WO2012116982A1

WO2012116982A1 - Method of regulating traffic on a main traffic route, system and computer program product corresponding thereto

Info

Publication number: WO2012116982A1
Application number: PCT/EP2012/053354
Authority: WO
Inventors: Michel Fliess; Hassane ABOUAISSA; Cédric Join
Original assignee: Ecole Polytechnique; Universite De Lorraine; Universite D'artois; Centre National De La Recherche Scientifique (Cnrs)
Priority date: 2011-02-28
Filing date: 2012-02-28
Publication date: 2012-09-07
Also published as: FR2972067A1; FR2972067B1; EP2681725A1

Abstract

The invention relates to a method of regulating traffic on a main traffic route (2) through the control of an actuator (3) controlling access of vehicles to the main route (2) from an access route (7), in which: - at least one sensor (6) measures the density of vehicles on the main route (2), - a control device (4), linked to the sensor (6) and implementing a model‑free estimation of the behaviour of the flow of the vehicles on the main route (2), calculates a first control variable - the control device (4) determines a control variable by taking account of the first control variable; - the control device (4) despatches the control variable to the actuator (3); - the actuator (3) controls access by permitting or denying the access of the vehicles on the main route (2) as a function of the control variable, so that the traffic on the main route (2) is regulated.

Description

METHOD FOR CONTROLLING TRAFFIC ON A MAIN CIRCULATION AXIS, SYSTEM AND COMPUTER PROGRAM PRODUCT

CORRESPONDENTS GENERAL TECHNICAL AREA

The present invention relates to a method of regulating traffic on a main traffic axis by maintaining a vehicle density below a critical threshold, by controlling an actuator controlling the access of vehicles to the main axis from an access axis.

It also relates to a control system and a corresponding computer program product.

STATE OF THE ART

Figure 1 shows schematically a main axis 2 of circulation for vehicles, associated with an axis 7 of access to this main axis 2 of circulation. The main axis 2 of circulation can thus be for example a highway, a road for automobile, or a railway. Here we have identified a portion of length L of the main axis 2. The arrows Q _e and Q _s represent the flows of vehicles, respectively, entering and leaving the portion of the axis 2 of vehicles. The arrow Q _r represents the flow of vehicles returning to the portion of the axis 2 of circulation by the axis 7 access.

The regulation of traffic on such a main traffic axis is currently facing various problems, linked, inter alia, to the very complex modeling of the evolution of traffic flow on the main axis. Indeed, such a modeling must take into account the behavior of the drivers, the types of vehicles, the meteorological conditions as well as the geometry of the infrastructures. A type of modeling, called macroscopic modeling, is commonly used to model the circulation of vehicles on a given traffic axis. This modeling is mainly based on a vehicle conservation law for a portion of the axis 2 of circulation with an access axis 7, with a derivative of the density p (t) of vehicles on the corresponding traffic axis 2 to a formula of the type: p (t) = j ^ (Q _e - Q _s + Q _r ) with

- Q _s the flow of vehicles leaving the portion of the traffic axis 2, in vehicles per hour (veh / h),

- Q _e the flow of vehicles entering the portion by the axis 2 of circulation, in vehicles per hour (veh / h),

- Q _r the flow of vehicles entering the portion of the axis 2 of circulation by axis 7 access, in vehicles per hour (veh / h),

- L the length of the portion of the axis 2 of circulation, in kilometers,

- λ the number of traffic lanes of the traffic axis 2,

- the density of vehicles in vehicles per kilometer per lane.

The flow Q _s of vehicles leaving the portion of the axis 2 of circulation corresponds to

Q _s = ρ- ^ν -λ,

with v the average speed of vehicles.

The dynamics of the velocity is then described by: = v {p) - v (t)) + v (t) (v ₀ (t) - v (t)) - P

and the relation between the density p of vehicles of the section considered and an equilibrium speed V _e representing the average speed of the vehicles by one of the possible relations of equilibrium:

- pi (t) is a density of vehicles downstream of the access axis, beyond 1 section considered,

- α, Θ, K and υ parameters of the model, determined empirically and whose calibration is a very delicate task, - by a critical density of vehicles,

- Vf a free speed, speed of the flow when the vehicles do not interfere with each other.

Other types of equilibrium relationships can be used. This is given as an example as a commonly used equilibrium relationship.

FIG. 2 is a diagram whose curve 11 represents the flow rate Q _s of vehicles at the exit of a portion of a circulation axis 2, as a function of the density p of vehicles on this portion of the circulation axis 2. There is a critical density p _cr below which the circulation of vehicles is fluid, defining a zone 12 of fluidity, while a density p greater than p _{cr will} cause congestion of the circulation, defining a zone 13 of congestion.

The dotted line 14 represents the tangent at the origin of the curve, and is indicative of the ideal flow when the density p is sufficiently low so that the vehicles do not interfere with each other during their circulation. This line therefore shows the flow rate in the case where all the vehicles are traveling at the free speed Vf.

The fact that the curve 11 deviates from this line 14 indicates that as the density p of vehicles increases on the axis 2 of circulation, the average speed v of the vehicles decreases because of the interaction of the vehicles with each other. . This interaction can take many forms affecting the speed of vehicles. For example, a vehicle may decrease its speed in order to maintain a safe distance from the vehicle in front of it. In the zone 12 of fluidity, the decrease of this average speed v does not prevent the increase of the flow rate Q _s coming out of vehicles. On the other hand, when the density p of vehicles on the main circulation axis 2 exceeds the critical density p _cr , the average speed v decreases sufficiently for the outgoing flow Q _{s to} decrease despite the increase in the density p of vehicles on the road. axis of circulation. The axis 2 of circulation is no longer operated at full capacity since the flow rate Q _s out is not maximum. The determination of the critical threshold p _{cr as} well as the free speed Vf is already one of the problems encountered in practice.

The main problem to be solved by the regulation of traffic on a main axis 2 of circulation consists in regulating the access of the vehicles to this main axis of circulation 2 so that the traffic density p remains below a critical density p _r beyond which the circulation of vehicles will no longer be fluid.

The problem really arises when there is, or is desirable to have, only means of partial control of the access of vehicles to a main axis 2 of circulation. This is for example the case in the example illustrated in Figure 1, where the access of the vehicles to the portion of the main axis 2 of circulation is freely from the upstream of the main axis 2, to the like their exit downstream. In other words, there is no means to act directly on the incoming flow Q and outgoing _e Q _s. This is a frequent case, for example in the case of a motorway, where the traffic on the main axis 2 is not impeded.

The regulation possibilities are thus reduced to the control of access to the main axis 2 of circulation from an access axis 7. An actuator 3 controlling the access of the vehicles is then placed at the interface between the access axis 7 and the main axis 2 in order to regulate the traffic on the main axis 2 by controlling the access to the vehicle. main axis 2 of access axis 7 vehicles.

A light signaling light is an example of actuator 3 commonly used to regulate vehicle access to the main axis 2 from the axis 7 access. It can also be a mobile barrier or any other suitable device.

Such an actuator 3 usually controls the access by authorizing or not the access of the vehicles. By taking the example of a luminous signaling light, this light works by successively lighting indicators each having a meaning, for example:

- a red indicator light prohibits the passage of vehicles,

- a green indicator light allows the passage of vehicles,

- an orange indicator light indicates the imminent illumination of the red indicator light and prohibits the passage of vehicles unless they can not stop safely. The control of the operation of such an actuator 3 must then be constructed in order to regulate the traffic. In the case of a light signaling light, a common way of adapting its operation to the density p of vehicles is to change the ignition times of the indicators. For a density p of vehicles given on the access axis 7, increasing the proportion of the cycle time when the red indicator is lit leads for example to reducing the flow rate Q _r of vehicles returning on the main axis 2, while increasing the proportion of the cycle time when the green indicator is lit leads to increase the flow rate Q _r of vehicles returning on the main axis 2.

The control of an actuator 3 may be fixed, in which case the operating parameters of the actuator 3 are set according to a scenario, and such a control does not make it possible to adapt the operation of the actuator 3 to conditions which may be very different from those taken into account in that scenario.

The control of an actuator 3 can be variable and then depend on the measurement of an external parameter. For example, the regulation may be different as a function of time, to take into account time slots where the density p of vehicles is supposed to be greater than in other time slots. However, it is a system whose calibration is empirical and very delicate, and which does not take into account any changes in traffic conditions, such as the occurrence of an accident, weather conditions, or the presence of works.

A more advanced system consists in having sensors measuring parameters making it possible to deduce a density p of circulation on the main axis, and using this knowledge to determine a command of an actuator 3. The determination of the command sets most often implemented a modeling of the density p of vehicles on the axis of circulation as described above.

The determination of such a command will inherit the difficulties of determining the parameters of the model, its inaccuracy due to simplifications and its empirical modeling and calibration.

Historically, it has been noted that control systems have rarely achieved optimal performance or at the cost of a time of design and development. high focus. Indeed, such advanced control methods prove to be costly, insofar as they require:

- high level skills in automatic and mathematics,

a mathematical model representative of the arrangement constituted by the main axis 2 of circulation and the axis 7 of access, which is long and expensive to produce since it requires numerous measurements and readings,

- a high number of parameters, which makes the commissioning tests difficult and the maintenance delicate. PRESENTATION OF THE INVENTION

The invention proposes to overcome at least one of these disadvantages, preferably all.

For this purpose, according to a first aspect of the invention, a method of regulating a traffic on a main axis of circulation by the control of an actuator controlling the access of vehicles to the main axis from an axis access, in which:

at least one sensor measures the density of vehicles on the main axis,

a control device, connected to the sensor and implementing a model-less estimation of the behavior of the traffic of the vehicles on the main axis, calculates a first control variable, the first control variable being determined by the control device; function:

a vehicle density instruction,

a measurement of density p measured by the sensor, and

- a history of control variables;

the control device determines a control variable taking into account the first control variable;

the control device sends the control variable to the actuator;

the actuator controls the access by authorizing or not allowing the access of the vehicles on the main axis from the access axis according to the control variable, so that the traffic on the main axis is regulated. The invention is advantageously completed by the following features, taken alone or in any of their technically possible combination:

the control device calculates the first control variable rjft according to the formula:

dp *

r _i (t) = - is [F] \ (t) + PI (e)

a dt

with: p * (t) the vehicle density setpoint;

isfFJ a model-free estimate of traffic F behavior;

a is a constant coefficient such that F and ar are of the same order of magnitude; PI (e) = K _p .e + K _t e (x) .d with e (t) = p (t) - p * (t), with

p (t) the measurement of the traffic density of vehicles to be regulated over time, coming from the sensor and

Kp and Ki constant gains;

the vehicle density reference p * (t) is determined as a function of:

- an estimate of density p _cr critical beyond which traffic is more fluid;

an estimate of a free speed Vf representing the speed of the vehicles on the axis of circulation when they do not interfere with each other,

said estimates of the critical density p _cr and the free speed V f being determined from the measurements of the density p and the average speed v of the vehicles on the axis of circulation;

- the values of

- the measurement of the density p (t) of vehicles on the axis of circulation,

the setpoint p * (t) of vehicle density,

- the estimate is [F] without a model of the F behavior of the traffic, and

the first variable ri (t) of command,

are periodically updated at a time interval Te defining a sampling period, and kept constant during these periods, so that at each moment t such that t = kT + e

with k natural integer and ε <Te,

p (t) = p (kT _e + E) = p (kT _e ),

p * ft) = p * (kT _e + e) = p * (kT _e ),

is [F] (t) = is [F] (kT _e ),

ri (t) = r _! (kT _e + e) = ri (kT _e );

the method also comprises a step of model-less estimation of the behavior of the main axis according to which the estimation without a model of the behavior of the circulation is calculated according to the formula:

the control device determines the control variable r (t) taking into account the first command variable ri (t) by taking:

r (t) = n (t);

the method further comprises the steps according to which:

a detector, connected to the control device, detects the presence or absence of vehicles at a measuring point on the access axis,

the control device determines a second variable r ₂ according to the detection of the presence or absence of vehicles at the point of measurement of the access axis by the detector,

the control device determines the control variable r (t) by taking into account the first control variable rt) and the second control variable r ₂ (t) by taking:

r (t) = max (r i (t); r2 (t));

the actuator controls the access by authorizing the access of the vehicles on the main axis for a duration g of access authorization proportional to the duration Te of a cycle C, according to the ratio between the variable r ( t) and a control parameter Q t of _the saturation of the access axis by: laughs) laughs)

- ^S1 d _min <<d _max , ^then g = - ^ - T _c , with 0 <d _mm <d _max ≤l;

z sat z sat

if ¹ ^ -≥d _max , then g = T _c ; laughs)

- if - ≤d _min , then g = 0, and prohibiting vehicular access on the main axis from the access axis during the rest of the C cycle.

The invention also relates, according to a second aspect, to a system for regulating a traffic on a main axis of circulation by controlling an actuator controlling the access of vehicles to the main axis from an access axis, said system comprising:

at least one sensor for measuring the density p (t) of vehicles on the main axis,

a control device, said control device being connected to the sensor and being adapted to

- to implement an estimation is [F] without model of the behavior F of the circulation of vehicles on the axis of circulation,

calculating a first control variable r t) to the actuator, the first control variable ri (t) being determined by the control device as a function of:

a density instruction p * (t) of vehicles,

a measurement of density p (t) measured by the sensor, and

a history of control variables, and

an actuator adapted to control access by authorizing or not allowing access of the vehicles on the main axis by taking into account the first variable r t) of control, so that the traffic on the main axis is regulated.

The system according to the invention is advantageously completed by the following features, taken alone or in any of their technically possible combination: the system further comprises a detector connected to the control device and adapted to detect the presence or absence of vehicles at a measuring point on the access axis, the device being further adapted to:

determining a second control variable r ₂ ft) as a function of the detection of the presence or absence of vehicles at the point of measurement of the access axis by the detector,

determining the control variable r (t) taking into account the first control variable rjft) and the second control variable r ₂ (f) by taking:

r (t) = max (r (t); r ₂ (t)), and

- send the control variable r (t) to the actuator.

According to a third aspect, the invention also relates to a computer program product for carrying out the steps of the method comprising program code instructions for executing the steps of the method according to the invention, when said program is executed on a program. system according to the invention.

The invention has many advantages.

The invention implements a control called without a model by the person skilled in the art who, by its robustness, adaptability and simplicity properties, provides outstanding performance not equaled by known devices, in a development time very short, adapted to the industrial environment. The invention therefore makes it possible to control the time spent in optimally adjusting the parameters for implementing the regulation method, whether in a simulation study or on site.

The invention has the advantage of not requiring the knowledge of a mathematical model of the layout of the main axis of circulation or of the axis of access thereto, resulting in an even greater gain in time for the design of the control system (there is no need to identify the parameters of the model).

The invention can also be implemented on existing systems with very few adaptations, thanks to the computer program product according to the invention, which, once loaded onto the adapted system, allows the implementation of a process according to the invention.

The load of the controller is minimal, (simple processing and low number of operations). PRESENTATION OF FIGURES

Other features, objects and advantages of the invention will emerge from the description which follows, which is purely illustrative and nonlimiting, and which should be read with reference to the appended drawings among which:

- Figure 1, already commented, is a diagram illustrating a known main axis of traffic with an access axis;

FIG. 2, already commented on, is a diagram representing the flow rate Q _s of vehicle at the exit of a portion of a traffic axis as a function of the density p of vehicles on this portion of the traffic axis;

FIG. 3 is a diagram illustrating a main axis of circulation with an access axis provided with the system according to the invention;

FIG. 4 diagrammatically represents a principle of a possible embodiment of a system according to the invention;

- Figure 5 schematically shows a principle of a possible embodiment of a control device of a system according to the invention; and

- Figure 6 schematically shows the main steps of a method according to the invention.

In all the figures, similar elements bear identical reference numerals.

DETAILED DESCRIPTION FIG. 4, in combination with FIG. 3, schematically represents a principle of a possible embodiment of a system for regulating a traffic on a main axis 2 of circulation, according to the invention.

The control system mainly comprises a device 4 for controlling at least one actuator 3 controlling the access of vehicles to the main axis 2 from an access axis 7 for regulating the traffic on the main axis 2, the control device 4 acting as a regulator. The actuator 3 is for example a light signaling light, a movable barrier or any other device able to control the access of vehicles. The main axis 2 is for example a highway, a road for automobile, or a railway.

The system further comprises a sensor 6 measuring the density p of vehicles on the main axis 2 by means for example of the occupancy rate (in percentage) of a circulation space of the axis 2 by the vehicles circulating there. The density p can be deduced from this occupancy rate, for example by means of assigning a standard occupancy value to each vehicle.

Other variables can be measured such as the average vehicle flow and the average vehicle speed on the main axis 2, depending on the type of sensor used.

These variables can also be used to determine other parameters than the density p of vehicles. As will be detailed below, flow and average speed can be used to estimate

a critical density p _crc beyond which the circulation is no longer fluid and

a free speed Vf representing the speed of the vehicles on the axis 2 of circulation when they do not interfere with each other,

to establish a p * density setpoint.

The sensor 6 preferably comprises magnetic loops in order to determine both the occupancy rate, the speed and the average speed of the vehicles. It can still be counting vehicles, video and image processing. In addition, several sensors can be implemented to determine different variables.

Preferably, the sensor 6 is positioned at the access axis 7 (FIG. 3) so that the vehicles accessing the circulation axis 2 are taken into account in the measurement of the density (or the 'occupation).

The control device 4 is adapted to calculate and send a control variable r to the actuator 3 by implementing an estimate [F] without a model of the behavior F of the vehicular traffic on the main axis 2, said control without model being described in more detail in the remainder of this description, with reference to FIG. 4 relating to a possible embodiment. The variable r of control is determined by the control device 4 by taking into account a first variable ri control previously calculated by said device 4 control. At first, we will consider the simplest case where

r (t) = n (t).

The following developments concern traffic regulations on the main traffic axis 2, by controlling the flow Q _r of vehicles returning on the main axis 2 of circulation from the axis 7 of access.

The principle of the invention is based on a local model of the behavior of the traffic on the main axis 2 of circulation, recalculated and updated periodically at each step of a time interval T _e defining a sampling period. In other words, it is not necessary to have a complete physical model of the behavior of the traffic on the main axis 2.

We therefore avoid a modeling as complete as possible, and always difficult to obtain, taking as a local model of the behavior of the traffic on the main axis 2 the differential equation of order n:

d "p (t),,

- ^ - ^ = b + a.r (t)

dt ⁿ

where r (t) and p (t) are respectively the control variable for the actuator 3, and a measurement of the density p (t) of vehicles, evolving over time;

F is a variable which represents the not precisely determined behavior of vehicular traffic on the main axis 2: it is a variable updated at each time step T _e of a sampling period, a period that can be variable;

a is a constant coefficient, determined so that F and a.r (t) are of approximately the same order of magnitude, and preferably of the same order of magnitude.

Known models of traffic axes are based on partial differential equations that intuitively suggest taking a high n value.

The inventors have however found that, for an industrial application to the regulation of traffic on a main axis 2 of circulation, the choice of n = 1 is sufficient and even advantageous because it also makes it possible to reduce the number

of operations performed by the regulatory system.

Advantageously, the control device 4 calculates an estimate [F] of the variable, set at each time step T _e , according to the formula:

possible from the derivative

temporal measurement of the density p of traffic vehicles to be regulated. Other derivative estimators known to those skilled in the art can be used if desired.

The estimate thus proposed greatly simplifies the calculations, and makes it possible to obtain a good estimate of the numerical value of each instant T _e , despite the high sampling period (of the order of several seconds, for example varying between T _e = 10 s and T _e = 20 s).

We deduce a first control variable ri (t) from the control device 4, which the skilled person also calls Intelligent Proportional-Integral Corrector, by:

where PI (e) = K _p .e + K \ β (τ) άτ with e = p (t) - p * (t), and

Jo

where the gains K _p and K _t are those usual for this type of corrector. It is thus reduced to the setting of a simple integrator in the control device 4. The setting of the integrator of the control device 4 is therefore particularly easy and extremely robust, since it is insensitive to model variations.

In addition, p * is a vehicle density setpoint on the main axis 2 of circulation, preferably chosen to be less than a density p _cr of critical vehicles beyond which the circulation is no longer fluid. However, any p * density setpoint can be used depending on the objectives and constraints to be met by the control system. Thus, the vehicle density reference p * (t) can be determined as a function, for example, of two parameters, for example as a function of:

- an estimate of critical _cr p density beyond which traffic is more fluid, and

an estimate of a free speed Vf representing the speed of the vehicles on the traffic axis 2 when they do not interfere with each other,

The invention may then first include a step of estimating these two parameters. Indeed, these parameters can be expressed as a function of measured quantities, such as measurements of the density p and the average speed v of the vehicles on the circulation axis 2, and their derivatives, up to order 2 , in relation to time. The derivatives can also be estimated similarly to the estimate of the time derivative of the measurement of the density p of traffic vehicles to be regulated presented above.

Thus, estimates of the critical _pcr density and the free Vf velocity can be determined from measurements of the p density and the average vehicle velocity on the traffic axis 2.

For example, the estimates of the critical _pcr density and the free Vf velocity can be determined from measurements of the p density and the mean v velocity according to:

- an equilibrium relation written from measurements of density and speed,

with V _e (p) a speed of equilibrium,

and K and has values to be determined, and

p _cr the critical density;

a logarithmic derivation, denoted IV, of Ve with respect to p:

a logarithmic derivation of W with respect to p: dW

dp a - 1

~ W ^{~ ~} p ^{~ '}

From these relations, the value a can then be estimated according to:

dW

dp,

a = p- ^! - + l.

W

Similarly, the value K can be estimated according to

- Wp ' ^{~ a}

K = ^P

at

from where

and

v _f = Ve (p) exp (Kp ^a ).

An example of density setpoint p * can then be the critical density p _cr , with the objective of establishing a setpoint maximizing the flow rate Q _s of vehicles, since it is maximum at the critical density p _cr while ensuring fluid circulation.

The critical _pcr density determination method presented above is highly efficient, but requires data of great richness, which currently existing sensors can not always provide.

A second method of determining the critical density p _cr presented below allows to afranchir this constraint by using measurements of the speeds supplied by stations of automatic data collection (RAD). The measurements provided by the RAD stations relate to three traffic variables:

flow (vehicles / hour),

occupancy rate (in%) (translated in terms of vehicle density), the average speed in kilometers per hour.

The objective is, from these data, to be able to distinguish congested areas from fluids. In other words, to implement or not the regulation systems. Indeed, in fluid mode, there is no need to activate the control access. It is activated when a congestion occurs. The command will start as soon as the average speed is higher than or equal to a threshold speed Vseuii- In the case of a motorway network, an example of a threshold speed v _seu ii is 85 km / h.

The flow capacity corresponds to the maximum flow of vehicles that have been flown for one hour before switching to the saturation situation. Thus, congestion is defined as the inconvenience due to an accumulation of vehicles in circulation on a traffic axis. It covers situations of relatively mild discomfort to the harshest situations.

Saturation is the hardest form of congestion. It appears when the demand exceeds the flow capacity of the traffic axis, and is characterized by the formation of permanent queues over a certain period of time and a sharp fall in the average speed v. In practice, on major roads such as motorways and national roads, saturation is reached when the average speed v reaches 30km / h. In this case, the service level of the section in question is very badly degraded and the congestion is at its maximum.

When the average speed on the motorway is v = 85 km / h, the threshold of inconvenience is reached (v = 45 km / h on national road). The level of service is, then, strongly degraded, resulting in a prolonged slowdown for vehicles.

The proposed approach is based on the following points:

- analysis of a recording of the speed in open loop,

- deduction of the average speed v,

- Translation in terms of p * density reference, which can be adapted online.

Knowing that the traffic is strongly degraded when v <v _seu ii, the estimation algorithm proposed revolves around the following points.

Let p _{cr be} the critical density, p; the initial value, β and δ of the values allowing to define the threshold:

if v> v _S euii then p _cr = p _; + β;

- otherwise p _cr = pi - δ.

It is necessary to saturate the critical densities thus calculated. So :

if p _cr > Pmax, then p _cr = Pmax, Pmax being the maximum density; if p _cr <0, then p _cr = 0.

The set density p * can then be critical density p _cr, defined using only data rate and occupancy rates collected by ARD stations and taking into account a threshold speed v _seu ii discomfort.

As explained above, the values

- measuring the density p (t) of vehicles,

the setpoint p * (t) of vehicle density,

- the estimate is [F] without model of the behavior (F) of the circulation, and

the first variable ri (t) of command,

are periodically updated at a time interval Te defining a sampling period, and kept constant during these periods, so that at each moment t such that

t = k _e + s

with k natural integer and ε <Te,

p (t) = p (kT _e + e) = p (k ^' ,

p * (t) = p * (kT _e + e) = p * (kT _e ),

is [F] (t) = is [] (kT _e ),

n (t) = n (kT _e + z) = r ₁ (k _e )

The sampling period may be fixed or variable, for example to adapt to the speed of variation of the parameters.

Once the control variable r determined by one of the previously detailed methods, this control variable r is used to control the actuator 3, in order to control the flow Q _r of vehicles returning on the main axis 2 of circulation since the axis 7 of access by allowing or not the access of vehicles, so that the traffic of vehicles on the main axis 2 is regulated.

The control of the actuator 3 depends in particular on the type of actuator used.

Preferably, such an actuator 3 controls the access of the vehicles in a cycle C, allowing the passage of vehicles during a period g of the cycle C, and prohibiting the passage of vehicles during the rest of the cycle C. A use of the variable r is to determine the duration of access authorization as as a proportion of the duration Te of the cycle C, as a function of the ratio between the variable r of control and a variable Q _sat saturation of the access axis 7 according to:

and thus control the actuator 3 so that the access to the axis 2 of circulation for the vehicles of the access axis 7 is restricted to the duration g of the cycle C.

In the case where the actuator 3 is a luminous signaling light, the duration Te of the cycle C may then for example be 40 seconds, and the duration g will then represent the duration of illumination of the green indicator light allowing the access of vehicles.

Insofar as such a luminous signaling light must usually respect a preceding duration of lighting of the orange indicator, for example of a few seconds, before the illumination of the red indicator prohibiting the access of the vehicles, a duration g Too long would prevent the red indicator from being lit, which can be disruptive to the drivers of the vehicles.

Consequently, in the case where the duration g exceeds a maximum proportion d _max of the duration Te of the cycle C, it is preferable for the green indicator light to remain alone during the entire duration Te of the cycle C.

Similarly, in the case where the duration g is below a minimum proportion d _min of the duration Te of the cycle C, it is preferable that the red indicator light remains alone during the entire duration T _c of the cycle C.

It is the same for most actuators 3, for example a movable barrier.

The control of the actuator 3 then becomes:

r r

if d mm <<d _max , then g T _c , with 0 <d _mm <d _max <1;

^■ sat Q ^: sat - if - -≥d _ace , sleep g = T _c ;

- if ^■ j- <d _nnn , then g = 0.

However, taking into account the only density p of vehicles on the main axis 2 of circulation may not be satisfactory. Indeed, if the access axis 7 becomes cluttered with vehicles, for example in the case where the density p of vehicles on the main axis 2 becomes too high and the actuator 3 allows the passage of vehicles for a period g insufficient for the flow Q _r of vehicles returning to the main axis 2 from the access axis 7 exceeds the flow of vehicles on the axis 7 access to enter the main axis 2.

There is a gradual accumulation of vehicles, which can take the form of a queue of vehicles. Such a phenomenon is all the more troublesome as it frequently happens that the access axis 7 has a vehicle storage capacity waiting relatively low compared to the arrival of vehicles on this axis 7 access.

It can then very quickly result in a queue of vehicles extending along the access axis 7, or even upstream of this axis 7 access. In the case where for example a roundabout is upstream of the access axis 7, the extension of such a queue can block the roundabout, which is highly undesirable.

In order to limit the range of the vehicle queue, it is provided in a preferred embodiment, to have a detector 9, connected to the control device 4 and able to detect the presence or absence of a vehicle at a measurement point M, on the axis 7 of access.

The detector 9 is provided so that the detection of the presence or absence of a vehicle by the detector 9 is representative of the existence of a queue of vehicles stretching at least from the actuator 3 to the point M of measurement. It may be a presence detector of the type known to those skilled in the art, or a system similar to the sensor 6. The detector 9 may for example send a detection variable m representative of the presence or absence of vehicles at point M of measurement.

A first control variable r _} is determined according to the method described above, said control variable r _} being determined to regulate the density p of vehicles on the main axis 2 of circulation according to:

a density instruction p * of vehicles,

a measurement of density p (t) measured by the sensor (6), and

- a history r ((kl) T _e ) of control variables. A second control variable r ₂ is then determined as a function of the value taken by the detection variable m, thus depending on the presence or absence of a vehicle at the measurement point M detected by the detector 9.

The second control variable r ₂ is intended to resorb the queue of vehicles on the access axis 7 and its determination will therefore depend on the configuration of the access axis 7 (position of the measurement point M , average flow of vehicles on the access axis 7 ...). The second variable r ₂ can thus be fixed, for example, r ₂ = 0.3 Q _sat , or be calculated dynamically.

With reference to FIG. 5, and as an example, the second control variable r ₂ is generated by a data processing module 8 in response to the detection variable m.

The first command variable ri and the second command variable r ₂ are transmitted to a selector 5 which selects that of the first command variable r ₁ and the second command variable ₂ which has the highest value to be the variable r command that is sent to the actuator 3.

The control of the actuator 3 will then be done by means of a control variable r chosen as the maximum between r ₁ and r ₂ :

r (t) = πιαχ (ν _γ ; ν ₂ ).

In this way, even in the case of high density p of vehicles on the main axis 2 of circulation (for example if p> p *), the actuator 3 controls the access of the vehicles of the access axis 7 to the main axis 2 not only to regulate the density p of vehicles on the main axis 2, but also to avoid the overflow of the vehicle queue on the access axis 7 beyond the point M of measured.

Preferably, the point M is chosen at a distance adapted so that the detection of the presence of a vehicle at this point M, that is to say when the queue reaches said point M, can cause the capture in consideration of the resorption of the queue of vehicles in the control of the actuator 3, as explained above, before it extends to an area upstream of the axis 7 of access where the presence of a queue of vehicles is undesirable, such as a roundabout. At the same time, it is desirable that the point M be chosen at a sufficient distance from the actuator 3 so that the vehicle queue can reach a length that is suitable for storing a sufficient number of vehicles so that the regulation has an effect on the density p of vehicles on the main axis 2.

As shown in FIG. 6, a method of regulating traffic on a main axis 2 of circulation by the control of an actuator 3 controlling the access of the vehicles to the main axis 2 from an axis 7 of access, comprises the steps according to which:

at least one sensor 6 measures IF the density p (t) of vehicles on the main axis 2; a control device 4 connected to the sensor 6 and implementing an estimation is [F] without a model of the behavior F of the circulation of the vehicles on the main axis 2, calculates S2 a first variable rt) of control, the first control variable ri (t) being determined by the control device 4 according to:

- a density instruction (p * ft)) of vehicles,

a density measurement (pft) measured by the sensor (6), and

- a history r ((kl) T _e ) of control variables;

the control device 4 determines S3 a control variable r (t) taking into account the first control variable ri (t);

the control device 4 sends S4 the control variable r (t) to the actuator 3;

the actuator 3 controls S5 the access by authorizing or not the access of the vehicles on the main axis 2 according to the variable r (t) of control, so that the traffic of vehicles on the main axis 2 be regulated.

Optionally, the method may further comprise the steps of:

a detector 9 connected to the control device 4 detects S6 the presence or absence of vehicles at a measurement point M on the access axis 7,

the control device 4 determines S7 a second control variable r ₂ (t) as a function of the detection of the presence or absence of a vehicle at the measuring point M, the control device 4 determines S3 the control variable r (t) taking into account the first control variable ri (t) and the second control variable r ₂ (t) according to:

r (t) = max (r (t); r ₂ (t)).

where max is the mathematical operator taking the maximum of the two values ri (t) and r ₂ (t).

Also, optionally, the method may include the steps of:

the sensor 6 measures S8 an average speed of the vehicles on the axis 2 of circulation,

a critical _pcr density beyond which the circulation is no longer fluid and a free speed Vf representing the speed of the vehicles on the circulation axis 2 when they do not interfere with each other are estimated S9 in function measurements of density p and velocity v,

the setpoint p * (t) of vehicle density is determined S 10 as a function of the estimates of the critical density p _cr and the free speed V f.

As has been said, the invention also relates to a computer program product which, when loaded onto a suitable system, allows the implementation of a method according to the invention.

Claims

claims

A method of regulating a traffic on a main axis (2) of circulation by the control of an actuator (3) controlling the access of vehicles to the main axis (2) from an axis (7) of access, in which:

at least one sensor (6) measures (SI) the density (pft)) of vehicles on the main axis (2),

a control device (4), connected to the sensor (6) and implementing an estimation (estfFJ) without a model of the behavior (F) of the circulation of the vehicles on the main axis (2), calculates (S2) a first control variable (rift)), the first control variable (rift) being determined by the control device (4) based on:

- a density instruction (p * ft)) of vehicles,

a density measurement (pft) measured by the sensor (6), and

- a history (r ((kl) T _e )) of control variables;

the control device (4) determines (S3) a control variable (rft)) taking into account the first control variable (rift));

the control device (4) sends (S4) the control variable (rft)) to the actuator (3);

the actuator (3) controls (S5) the access by authorizing or not allowing the access of the vehicles on the main axis (2) according to the control variable (rft), so that the traffic on the main axis (2) is regulated.

2. Method according to claim 1, wherein the control device (4) calculates (S2) the first command variable rift according to the formula:

with: p * (t) the vehicle density setpoint;

is [F] a model-free estimate of the F behavior of the circulation; a is a constant coefficient such that F and ar are of the same order of PI (e) = K _p .e + K _r the (T) .dT with e (t) = p (t) - p * (t), with p (t) the measurement of the density of traffic vehicles at regulate over time, from the sensor (6) and

Kp and Ki constant gains.

3. Method according to any one of claims 1 or 2, wherein the target (p * ft)) of vehicle density is determined (S 10) according to:

an estimate (S9) of a critical density (p _cr ) beyond which the circulation is no longer fluid;

an estimate (S9) of a free speed (v /) representing the speed of the vehicles on the axis (2) of circulation when they do not interfere with each other, said estimates of the density (p _cr ) and critical velocity (v /) free being determined from the measurements of the density (p) (SI) and the average speed (v) (V) of vehicles on the axis (2) of circulation.

4. Method according to one of claims 1 to 3, wherein the values of

the measurement of the density p (t) of vehicles on the axis (2) of circulation,

the setpoint p * (t) of vehicle density,

the estimate is [F] without model of the behavior (F) of the circulation, and

the first variable ri (t) of command,

t = k _e + s

with k natural integer and ε <Te,

p (t) = p (kT _e + e) = p (kT _e ),

p * (t) = p * (kT _e + e) = p * (kT _e ),

is [F] (t) = is [F] (kT _e ),

n (t) = n (kT _e + E) = _ri (kT _e )

5. Method according to claim 4, further comprising a model-less estimation step of the behavior (F) of the main axis (2) according to which the estimate is calculated without a model of the behavior (F) of the circulation according to the formula:

Method according to any one of claims 1 to 5, wherein the control device (4) determines (S3) the control variable r (t) taking into account the first control variable ri (t) by taking:

r (t) = n (t).

7. Control method according to one of claims 1 to 5, wherein:

a detector (9), connected to the control device (4), detects (S6) the presence or absence of vehicles at a measurement point (M) on the access axis;

(7)

the control device (4) determines (S7) a second control variable r ₂ as a function of the detection of the presence or absence of vehicles at the point (M) of measurement of the access axis (7); ) by the detector (9), - the control device (4) determines (S3) the control variable r (t) taking into account the first control variable ri (t) and the second variable r ₂ (t) of order by taking:

r (t) = max (r (t); r ₂ (t)).

Method according to any one of claims 1 to 7, wherein the actuator (3) controls (S5) access by allowing access of the vehicles on the main axis (2) for a duration g of authorization d access proportional to the duration Te of a cycle C, depending on the ratio between the variable r (t) and a control parameter Q t _its saturation of the axis (7) access by: r (t) rft)

d _min <<d _max , ^then g = - ^ - T _c , with 0 <d _mm <d _max ≤l; then g = T _c ;

• r (t),

then g = 0,

^ sat

and prohibiting vehicular access to the main axis (2) from the access axis (7) during the remainder of the cycle C.

9. System for regulating a traffic on a main axis (2) of circulation by the control of an actuator (3) controlling the access of vehicles to the main axis (2) from an axis (7) of access, said system comprising:

at least one sensor (6) for measuring the density (pft) of vehicles on the main axis (2),

a control device (4), said control device (4) being connected to the sensor (6) and being adapted to

- implement an estimation (is []) without model of the behavior (F) of vehicle traffic on the axis (2),

calculating a first control variable (ri (t)) to the actuator (3), the first variable {ri (t)) being determined by the control device (4) based on:

- a density instruction (p * ft)) of vehicles,

a density measurement (pft) measured by the sensor (6), and

a history (r ((kl) T _e )) of control variables, and

an actuator (3) adapted to regulate access by authorizing or not allowing vehicles to access the main axis (2) taking into account the first control variable (rift), so that traffic on the main axis (2) is regulated.

10. System according to claim 9, characterized in that the system further comprises a detector (9) connected to the control device (4) and adapted for detecting the presence or absence of vehicles at a measurement point (M) on the access axis (7), the device (4) being further adapted to:

determining a second control variable r ₂ ft) as a function of the detection of the presence or absence of vehicles at the point (M) of measurement of the access axis (7) by the detector (9),

r (t) = max (r (t); r ₂ (t)), and

- send the control variable r (t) to the actuator (3).

A computer program product, comprising program code instructions for performing the steps of a method according to any one of claims 1 to 8, when said program is executed on a system according to one of the claims. 9 or 10.