CN108154688B - The through street On-ramp Control method and system of iterative learning under packet loss environment - Google Patents
The through street On-ramp Control method and system of iterative learning under packet loss environment Download PDFInfo
- Publication number
- CN108154688B CN108154688B CN201711336908.XA CN201711336908A CN108154688B CN 108154688 B CN108154688 B CN 108154688B CN 201711336908 A CN201711336908 A CN 201711336908A CN 108154688 B CN108154688 B CN 108154688B
- Authority
- CN
- China
- Prior art keywords
- traffic
- control
- street
- error
- function
- 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.)
- Expired - Fee Related
Links
- 238000000034 method Methods 0.000 title claims abstract description 30
- 230000006870 function Effects 0.000 claims abstract description 85
- 230000005540 biological transmission Effects 0.000 claims abstract description 19
- 230000003044 adaptive effect Effects 0.000 claims abstract description 14
- 238000001514 detection method Methods 0.000 claims description 35
- 230000004907 flux Effects 0.000 claims description 35
- 238000004891 communication Methods 0.000 claims description 19
- 238000005070 sampling Methods 0.000 claims description 3
- 238000012544 monitoring process Methods 0.000 claims 1
- 238000011160 research Methods 0.000 description 4
- 230000008859 change Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 230000006399 behavior Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000452 restraining effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/07—Controlling traffic signals
- G08G1/075—Ramp control
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/07—Controlling traffic signals
- G08G1/08—Controlling traffic signals according to detected number or speed of vehicles
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
- Traffic Control Systems (AREA)
Abstract
The present invention provides a kind of through street On-ramp Control method and system of iterative learning under packet loss environment.The actual traffic current density value and desired traffic current density that the method needs not lose obtain error current function, by error function and iterative learning control law and learning gains based on switching system feature under the adaptive forgetting factor setting transmission packet loss environment based on error.Iterative learning control law based on switching system feature under transmission packet loss environment is applied in the urban expressing system with ring road entrance, makes it possible to that through street is enable to reach desired traffic current density in certain the number of iterations.The method of the present invention not only can solve the problem of through street control with ring road crossing under transmission packet loss environment with switching system characteristic, but also can improve the problem that traffic control convergence rate is slack-off as caused by data packetloss, and preferably gear to actual circumstances needs.
Description
Technical field
The present invention relates to intelligent transportation fields, more particularly, to the through street entrance circle of iterative learning under packet loss environment
Channel control method and system.
Background technique
Through street On-ramp Control is the important component of intelligent transportation field.Traffic flow model has repetition daily
Property, but influenced since the factors such as the movable flow of personnel of regional society, weather, road conditions change, in practice, through street model system
System parameter changes therewith, and urban expressing system has switching system characteristic, therefore through street On-ramp Control is very at this stage
It is mostly based on the research under switching system characteristic.
Expressway Traffic control belongs to long-range control scene, and actually physical distance is distant between through street and control centre
Far, and the measuring device of through street often lacks maintenance and exposure field, so that measurement data output is likely to occur with transmission
The case where packet loss or delay.Many research achievements are all based on the control ideally for not having to transmit packet loss at this stage
Scene, cannot preferably gear to actual circumstances needs.Therefore the through street On-ramp Control practicability based on iterative learning reduces.
Summary of the invention
Primary and foremost purpose of the present invention be to provide under a kind of packet loss environment the through street On-ramp Control method of iterative learning and
System, to solve the control problem of the through street with ring road crossing under packet loss environment with switching system characteristic.
The present invention also provides a kind of through street Entrance ramps with unilateral transmission packet drop iterative learning remotely to control
System.
In order to solve the above technical problems, technical scheme is as follows:
The through street On-ramp Control method of iterative learning under packet loss environment, comprising the following steps:
S1: control centre obtains current Expressway Traffic Flow density value by communication module；
S2: control centre obtains the traffic flow control that expectation the traffic flow density value and ring road entrance of middle through street is locally stored
Amount processed；
S3: current error function is obtained according to the actual traffic current density value and desired traffic flow density value that receive
ek(t)=αk(t)(yd(t)-yk(t))
Wherein ydIt (t) is desired traffic flow density, yk(t) the actual traffic stream that control centre receives when iteration secondary for kth
Density, αkIt (t) is the y of kth time iterationk(t) sequence of events of random loss, the biography of quick terminal to control centre end do not occur
Defeated drop probabilities are
S4: being based on error function, calculates forgetting factor value, and calculation formula is as follows
Wherein γ (k) to be in the kth time iteration error sequence of function be not 0 element number, θmaxWith θminFor θ (k) function
Dividing value up and down, φ (k) is maximum selection rule function, for having selected from the 1st iteration to kth maximum error since time iteration
Function mean-square value；
S5: according to the error function and forgetting factor, it is arranged based on changing with switching characteristic under transmission packet drop
Generation study control rate and iterative learning gain.Wherein iterative learning control rate is expressed as follows:
uk+1(t)=(1- αk(t+1)θ(k+1))sat[uk(t)]+Γiek(t+1)
In formula, uk+1(t) traffic flow of the ring road entrance of+1 iteration of kth, u are indicatedk(t) circle of kth time iteration is indicated
The traffic flow of road entrance, θ (k) are adaptive forgetting factor function, error function ek(t+1) the kth time iteration t+1 moment is indicated
Output error, and ΓiFor the control gain of i-th of switching system subsystem, saturation function sat [], expression formula was as follows:
Wherein
Wherein j ∈ { 1,2 ..., K }, K are the ring road number in one section of through street,Ring road j when iteration secondary for kth
Traffic flux detection amount, u(j),max(t) and u(j),min(t) be respectively ring road j Traffic flux detection amount saturation bound；
S6: control centre generates fast road ramp entrance in conjunction with the iterative learning control law and iterative learning gain
Traffic flux detection signal, and quick terminal is sent the signal to by reliable communication mode；
S7: quick terminal receives the Traffic flux detection signal of ring road entrance by reliable communication mode；
S8: Traffic flux detection signal is applied to the control of the urban expressing system with ring road entrance, is made it possible to certain
Through street is set to reach desired traffic current density in the number of iterations；
S9: Expressway Traffic density current signal is sent to control centre by quick terminal.
The through street Entrance ramp tele-control system of iterative learning under a kind of packet loss environment, comprising:
Current flows obtain module: it is close to obtain current through street actual traffic stream by correspondence for control centre
Degree；
Traffic information acquisition module: it is obtained for control centre and it is expected traffic flow, ring road entrance in local memory device
Traffic flux detection amount；
Error function computing module: close according to the currently practical traffic flow received for calculating error current function
Angle value and desired traffic flow density, which calculate, obtains error current function
ek(t)=αk(t)(yd(t)-yk(t))
Wherein ydIt (t) is desired traffic flow density, yk(t) the actual traffic stream that control centre receives when iteration secondary for kth
Density, αkIt (t) is the y of kth time iterationk(t) sequence of events of random loss, the biography of quick terminal to control centre end do not occur
Defeated drop probabilities are
Forgetting factor computing module: for calculating forgetting factor value according to error function, calculation formula is as follows:
Wherein γ (k) to be in the kth time iteration error sequence of function be not 0 element number, θmaxWith θminFor θ (k) function
Dividing value up and down, φ (k) is maximum selection rule function, for having selected from the 1st iteration to kth maximum error since time iteration
Function mean-square value；
Iterative learning control law and learning gains setup module: for being transmitted according to error function and adaptive forgetting factor setting
Iterative learning control law and iterative learning gain under packet drop based on switching system feature, wherein iterative learning control
Rule is expressed as follows:
uk+1(t)=(1- αk(t+1)θ(k+1))sat[uk(t)]+Γiek(t+1)
In formula, uk+1(t) traffic flow of the ring road entrance of+1 iteration of kth, u are indicatedk(t) circle of kth time iteration is indicated
The traffic flow of road entrance, θ (k) are adaptive forgetting factor function, error function ek(t+1) the kth time iteration t+1 moment is indicated
Output error, and ΓiFor the control gain of i-th of switching system subsystem, saturation function sat [], expression formula was as follows:
Wherein
Wherein j ∈ { 1,2 ..., K }, K are the ring road number in one section of through street,Ring road j when iteration secondary for kth
Traffic flux detection amount, u(j),max(t) and u(j),min(t) be respectively ring road j Traffic flux detection amount saturation bound；
Control signal output module: for the Traffic flux detection signal of control centre's generation fast road ramp entrance, and with
Reliable communication mode is sent to quick terminal；
Control signal acquisition module: Traffic flux detection signal is obtained by reliable communication mode for quick terminal；
Traffic control module: for Traffic flux detection signal to be applied to the urban expressing system control with ring road entrance,
Make it possible to that through street is enable to reach desired traffic current density in certain the number of iterations；
Traffic current density output module: pass through communication mode for current Expressway Traffic Flow density signal for quick terminal
It is sent to control centre.
Compared with prior art, the beneficial effect of technical solution of the present invention is: the present invention is provided under a kind of packet loss environment
Through street On-ramp Control method, the switching system characteristic based on through street Entrance ramp, according to not transmitting loss actually
Traffic flow density value and it is expected traffic flow density value acquisition error current function；According to error function with based on the adaptive of error
Answer the iterative learning control law under forgetting factor setting transmission packet loss environment based on switching system feature and iterative learning gain；
The control signal that the iterative learning control law transmitted under packet loss environment based on switching system feature generates is sent to ring road
In the urban expressing system of entrance, through street can be enabled to reach desired traffic current density in certain the number of iterations.This hair
Bright method can not only solve the through street control problem with ring road crossing under packet loss environment with switching system characteristic, together
When can improve the problem that traffic control convergence rate is slack-off as caused by data packetloss, preferably gear to actual circumstances needs.
Detailed description of the invention
Fig. 1 is the flow chart of the through street On-ramp Control method of iterative learning under 1 packet loss environment of embodiment.
Fig. 2 is through street illustraton of model in embodiment 1.
Fig. 3 be in embodiment 1 under an application scenarios under the packet loss environment iterative learning control law block diagram.
Fig. 4 be in embodiment 1 under an application scenarios under the packet loss environment urban expressing system of iterative learning one kind
Switching law.
Fig. 5 be in embodiment 1 under an application scenarios in urban expressing system each section ring road entrance vehicle demand.
Fig. 6 be in embodiment 1 under an application scenarios in urban expressing system each section ramp exit vehicle flowrate.
Fig. 7 is the traffic flow output error index analysis figure in each section under an application scenarios of the invention.
Fig. 8 is the output error index of the Expressway Traffic of two kinds of control methods and and reason under an application scenarios of the invention
Think the output error Indexes Comparison figure of the Expressway Traffic controlled under no packet loss environment.
Fig. 9 is for the embodiment of the present invention 1 with unilateral transmission packet drop based on the fast of switching system characteristic iterative learning
The schematic diagram of fast road Entrance ramp tele-control system.
Specific embodiment
The attached figures are only used for illustrative purposes and cannot be understood as limitating the patent；
In order to better illustrate this embodiment, the certain components of attached drawing have omission, zoom in or out, and do not represent actual product
Size；
To those skilled in the art, it is to be understood that certain known features and its explanation, which may be omitted, in attached drawing
's.
The following further describes the technical solution of the present invention with reference to the accompanying drawings and examples.
Embodiment 1
As shown in Figure 1, under a kind of packet loss environment iterative learning through street On-ramp Control method, including following step
It is rapid:
101, the actual traffic current density of current through street is obtained；Firstly, control centre is received currently by communication mode
The reality output traffic current density of the number of iterations；
102, the ring road entrance traffic flow control of the through street expectation traffic flow density value and current iteration that are locally stored is obtained
Amount processed；
103, the traffic current density and the expectation traffic flow density value that are an actually-received according to and obtain the side
The error current function e of methodk(t)；
ek(t)=αk(t)(yd(t)-yk(t))
Wherein ydIt (t) is desired traffic flow density, yk(t) the actual traffic stream that control centre receives when iteration secondary for kth
Density, αkIt (t) is the y of kth time iterationk(t) sequence of events of random loss, the biography of quick terminal to control centre end do not occur
Defeated drop probabilities are
After obtaining the expectation traffic flow density value of actual traffic current density value and through street of current through street, according to fast
The difference of the actual traffic current density value of the expectation traffic flow density value and current through street on fast road, is as a result multiplied by αk(t), it can obtain
To the error function e for working as previous iterationk(t)；
104, according to the error function e of current iterationk(t) forgetting factor θ (k+1) is calculated, calculation formula is as follows
Wherein γ (k) to be in the kth time iteration error sequence of function be not 0 element number, θmaxWith θminFor θ (k) function
Dividing value up and down.φ (k) is maximum selection rule function, for having selected from the 1st iteration to kth maximum error since time iteration
Function mean-square value；
105, according to the error function and forgetting factor setting iterative learning control law and iterative learning gain, so that described
The error criterion of method is restrained in certain the number of iterations, wherein is being obtained after previous iteration error function, according to the mistake
Difference function provides the iterative learning control law based on switching system under transmission packet loss environment and iterative learning gain is arranged
uk+1(t)=(1- αk(t+1)θ(k+1))sat[uk(t)]+Γiek(t+1)
In formula, uk+1(t) traffic flow of the ring road entrance of+1 iteration of kth, u are indicatedk(t) circle of kth time iteration is indicated
The traffic flow of road entrance, θ (k) are adaptive forgetting factor function.Error function ek(t+1) the kth time iteration t+1 moment is indicated
Output error, and ΓiFor the control gain of i-th of switching system subsystem, saturation function sat [], expression formula was as follows:
Wherein
Wherein j ∈ { 1,2 ..., K }, K are the ring road number in one section of through street.Ring road j when iteration secondary for kth
Traffic flux detection amount, u(j),max(t) and u(j),min(t) be respectively ring road j Traffic flux detection amount saturation bound；
106, according to the Traffic flux detection signal of the law of learning and iterative learning Gain generating fast road ramp entrance, lead to
It crosses reliable communication and is transferred to Expressway Traffic；
107, the control signal that control centre sends is obtained, quick terminal receives ring road entrance by reliable communication mode
Traffic flux detection signal；
108, the Traffic flux detection signal received is applied to the control of the urban expressing system with ring road entrance, made it possible to
It is enough that through street is enable to reach desired traffic current density in certain the number of iterations；
109, Expressway Traffic density current signal is sent to control centre by quick terminal.
In the present embodiment, firstly, obtaining the actual traffic current density of current through street；Secondly, control centre obtains quickly
The Traffic flux detection amount of the ring road entrance of the expectation traffic flow density value and current iteration number on road；Then, according to described current
The traffic current density and the expectation traffic flow density value that are an actually-received obtain the error function of the method；Then, according to
The error function and adaptive forgetting factor setting iterative learning control law and iterative learning gain based on error；Followed by,
It is transferred to quick terminal according to the control signal of the law of learning and iterative learning Gain generating, is applied to special based on switching system
The through street On-ramp Control of property, so that the error criterion of the method is restrained in certain the number of iterations；Finally, by quick
Road traffic density stream signal is sent to control centre.The present embodiment is based on iterative learning control method in existing urban expressing system
Further furtherd investigate in research so that quickly path control system consider practical factor, i.e., by weather, road conditions etc. because
Element, through street model system parameter change therewith.Furthermore physical distance is remote between through street and control centre, and quickly
The measuring device on road, which often lacks, to be safeguarded and exposes field.Therefore urban expressing system meets switching system characteristic simultaneously and may have survey
The practical problem of amount output packet loss, more meets the needs of practical application.
For ease of understanding, according to the embodiment of Fig. 1, below with a practical application scene to one in the embodiment of the present invention
The iterative learning through street On-ramp Control method based on switching system characteristic is described under kind packet loss environment:
Fig. 2 is the traffic flow model that the present invention uses, which is divided into a described through street more
A section, there are an Entrance ramp and one outlet ring road in each section, as shown, traffic flow model is as follows:
q(j)(t)=ρ(j)(t)v(j)(t)；
Wherein, T is sampling period (hour), and t indicates the sampling interval, and K indicates that the section is divided into K sub- sections, j ∈
{ 1,2 ..., K } indicate the label on each subsegment road.The meaning of other model variables is as follows: ρ(j)(t) (veh/lane/km) table
Show the averag density of jth section；v(j)(t) (km/h) indicates the average speed of jth section；q(j)(t) (veh/h) indicate from jth section into
Enter+1 section of jth of vehicle flowrate；r(j)(t) (veh/h) indicates the vehicle flowrate that the ring road entrance of jth section enters；s(j)(t)(veh/h)
Indicate the vehicle flowrate of jth section ramp exit outflow；L(j)(km) length of jth Duan Lu is indicated；vfreeIndicate free stream velocity；ρjam
The maximum potential density in single lane；τ, υ, κ, l, m are normal parameters, reflect the road geometrical feature of special traffic system, drive
Member behavior and driving vehicle feature etc..
In this practical application scene, each section of section all includes a ring road entrance and a ramp exit.And it hands over
Variable-definition in through-flow model is
X (t)=[v(1)(t) v(2)(t) ... v(K)(t)]T, y (t)=[ρ(1)(t) ρ(2)(t) ... ρ(K)(t)]T,
U (t)=[r(1)(t) r(2)(t) ... r(K)(t)]T, s (t)=[s(1)(t) s(2)(t) ... s(K)(t)]T,
Wherein,j∈{1,
2,...,K}.Through street model has repeat property on t ∈ { 0,1,2 ..., N }, can be expressed as following state space table
Up to formula:
yk(t+1)=A (xk(t),t)yk(t)+B(sat[uk(t)])+η(xk(t))-Bs(t)
xk(t+1)=C (xk(t),t)+D(yk(t),t)
Wherein, k is the number of iterations, and C (x (t), t), D (x (t), t) are two nonlinear functions.Due in finite time section
The vehicle flowrate of entrance is limited and cannot be negative, therefore uk(t) saturation by ring road physical condition is limited.In real system
In, the parameter of urban expressing system can change with environmental factor, and in these parameters, being affected maximum is vfree,
ρjam.Axis changes the two parameters at any time, has repeat property on iteration axis, therefore meets switching system characteristic, expresses
Formula can indicate following:
Wherein i=i (t) be in t ∈ { 0,1,2 ..., N } switching law, value in finite sequence P={ 1,2 ..., m },
M is the number of subsystem.At t ∈ { 0,1,2 ..., N }, although because the outflow of ring road vehicle flowrate has some model uncertainties
And interference, system inputSo that system exportsAlso it can reach expectation
Export yd(t)=[yd,(1)(t) yd,(2)(t) ... yd,(K)(t)]T。
And for the traffic flow density signal y of outputk(t), the packet loss for being transmitted to control centre end from quick terminal is general
Rate isThen control centre end is according to the actual traffic current density signal received and the expectation traffic current density being locally stored
It is worth error letter obtained to be defined as
ek(t)=ak(t)(yd(t)-ak(t)yk(t))=ak(t)(yd(t)-yk(t))
Define αkIt (t) is the y of kth time iterationk(t) sequence of events of random loss does not occur.
Fig. 3 is being obtained after previous iteration error function, provides base under transmission packet loss environment according to the error function
In the iterative learning control law of switching system.
uk+1(t)=(1- αk(t+1)θ(k+1))sat[uk(t)]+Γiek(t+1)
K=12 indicates the traffic flow of the ring road entrance of+1 iteration of kth, uk(t) the ring road entrance of kth time iteration is indicated
Traffic flow, θ (k) are adaptive forgetting factor function.αkIt (t) is the y of kth time iterationk(t) the event sequence of random loss does not occur
Column, error function ek(t+1) output error at expression kth time iteration t+1 moment, and ΓiFor i-th switching system subsystem
Gain is controlled, saturation function sat [], expression formula was as follows:
Wherein
Wherein j ∈ { 1,2 ..., K }, K are the ring road number in one section of through street.Ring road j when iteration secondary for kth
Traffic flux detection amount, u(j),max(t) and u(j),min(t) be respectively ring road j Traffic flux detection amount saturation bound.
Under this application scene, we have K=5and L(j)=500m, j ∈ { 1,2 ..., K }, yk(t) losing probability is
0.3, there are a ring road entrance and a ramp exit in each section.The upper and lower of forgetting factor is set and is limited to θmin=0.0, θmax
=0.03, the tolerable error for restraining index is 0.3.
Fig. 4 is the switching law under this application scene.Because in practical applications, being influenced by factors such as weather, road conditions, fastly
Fast road model system parameter changes therewith, and in these parameters, being affected maximum is vfree, ρ jam.Should use
Under scape, urban expressing system switches between two systems on a timeline, switching law such as Fig. 4.
Fig. 5 is the ring road entrance vehicle demand in each section in urban expressing system under this application scene.
Fig. 6 is the vehicle flowrate of the ramp exit in each section in urban expressing system under this application scene.
Fig. 7 is the output error indicatrix in each section under this application scene.In this application scene, it can be seen that error refers to
It being marked in limited times and rapidly converges within tolerable error, error criterion is,J=1,2,3,4,5, that is, error criterion is that the output in each section misses
Mean square error of the difference on set time section.
It can be seen from figure 7 that the 3rd section, the 4th section with the error criterion of the 5th section of through street the 30th iteration convergence extremely
Within tolerable error, i.e., this two sections of wagon flow metric densities reach approximately desired wagon flow metric density.Then, paragraph 1 and the 2nd section of through street
Error criterion according to the 40th iteration convergence to tolerable error within, i.e., this two sections of wagon flow metric densities reach approximately desired vehicle flowrate
Density.
Fig. 8 is for the Expressway Traffic output error index of two kinds of control methods under this application scene and without transmission packet loss environment
The Expressway Traffic output error Indexes Comparison figure of lower control.In this application scene, error criterion is,J=1,2,3,4,5.It can be seen that, use the transmission packet loss ring
Iterative learning through street On-ramp Control method under border based on switching system characteristic, the error criterion of through street is at the 40th time
Within iteration convergence to tolerable error, i.e., Expressway Traffic wagon flow metric density reaches approximately desired wagon flow metric density.It is than transmission
Less iteration is needed using the Expressway Traffic output error index convergence of conventional iterative learning control method under packet loss environment
Number, and it is close using the number of iterations of conventional iterative learning control method under packet loss environment with without transmitting.It can be seen that institute
State method not only can enable through street to reach desired traffic current density in certain the number of iterations, while can improve due to number
According to the problem that traffic control convergence rate caused by packet loss is slack-off.
Embodiment 2
As shown in figure 9, a kind of through street based on switching system characteristic iterative learning with unilateral transmission packet drop
Entrance ramp tele-control system, comprising:
Current flows obtain module 901: obtaining current through street actual traffic stream by correspondence for control centre
Density；
Traffic information acquisition module 902: it is obtained for control centre and it is expected traffic flow, ring road entrance in local memory device
Traffic flux detection amount；
Error function computing module 903: for calculating error current function.It is according to the currently practical traffic flow received
Density value and desired traffic flow density, which calculate, obtains error current function；
ek(t)=αk(t)(yd(t)-yk(t))
αkIt (t) is the y of kth time iterationk(t) sequence of events of random loss, quick terminal to control centre end do not occur
Transmitting drop probabilities is
Forgetting factor computing module 904: for calculating forgetting factor value according to error function, calculation formula is as follows
Wherein γ (k) not to be in the kth time iteration error sequence of function be not 0 element number.
Iterative learning control law and learning gains setup module 905: for being arranged according to error function and adaptive forgetting factor
Transmit the iterative learning control law under packet drop based on switching system feature and iterative learning gain, wherein iterative learning
Control law is expressed as follows:
uk+1(t)=(1- αk(t+1)θ(k+1))sat[uk(t)]+Γiek(t+1)
In formula, uk+1(t) traffic flow of the ring road entrance of+1 iteration of kth, u are indicatedk(t) circle of kth time iteration is indicated
The traffic flow of road entrance, θ (k) are adaptive forgetting factor function, error function ek(t+1) the kth time iteration t+1 moment is indicated
Output error, and ΓiFor the control gain of i-th of switching system subsystem, saturation function sat [], expression formula was as follows:
Wherein
Wherein j ∈ { 1,2 ..., K }, u(j),max(t) and u(j),min(t) be respectively gate j saturation function bound；
It controls signal output module 906: generating the Traffic flux detection signal of fast road ramp entrance for control centre, and
Quick terminal is sent in a manner of reliable communication；
Control signal acquisition module 907: Traffic flux detection signal is obtained by reliable communication mode for quick terminal；
Traffic control module 908: for Traffic flux detection signal to be applied to the urban expressing system control with ring road entrance
System, makes it possible to that through street is enable to reach desired traffic current density in certain the number of iterations；
Traffic current density output module 909: pass through communication mode for current Expressway Traffic current density for quick terminal
Signal is sent to control centre.
In the present embodiment, firstly, the current flows of traffic current density reality output obtain module 901, it is fast for obtaining
The actual traffic current density for the current main road that fast road traffic is sent；Secondly, traffic information acquisition module 902, for obtaining expectation
Traffic flow density value with and current iteration ring road entrance Traffic flux detection amount；Then, error function obtains module 903, uses
The error current of the method is obtained in the traffic current density and the expectation traffic flow density value being an actually-received according to
Function；Then, forgetting factor computing module 904, for calculating forgetting factor value according to error function；Followed by iteration
Rule and learning gains setup module 905 are practised, for iterative learning control law to be arranged according to the error function and adaptive forgetting factor
Iterative learning gain and transmission packet loss environment under the iterative learning control law based on switching system；Followed by control signal
Output module 906: the fast road ramp entrance Traffic flux detection signal of generation is sent to through street by reliable communication mode
Traffic；Traffic flux detection signal is obtained by reliable communication mode followed by control signal acquisition module 907: quick terminal；Most
Eventually, traffic control module 908: Traffic flux detection signal is input in the urban expressing system with ring road entrance, is made it possible to
It is enough that through street is enable to reach desired traffic current density in certain the number of iterations；In addition, traffic current density output module 909:
For sending current Expressway Traffic Flow density signal to control centre.The present embodiment is based on iteration in existing urban expressing system
It is further furtherd investigate in learning control method research, so that quickly path control system considers practical factor, i.e., by day
The factors such as gas, road conditions.Furthermore physical distance is remote between through street and control centre, and the measuring device of through street often lacks
It safeguards and exposes field.Therefore urban expressing system meets switching system characteristic simultaneously and may have actually asking for measurement output packet loss
Topic, is controlled the needs for more meeting practical application in this case.
The same or similar label correspond to the same or similar components；
The terms describing the positional relationship in the drawings are only for illustration, should not be understood as the limitation to this patent；It is aobvious
So, the above embodiment of the present invention be only to clearly illustrate example of the present invention, and not be to reality of the invention
Apply the restriction of mode.For those of ordinary skill in the art, it can also make on the basis of the above description other
Various forms of variations or variation.There is no necessity and possibility to exhaust all the enbodiments.It is all in spirit of the invention
With any modifications, equivalent replacements, and improvements made within principle etc., the protection scope of the claims in the present invention should be included in
Within.
Claims (5)
1. a kind of iterative learning through street On-ramp Control method under packet loss environment, which comprises the following steps:
S1: control centre obtains current Expressway Traffic Flow density value；
S2: control centre obtains the Traffic flux detection that expectation the traffic flow density value and ring road entrance of middle through street is locally stored
Amount；
S3: current error letter is obtained according to the through street actual traffic current density value and desired traffic flow density value that receive
Number,
ek(t)=αk(t)(yd(t)-yk(t))
Wherein, t is the discrete sampling time, is determined by the monitoring cycle of actual traffic system；ydIt (t) is desired traffic flow density, yk
(t) the actual traffic current density that control centre receives when iteration secondary for kth, αkIt (t) is the y of kth time iterationk(t) do not occur with
The sequence of events that machine is lost, the transmission drop probabilities of quick terminal to control centre end are
S4: being based on error function, calculates forgetting factor value, and calculation formula is as follows
Wherein γ (k) to be in the kth time iteration error sequence of function be not 0 element number, θmaxWith θminFor the upper of θ (k) function
Floor value, φ (k) are maximum selection rule function, for having selected from the 1st iteration to kth maximum error function since time iteration
Mean-square value；
S5: according to the error function and forgetting factor, it is arranged based on changing with switching system characteristic under transmission packet drop
Generation study control law and iterative learning gain, wherein iterative learning control law is expressed as follows:
uk+1(t)=(1- αk(t+1)θ(k+1))sat[uk(t)]+Γiek(t+1)
In formula, uk+1(t) traffic flow of the ring road entrance of+1 iteration of kth, u are indicatedk(t) indicate that the ring road of kth time iteration enters
The traffic flow of mouth, θ (k) are adaptive forgetting factor function, error function ek(t+1) output at kth time iteration t+1 moment is indicated
Error, and ΓiFor the control gain of i-th of switching system subsystem, saturation function sat [], expression formula was as follows:
Wherein
Wherein j ∈ { 1,2 ..., K }, K are the ring road number in one section of through street,The traffic of ring road j when iteration secondary for kth
Flow control amount, u(j),max(t) and u(j),min(t) be respectively ring road j Traffic flux detection amount saturation bound；
S6: control centre generates the friendship of fast road ramp entrance in conjunction with the iterative learning control law and iterative learning gain
Logical flow control signals, and quick terminal is sent the signal to by reliable communication mode；
S7: quick terminal receives the Traffic flux detection signal of ring road entrance by reliable communication mode；
S8: Traffic flux detection signal is applied to the control of the urban expressing system with ring road entrance, is made it possible in certain iteration
Through street is set to reach desired traffic current density in number；
S9: Expressway Traffic density current signal is sent to control centre by quick terminal.
2. the through street On-ramp Control method of iterative learning, feature exist under packet loss environment according to claim 1
In step S4 carries out the calculating of forgetting factor by error function.
3. the through street On-ramp Control method of iterative learning, feature exist under packet loss environment according to claim 1
In the convergent the number of iterations of error function is the convergent iterations number of error criterion.
4. the through street Entrance ramp tele-control system of iterative learning under a kind of packet loss environment characterized by comprising
The acquisition module of current flows: current through street actual traffic current density is obtained by correspondence for control centre；
Traffic information acquisition module: it is obtained for control centre and it is expected traffic flow, the traffic of ring road entrance in local memory device
Flow control amount；
Error function computing module: for calculating error current function, according to the currently practical traffic flow density value received
It is calculated with desired traffic flow density and obtains error current function；
ek(t)=αk(t)(yd(t)-yk(t))
Wherein ydIt (t) is desired traffic flow density, yk(t) the actual traffic stream that control centre receives when iteration secondary for kth is close
Degree, αkIt (t) is the y of kth time iterationk(t) sequence of events of random loss, the transmission of quick terminal to control centre end do not occur
Drop probabilities are
Forgetting factor computing module: for calculating forgetting factor value according to error function, calculation formula is as follows
Wherein γ (k) to be in the kth time iteration error sequence of function be not 0 element number, θmaxWith θminFor the upper of θ (k) function
Floor value, φ (k) are maximum selection rule function, for having selected from the 1st iteration to kth maximum error function since time iteration
Mean-square value；
Iterative learning control law and learning gains setup module: for according to error function and adaptive forgetting factor setting transmission packet loss
In the case of iterative learning control law based on switching system feature and iterative learning gain, wherein iterative learning control law table
Show as follows:
uk+1(t)=(1- αk(t+1)θ(k+1))sat[uk(t)]+Γiek(t+1)
In formula, uk+1(t) traffic flow of the ring road entrance of+1 iteration of kth, u are indicatedk(t) indicate that the ring road of kth time iteration enters
The traffic flow of mouth, θ (k) are adaptive forgetting factor function, error function ek(t+1) output at kth time iteration t+1 moment is indicated
Error, and ΓiFor the control gain of i-th of switching system subsystem, saturation function sat [], expression formula was as follows:
Wherein
Wherein j ∈ { 1,2 ..., K }, K are the ring road number in one section of through street,The traffic of ring road j when iteration secondary for kth
Flow control amount, u(j),max(t) and u(j),min(t) be respectively ring road j Traffic flux detection amount saturation bound；
It controls signal output module: generating the Traffic flux detection signal of fast road ramp entrance, and for control centre with reliable
Communication mode is sent to quick terminal；
Control signal acquisition module: Traffic flux detection signal is obtained by reliable communication mode for quick terminal；
Traffic control module: for Traffic flux detection signal to be applied to the urban expressing system control with ring road entrance, so that
Through street can be enable to reach desired traffic current density in certain the number of iterations；
Traffic current density output module: current Expressway Traffic Flow density signal is sent by communication mode for quick terminal
To control centre.
5. the through street Entrance ramp tele-control system of iterative learning under a kind of packet loss environment according to claim 4,
It is characterized in that, the convergent the number of iterations of error function is the convergent iterations number of error criterion.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711336908.XA CN108154688B (en) | 2017-12-14 | 2017-12-14 | The through street On-ramp Control method and system of iterative learning under packet loss environment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711336908.XA CN108154688B (en) | 2017-12-14 | 2017-12-14 | The through street On-ramp Control method and system of iterative learning under packet loss environment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108154688A CN108154688A (en) | 2018-06-12 |
| CN108154688B true CN108154688B (en) | 2019-07-02 |
Family
ID=62466215
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711336908.XA Expired - Fee Related CN108154688B (en) | 2017-12-14 | 2017-12-14 | The through street On-ramp Control method and system of iterative learning under packet loss environment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108154688B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101045863B1 (en) * | 2011-02-16 | 2011-07-01 | 공주대학교 산학협력단 | A highway ramp metering system for solving non-recurrent congestion |
| CN103606284A (en) * | 2013-11-14 | 2014-02-26 | 北京交通大学 | Intelligent ramp entrance traffic demand combined control system and method |
| CN104240521A (en) * | 2014-09-30 | 2014-12-24 | 天津通翔智能交通***有限公司 | Intelligent ramp traffic control machine, system and method |
| CN106355905A (en) * | 2016-10-28 | 2017-01-25 | 银江股份有限公司 | Control method for overhead signal based on checkpoint data |
| CN106600991A (en) * | 2017-01-17 | 2017-04-26 | 河北工业大学 | City expressway multi-ramp coordination control method based on chaos |
-
2017
- 2017-12-14 CN CN201711336908.XA patent/CN108154688B/en not_active Expired - Fee Related
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101045863B1 (en) * | 2011-02-16 | 2011-07-01 | 공주대학교 산학협력단 | A highway ramp metering system for solving non-recurrent congestion |
| CN103606284A (en) * | 2013-11-14 | 2014-02-26 | 北京交通大学 | Intelligent ramp entrance traffic demand combined control system and method |
| CN104240521A (en) * | 2014-09-30 | 2014-12-24 | 天津通翔智能交通***有限公司 | Intelligent ramp traffic control machine, system and method |
| CN106355905A (en) * | 2016-10-28 | 2017-01-25 | 银江股份有限公司 | Control method for overhead signal based on checkpoint data |
| CN106600991A (en) * | 2017-01-17 | 2017-04-26 | 河北工业大学 | City expressway multi-ramp coordination control method based on chaos |
Non-Patent Citations (5)
| Title |
|---|
| A new PI controller for freeway ramp metering based on fuzzy logic;Chi Ronghu Et al.;《2013 25th Chinese Control and Decision Conference (CCDC)》;20130718;第4502-4506页 |
| An iterative learning approach for density control of freeway traffic flow via ramp metering;Hou Zhongsheng Et al.;《Transportation Research Part C: Emerging Technologies》;20080228;第16卷(第1期);第71-97页 |
| 受限迭代学习入口匝道控制收敛性分析;赵明等;《交通运输***工程与信息》;20110430;第11卷(第2期);第166-173页 |
| 带有迭代学习前馈的快速路无模型自适应入口匝道控制;侯忠生等;《自动化学报》;20090531;第35卷(第5期);第588-595页 |
| 快速路交通入口匝道的智能学习控制方法;刘祥鹏;《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》;20140715(第7期);C034-402 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108154688A (en) | 2018-06-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Gregoire et al. | Capacity-aware backpressure traffic signal control | |
| US20200334979A1 (en) | Predictive, integrated and intelligent system for control of times in traffic lights | |
| CN102436751B (en) | Short-time forecasting method for traffic flow based on urban macroscopic road network model | |
| CN104504941B (en) | A kind of Solving Flight Conflicts method of air traffic control system | |
| CN106530749B (en) | Signal-control crossing queue length estimation method based on single section low frequency detection data | |
| CN104616078B (en) | Photovoltaic system electricity generation power Forecasting Methodology based on Spiking neutral nets | |
| Li et al. | Traffic signal timing optimization in connected vehicles environment | |
| CN106846830A (en) | Through street On-ramp Control method and system based on switching system characteristic | |
| CN106935033A (en) | The ofaiterative, dynamic linearisation of Expressway Traffic system and Learning Control Method | |
| CN104408914A (en) | Signal intersection single vehicle stopping delay time estimating method and system based on GPS data | |
| Chow et al. | Modelling urban traffic dynamics based upon the variational formulation of kinematic waves | |
| CN112991750B (en) | Local traffic optimization method based on reinforcement learning and generation type countermeasure network | |
| CN113053120B (en) | Traffic signal lamp scheduling method and system based on iterative learning model predictive control | |
| Thianniwet et al. | Classification of road traffic congestion levels from GPS data using a decision tree algorithm and sliding windows | |
| CN109581928A (en) | A kind of end-to-end decision-making technique of intelligent vehicle towards highway scene and system | |
| Huang et al. | Traffic congestion level prediction based on recurrent neural networks | |
| Zhao et al. | Traffic signal control with deep reinforcement learning | |
| CN109410608B (en) | Picture self-learning traffic signal control method based on convolutional neural network | |
| CN108154688B (en) | The through street On-ramp Control method and system of iterative learning under packet loss environment | |
| TWI584236B (en) | Dynamic congestion information prediction system and method | |
| CN108417053A (en) | A kind of photoelectric detection system and control method suitable for T-junction mouth | |
| Ma et al. | A method for queue length estimation in an urban street network based on roll time occupancy data | |
| Wu et al. | Development of platoon‐based actuated signal control systems to coordinated intersections: application in corridors in Houston | |
| Li et al. | Development of analytical procedure for selection of control measures to reduce congestions at various freeway bottlenecks | |
| Thonhofer et al. | A flexible, adaptive traffic network simulation with parameter estimation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20190702 Termination date: 20201214 |