control program at target crossing at step C12 and allow threshold value to judge to start the traffic plan switching time among transport information storage and the processing layer A2, when wave amplitude exceeds the permission threshold value, think that idiographic flow is as follows because the control traffic state at road cross has occured to change so that existing traffic signals control program can't satisfy the traffic control demand:
Step SB1: n control performance parameter being obtained target crossing traffic signals control program i under current traffic scene at t crossing traffic evaluation time point by the crossing checkout equipment, such as the crossing vehicle mean delay time, the average passage rates of crossing vehicle etc. are designated as
Step SB2: described n control performance parameter carried out normalized, carry out according to following formula:
In the following formula
With
Be respectively the upper limit and the lower limit of d item performance index,
Score for the current traffic signals control program i at target crossing t crossing traffic evaluation time point on d item performance index;
Step SB3: the weighted sum of crossing traffic overall control performance, carry out according to following formula:
W in the following formula
dBe the weight of d item performance index,
It is the target crossing control performance PTS of t crossing traffic evaluation time point target crossing traffic signals control program i under current traffic scene;
Step SB4: calculate described control performance fluctuating range, carry out according to following formula:
In the following formula
When giving this target crossing for the current demand signal control program at target crossing is recommended, the PTS of the control performance of current demand signal control program i being assessed out for commending system under the at that time traffic scene at this target crossing.
Content-based recommendation is at traffic intersection and the typing of traffic control strategy feature and large class decision tree generation (the step B11～B13) finish by characteristic matching (step C13) under the support of traffic control scheme.The feature that feature extraction phases is extracted is structured message, is designated as respectively crossing feature IAttr={IAttr
1..., IAttr
nAnd scheme feature PAttr={PAttr
1..., PAttr
m, wherein n and m are the feature maximum number, and this step is namely finished the large class decision tree of traffic intersection and the large class decision tree stage of obtaining of traffic control scheme, and decision tree obtains all and constructs the most shallow decision tree by the ID3 algorithm.When the transport information assessment triggers content-based traffic recommending module with processing module, the traffic scene data to be switched crossing will be injected the large class decision tree of traffic control scheme and obtain the suitable large class of traffic control scheme, finish the characteristic matching stage.The large class of traffic control scheme that obtains will trigger based on the traffic recommending module of collaborative filtering further to be processed.
On-the-spot execution level workflow diagram in the embodiment of the invention as shown in Figure 5, wherein:
Step e 11: by crossing detector acquisition crossing traffic information;
Step e 12: submit transport information to processing layer A2 to the transport information storage;
Whether step e 13: intercepting simultaneously has from the transport information storage renewal crossing control program signal with processing layer A2, if having, then transmitting to step e 14 needs the new traffic signals control program switched number, if nothing continues to intercept;
Step e 14: judging whether needs to switch current traffic signals control program, if do not need, execution in step E15 if need, then begins the switching flow that carries into execution a plan, and namely begins execution in step E16;
Step e 15: the crossing traffic information that obtains as input, is carried out existing traffic signals control program, produce control signal.Local controller begin control in, existing traffic signals control program is default traffic signals control program;
Step e 16: with the crossing state of a control that records in the former traffic signals control program, such as current phase place, a upper phase place, when current phase place is green etc., pass to new traffic signals control program, finish the control program initial work;
Step e 17: former traffic signals control program will be controlled work handover to new traffic signals control program, and new traffic signals control program will be obtained crossing traffic information as input, produce control signal;
Step e 18: according to the physical communication agreement crossing control signal is finished the lamp control signal coding, realize traffic signals control function;
Above-described specific embodiment; purpose of the present invention, technical scheme and beneficial effect are further described; institute is understood that; the above only is specific embodiments of the invention; be not limited to the present invention; within the spirit and principles in the present invention all, any modification of making, be equal to replacement, improvement etc., all should be included within protection scope of the present invention.
Claims (10)
1. one kind based on the real-time choosing method of traffic signals control program of mix recommending, and the method comprises:
Off-line knowledge acquisition layer module construction standard crossing model and standard traffic scene, and obtain the traffic behavior parameter of each suitable traffic signals control program under the standard traffic scene, and described traffic behavior parameter and corresponding standard traffic scene data are sent to transport information store and the processing layer module;
Transport information storage and processing layer module are according to the traffic scene that calculates each traffic signals control program from the resulting described traffic behavior parameter of off-line knowledge acquisition layer module and standard traffic scene data-scheme score, also continue to fill and revise traffic scene-scheme call according to the real crossing traffic information that obtains from on-the-spot execution level module, whether the traffic signals control program of judging simultaneously real crossing needs to switch, and sends new traffic signals control program when needing to switch to on-the-spot execution level module;
On-the-spot execution level module is obtained real crossing traffic information, and the real crossing traffic information that obtains is sent to transport information storage and processing layer module; When receiving the renewal traffic signals control program message of transport information storage and the transmission of processing layer module, existing signal timing plan is switched to new traffic signals control program, and according to the physical communication agreement crossing control signal is carried out lamp control signal and encode, carry out described new traffic signals control program.
2. the method for claim 1 is characterized in that: off-line knowledge acquisition layer module also according to the crossing characteristic of input, generates the large class decision tree of traffic intersection by the ID3 decision trees; Crossing type and traffic scene data that the transport information storage is applicable according to described traffic signals control program with the processing layer module generate the large class decision tree of traffic signals control program by the ID3 decision trees; Wherein, the large class decision tree of described traffic signals control program is used for the traffic signals control program is sorted out, and the large class decision tree of described traffic intersection is used for all controlled crossings of traffic network are sorted out.
3. method as claimed in claim 2, it is characterized in that: off-line knowledge acquisition layer module makes up standard crossing model according to the large class in the crossing in the large class decision tree of traffic intersection, also obtain each traffic signals control program that is fit under the described standard traffic scene by the large class decision tree of the data based traffic signals control program of standard traffic scene, then go out the standard traffic scene and corresponding traffic signals control program is implemented simulation work based on standard crossing model construction, and the traffic behavior parameter that obtains and corresponding standard traffic scene data are sent to transport information store and the processing layer module.
4. the method for claim 1 is characterized in that: described transport information storage and processing layer module are calculated the traffic scene of traffic signals control program-scheme score by following formula:
Score in the following formula
iBe the traffic scene score of traffic signals control program i, γ is discount factor, span (0,1); T is crossing traffic evaluation time point in service, and span is [0, K]; The traffic behavior assessment comprises n crossing traffic Control performance standard, w
dBe the weight of d item traffic control performance index, d is the index numbering, and span is [1, n], wherein
For described traffic signals control program i t crossing traffic evaluation time the score o'clock on d item traffic control performance index, following calculating:
In the following formula
With
Be respectively the upper limit and the lower limit of d item performance index,
For put the d item control performance parameter value of the target crossing traffic signals control program i under current traffic scene that is obtained by the crossing checkout equipment t crossing traffic evaluation time.
5. the method for claim 1 is characterized in that: described transport information storage and processing layer module record standard traffic scene data and the real traffic scene data that described off-line knowledge acquisition layer module and on-the-spot execution level module provide in the traffic knowledge base; And described transport information storage judges with the processing layer module whether the traffic signals control program at real crossing needs switching specifically to comprise:
Step 1, basis judge from the transport information at the target crossing that on-the-spot execution level module is obtained whether the performance inconsistency amplitude of the signal timing plan at target crossing surpasses certain threshold value, if surpass then execution in step 2;
Step 2, obtain being fit to the large class of traffic signals control program of described traffic scene data from the large class decision tree of described traffic signals control program according to the current traffic scene data in described target crossing;
Step 3, judge whether the traffic signals control program that relates in the large class of described traffic signals control program unique, if unique then directly with this traffic signals control program as recommending the traffic signals control program, and execution in step 7; If not unique, then execution in step 4, to start the collaborative filtering recommending flow process;
The traffic scene data of storing in step 4, traffic scene data that described target crossing is current and the traffic knowledge base are mated, and obtain similar scene;
Step 5, find out the highest traffic signals control program of traffic scene-scheme score in all traffic signals control programs corresponding to described similar scene;
Step 6, judge that whether the score of the traffic signals control program that described score is the highest surpasses threshold value, if surpass, then the traffic signals control program that this score is the highest is as recommending the traffic signals control program, and execution in step 7;
Step 7, the existing traffic signals control program of carrying out on described recommendation traffic signals control program and the target crossing is compared, if inconsistent, then send upgrade the traffic signals control program message to transport information storage and processing layer module; If consistent, then do not need the switching signal control program.
6. method as claimed in claim 5 is characterized in that:
When the score of judging the traffic signals control program that described score is the highest does not surpass threshold value, carry out following steps in the described step 6:
Step 61: whether the neighborhood range size of judging described similar scene can increase, if can increase, then execution in step 62, if can not increase, then the traffic signals control program that described score is the highest is as recommending the traffic signals control program, and execution in step 7;
Step 62: in the traffic knowledge base, according to the traffic scene of different traffic signals control programs under the traffic scene in the neighborhood scope that increases-scheme score, search for the new traffic scene the highest to described similar scene similarity;
Step 63: obtain the traffic scene of each traffic signals control program corresponding with described new traffic scene under described similar scene-scheme score, and find out the top score scheme, then execution in step 6.
7. method as claimed in claim 5, it is characterized in that: the performance inconsistency amplitude computation process of the traffic signals control program at target crossing described in the described step 1 is as follows:
Step 11, obtained n the crossing traffic Control performance standard parameter of target crossing traffic signals control program i under current traffic scene by the crossing checkout equipment at t crossing traffic evaluation time point, be designated as
Step 12: carry out performance index parameter normalized according to following formula:
Wherein
With
Be respectively the upper limit and the lower limit of d item performance index parameter,
Be the current traffic signals control program i at the target crossing score at t crossing traffic evaluation time point on d item performance index;
Step 13: according to following formula to the weighted sum of crossing traffic overall control performance:
W wherein
dBe the weight of d item performance index,
Be the target crossing control performance PTS at t crossing traffic evaluation time point target crossing traffic signals control program i under current traffic scene;
Step 14: calculate described performance inconsistency amplitude, carry out according to following formula:
Wherein
When giving this target crossing for the current traffic signals control program at target crossing is recommended, the control performance PTS at this target crossing.
8. method as claimed in claim 5, it is characterized in that: when obtaining similar scene in the described step 4, utilize Pearson's correlation calculations to obtain the degree of correlation of the traffic scene stored in the current traffic scene at described target crossing and the traffic knowledge base, and then determine described similar scene, the traffic scene of storing in the wherein said traffic knowledge base comprises standard traffic scene and real traffic scene.
9. method as claimed in claim 6 is characterized in that: search for the new traffic scene the highest to similar scene similarity based on this coefficient of paddy method for measuring similarity in the described step 62.
10. one kind based on the real-time selecting device of traffic signals control program of mix recommending, and it comprises:
Off-line knowledge acquisition layer module, it comprises that the traffic simulation scene makes up module, scene simulation module and artificial transport information acquisition module, wherein the traffic simulation scene makes up module construction standard crossing model and standard traffic scene, each traffic signals control program that is fit under the scene simulation modular simulation various criterion traffic scene, artificial transport information acquisition module obtains the traffic behavior parameter in the simulation process, and described traffic behavior parameter and corresponding standard traffic scene data are sent to transport information storage and processing layer module;
Transport information storage and processing layer module, it comprises transport information assessment and processing module and traffic signals control program administration module; The assessment of wherein said transport information and processing module are according to the traffic scene that calculates each traffic signals control program from the resulting described traffic behavior parameter of off-line knowledge acquisition layer module and standard traffic scene data-scheme score, described transport information assessment also continues to fill and revise traffic scene-scheme call according to the real crossing traffic information that obtains from on-the-spot execution level module with processing module, whether the traffic signals control program of judging simultaneously real crossing needs to switch, if need to switch, then traffic signals control program administration module sends new traffic signals control program to on-the-spot execution level module;
On-the-spot execution level module, it comprises real crossing traffic information acquisition module, signal controlling execution module and signal timing plan handover module; Wherein said real crossing traffic information acquisition module obtains real crossing traffic information, and the real crossing traffic information that obtains is sent to transport information storage and processing layer module; Described signal timing plan handover module switches to new traffic signals control program with existing traffic signals control program when receiving the renewal traffic signals control program message of transport information storage and the transmission of processing layer module; Described signal controlling execution module carries out the lamp control signal coding according to the physical communication agreement to the crossing control signal, carries out the traffic signals control program of upgrading.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210568417.9A CN103050016B (en) | 2012-12-24 | 2012-12-24 | Hybrid recommendation-based traffic signal control scheme real-time selection method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210568417.9A CN103050016B (en) | 2012-12-24 | 2012-12-24 | Hybrid recommendation-based traffic signal control scheme real-time selection method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103050016A true CN103050016A (en) | 2013-04-17 |
| CN103050016B CN103050016B (en) | 2015-03-25 |
Family
ID=48062638
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210568417.9A Expired - Fee Related CN103050016B (en) | 2012-12-24 | 2012-12-24 | Hybrid recommendation-based traffic signal control scheme real-time selection method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103050016B (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104050319A (en) * | 2014-06-13 | 2014-09-17 | 浙江大学 | Method for realtime online verification of complex traffic control algorithm |
| CN105678430A (en) * | 2016-02-29 | 2016-06-15 | 大连大学 | Improved user recommendation method based on neighbor project slope one algorithm |
| CN107730890A (en) * | 2017-11-09 | 2018-02-23 | 石数字技术成都有限公司 | A kind of intelligent transportation method based on wagon flow speed prediction under real-time scene |
| CN108648446A (en) * | 2018-04-24 | 2018-10-12 | 浙江工业大学 | A kind of road grid traffic signal iterative learning control method based on MFD |
| CN109299531A (en) * | 2018-09-12 | 2019-02-01 | 清华四川能源互联网研究院 | Electromagnetical transient emulation method and device |
| CN109859475A (en) * | 2019-03-14 | 2019-06-07 | 江苏中设集团股份有限公司 | A kind of intersection signal control method based on DBSCAN Density Clustering, apparatus and system |
| CN110047284A (en) * | 2019-04-16 | 2019-07-23 | 浙江工业大学 | A kind of Traffic Signal Timing Optimal Decision-making support method based on expert system |
| CN112037539A (en) * | 2020-07-31 | 2020-12-04 | 银江股份有限公司 | Method and system for recommending signal control scheme for saturated urban traffic network |
| CN113450559A (en) * | 2020-03-27 | 2021-09-28 | 华为技术有限公司 | Method and device for auditing control scheme of traffic light |
| CN113793527A (en) * | 2021-09-14 | 2021-12-14 | 北京石油化工学院 | Test verification system for urban traffic active control |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2007122584A (en) * | 2005-10-31 | 2007-05-17 | Sumitomo Electric Ind Ltd | Traffic signal control system and control method of traffic signal control system |
| CN101464923A (en) * | 2009-01-16 | 2009-06-24 | 天津大学 | Traffic control, inducement and cooperation oriented simulation intelligent traffic system |
| CN101470955A (en) * | 2007-12-26 | 2009-07-01 | 奥城同立科技开发（北京）有限公司 | Integrated control system for road junction traffic |
| CN101710451A (en) * | 2009-12-18 | 2010-05-19 | 浙江富阳市新源交通电子有限公司 | Control method and control device of traffic pass signals |
| US20120038490A1 (en) * | 2007-06-29 | 2012-02-16 | Orion Energy Systems, Inc. | Outdoor lighting fixtures for controlling traffic lights |
-
2012
- 2012-12-24 CN CN201210568417.9A patent/CN103050016B/en not_active Expired - Fee Related
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2007122584A (en) * | 2005-10-31 | 2007-05-17 | Sumitomo Electric Ind Ltd | Traffic signal control system and control method of traffic signal control system |
| US20120038490A1 (en) * | 2007-06-29 | 2012-02-16 | Orion Energy Systems, Inc. | Outdoor lighting fixtures for controlling traffic lights |
| CN101470955A (en) * | 2007-12-26 | 2009-07-01 | 奥城同立科技开发（北京）有限公司 | Integrated control system for road junction traffic |
| CN101464923A (en) * | 2009-01-16 | 2009-06-24 | 天津大学 | Traffic control, inducement and cooperation oriented simulation intelligent traffic system |
| CN101710451A (en) * | 2009-12-18 | 2010-05-19 | 浙江富阳市新源交通电子有限公司 | Control method and control device of traffic pass signals |
Non-Patent Citations (2)
| Title |
|---|
| 于泉，等: "信号交叉口固定配时控制方案可靠度研究", 《北京工业大学学报》, vol. 33, no. 10, 31 October 2007 (2007-10-31) * |
| 高云峰，等: "交叉口信号控制方案评价指标动态估计模型", 《同济大学学报（自然科学版）》, vol. 39, no. 6, 30 June 2011 (2011-06-30) * |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104050319B (en) * | 2014-06-13 | 2017-10-10 | 浙江大学 | A kind of method of the complicated traffic control algorithm of real-time online checking |
| CN104050319A (en) * | 2014-06-13 | 2014-09-17 | 浙江大学 | Method for realtime online verification of complex traffic control algorithm |
| CN105678430A (en) * | 2016-02-29 | 2016-06-15 | 大连大学 | Improved user recommendation method based on neighbor project slope one algorithm |
| CN107730890A (en) * | 2017-11-09 | 2018-02-23 | 石数字技术成都有限公司 | A kind of intelligent transportation method based on wagon flow speed prediction under real-time scene |
| CN107730890B (en) * | 2017-11-09 | 2021-04-20 | 一石数字技术成都有限公司 | Intelligent transportation method based on traffic flow speed prediction in real-time scene |
| CN108648446B (en) * | 2018-04-24 | 2020-08-21 | 浙江工业大学 | Road network traffic signal iterative learning control method based on MFD |
| CN108648446A (en) * | 2018-04-24 | 2018-10-12 | 浙江工业大学 | A kind of road grid traffic signal iterative learning control method based on MFD |
| CN109299531A (en) * | 2018-09-12 | 2019-02-01 | 清华四川能源互联网研究院 | Electromagnetical transient emulation method and device |
| CN109859475A (en) * | 2019-03-14 | 2019-06-07 | 江苏中设集团股份有限公司 | A kind of intersection signal control method based on DBSCAN Density Clustering, apparatus and system |
| CN110047284A (en) * | 2019-04-16 | 2019-07-23 | 浙江工业大学 | A kind of Traffic Signal Timing Optimal Decision-making support method based on expert system |
| CN113450559A (en) * | 2020-03-27 | 2021-09-28 | 华为技术有限公司 | Method and device for auditing control scheme of traffic light |
| CN112037539A (en) * | 2020-07-31 | 2020-12-04 | 银江股份有限公司 | Method and system for recommending signal control scheme for saturated urban traffic network |
| CN112037539B (en) * | 2020-07-31 | 2021-11-16 | 银江技术股份有限公司 | Method and system for recommending signal control scheme for saturated urban traffic network |
| CN113793527A (en) * | 2021-09-14 | 2021-12-14 | 北京石油化工学院 | Test verification system for urban traffic active control |
| CN113793527B (en) * | 2021-09-14 | 2022-11-18 | 北京石油化工学院 | Test verification system for urban traffic active control |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103050016B (en) | 2015-03-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103050016B (en) | Hybrid recommendation-based traffic signal control scheme real-time selection method | |
| Lipscomb et al. | Development effects of electrification: Evidence from the topographic placement of hydropower plants in Brazil | |
| Guo et al. | Efficiency assessment of transit-oriented development by data envelopment analysis: Case study on the Den-en Toshi line in Japan | |
| Gallo et al. | The transit network design problem with elastic demand and internalisation of external costs: An application to rail frequency optimisation | |
| Guo et al. | Modeling parking behavior under uncertainty: a static game theoretic versus a sequential neo-additive capacity modeling approach | |
| CN101325004B (en) | Method for compensating real time traffic information data | |
| Kim | Integrated urban systems modeling: theory and applications | |
| CN106447119A (en) | Short-term traffic flow prediction method and system based on convolutional neural network | |
| CN102610092A (en) | Urban road speed predication method based on RBF (radial basis function) neural network | |
| CN104915897B (en) | A kind of computer implemented method of Electric Power Network Planning evaluation assignment | |
| CN101930481A (en) | Method used for generating CIM model describing power grid change in designated time slot and system thereof | |
| CN110956807A (en) | Highway flow prediction method based on combination of multi-source data and sliding window | |
| Wegener | Transport network equilibrium and regional deconcentration | |
| CN108985518A (en) | A kind of traffic congestion charging policy optimization method based on VISSIM emulation | |
| CN104537446A (en) | Bilevel vehicle routing optimization method with fuzzy random time window | |
| Pfaffenbichler | The strategic, dynamic and integrated urban land use and transport model MARS (Metropolitan Activity Relocation Simulator) | |
| Wei et al. | A two-layer network dynamic congestion pricing based on macroscopic fundamental diagram | |
| CN104680782B (en) | Traffic control cloud system | |
| Song | Infrastructure and Urban development: Evidence from Chinese cities | |
| Li et al. | Express Parcel Packaging Waste Recycling Platform | |
| Yang et al. | An adaptive workflow to generate street network and amenity allocation for walkable neighborhood design | |
| Yu et al. | Optimization of urban bus operation frequency under common route condition with rail transit | |
| Jin | Modelling the influence of neighbourhood design on daily trip patterns in urban neighbourhoods | |
| Oubnaki et al. | Energy Consumption in the Transport Sector: Trends and Forecast Estimates in Morocco | |
| İnaç et al. | Smartness and Strategic Priority Assessment in Transition to Mobility 4.0 for Smart Cities |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20150325 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |