EP0821334A1 - Traffic control method and device - Google Patents

Abstract

The traffic regulation method uses at least one traffic signal (2) and an associated control device (3,5). The device is coupled to units (4) for detecting a traffic flow. The device has a processor which uses the detector signals (DS) and the traffic control signals (VS) for providing a traffic model, by measuring the movement and the position (BA) of the individual vehicles. The velocity of the individual vehicles is simulated by the control device. The traffic information is fed to a regulator (3) which provides the control signals for the traffic signal.

Description

The invention relates to a method and Device for traffic control according to the preamble of the independent claims.

Conventional control and regulation of Crossroads through traffic lights uses the theoretically possible flow across the intersection is not optimal; it arise easy idle times with green phases for access, for which there are no vehicles.

If you provide an intersection with vehicle detectors, so the regulation can adapt to the current Adapt traffic situation. The measurements of the detectors are usually interpreted by counters, which quantify the inflows to the intersection.

In practice, however, also achieve with Systems equipped with detectors ensure optimal traffic flow by far not. They are also hardly suitable, others Optimize target values, e.g. the entire Energy consumption or waiting times for the public Traffic.

It is therefore the task of a device or a procedure of the aforementioned type provide that allow better traffic regulation i.e. with which a better optimization of a desired one Target size is possible.

This task is the subject of independent Claims met.

According to the invention, the detectors coming signals not in their raw form Traffic control or regulation used, but they are used to create a model that matches the locations and the movement (e.g. speed) of individuals Vehicles, preferably essentially all vehicles, simulated. In this way, from the detector signals diverse and rich information about the Traffic can be determined. With this information can then a more efficient control of the Traffic can be carried out.

The detector signals are preferably used for this purpose used to track sizes used in the model, especially to create a vehicle in the model that Position and / or speed of an existing vehicle to correct or an existing vehicle delete the model. Furthermore, by means of the detector signals and / or with data from the model typical Values of individual sizes, e.g. the speed of a vehicle entering the simulation will. These sizes can be used to e.g. the initial speed of a new one created in the model Vehicle. These sizes are preferred constantly updated, so that a change in traffic behavior (e.g. due to rain or fog) leads to an adaptation of the model.

To the establishment of an algorithm for calculation of the model will be its parameters preferably at least partially as a graphic Representation entered, which then in one of the control program readable or executable form of the observer being transformed.

Fig. 1 eine Steuerung bzw. Regelung gemäss dem Stand der Technik,

Fig. 2 eine erfindungsgemässe Regelung,

Fig. 3 eine einfache Streckenanordnung,

Fig. 4 den Synchronisationsalgorithmus einer ersten Ausführung der Erfindung bei einem Signal von einem ersten Detektor,

Fig. 5 den Synchronisationsalgorithmus bei einem Signal von einem zweiten Detektor,

Fig. 6 den Synchronisationsalgorithmus bei einem Signal von einem dritten Detektor,

Fig. 7 eine graphische Darstellung eines Modells,

Fig. 8 eine Detailansicht eines Synchronisationssymbols von Fig. 7 und

Fig. 9 - 12 Synchronisationsschritte gemäss einer zweiten Ausführung der Erfindung.

Further applications and advantages of the present invention result from the following description of a conventional system and an embodiment of a system according to the invention using the figures. Show:

1 is a control or regulation according to the prior art,

2 shows a regulation according to the invention,

3 shows a simple route arrangement,

4 shows the synchronization algorithm of a first embodiment of the invention for a signal from a first detector,

5 shows the synchronization algorithm for a signal from a second detector,

6 shows the synchronization algorithm for a signal from a third detector,

7 is a graphical representation of a model,

Fig. 8 is a detailed view of a synchronization symbol of Fig. 7 and

9-12 synchronization steps according to a second embodiment of the invention.

To illustrate the invention, first 1 shows a conventional system. This System is to control traffic lights 2 at an intersection 1 provided. A controller is used to control traffic lights 2 3 provided that receives signals DS from detectors 4. The detectors 4 are generally simple vehicle detectors, which give a signal when a vehicle passes through its measuring range, e.g. Induction loops, Ultrasonic detectors, infrared detectors or even Cameras with image processing.

The controller 3 controls by means of traffic control signals VS the traffic lights 2 based on the detector signals DS. The traffic lights 2 can be controlled directly, however, there is usually between controller 3 and the traffic lights 2 to ensure the necessary safety a local traffic light control is arranged.

Due to the detector signals DS the Controller provides feedback on the current traffic situation. Because controller 3 has this feedback in its algorithms considered, it is a regulation in the sense the automatic. Instead of controller 3, however, can also simple control can be provided. One control a route (an object to be controlled, here the Intersection or a system of several intersections) works without feedback; contrary to that the regulation of a route on a feedback. Feedback notice the result of an action, not only their requirements. With feedback it is possible to strive for a desired goal and to be verifiable to reach. The feedback is done by measuring the Condition of the route to be regulated.

Traffic junctions are controlled in conventional systems mostly through cyclical Sequence of phases that are displayed on the traffic lights. Measure to control a traffic intersection Detectors the traffic around the intersection (Private transport, public transport, pedestrians). Based on the measurements by the detectors, driving can certain road sections recognized and it can Vehicles are counted. According to criteria used in the design the rules have been chosen, the traffic lights Allotted phase lengths and sequences.

For the purpose of coordination with neighboring Intersections are usually aimed for here, too the sequence of phases within a cycle in certain Limits can be chosen freely. But it is also a free allocation of the phases, depending on traffic volume, conceivable.

It is possible to systematically close the controller design so that it optimizes a given objective function. This can maximize the flow capacity of a Crossing or minimizing pollutants Exhaust gases. Such a systematic controller can be used Build the help of mathematics and its algorithms.

On the one hand, it is important that the regulator is provided with reliable information, on the other hand, that he gets the information in a form which he understands.

As it now shows, a control algorithm can or a control algorithm on the appearance of Detector pulses are difficult to build up, but rather on information about the number of vehicles and their speeds and positions in the driveways to the intersection: only when a controller knows these sizes can guide traffic in the best possible way. To this The purpose of the inventive execution of 2 realized device.

In this device there is an observer 5 provided that processes the detector signals DS and controller 3 (or a controller) in more detail Information in the form of observer information BA forwards. The observer is one Computer or program unit that is logical and / or physically between the detectors 4 and the controller 3 is switched. The task of the observer 5 lies in designing a model of traffic. In the The formation and updating of the model is based on the Observers on the continuously incoming detector signals, previous detector signals and the positions of the traffic lights. It also draws on past experience Measurements or simulations, from which he in particular typical speeds and norms of behavior of the Vehicles determined. The model that consists of these values is generated, the positions and are continuously simulated Speeds of individual vehicles in the range of Intersection or route. The determined from the model Values, especially e.g. the positions and speeds, but possibly also derived from it Information such as a column length in front of a traffic light, are then passed to the controller 3, which they for Calculation of traffic lights used at the intersection.

Thus, unlike the conventional one 1, in the inventive 2 the controller 3 essential more meaningful input signals are made available, that enable more precise and efficient traffic regulation. It won't just be the presence of one Vehicle in a section of the route, but its location and movement.

The mode of operation of the observer 4 is now intended explained in more detail below using a simple example will. The one underlying this example Situation is illustrated in Fig. 3. It is a matter of a one-way street 8 (with traffic in the figure from the left to the right), with a simple entrance 9 at one Intersection 1. At the intersection is a traffic light to be regulated 2 arranged. Before the intersection are along the Street 8 three detectors 4a, 4b, 4c at the positions xa, xb and xc. The last position xc lies on or after the traffic light stop bar 2.

In a simplest version, the Observer 5 lists the vehicles according to the "vehicle table", which according to its model in the area the street 8 between the positions xa and xc. In this table there is one for each simulated vehicle Speed and a position entered. Further the table contains further information, e.g. if a certain vehicle has already been detected by a detector has been. In another version, the one below the observer maintains several vehicle tables for individual sections of the route.

If detector 4a reports a vehicle, step 20 according to FIG. 4 takes place in the observer. Based on the signal from detector 4a, the observer assumes that a new vehicle has entered the simulated area. He therefore adds a new vehicle to the vehicle table, which initially shows the position x = xa and the speed v = v0 owns. The initial speed v0 is an empirical value which, as described below, can be obtained from previous measurements or simulations or can be predetermined.

If the detector 4b reports a vehicle, it checks the observer 5 in step 22 of FIG. 5 first of all whether in a vehicle is entered in the vehicle table, which currently in an environment Eb to Ab (see Fig. 3) from detector 4b but should still be from detector 4b has not been detected. The use of a tolerance range Eb to Ab takes into account the fact that the speeds assumed by the observer 5 and the positions calculated from this are usually not correspond exactly to reality.

If a vehicle that has not yet been detected in the Range Eb to Ab found, observer 5 knows that it is now exactly at point xb and corrected its position accordingly. He also leads depending on the model used, including its speed after, e.g. by calculating a new speed, which is a function of time dt that the vehicle from detector xa to detector xb (step 24).

From the time dt, the observer can also determine at what average speed the vehicle moves between positions xa and xb Has. This average speed is for each Vehicle identified and used to be a typical Value for the initial speed v0 at position xa too calculate.

If the observer finds 5 after a signal from Detector 4b in the vehicle table is not a suitable vehicle, so he assumes that the vehicle is between Position xa and xb newly inserted in the traffic. Accordingly, he adds a new vehicle to the vehicle table on, with an initial position xb and an initial velocity v1 (step 26). The initial speed v1 can in turn be from previous measurements determined value or a predetermined size.

6 finally shows the sequence a signal from detector 4c. Here, similar to in Fig. 5, first checked in step 28 whether a vehicle in the vehicle table is listed, which is from detector 4c has not yet been detected and is in an area Ec to xc is located. In this case, an asymmetrical Tolerance range selected because the "simulation path" of the observer only extends to xc.

If such a vehicle is found (step 30), the observer knows that this vehicle is now run over the stop bar and thus the simulation route has left. The vehicle turns off accordingly removed from the vehicle table.

If the observer does not find a suitable one in step 28 Vehicle, so it ignores the signal from the detector 4c, or he creates a new vehicle, which shortly afterwards, however, because it leaves the simulated area.

In this way it is possible for the observer create a model based on the detector signals and track which the individual vehicles in Range xa to xc simulated. The detector signals are used to create, delete individual vehicles or correct their data.

In addition to those shown in Figures 4-6 The observer 5 carries out synchronization steps the values in the vehicle table by looking in simulation steps the vehicle positions due to the Vehicle speeds and other parameters continuously recalculated. The speeds will change accordingly track real or simulated information.

For example, there is a simulated one Vehicle just before traffic light 2, the observer checks whether traffic light 2 is red. If so, he reduces it its speed or set it to zero. Poses the observer based on that stored in the vehicle table Data shows that there is a traffic light in front of traffic light 2 Column has formed, he determines the end of it. Approaching If a simulated vehicle gets to this end, then it continues its speed back.

The observer determines that a simulated Vehicle exceeds the Ab position without it was detected by detector 4b, it assumes that the vehicle has left traffic (e.g. has been parked) and deletes it from the vehicle table. In another version, he can continue the vehicle keep in the model and e.g. only leave when it is also not detected by a next detector.

In the simple embodiment described so far 3 only those on the street 8 moving vehicles simulated. In a normal facility However, the vehicles in the access road would also be preferred 9 detected and simulated with detectors.

In an improved model you can also different vehicle types can be distinguished. So can e.g. a further detector 4d is arranged at position xa public transport (or another, special vehicle type, e.g. Trucks, two-wheelers, etc.) detected. Then in the vehicle table the type of each vehicle is also listed. Thereby can provide additional information for the controller 3 to be provided. The vehicle type can also be used for a more individual simulation of the movements.

The number and position of the detectors is the adapted to the respective circumstances. The more detectors there are the greater the amount of measurable information, and the more accurate the state of the system be modeled. Also with the increase in measuring points the observer becomes more tolerant of faulty ones Measurements.

Is a multi-lane road simulated and the detectors are next to each other in different Lanes, such as a lane change or inaccurate driving two detectors detected by the same vehicle will.

The synchronization of the simulated vehicles by means of the detector signals enables the vehicles track individually through the system and also their speeds to determine. The determined speed serves the observer to estimate the speed of vehicles to adapt to the first Times to be noted (during generation of vehicles in the observer).

The fact that each vehicle with position and Speed can be tracked through the entire system can also be a subsequent evaluation of parameters possible that directly through the detectors not are measurable (so the average speed or the length of the queue in a certain access to Crossing).

In summary, the observer leads both Simulation and synchronization steps by. The observer uses in the simulation steps the data stored in the vehicle table, and Information about the current state of the street or crossing (e.g. the traffic light state) to the movements to calculate the vehicles of his model. In the synchronization steps (as e.g. in Figures 4-6 and 9-12, see below) leads the observer his model based on the signals from the detectors or correct it.

• Die Zahl der Fahrzeuge, die voraussichtlich in einem nächsten Zeitintervall bei der Ampel 2 eintreffen werden;
• Die Länge einer allfälligen Kolonne vor der Ampel 2;
• Die Anwesenheit, baldige Ankunft oder der geschätzte Ankunftszeitpunkt eines öffentlichen Verkehrsmittels im Bereich kurz vor der Ampel.
• Die Position und Geschwindigkeit jedes Fahrzeugs.
• Die Beschleunigung jedes Fahrzeugs.
• Die voraussichtliche Route eines Fahrzeugs bei Verzweigungen.
• Die Feststellung eines ungewöhnlichen Zustands, z.B. bei einem Unfall.

On the basis of the data stored in the vehicle table, the observer 5 can now make detailed information about the traffic situation in front of the traffic light 2 at any time and transmit this to the controller 3. The type of observer information BA which the observer 5 transmits to the controller 3 can be adapted to the algorithm used in the controller 3. Typical observer information supplied to the controller are, for example:

• The number of vehicles that are expected to arrive at traffic light 2 in a next time interval;
• The length of a possible column before the traffic light 2;
• The presence, early arrival, or estimated time of arrival of public transport in the area just before the traffic light.
• The position and speed of each vehicle.
• The acceleration of every vehicle.
• The expected route of a vehicle when it branches.
• The detection of an unusual condition, for example in an accident.

Depending on the controller algorithm, all or only part of this (or further) information to the Controller are sent. In addition, the observer is like that designed that if the simulation fails or turns on Losing value loses instead of the effective values Passes empirical values on the flow of traffic to the controller. This can e.g. in the event of a large-scale traffic jam, the extends over the range of all detectors that Case.

3 in the present example a fairly simple arrangement with only one one-way street discussed. However, the invention can also in much more complex arrangements are used that a variety of traffic lights, intersections, entrances and exits, Roadways, etc., including more complicated Arrangements are referred to here as "route" should. To monitor such complex routes, must the number of detectors will be increased. Since both the Simulation as well as synchronization in one local frameworks can also be carried out in such Models essentially continue to the ones described simple simulation and synchronization steps be used.

Programming observers for more complex Systems becomes pretty with conventional methods complex. A method was therefore developed with which an observer in a simple way defined, i.e. can be manufactured. This method is described below.

The process is based on the streets, Crossings, e.g. the positions of the detectors, the Signal paths and the desired synchronization, i.e. the Show route graphically.

An example of such a representation of the Street 8 (without the entrance 9) according to FIG. 3 is in Fig. 7 shown. The presentation comprises three areas: a simulation level 40, a synchronization level 42 and a detector plane 44.

The simulation level 40 defines the geometric structure of the route and the position of the detectors, traffic lights, etc. arranged thereon Street 8 itself is represented by a solid, directed line shown. Points 4a, 4b and 4c correspond to Positions of the detectors 4a - 4c and are to scale arranged along the street. The points Eb, Ab and Ec limit the tolerance ranges of the detectors.

In the present example, the geometric Structure of the route very simple. Because the intersection itself the representation does not exist from a single traffic lane, which from a line is represented with directional information. For more complex ones The geometric structure is more complicated and systems contains information about all traffic lanes, the directions of the traffic running on the tracks, the length the tracks, the crossings and the possibilities to one Lane change, the position of the traffic lights (traffic control signals), the locations and tolerance ranges of the detectors. she may also contain additional information such as e.g. any speed restrictions on the individual lanes or streets.

The synchronization level 42 of FIG. 7 defines the synchronization steps that the observer can perform. This level contains synchronization symbols 46a, 46b, 46c. As shown in Fig. 8 , each symbol has four inputs 50 - 53 for Parameters from the range 40 and a signal input 54. Their function is explained below.

The detector plane 44 of FIG. 7 defines finally the detectors 4a - 4c.

Starting from that shown in Fig. 7 The program of the observer is generated. This program can be carried out according to that shown in Figs. 4-6 Algorithms work. However, the following is an improved one Observation method described in is generated in the following way:

The information of the simulation level 40 are used to define the simulation steps. The simulation steps extend here each only over a single section between two Symbols (nodes) of the solid line of the simulation level 40. For this purpose, the observer creates a vehicle list for each section. If a vehicle enters a section, it will (with position and speed and possibly driving behavior) in the respective vehicle list entered. Leaves the vehicle the section, it will be deleted from the list. Entry and deletion (as well as any other Events) generate signals or messages which are sent via the dashed lines to the synchronization level 42 be passed on.

The synchronization level information 42 are used to perform the synchronization steps define. These also only extend over a local section of the route through the respective synchronization symbol 46a, 46b, 46c defined is. The synchronization steps are as follows explained briefly with reference to FIGS. 9-12.

The observer receives a detector signal on a detector input 54, as in FIG. 9 shown, a message via output 52 in step 60 give up. This message is then generated in the simulator section an entry of a vehicle in the vehicle list of the corresponding route section. The simulator checks whether the vehicle is an already simulated vehicle corresponds.

If there is a message on connection 51 (e.g. if a simulated vehicle drives over point Eb) is in the list belonging to the respective synchronization symbol entered a vehicle (step 62, FIG. 10). At a Message on connection 53 will be the corresponding vehicle (i.e. the first one in the list Vehicles) deleted (step 64, Fig. 11).

If there is a message on connection 52 (which in turn triggered by the sequence shown in FIG. 9 is) in the list of the respective synchronization symbol the corresponding vehicle is determined and its Position on each connected to connection 52 Point set. Furthermore, the speed of the vehicle determined, from the times when the Vehicle was at the point that connected to port 50 and at the point that connects to port 52 and the distance between the two points. To do this a global list of times for each vehicle, to which it is at the points shown in area 40 arrived. Connection 53 is not with the simulation level connected, the vehicle is already removed from the list here deleted.

If there is a message on connection 50, the Vehicle imprinted a speed based on experience based. (Measured travel time between Connection 50 and 52 for earlier vehicles.)

The processes according to FIGS. 9-12 are simplified shown. Included in a practical implementation they e.g. additional test steps, as in Figures 4 to 6 are shown to determine whether a vehicle is out of traffic or new in this has occurred.

• einen Eingabeteil, der es erlaubt, eine Darstellung gemäss Fig. 7 in graphischer Weise einzugeben; und
• einen Kompilierungsteil, der die in der Darstellung gemäss Fig. 7 enthaltenen Informationen in algorithmische Anweisungen oder in Parameter umwandelt, welche sodann im Betrieb vom Beobachter 5 zur Berechnung des Modells ausgeführt bzw. berücksichtigt werden.

The definition of the observer 5 by means of a representation according to FIG. 7 can be done on a computer. This computer essentially comprises the following parts or programs or program parts:

• an input part which allows a representation according to FIG. 7 to be entered graphically; and
• a compilation part, which converts the information contained in the representation according to FIG. 7 into algorithmic instructions or into parameters, which are then executed or taken into account by the observer 5 for the calculation of the model.

In one possible version, the works Compilation part not as a compiler in the classic sense but as an interpreter, i.e. he translates that in Fig. 7 shown representation during the operation of the Observer.

The instructions generated by the compilation part are stored in the observer 5. You define the synchronization and simulation steps that be carried out by this.

In the present embodiment, the Observer 5 and controller 3 designed as separate computers, on which suitable programs are running. It is also conceivable, the controller and the observer as two separate Programs on a computer or as part of a individual program to implement.

The programming of controller 3 can also via a graphical user interface that can be used as a supplement 7 is implemented for the definition according to FIG.

In the embodiment described here Invention, traffic was directed via traffic lights. The inventive Concept can however also be applied if instead of or in addition to the traffic lights, others traffic control signals actuable by the controller are used such as adjustable speed limits, Detour boards or barriers.

Claims (17)

Method for regulating or controlling the road traffic of vehicles on a route (8) by means of at least one traffic control signal (2), a control device (3, 5) for checking the at least one traffic control signal (2) and detectors (4) for detecting vehicles, characterized by the steps: Creation of a model based on signals (DS) from the detectors (4), the locations and movements of individual vehicles on the route being simulated in the model, and generation of observer information (BA) derived from the model, and Use of the observer information (BA) to control the at least one traffic control signal (2). A method according to claim 1, characterized in that in the model the speeds of each Vehicles are simulated. Method according to one of the preceding claims, characterized in that the signals (DS) of the detectors used to be used in the model Track sizes, especially around a new simulated one Generate the data of a simulated vehicle Correct vehicle or a simulated vehicle too Clear. A method according to claim 3, characterized in that at least some of the detectors (4b, 4c) a tolerance range (Eb, Ab) is assigned, whereby with a detector signal from one of these detectors (4b, 4c) it is checked whether there is a simulated model Vehicle is in the respective tolerance range, and if yes, the position of the simulated vehicle essentially to the position of the corresponding detector is set. Method according to one of the preceding claims, characterized in that for the calculation of the Model at least the following parameters available the geometric structure of the route (9), the arrangement of the detectors (4) on the route, the Arrangement of traffic control signals (2). A method according to claim 5, characterized in that the parameters are at least partially are encoded as a graphic representation and that the graphical representation is converted into instructions to calculate the model. A method according to claim 6, characterized in that the route (9) as one or more directional lines is displayed. Method according to claim 4 and one of the Claims 5 to 7 characterized in that at least a synchronization symbol for some of the detectors (4a, 4b, 4c) (46a, 46b, 46c) is assigned to which Points (Eb, Ab) on the route to determine the Tolerance range can be assigned. Method according to one of the preceding Claims, characterized in that the observer information (BA) include one or more of the following data: the position of simulated vehicles that Speed of simulated vehicles, an expected one Driving behavior of at least part of the simulated vehicles. Method according to one of the preceding Claims, characterized in that due to the Detector signals typical values of individual sizes, in particular Initial parameters for the movement of the vehicles, can be determined, in particular for simulation the movement of a new one appearing in the simulation Vehicle are used, these sizes preferably be constantly updated. Method according to one of the preceding Claims, characterized in that in the model between distinguished at least two different vehicle types becomes. Method according to one of the preceding claims, characterized in that the generation and updating of the model comprises the following steps: Creation of at least one vehicle list of the simulated vehicles on the basis of the signals (DS) from the detectors, with a location and a speed being stored for each vehicle that is in the route or a partial area of the route, Tracking the location of each vehicle based on its speed, Correction of the location of vehicles based on the signals (DS) from the detectors (4). A method according to claim 12, characterized in that acceleration to every vehicle is saved and updated. Method according to one of the preceding Claims, characterized in that for tracking of the model the state or states of the at least a traffic control signal (2) are taken into account. Device for regulating the traffic of vehicles on a route (8) comprising at least one traffic control signal (2), a control device (3, 5) for checking the at least one traffic control signal (2) and detectors (4) for detecting vehicles, characterized in that that the control device comprises an observer part (5) and a controller part (3), the observer part (5) creates a model on the basis of signals (DS) from the detectors (4), the locations and movements of individual vehicles on the route being simulated in the model, and deriving observer information (BA) from the model, and wherein the controller part (3) regulates or controls the traffic on the route based on the observer information (BA). Apparatus according to claim 15, characterized in that the observer part (5) has a memory with at least one vehicle list, whereby to every simulated vehicle position and speed and preferably an acceleration is stored. Device according to claim 16, characterized in that that to every simulated vehicle Vehicle type is stored, and that at least one Part (4d) of the detectors (4) is a vehicle type-dependent Signal delivers.

Images