EP1480184A2 - Method for detecting road traffic characteristics at access points - Google Patents

Abstract

Points in comparative fundamental diagrams (CFD) are made available along with points for each CFD in a parameter diagram to include a traffic parameter value at each CFD point. A detector's data produces points in a fundamental diagram (FD) for a control point. The FD points are automatically compared with each CFD's points according to a set criterion until a preset matching condition is met. Independent claims are also included for the following: (a) A method for calibrating a Pulk model with outset parameters for a control station; (b) and for a method for controlling a control point so as to move off single moving vehicles; (c) and for a computer program product with a computer-readable medium to carry out the stages in the method of the present invention.

Description

The invention relates to a method for determining traffic parameters Operator stations for handling individually moving units with alternating ones Blocking and pass phases and with one in front of the operator station arranged detector.

The operating stations mentioned, such as traffic lights or Locks usually serve to block the traffic of individually moving units, such as motor vehicles. The operator stations have handling phases on that from a blocking phase with a certain, but possibly variable blocking period and a pass phase with a certain, but there may be variable pass times.

Especially with today's traffic on the streets, it is of great importance The importance of directing and controlling traffic flows in an optimal manner. Around For example, it is suitable to set and control traffic light systems necessary, statements about the traffic quality of signaled sections of the route or light signal systems (LSA) access.

For measuring the quality of traffic flow in LSA access roads or The quality of "green waves" has so far not been used in practice, that with the help of simple detection traffic parameters of LSA accesses, such as capacity (number of vehicles that can be handled in a handling phase), Quality of the coordination of successive LSA depending of utilization, information about the time extension and Frequency of disturbances and overloads. traffic parameters Up to now, the aforementioned type could only be carried out using complex test drives or by simulating the flow of traffic. In particular, it was not possible to use traffic parameters in automated Determine shape.

The methods known from the prior art have various Disadvantages. Firstly, the assessment of the quality of "green waves" subjective criteria. The optimization of "green waves" still requires a high level of planning and IT support, with the corresponding Results are nevertheless questioned by practitioners. Besides, is a systematic quality management of traffic signal systems only with high Measurement effort possible (e.g. measurement runs). Also the comparison of performance (Performance) various control methods for traffic signal systems can only be carried out with great effort.

It is therefore the object of the invention to provide a method with which Traffic parameters can be determined at operator stations with little effort can.

a)

Bereitstellen der Punkte einer Mehrzahl von Vergleichsfundamentaldiagrammen und Bereitstellen der Punkte eines Kenngrößendiagramms, das zu jedem Punkt eines Vergleichsfundamentaldiagramms einen Verkehrskenngrößenwert umfasst, für jedes Vergleichsfundamentaldiagramm,

b)

Bereitstellen der Punkte eines Fundamentaldiagramms für die Bedienstation unter Verwendung von Detektordaten,

c)

automatisches Vergleichen von Punkten des Fundamentaldiagramms mit Punkten jeweils eines der Vergleichsfundamentaldiagramme nach einem vorbestimmten Kriterium, bis eine vorbestimmte Ähnlichkeitsbedingung erfüllt ist.

This object is achieved by a method according to claim 1. Accordingly, the invention provides a method for determining traffic parameters at operating stations for handling individually moving units with alternating blocking and passage phases and with a detector arranged in front of the operating station, with the steps:

a)

Providing the points of a plurality of comparative fundamental diagrams and providing the points of a characteristic quantity diagram, which comprises a traffic characteristic value for each point of a comparative fundamental diagram, for each comparative fundamental diagram,

b)

Providing the points of a fundamental diagram for the operator station using detector data,

c)

automatic comparison of points in the fundamental diagram with points in each case one of the comparison fundamental diagrams according to a predetermined criterion until a predetermined similarity condition is met.

A fundamental diagram shows the relationship between traffic levels (Number of units per unit of time) and traffic density (number of units per path length). Both sizes, traffic volume and traffic density, are usually used for predetermined time intervals (sometimes also as measuring intervals or aggregation periods) and then for each time interval represented as a point in the fundamental diagram. In general So a fundamental diagram includes a number of discrete points. Everyone Point of a fundamental diagram consists of two values (coordinates) and gives the traffic volume and traffic density during a time interval on. Both quantities can be standardized in a suitable manner, for example by be divided by a time and / or length interval. In the graphic Representation of a fundamental diagram can show the traffic volume over the traffic density be applied; alternatively, the axes can also be interchanged (in the latter case, the coordinates of the points are reversed).

There are different options, traffic volume and traffic density and determine the points for a fundamental diagram. For example, the traffic density can be determined by explicit observation and counting of units can be determined. Alternatively, the traffic density can also be through the occupancy time of a detector (sum of the occupancy times) can be specified within a (measurement) interval. The occupancy period can can also be divided by the duration of the measurement interval, so that the Traffic density is indicated by the so-called occupancy rate. In a graphical representation of a fundamental diagram in these cases, As already mentioned before, traffic volume over occupancy time or Degree of occupancy or with interchanged axes.

The points of a characteristic diagram consist of three values (coordinates), where two of the coordinates correspond to the coordinates of the corresponding point correspond to the fundamental diagram; the third value is the corresponding one Parameter value. Possible traffic parameters that are indicated by the points of a Characteristic diagram, the maximum stowage length, the waiting time per vehicle, the average stowage length, the total waiting time of all vehicles, the degree of utilization or the like; these values are therefore everyone Orbit (which corresponds to a point in the fundamental diagram).

According to the method according to the invention, points (all or a subset thereof) of the (measured) fundamental diagram with points of the comparative fundamental diagrams automatically compared. When comparing the points are compared directly or indirectly; the latter for example by first transforming the points in a suitable manner become. With automatic comparison, a comparison fundamental diagram, the fundamental diagram - according to a predetermined similarity condition - "As similar as possible" can be determined in a simple manner. Via the characteristic diagram corresponding to this comparative fundamental diagram Corresponding traffic parameters for the (measured) Determine the fundamental diagram and thus the operator station.

By comparing with comparison foundation diagrams it is still possible to recognize faulty detectors, for example by corresponding ones Comparative fundamental diagrams are provided.

Depending on the criterion and the similarity condition, the comparison with all comparison foundation diagrams can be made carried out; alternatively, it can only work with some be carried out if the similarity condition has already been met beforehand should.

According to an advantageous development, step c) can include the steps:

Providing lattice points of a lattice section that are to be compared Points of the fundamental diagram or the comparative fundamental diagram covers,

Determine the values of a density diagram for the fundamental diagram and the at least one comparative fundamental diagram, wherein a one Lattice point assigned density value from the sum of the distances between the grid point of the grid section and the points of a subset of the Fundamental diagram or the comparative fundamental diagram according to a predetermined distance measure results.

For example, a section of a rectangular grid can be used as a grid section be used, through which the area with the points of the Fundamental diagram or the comparison fundamental diagrams covered becomes. In such a case, the base vectors of the grid are preferred aligned parallel to the axes of the fundamental diagram. Such one Density diagram therefore corresponds to a three-dimensional representation in which a density value is assigned to each (two-dimensional) grid point; the points of a density diagram consist of three coordinate values. The one The density value assigned to the lattice point results from the distance of the Points of a subset of the fundamental diagram calculated at this point and then summed up. The subset can be part or all of the points Fundamental diagram.

Various alternatives are possible as a distance measure. For example the Euclidean distance can be used. Can also be an exponential function a power of Euclidean distance can be taken as a measure of distance. Farther the density values can be standardized, for example by the sum of all density values of a comparison fundamental diagram can be divided.

By indirectly comparing points of the fundamental diagram with Points of the comparative fundamental diagram are made using density diagrams also points of the comparison fundamental diagrams in the distance or similarity investigation involved, even if there are no immediately corresponding ones Points in the fundamental diagram there. Continue on this Differentiated cases, in which in each case in the same area There are points, but these differ in frequency or density exhibit.

Zuordnen der Punkte des Vergleichsfundamentaldiagramms jeweils einem untersättigten oder einem übersättigten Zustand gemäß einem vorbestimmten Kriterium,

Bestimmen einer ersten Summe der Abstände zwischen einem Gitterpunkt des Gitterausschnitts und allen Punkten des untersättigten Bereichs des Vergleichsfundamentaldiagramms gemäß dem vorbestimmten Abstandsmaß,

Bestimmen einer zweiten Summe der Abstände zwischen dem Gitterpunkt des Gitterausschnitts und allen Punkten des übersättigten Bereichs des Vergleichsfundamentaldiagramms gemäß dem vorbestimmten Abstandsmaß,

Summieren der mit einem ersten Faktor aus einer ersten vorbestimmten Menge multiplizierten ersten Summe und der mit einem zweiten Faktor aus einer vorbestimmten zweiten Menge multiplizierten zweiten Summe, um einen dem Gitterpunkt zugeordneten Dichtewert zu erhalten.

According to an advantageous further development, the determination of the points of a density diagram for each comparison fundamental diagram can comprise the steps:

Assigning the points of the comparison fundamental diagram to an undersaturated or an oversaturated state according to a predetermined criterion,

Determining a first sum of the distances between a grid point of the grid section and all points of the undersaturated area of the comparison fundamental diagram in accordance with the predetermined distance measure,

Determining a second sum of the distances between the grid point of the grid section and all points of the supersaturated region of the comparison fundamental diagram according to the predetermined distance measure,

Sum the first sum multiplied by a first factor from a first predetermined amount and the second sum multiplied by a second factor from a predetermined second amount to obtain a density value assigned to the grid point.

One speaks of an oversaturated traffic condition when the outflow of the units at an operator station is not complete during a pass phase he follows. This may be due to the passage time being too short or due to a disability be the case. In the event of a disability, the outflow of the units is only restricted or no longer possible. This may be due to, for example Backflow of a downstream operator station may be the case. On Undersaturated traffic condition exists when the drain at an operator station completely during a pass phase, d. H. there is no overload. Each point in a fundamental diagram corresponds to one revolution, i.e. H. a blocking and a pass phase at an operator station, and can therefore can be assigned to an undisturbed or a disturbed state.

In general, an operator station (with exceptions) always occurs also an undersaturated traffic condition while an oversaturated condition not necessarily (even with a longer observation period) have to be. By dividing the point set of the fundamental diagrams or the assignment of these points to one of the two states is made possible that a fundamental diagram with comparative fundamental diagrams can be compared, although similar in terms of the undersaturated area are, but have different supersaturated areas. For example a fundamental diagram has no oversaturated area at all, however, a comparison fundamental diagram, which is also a supersaturated one Area is very similar with respect to the points of the undersaturated area be and thus have a similar traffic characterization.

In the further development described, the density diagram is determined in addition to the distinction between oversaturated and undersaturated The contributions of these areas are weighted differently by the two areas Contributions multiplied by one, in particular different, factor become. In particular, one of the two factors can be 1. By such a weighting can be the influence of different traffic levels in the various fundamental diagrams or comparative fundamental diagrams be taken into account and corrected if necessary. each Fundamental diagram is a collection of orbits over a certain Period. During this period, the traffic level fluctuates after one for the examined route section typical pattern. When measuring traffic volume one speaks over a whole day. B. from "hydrographs". These traffic intensity profiles affect the density distribution in the fundamental diagrams so that a comparison of fundamental diagrams with different Gradients or curve lines can lead to distorted results, which is corrected by the weighting described above.

Preferably, the summation of the first sum and the second sum can for all grid points that are on a straight line parallel to the traffic density axis of the Comparative fundamental diagram, with a variety of different Factors are repeated from the predetermined first and second set, and the density values obtained in each case with the corresponding density values of Density diagram of the fundamental diagram according to a predetermined criterion be compared until a predetermined similarity condition is met.

This means that a comparison foundation diagram is initially divided into two comparison foundation diagrams or partial foundation diagrams that correspond to the undersaturated and the oversaturated region. These two sub-fundamental diagrams then become density diagrams composed, with the contributions of the undersaturated and the oversaturated Range are weighted differently and this weighting varies becomes. In particular, the weighting is not chosen to be the same for all points; instead, the weighting is used for subsets of the two density diagrams carried out separately and by comparison with the density diagram of the Fundamental diagram optimized. Such a subset includes all the points that lie on a straight line parallel to the traffic density axis. As in the previous ones Further developments described here can also be the density values of the density diagrams be standardized (for example, in each case via the sum of all density values on one of the mentioned straight lines), in order to facilitate the comparison.

Normieren der Dichtewerte des Dichtediagramms des Fundamentaldiagramms, die auf einer Geraden parallel zur Verkehrsdichte liegen,

Normieren der Dichtewerte des Dichtediagramms des Vergleichsfundamentaldiagramms, die auf einer Geraden parallel zur Verkehrsdichte liegen,

wobei die vorbestimmte Ähnlichkeitsbedingung erfüllt ist, wenn die Summe der Differenzbeträge von sich entsprechenden Werten der Dichtediagramme minimal oder kleiner als ein vorbestimmter Schwellwert ist.Advantageously, the comparison of density values lying on a straight line parallel to the traffic density axis can comprise the steps for each factor from the predetermined first and second set:

Normalizing the density values of the density diagram of the fundamental diagram, which lie on a straight line parallel to the traffic density,

Normalizing the density values of the density diagram of the comparison fundamental diagram, which lie on a straight line parallel to the traffic density,

wherein the predetermined similarity condition is met if the sum of the difference amounts of corresponding values of the density diagrams is minimal or less than a predetermined threshold value.

By comparing the subsets of the density diagrams in this way, the Factors and thus the resulting density diagram of the comparison fundamental diagram be optimized in a simple and suitable manner.

Preferably, comparing density values that are parallel on a straight line lie to the traffic density axis, for all straight lines parallel to the traffic density axis, on which there is a point on the fundamental diagram are repeated, to determine an optimized density diagram.

By repeating the comparison, it is thus for each comparison fundamental diagram a density diagram determines that composed of parts like this is that it is according to the predetermined similarity condition comes as close as possible to the density diagram of the fundamental diagram.

According to an advantageous development of the method described above can in step c) for the density diagram of the fundamental diagram and the Density diagrams of the comparative fundamental diagrams show the differences in the distance values of the corresponding points are formed, and continue Density diagram of the comparison fundamental diagrams as a reference density diagram be determined for which the sum or the average of the difference amounts is minimal or less than a predetermined threshold.

In this way, one obtains from the (possibly optimized) density diagrams the comparison fundamental diagrams a reference density diagram, that of the density diagram of the fundamental diagram according to the similarity condition comes as close as possible. A comparison fundamental diagram is after of this training "as close as possible" to the fundamental diagram if - according to the 1st alternative - the sum of the difference amounts (and thus also the Average of the difference amounts) minimal compared to all other comparative foundation diagrams is. To do this, the fundamental diagram must be included all or a predetermined subset of comparison foundation diagrams be compared. According to the second alternative, the predetermined similarity condition be satisfied if the sum of the difference amounts (and thus also the average of the difference amounts) smaller than a predetermined one Is threshold. In this case, it is sufficient to use the fundamental diagram with comparative fundamental diagrams to compare until this threshold is undershot becomes. This can speed up the process.

d)

automatisches Bestimmen der Punkte eines Dichtediagramms für die Punkte des dem Referenzdichtediagramm entsprechenden Kenngrößendiagramms, wobei ein einem Gitterpunkt zugeordneter Dichtewert aus der Summe der Abstände zwischen dem Gitterpunkt des Gitterausschnitts und den Punkten einer Teilmenge des Kenngrößendiagramms gemäß einem vorbestimmten Abstandsmaß multipliziert mit dementsprechendem Verkehrskenngrößenwert resultiert,

e)

automatisches Bestimmen des Dichtewertes des Dichtediagramms des Kenngrößendiagramms, der einem Punkt des Fundamentaldiagramms entspricht.

The method can preferably further comprise the steps:

d)

automatic determination of the points of a density diagram for the points of the characteristic diagram corresponding to the reference density diagram, a density value assigned to a grid point resulting from the sum of the distances between the grid point of the grid section and the points of a subset of the characteristic diagram according to a predetermined distance measure multiplied by the corresponding traffic characteristic value,

e)

automatic determination of the density value of the density diagram of the characteristic diagram which corresponds to a point of the fundamental diagram.

With these further steps, the parameters are thus advantageously (or their density), which correspond to the determined reference density diagram, the searched traffic parameters (or their density) for the examined Fundamental diagram determined. The subset can be part or all Include points on the characteristic diagram.

Zuordnen der Punkte des dem Referenzdichtediagramm entsprechenden Vergleichsfundamentaldiagramms jeweils einem untersättigten oder einem übersättigten Zustand gemäß einem vorbestimmten Kriterium, um zwei Teilfundamentaldiagramme zu erhalten,

Bestimmen der Punkte eines Dichtediagramms für die jeweils einem Teilfundamentaldiagramm entsprechenden Kenngrößendiagramme,

Zuordnen der Punkte des Fundamentaldiagramms jeweils einem untersättigten oder einem übersättigten Zustand gemäß einem vorbestimmten Kriterium,

und wobei in Schritt e) das Dichtediagramm von den beiden Dichtediagrammen des Kenngrößendiagramms verwendet wird, das dem Zustand des Punktes des Fundamentaldiagramms entspricht.

Step c) can preferably comprise the steps:

Assigning the points of the comparison fundamental diagram corresponding to the reference density diagram in each case to an undersaturated or an oversaturated state in accordance with a predetermined criterion, in order to obtain two partial foundation diagrams,

Determining the points of a density diagram for the characteristic diagrams corresponding to a partial foundation diagram,

Assigning the points of the fundamental diagram to an undersaturated or an oversaturated state according to a predetermined criterion,

and wherein in step e) the density diagram from the two density diagrams of the characteristic diagram is used, which corresponds to the state of the point of the fundamental diagram.

This distinction between an undersaturated and an oversaturated state both in the comparison fundamental diagram corresponding to a reference density diagram as well as the examined fundamental diagram can be used for a point on the fundamental diagram with even greater accuracy a statement about one or more desired traffic parameters to meet.

For assigning the points of the fundamental diagram, the density value of the reference density diagram corresponding to a point for the undersaturated state with the density values of the reference density diagram corresponding to the point for the supersaturated state according to a predetermined one Criterion to be compared.

Such a comparison offers a simple way of making an assignment hold true. For example, the two values can be subtracted from each other, whereupon belonging to a state arises from the sign of the Difference results.

Alternatively or additionally, you can use the fundamental diagram to map the points the differences between two successive points of the fundamental diagram with the differences of two successive ones Points of the reference fundamental diagrams for the undersaturated and compare the supersaturated state according to a predetermined criterion become.

A temporal "gradient" determined in this way also allows one Achieve assignment in a suitable manner. In particular, the results both comparisons can be combined.

Bestimmen eines Vergleichsfundamentaldiagramms als Referenzfundamentaldiagramm, für das die Summe oder der Durchschnitt der Abstände der Punkte des Fundamentaldiagramms zu den jeweils nächstliegenden Punkten in dem Vergleichsfundamentaldiagramm minimal oder kleiner als ein vorbestimmter Stellwert ist,

Bestimmen eines Verkehrskenngrößenwerts, der einem Punkt des dem Referenzfundamentaldiagramm entsprechenden Kenngrößendiagramms entspricht.

According to an advantageous variant of the method, in step c) for each point of the fundamental diagram a closest point in the comparison fundamental diagram can be determined according to a predetermined distance measure, and the method can further comprise the steps:

Determining a comparative fundamental diagram as a reference fundamental diagram for which the sum or the average of the distances of the points of the fundamental diagram from the respectively closest points in the comparative fundamental diagram is minimal or smaller than a predetermined manipulated value,

Determine a traffic parameter value that corresponds to a point of the parameter diagram corresponding to the reference fundamental diagram.

This represents a possible and advantageous variant, as from the plurality of Comparative Foundation Diagrams is a reference foundation diagram that is as similar as possible can be determined. As a measure of distance, as above, for example the Euclidean distance can be used. Also, for example an exponential function of a power of the Euclidean distance as a measure of distance be taken. A comparative fundamental diagram is the fundamental diagram "as similar as possible" if - according to the 1st alternative - the sum of the Distances (and thus the average of the distances) minimal among all Comparative fundamental diagrams. To do this, the fundamental diagram with all or a predetermined subset of comparison foundation diagrams be compared. According to the second alternative, the predetermined one Similarity condition must be met if the sum of the distances (and thus also the average of the distances) smaller than a predetermined threshold is.

Bestimmen eines Vergleichsfundamentaldiagramms als Referenzfundamentaldiagramm, für das die Summe oder der Durchschnitt der Abstände aller Punkte des Fundamentaldiagramms zu allen Punkten des Vergleichsfundamentaldiagramms minimal oder kleiner als ein vorbestimmter Schwellwert ist,

Bestimmen eines Verkehrskenngrößenwerts, der einem Punkt des dem Referenzfundamentaldiagramm entsprechenden Kenngrößendiagramms entspricht.

According to another advantageous variant, the distances of each point of the fundamental diagram to all points of the at least one comparison fundamental diagram can be determined according to a predetermined distance measure in step c), and the method further comprises the steps:

Determining a comparative fundamental diagram as a reference fundamental diagram for which the sum or the average of the distances of all points of the fundamental diagram from all points of the comparative fundamental diagram is minimal or less than a predetermined threshold value,

Determine a traffic parameter value that corresponds to a point of the parameter diagram corresponding to the reference fundamental diagram.

In this alternative too, a wide variety of spacing dimensions can be used, for example the aforementioned can also be used. Compared to the one previously described Variant is avoided here that a point of the fundamental diagram, if several points of a comparison fundamental diagram are equally far away from this, randomly one of the points of the comparison fundamental diagram is assigned because of the criterion and the similarity condition for each point of the fundamental diagram, all points of a comparison fundamental diagram be taken into account.

In step c), all of the methods described above can preferably use the points of the fundamental diagram and the comparative fundamental diagrams common blocking and / or transmission periods are transformed. You can do this the points of only one of the two diagrams (fundamental diagram or Comparison fundamental diagram) or both diagrams actually modified become. In this way, the fundamental diagram and the comparative fundamental diagrams can be created compare particularly easily. In particular can the diagrams in this way regardless of the circulation, blocking and / or Pass time can be made. This allows the method to be applied to any LSA controls, even with strongly fluctuating circular or release signs, are used (e.g. public transport-prioritized signal systems or adaptive LSA controls). For example, in the case of fundamental diagrams for a Light signal system the points are transformed such that the orbital period at 1 and the red time is 0.5.

According to an advantageous development, the comparison can only be carried out in step c) for points within a predetermined traffic range become. It can also be used for fundamental diagrams that only have a specific one Show time range of a day (e.g. only with high traffic volumes), evaluated become.

According to an advantageous development of all pre-existing methods the comparison fundamental diagrams through simulations, through real measurements and / or manually, especially taking into account geometric or operational Special features of the operator station and its inflow become.

In the case of simulations in particular, special geometric boundary conditions can be used (for example, short lanes in front of traffic lights) or operational Special features (construction sites, fluctuating release times or passage times due to measures to prioritize buses or trains on traffic lights) be taken into account and adjusted. Real measurements or manually designed comparison fundamental diagrams can be used especially for recognition technical errors, for example multiple trigger detectors or Detectors with incorrect occupancy measurement are used.

According to an advantageous development, in the methods described above step c) and / or step d) and / or step e) with the aid of Matrices are performed. This enables a particularly simple implementation of the procedure. For example, when using a rectangular grid in training with density diagrams, the points of the Display density diagram in matrix form particularly well.

Bereitstellen eines Pulkmodells mit Anfangsparametern für eine Bedienstation,

Bestimmen von Verkehrskenngrößen an der Bedienstation gemäß einem der zuvor beschriebenen Verfahren,

Anpassen der Parameter des Pulkmodells in Abhängigkeit der bestimmten Verkehrskenngrößen.

The present invention also provides a method of calibrating bulk models comprising the steps of:

Provision of a bulk model with initial parameters for an operator station,

Determining traffic parameters at the operating station according to one of the methods described above,

Adjusting the parameters of the bulk model depending on the specific traffic parameters.

Bulk models form the behavior of currents from individually moving units between operator stations. The method according to the invention thus offers a Possibility to calibrate such a bulk model with little effort, what a theoretical description (modeling) of traffic at the operator station allows. This means that the setting of successive operator stations can also be set be optimized.

Bestimmen von Verkehrskenngrößen an der Bedienstation gemäß einem der zuvor beschriebenen Verfahren,

Steuern der Bedienstation in Abhängigkeit der bestimmten Verkehrskenngrößen.

Furthermore, the invention provides a method for controlling an operating station for handling individually moving units, with the steps:

Determining traffic parameters at the operating station according to one of the methods described above,

Control of the operator station depending on the specific traffic parameters.

Such control can be done offline or online. In the former case determined the data over a certain period of time and then using evaluated the procedure. With online control, for example The points of a fundamental diagram are determined and evaluated once at the beginning become. Then points for the fundamental diagram continue is calculated and the associated traffic parameter is calculated for each point or the traffic parameters. This means in particular that if the traffic quality deteriorates, for example if it occurs of a traffic jam, the operating station is actuated in a suitably modified manner can be.

The invention also provides a computer program comprising at least one computer-readable medium with which the steps of the previous described methods can be performed.

Figur 1

stellt schematisch die Anordnung eines Detektors vor einer Lichtsignalanlage dar;

Figur 2

ist eine graphische Darstellung eines Fundamentaldiagramms zur Verwendung in dem erfindungsgemäßen Verfahren;

Figur 3

ist eine graphische Darstellung eines auf eine bestimmte Sperr- und Durchlassdauer transformierten Fundamentaldiagramms;

Figur 4

illustriert die Transformation des Fundamentaldiagramms von Figur 2;

Figur 5

stellt ein Dichtediagramm eines Fundamentaldiagramms dar;

Figur 6

stellt eine graphische Darstellung eines transformierten Fundamentaldiagramms dar, bei dem eine zeilenweise Betrachtung mit getrennten Bereichen durchgeführt wird;

Figur 7

illustriert ein Beispiel des erfindungsgemäßen Verfahrens zur Bestimmung der maximalen Staulänge und der Wartezeit je Fahrzeug;

Figur 8

illustriert graphisch die Zugehörigkeit anhand der Dichte eines Punktes zum untersättigten und übersättigten Bereich;

Figur 9

illustriert graphisch die Zugehörigkeit anhand des Gradienten zum untersättigten und übersättigten Bereich;

Figur 10

illustriert graphisch eine Kombination der Zugehörigkeitskriterien aus Figur 8 und 9;

Figur 11

illustriert graphisch die geglättete Zugehörigkeit zum untersättigten und übersättigten Bereich.

Further features and advantages of the invention are explained using the following exemplary embodiments and drawings.

Figure 1

schematically shows the arrangement of a detector in front of a traffic light system;

Figure 2

Figure 3 is a graphical representation of a fundamental diagram for use in the method of the invention;

Figure 3

Fig. 3 is a graphical representation of a fundamental diagram transformed to a particular hold and pass time;

Figure 4

illustrates the transformation of the fundamental diagram of Figure 2;

Figure 5

represents a density diagram of a fundamental diagram;

Figure 6

is a graphical representation of a transformed fundamental diagram in which line-by-line viewing is performed with separate areas;

Figure 7

illustrates an example of the method according to the invention for determining the maximum stowage length and the waiting time per vehicle;

Figure 8

illustrates graphically the affiliation based on the density of a point to the undersaturated and oversaturated area;

Figure 9

illustrates graphically the affiliation based on the gradient to the undersaturated and oversaturated area;

Figure 10

graphically illustrates a combination of the membership criteria of Figures 8 and 9;

Figure 11

graphically illustrates the smoothed membership of the undersaturated and oversaturated area.

FIG. 1 illustrates an example of a geometry on which the method is based in the case of traffic lights as an operator station on a street. Along a road 1 are two operator stations, in this case traffic lights 2 and 2 'arranged. In each traffic light system there is a stop line 3 or 3 'on the Street arranged. The street shown is a one-way street, on which vehicles 4 move from left to right, as through the arrow is indicated. At a predetermined distance D before the stop line a detector 5 is arranged.

On traffic signal systems such as the LSA 2 shown, coordination with a upstream system (LSA 2 ') has a significant influence on the point distribution of the fundamental diagram. Depending on the arrival time of a group of Vehicles within one round (handling phase) face a small number or long occupancy of the detector. These occupancy periods, which in Fundamental diagram evaluated in connection with the traffic load conversely, statements about the operating quality can be made derive approaching vehicles. This can also be used to make statements about coordination or via a "green wave" between the LSA 2 and the one upstream Appendix 2 'are made possible.

An example of the graphical representation of a fundamental diagram for Light signaling systems are shown in FIG. 2. In this representation, the traffic volume Z (in vehicles per round trip time) over the occupancy period B (in seconds) shown. However, as mentioned earlier, it is also a mix up of the axes or a standardization of the sizes shown possible. In particular the traffic density can also be in a different form than the duration of the occupancy can be specified.

To carry out the method according to the invention, comparison fundamental diagrams must be used to be provided. This can be done in different ways. On the one hand, microscopic traffic flow simulations can be carried out with which the most varied traffic situations are reproduced and have the resulting fundamental diagrams created. The measuring process is determined by one or more detectors at defined intervals simulated to the stop line. Preferably, the data from each individual detector collected separately. Different types are used for such traffic flow simulations Parameters set; this includes the total duration of the simulation (for example 20 hours), the round trip time, the pass time (both the operator station itself as well as optionally arranged upstream and / or downstream Operating stations), the offset time and, if applicable, the proportions of benders and turning, the course of traffic demand or traffic intensity (for example increasing linearly until 3 p.m., then constant) etc.

To create the comparison fundamental diagrams (in Sometimes called the "pattern" below) the complete load profile up to a "full closure" (no more drainage possible) simulated.

In particular through simulations, special geometric boundary conditions, like short lanes, or operational peculiarities of a Operating station, for example due to construction sites or public transport priorities, are adjusted and corresponding patterns are generated. Then be for example, only undefined parameters of the real situation, such as the proportion of turners or turners varies. This way the catalog patterns are gradually added.

Furthermore, comparison fundamental diagrams can also be taken from real measurements to be created. This can be particularly useful when patterns are out of order Detectors originate, are to be generated. Depending on the defect (for example Flutter of a relay and the resulting multiple counts) there are other patterns, which then (contained in a regular traffic pattern or as an additional comparison fundamental diagram) for the detection of technical Errors can be used. This results in a well configurable and at the same time a highly efficient procedure for plausibility checking and Fault detection of detectors.

Comparative fundamental diagrams can also be used, especially for recognition technical errors, generated manually. The result of the evaluation is that such a comparative fundamental diagram is the measured fundamental diagram the most similar may be due to a corresponding detector defect getting closed.

Both comparison fundamental diagrams and measured fundamental diagrams depend on the period of circulation and passage. In particular can These durations vary over the course of a measurement or simulation. Around a fundamental diagram easier with the comparative fundamental diagrams To be able to compare, the points of the diagrams are preferably more suitable Transformed way. This results in a transformed fundamental diagram or comparative fundamental diagram, as it is for example in Figure 3 is shown.

If the comparison fundamental diagrams all with the same blocking and passage times have been generated, it may be sufficient to use only the points of the transform (measured) fundamental diagram.

A possible transformation is based on various characteristics of Fundamental diagrams, which are described below with reference to FIG. 4.

In general, various straight lines can be identified in the graphic representation of a fundamental diagram. The straight line a intersects the origin and has an incline that corresponds to the reciprocal of the occupancy on the detector that arises for each vehicle during (slow) travel: m = 1 / B _Fz . A favorable value for B _Fz is one second per vehicle. The quantity B _Fz (for example B _Fz = 1 ^s / _Fz ) represents the average occupancy time of the detector when a vehicle is crossed.

The straight line b is created by a parallel shift up to r ', a modified blocking time (red time). This results as r '= r - k · x, where x is the distance between the detector and stop line, r the red time and k is a reduction factor that takes into account the reduced stop time at a greater distance from the stop line. The further the vehicles are from the stop line, the shorter they are. This results from the fact that exhibited vehicles start again faster (approximately one vehicle per second) than they line up (approximately 0.5 vehicles per second). A possible value for k is the reciprocal of the vehicle length, for example with 6.4 m as the average vehicle length.

The horizontally running straight line d represents the maximum traffic volume Z _max , which results in free traffic flow for a given green time. In this view, it is proportional to the modified green time g '= tu - r' as well as to the saturation traffic level S (number of the maximum number of units that can be dispatched per opening period); tu denotes the period of circulation.

The straight line c starts on the occupancy axis from the point of the interval length (round trip time tu) and meets the intersection of the straight lines b and d.

The points in the area of the straight line a correspond to freely passing vehicles. The points in the area of route b correspond to orbits in which vehicles stand on the detector throughout the red season; there is one poor coordination of successive operator stations. Points in the area of the distance c correspond to the area of supersaturation (OS) that due to overload (too short green time) or disturbances in the drain. in the Intersection of line b and c meet the areas of poor coordination and supersaturation.

With constant B _Fz and S, the straight lines or characteristics described in the fundamental diagram are only dependent on the period of circulation and the blocking period.

A transformation of a fundamental diagram to compare for Fundamental and comparative fundamental diagrams common locking and Having pass times should preferably be unambiguous, so with new ones Reverse transformation is possible for green and round trip times. Furthermore, should the positions between the straight lines b and c are as unchanged as possible in relation to one another remain as that area most of the information of the fundamental diagram contains.

In the exemplary transformation described below, the Points first by changing the red time r 'to half the orbital time tu transformed. The x and y axes are then transformed into In relation to the round trip time, so that this becomes 1.

Depending on where a point is in the fundamental diagram, it becomes one subjected to different transformation. For this, the in Figure 4 shown areas I (left of line a), II (between line a and b) and III (to the right of line b).

The following factors are defined for the transformation: f r = r N r ' , tu . f G = tu N - r N tu - r ' · tu . where r _N = 0.5 and tu _N = 1 denote the normalized red and orbital period.

The coordinates of the points in area I are simply multiplied by the factor fg:

The transformation of the points of area II takes place via:

sodass sich dann für die Koordinaten der Punkte ergibt:

The following intermediate values are defined for the transformation of points in area III:

so that the coordinates of the points are:

sodass sich daraus die neuen Koordinaten jedes Punkts (B_N, Z_N) ergeben. In this transformation, the y-coordinates (traffic intensity Z) are always transformed in the same way, while the transformation of the x-coordinates (occupancy duration B) takes place differently. Finally, the coordinates of the points are divided by the orbital period:

so that the new coordinates of each point (B _N , Z _N ) result.

It goes without saying that not only the standardized red and round trip times r _N and tu _N , but also the entire transformation can be selected differently.

For the method according to the invention, it is furthermore not necessary for the rotund Orbital times are constant over an observation period; you can vary. In particular, through a suitably chosen transformation, it is possible show the (Z, B) values of these revolutions in a diagram. In order to the method can be applied to any LSA control system even with strongly fluctuating Round-trip or release times can be applied.

The following describes various methods that are used to create a fundamental diagram (also called measurement) with a comparative fundamental diagram (Pattern) can be compared and examined for similarity. there the points of the measurement (fundamental diagram) and the pattern (comparative fundamental diagram) represented as vectors, an entry of the Vector includes the two coordinates of a point.

berechnet werden (n_i bezeichnet die Zahl der Punkte des Fundamentaldiagramms).In the first method, for each point P _M (i) = (B _M (i), Z _M (i)) of the measurement P _M the point P _P ( j ) = ( B _P ( j ), Z _P ( j )), of a pattern P _{P is} sought which is closest to it according to a distance measure. The Euclidean distance in particular can be used as a distance measure d i, j = ( B M ( i ) - B P ( j )) 2 + ( Z M ( i ) - Z P ( j )) 2 be used. The distance of a point P _P ( j ) from the point P _M ( i ) under consideration is d _i . The similarity or distance of a comparison fundamental diagram to a considered fundamental diagram can then, for example, with

are calculated (n _i denotes the number of points on the fundamental diagram).

With this method, there may be several points on the comparison fundamental diagram the same distance to a measurement point of the fundamental diagram have, so in this case the coincidence about the assignment of the Measuring point for one of these points of the pattern decides.

To prevent this, the distance to all points can be taken for each measuring point of the comparison fundamental diagram as described in the following Procedure.

bei der die Einträge einer Zeile die Entfernung von einem Punkt der Messung zu allen n_j Punkten des Patterns und die Einträge jeder Spalte die Entfernung eines Punktes des Patterns zu den n_i Punkten der Messung angeben. Die Zahl der Punkte des Patterns n_j muss nicht gleich der Zahl der Punkte der Messung n_i sein.The starting point is the distance matrix

where the entries of a line indicate the distance from one point of the measurement to all n _j points of the pattern and the entries of each column the distance of a point of the pattern to the n _i points of the measurement. The number of points in the pattern n _j need not be equal to the number of points in the measurement n _i .

eingesetzt werden, mit dem eine besondere Gewichtung der Entfernung erzielt wird. Die Parameter a und b können geeignet gewählt werden, beispielsweise a = 20, b = 0,4. Für (euklidisch gesehen) nahe bei einander liegende Punkte geht der Wert des Abstandsmaßes u_i,j gegen 1, für weit entfernte Punkte geht er gegen 0. Damit erhält man die Gewichtungsmatrix

bei der Pattern-Punkte, die in etwa in gleicher Entfernung zu einem Messpunkt liegen auch mit etwa gleicher Gewichtung zur Geltung kommen.Instead of using the Euclidean distance directly, another distance measure can also be used as follows

can be used with which a special weighting of the distance is achieved. The parameters a and b can be selected appropriately, for example a = 20, b = 0.4. For (Euclidean) points close to each other, the value of the distance measure u _{i, j} goes towards 1, for points far away it goes towards 0. This gives the weighting matrix

the pattern points, which are roughly the same distance from a measurement point, also come into play with roughly the same weight.

bestimmt werden. Damit werden nahe Punkte sehr stark, weit entfernte Punkte (im euklidischen Sinne) praktisch gar nicht berücksichtigt. Liegen mehr Punkte des Pattern im Bereich eines Messpunktes wird das entsprechende Gewicht ebenfalls höher. Die Ähnlichkeit eines Fundamentaldiagramms und eines Vergleichsfundamentaldiagramms ergibt sich zu

je größer dieser Wert, umso ähnlicher werden das Fundamentaldiagramm und das Vergleichsfundamentaldiagramm angesehen. For each point i of the measurement, for example, a measure of the proximity to the pattern can be made

be determined. This means that close points are considered very strongly, points far away (in the Euclidean sense) are practically not taken into account. If there are more points of the pattern in the area of a measuring point, the corresponding weight will also be higher. The similarity of a fundamental diagram and a comparative fundamental diagram results in

the larger this value, the more similar the fundamental diagram and the comparative fundamental diagram are viewed.

Both variants described above can be further refined by only takes points of the pattern that are in the same area of traffic like the points of measurement. It can also be used for fundamental diagrams that only a certain time range of a day (for example, with high traffic volumes) map, be evaluated.

In the exemplary method described below, a transformed fundamental diagram spanned a grid. The grid points a value for the frequency of the points on the fundamental diagram is weighted assigned with the distances to the respective grid points. With that, the Get points from density graphs. The points of two density charts same resolution (i.e. with the same underlying grid) are with each other compared by the height for each grid point in both graphs subtracted from each other and the difference in amount is determined. Around the different frequency of certain traffic volumes in fundamental diagrams To compensate, this comparison can be made line by line (in matrix notation) be performed.

A further improvement is obtained if there are still undersaturated ones and supersaturated areas in the pattern diagram.

Analogous to the way in which a fundamental diagram is used as a vector as an example, a grid can also be used as a vector P _{G are} written, the entries of which represent the grid coordinates of the grid. The corresponding vector for a rectangular grid with 40 points in the B direction and 20 points in the Z direction has the form, for example

Analogous to the case described above, a weighting matrix is also determined here in order to determine the similarity or the distance of the fundamental diagram to the grid points (see equations (8), (11) and (12)). For this, the grid vector takes P _G the role of the measurement (indexed with i) and the fundamental diagram that of the pattern (indexed with j). The resulting weighting matrix U with the entries u _{i, j} (the same distance measure is used as above) now contains entries which give the weighting of the distance of all grid points (per grid point of a line) to the points of the fundamental diagram (one column per point of the fundamental diagram) ) specify. The dimension of the matrix is, as in the case above, n _i · n _j . The higher the values of the matrix entries in a row, the more points are close to the grid point that corresponds to this row.

berechnet. Ein entsprechendes Dichtediagramm ist in Fig. 5 dargestellt. Zusätzlich kann die Summe

bestimmt werden. Damit ergibt sich dann die normierte Dichte für jeden Gitterpunkt als

A density value can be assigned to each grid point as follows. First of all

calculated. A corresponding density diagram is shown in FIG. 5. In addition, the sum

be determined. This then gives the normalized density for each grid point as

wobei hier die Einträge beispielhaft über die Koordinaten der Gitterpunkte indiziert sind.Such a density diagram is preferably represented in the form of a matrix, the coordinates of the matrix entries corresponding to the grid coordinates. For a grid with 40 points in the B direction and 20 points in the Z direction, a 40 x 20 matrix is obtained, the entries of which each indicate the density value assigned to a grid point. Such a density matrix G _U thus has the form

where the entries are indexed by way of example via the coordinates of the grid points.

wobei von den Einträgen der Differenzmatrix jeweils der Betrag genommen wird und über alle Elemente (Einträge) summiert wird; dann kann ggf. noch durch die Zahl der Summanden geteilt werden, um den Durchschnitt zu erhalten. Mit diesem einfachen Ähnlichkeitskriterium kann die Suche nach dem ähnlichsten Muster bei zuvor berechneten Dichtematrizen sehr schnell durchgeführt werden. Das Ergebnis kann auch noch normiert werden.If density matrices G _{U, 1} and G _{U, 2} are available for two fundamental diagrams, the distance (or the similarity) between them can be calculated, for example, via

whereby the amount is taken from the entries of the difference matrix and is summed over all elements (entries); Then you can divide by the number of summands to get the average. With this simple similarity criterion, the search for the most similar pattern can be carried out very quickly with previously calculated density matrices. The result can also be standardized.

As mentioned before, one can compare the fundamental diagram and limit the comparative fundamental diagram to traffic volumes, actually in the (transformed) measurement (fundamental diagram) occurrence. This is always useful if the value range for the traffic volume Z in the measurement does not cover the entire range up to the saturation traffic level contains or if, for example, only during periods of high load is measured.

Such a limitation can be achieved by using all of the Z-values of the transformed Fundamental diagram are sorted in ascending order. Possibly can discard the first and last x% (percentile, e.g. x = 5) become. The first and last values in the list represent the limit values Zmin and Zmax. Then the density matrices of the fundamental diagram and the Comparative fundamental diagram limited to the rows whose traffic volumes Z lie between Zmin and Zmax. The similarity is then for these Equation (19) determined. The result can also be standardized here, for example about the number of grid rows with different Z.

An area-specific, line-by-line comparison can be used for further refinement can be carried out with variable weighting of supersaturation. Every measured Fundamental diagram is a collection of orbits over a larger one Period. The traffic volume usually fluctuates during this period a typical pattern for the section of the route examined. During measurements The traffic volume over a whole day is called "gangways".

Certain patterns of traffic intensity are also used as a basis for the patterns Service. Because the graphs are based on the density distribution in fundamental diagrams the comparison of a fundamental diagram with comparison fundamental diagrams with different guiglines to distorted results to lead.

Another important criterion is the distinction between undersaturated (US) and oversaturated round trips. Fundamentals generally always have an undersaturated area and if the green times are too short, they also show the upper one Section of the (OS) area (overload without disturbance in the drain, area around the intersection of straight lines b, c and d).

Drainage disorders leading to the right branch in a fundamental diagram lead with a negative slope, do not have to do so or not to the same extent to be available.

To simplify such fundamental diagrams with comparative fundamental diagrams To compare wisely, the following procedure can be used.

First, two density matrices G _{U, US} and G _{U, OS are} calculated, which contain only the undersaturated or the oversaturated points of the comparison fundamental diagram. The separation or assignment of the points of the pattern into undersaturated and oversaturated points can be carried out with the known traffic parameters in each round according to simple rules.

For example, it can initially be assumed for a point that it is the under saturated area. When the maximum back pressure in vehicles (Lmax) is greater than 1.5 Z (Z the traffic volume) and half of the green time is less than Lmax, then the point is assigned to the oversaturated area. If Lmax is greater than 1.2. Is Z and the point before it was oversaturated, the point is also assigned to the oversaturated area.

Then for each line of the grid matrix reduced to the traffic volumes actually to be taken into account, the measurement for all k with k = {0; 0.2; 0.4; ... 2.0 becomes a composite pattern matrix line by line G U . P . row ( k ) = G U . US ,Row + k · G U, OS ,Row calculated. The elements or entries of each line of the pattern matrix G _{U, P (k)} are standardized via the sum of the element values of the line. The elements of each line of the density matrix G _{U, M of} the fundamental diagram are also standardized by the sum of their element values. Then a line-by-line comparison of the grid matrix with the pattern matrix takes place as follows:

Then an average of the best similarities of all rows is formed.
This then represents the similarity or the distance between the fundamental diagram and the comparative fundamental diagram.

The line-by-line analysis with the two separate areas is shown in the Figure 6 illustrates the fundamental diagram shown.

A vector k _Best is defined, which has the k-value as entries for each row, which leads to the greatest similarity in the corresponding row of the matrices. From this, a matrix K _{Best is} generated, which has the dimension of the density matrices examined and whose elements of one row are identical to the corresponding value of the vector k _Best included. This results in G U . P Best = G U , US + K Best * G U , OS as the best approximation of the comparison fundamental diagram to the measurement, whereby the operation "*" means the element-by-element multiplication. This grid matrix is used below for determining traffic parameters.

With the comparison fundamental diagrams, especially from simulations, additional parameters are available in addition to the points of the fundamental diagram via the statistical evaluation of traffic data. The following table gives an example of such data for a simulation. B [s] Z [s] TU [s] TGR [s] Lmax [veh] WFZ [s] B_spec [s] dt [s] US / OS GradStoe , , , 24.66 8th 60 20:25 9 24.1 0 22:25 1 00:00 14.92 13 60 20:25 9 19.7 3 99 1 00:04 42.91 10 60 20:25 11 35.4 7 11.75 1 00:10 46.31 11 60 20:25 12 37.4 8th 11.75 1 00:11 46.32 11 60 20:25 14 38.1 10 11 1 00:14 43.51 14 60 20:25 14 44.2 6 13:25 1 00:09 44.66 11 60 20:25 14 50.3 8th 11 1 00:11 47.78 11 60 20:25 15 48.6 11 10:25 1 00:16 45.12 13 60 20:25 16 60.1 7 13:25 1 00:10 47.25 11 60 20:25 21 64.0 11 8.75 2 00:16 46.59 13 60 20:25 24 59.5 9 11.75 2 00:13 45.07 12 60 20:25 25 80.0 8th 11.75 2 00:11 43.63 14 60 20:25 17 83.9 6 13:25 2 00:09 42.84 14 60 20:25 16 54.4 5 13:25 1 00:07 46.21 12 60 20:25 17 54.7 9 11 1 00:13 43.88 14 60 20:25 18 53.9 U5 13:25 1 00:07 44.91 12 60 20:25 15 59.5 9 11 1 00:13 45.86 13 60 20:25 16 50.6 6 11.75 1 00:09 46.19 11 60 20:25 22 66.2 8th 11 2 00:11 46.56 11 60 20:25 22 68.3 10 11 2 00:14 44.11 13 60 20:25 22 81.4 7 13:25 2 00:10 43.61 14 60 20:25 19 74.0 6 13:25 2 00:09 43.94 13 60 20:25 17 65.6 6 13:25 2 00:09 42.85 14 60 20:25 13 45.6 5 13:25 1 00:07 43.99 14 60 20:25 14 46.8 5 13:25 1 00:07 46.59 12 60 20:25 13 44.9 10 11 1 00:14 46.01 13 60 20:25 16 43.0 9 11 1 00:13 46.11 12 60 20:25 16 42.5 8th 13:25 1 00:11 42.27 13 60 20:25 16 54.2 5 5 13.25 1 00:07 ...

definiert werden, wobei x der Abstand zwischen Detektor und Haltlinie ist.In addition to the quantities already mentioned, the special occupancy duration B_spec and the filling time dt (as used in EP1 276 085) and the degree of the disturbance GradStoe are shown in this table. When the flow of traffic comes to a standstill, the degree of disruption goes to 1. The degree of disruption can, for example, change

can be defined, where x is the distance between the detector and the stop line.

An example of a determination of traffic parameters is shown schematically in FIG shown.

In step (A), those provided in the comparative foundation diagrams B and Z data transformed as described above. A corresponding Transformation also takes place for the measured points of the fundamental diagram in step (D) instead.

The remaining steps of the method are explained below. In step (B), the traffic parameters to be determined (exemplary shown: maximum stowage length Lmax and waiting time per vehicle WFz) a transformation independent of the signal data (green / round trip time) and at Supersaturation made from the route length used in the simulation. As The result is one or more grid matrices, the elements of which are the mean values of the respective parameter for the (B, Z) point of the grid element represent (step (C)). These steps can be done offline.

For a transformed fundamental diagram, a time series of points is created d. H. the points are sorted according to their temporal occurrence, so that a path is rounded by rounding to the nearest grid points (in time Terms) by the lattice matrix or density matrix of the fundamental diagram results.

The values of the grid matrices calculated in step (C) for the traffic parameters are shown the respective points of the time series from step (E) time series of the transformed state variables (step (F)). Will the Values of these time series according to the measurement conditions (tu, tgr (green time), route length) back-transformed (step (H)) and optionally smoothed the result is the correspondingly determined traffic parameters.

In step (G) it is checked for each point of the measurement whether it is the undersaturated one or saturated traffic condition. The assignment takes place depending on this a transformed characteristic value from US or OS and depending on it also the reverse transformation in step (H).

Examples of these steps are now described in detail below.

The aim of the standardization step (B) are grid matrices analogous to the density matrix G _U , only the matrix values here contain maximum _{accumulation lengths} (at G _Lmax ) or average waiting times (at G _WFz ) .

The normalization of Lmax and W over green and orbital time is dependent performed from the state of under- or oversaturation.

For each measuring point of the pattern that is in the undersaturated state, normalization is made via tgr and tu as follows:

This is based on the connection that the maximum accumulation length with varying Green and circulation times essentially proportional to the green time, the waiting time is essentially proportional to the red time.

In the supersaturated state, the length of the entire route lying in the inflow of the LSA (for example in FIG. 1 the distance between the stop lines 3 and 3 ') is divided: L Max, OS, N = L Max L route

In a first approximation, the congestion extends after a short time with oversaturation the entire route in the inflow of the LSA. A certain "modulation" So depending on traffic volume and duration of occupancy, can with the shown standardization can be achieved.

The approach for the calculation of the waiting time in the oversaturated state in (G) is required only measured values, so that no standardization is required in (B) is.

Using a grid vector P _G , which contains all the points of the grid (see formula (15)), is used in step (C) in conjunction with the points (B, Z) of the pattern P _P according to formulas (8), (11), (12)) creates a general weighting matrix W 'analogous to the above. If this is still standardized line by line (sum of the elements per line is 1), the weighting matrix W is obtained.

If you take only the subset from the points of the pattern P _{P, US} , which includes the points in the undersaturated state, and carries out the above-mentioned calculation, a weighting matrix W _US is obtained in the same way and, accordingly, a weighting matrix W _OS for the points in the oversaturated state .

und einer anschließenden Anordnung in Matrixform analog zu G_u oben Gittermatrizen G_Lmax,US und G_Lmax,OS der maximalen Staulängen je Gitterpunkt berechnet werden.The normalized maximum accumulation lengths for the undersaturated and the oversaturated state can be sorted according to their temporal occurrence, so that vectors L _{max, US , N} and L _{Max, UOS , N} analog to P _{P , US} and P _{P , OS} receives. So that over

and a subsequent arrangement in matrix form analogous to G _u above _{grid matrices} G _{Lmax, US} and G _{Lmax, OS of} the maximum _{accumulation lengths} per grid point.

Similarly, from the time series of waiting times without oversaturation, W F _{Z US, N} , a lattice _matrix G _{WFz, US} can be determined.

If grid matrices are available for queue lengths or waiting times, each point of the measurement of the fundamental diagram in step (E) can be approximated by a position in the grid. Separated according to the time series of the standardized assignment B _{M , N} and the normalized traffic volume Z _{M , N} can be the index vectors for columns k _B and lines k _{Z of} the lattice matrices can be calculated as follows.

Operation ones () creates a 1 vector (vector only with ones as entries) the dimension of the argument; rounding () rounds the argument. The values of the result vectors can be due to software reasons be limited to the range of values used as the index for addressing elements of matrices is allowed. Set the integer values of these vectors now represent numbers of the columns or rows of the grid matrix.

In step (F), the queue lengths or waiting times can be determined from the corresponding grids for each cycle as follows:

The traffic parameters are thus for a measured point of a fundamental diagram "Maximum stowage length" and "waiting time per vehicle" determined Service.

Equation (28) applies to matrices that have unsaturated or oversaturated traffic conditions describe. As described, the corresponding matrix is for the waiting time not necessary in the case of oversaturation, and thus the corresponding arithmetic operation.

The distinction or assignment of undersaturated and oversaturated points (Step (G)) can, for example, already be based on the position in the fundamental diagram the measurement itself (see Fig. 4): the points are clear To the left of route b, if they represent unsaturated round trips, they are sufficient to the right of this stretch, oversaturated. This can be done using suitable distance criteria be implemented.

In the vicinity of the intersection of lines b and c, undersaturation can occur Points from poor coordination with orbits involving oversaturation ruled, overlay.

For such points, the affiliation can be checked in two ways and the results are combined. The basis is always the comparison fundamental diagram, which is most similar to the measurement. Measured the distance of each measuring point to the undersaturated and the oversaturated Partial pattern.

On the one hand, the affiliation can be determined based on the location of the point. The density matrices G _{U , US ,} G _{U , OS are used for this} . Their matrix entries represent the density that the fundamental diagram has at the location of the entry (at the corresponding grid point). If the density of the matrix of undersaturated points at one point is higher than the density of the oversaturated matrix, the membership of the undersaturated state is higher than that of the oversaturated, and vice versa.

The matrix G _{U, P Best} can be used, which is defined in formula (22). The normalization is done via a matrix

The matrix S has the dimensions of the lattice matrices and is made up of vectors s together, the elements of which are the row-by-row sums of the density lattice matrix G _{U , P Best} represent. It can be used for standardization and takes into account the proportion of disturbances in the fundamental diagram of the measurement for each traffic level.

The density of the undersaturated subset of the pattern (standardized on the basis of the measurement) is calculated G U, US, N = S * G U , US that of the supersaturated subset G U . OS . N = S * K Best * G U, OS

The affiliation of the measuring points to the undersaturated and oversaturated subset can be determined within the grid as follows:

The only difference between the two matrices is the sign of their elements. ever the lower the value of an element, the greater the affiliation with it corresponding area.

Transferring the membership to the individual points in the time series of the measurement using formula (27) analogously to equation (28) results in:

8 shows such a course using an example.

A determination of membership based on the gradient can be found in particular in the cases in which measuring points result from poor coordination and from oversaturation be in the same area of the fundamental diagram.

The "movement" of the points is used as a criterion. For this, the change (difference) to the previous point of the time series is calculated for each point of the standardized measurement and the (standardized) pattern. The starting point is then the respective vector for both time series:

The individual steps are as follows:

bestimmt. Anschließend erfolgt eine Umwandlung gegengerichteter Bewegungen für alle Elemente:

First, a difference vector

certainly. Subsequently, counter-movements are converted for all elements:

wobei beispielsweise α = 0.1.Optionally, smoothing can take place

where, for example, α = 0.1.

Then a notation can be made as a complex vector with j for the imaginary part:

Applied to the subsaturated and oversaturated subsets of the pattern, two gradient vectors Δ result for the pattern P _{P, US} and Δ P _{P, OS} . There is only one gradient vector Δ for the measurement P _M.

The gradient vectors of the pattern represent like the gradient vector of the measurement Time series represent, whereby the gradient vectors of the pattern from several Sections of different time periods can be composed.

It is now a matter of bringing all the points of the measurement together with points of the partial pattern which are similar in position in the fundamental diagram and comparing the property "gradient". The weighting matrices W _US and W _OS can be used for this, since they "project" the points of the pattern onto the points of the measurement. This results in analogy to formula (26)

ergeben sich die Zugehörigkeitsvektoren Γ_US und Γ_OS. Der Verlauf dieser Zugehörigkeitszeitreihen für das Beispiel aus Fig. 8 ist in Fig. 9 gezeigt.Vectors Δ P T / P, US and Δ P T / P, OS with the same dimension as the gradient vector of the measurement. about

the membership vectors Γ _US and Γ _{OS result} . The course of these membership time series for the example from FIG. 8 is shown in FIG. 9.

The following equation can be used to summarize both evaluations for each measuring point (vector corresponding to the time series of the measurement): p i = a 1 · Δ i + a 2 · Γ i where the factors are, for example, as a ₁ = 1; a ₂ = 1 can be selected. The lower the result, the greater the affiliation to a subset. On the basis of his calculation rule, the affiliation with regard to the position of the point Δ _{i can} also become negative. The membership based on the chronological sequence of points is represented by a (positive) Γ _i , which goes to zero if the membership is large. In the example of FIGS. 8 and 9, the course according to FIG. 10 results.

Based on this, each individual point of the measurement can be assigned to the undersaturated or oversaturated state:

was einen Verlauf wie in Fig. 11 ergibt.To avoid nervous estimation behavior, smoothing can be carried out:

which gives a course as in FIG. 11.

Finally, with a threshold value (1.5 in the example), the final classification of each measuring point into the US (1) or OS (2) range can be made:

If it is determined for each measuring point whether it is under- or oversaturated, the normalized accumulation lengths and waiting times can be assigned to the corresponding area and then de-normalized (step (H)). Corresponding to the formulas (24) and using the values from (28), the maximum stowage length finally results:

The average waiting time per vehicle can be determined as

L _{Fz is} the average vehicle length, e.g. 6.4m. Z _M (i) is the smoothed traffic volume of the measurement, calculated, for example, using exponential smoothing:

The degree of disturbance can also be determined for the circuits with supersaturation. Instead of the normalized maximum accumulation length, the degree of disturbance, as defined above, can be transferred into a corresponding grid G _{GradStoe, OS, P} and via the index vectors without further normalization k _Z ( i ) and k _B ( i ) are evaluated again with reference to the time series of the measured values. Additional traffic parameters can be determined in the same way using appropriate procedures for standardization and de-standardization.

If one takes from the measured values which are assigned to the oversaturated area, the degree of the disturbance does not exceed a certain limit value, for example 0.3, one obtains an approximation of the circulations in which saturated traffic flow took place without disturbances in the outflow. The saturation traffic intensity can then be determined from the traffic intensity and the release time:

Claims (20)

Method for determining traffic parameters at operator stations for handling individually moving units with alternating blocking and passage phases and with a detector arranged in front of the operator station, with the steps: a) providing the points of a plurality of comparative fundamental diagrams and providing the points of a characteristic variable diagram, which comprises a traffic characteristic value for each point of a comparative fundamental diagram, for each comparative fundamental diagram, b) providing the points of a fundamental diagram for the operator station using detector data, c) automatic comparison of points in the fundamental diagram with points in each case one of the comparison fundamental diagrams according to a predetermined criterion until a predetermined similarity condition is met. The method of claim 1, wherein step c) comprises the steps: Providing grid points of a grid section that covers the points of the fundamental diagram or the comparison fundamental diagram to be compared, Determining the points of density diagrams for the fundamental diagram and the at least one comparative fundamental diagram, wherein a density value assigned to a grid point results from the sum of the distances between the grid point of the grid section and the points of a subset of the fundamental diagram or the comparative foundation diagram according to a predetermined distance measure. The method of claim 2, wherein determining the points of a density diagram for each comparative fundamental diagram comprises the steps of: Assigning the points of the comparison fundamental diagram to an undersaturated or an oversaturated state according to a predetermined criterion, Determining a first sum of the distances between a grid point of the grid section and all points of the undersaturated state of the comparison fundamental diagram according to the predetermined distance measure, Determining a second sum of the distances between the lattice point of the lattice section and all points of the oversaturated state of the comparison fundamental diagram according to the predetermined distance measure, Summing the first sum multiplied by a first factor from a predetermined first quantity and the second sum multiplied by a second factor from a predetermined second quantity in order to obtain a density value assigned to the grid point. 4. The method of claim 3, wherein summing the first sum and the second sum for all grid points that are on a straight line parallel to Traffic density axis of the comparison fundamental diagram lie with a Plurality of different factors from the predetermined first and second quantity is repeated, and the density values obtained in each case with the corresponding density values of the density diagram of the fundamental diagram are compared according to a predetermined criterion, until a predetermined similarity condition is met. The method of claim 4, wherein comparing density values lying on a straight line parallel to the traffic density axis for each factor from the predetermined first and second set comprises the steps: Normalizing the density values of the density diagram of the fundamental diagram, which lie on a straight line parallel to the traffic density axis, Normalizing the density values of the density diagram of the comparison fundamental diagram, which lie on a straight line parallel to the traffic density axis, wherein the predetermined similarity condition is met if the sum of the difference amounts of corresponding density values of the density diagrams is minimal or less than a predetermined threshold value. The method of claim 4 or 5, wherein comparing density values, that lie on a straight line parallel to the traffic density axis, for everyone Straight lines parallel to the traffic density axis, on which a point of the fundamental diagram is repeated, an optimized density diagram is repeated to determine. Method according to one of claims 2-6, wherein in step c) for the Density diagram of the fundamental diagram and the density diagrams of the Comparative fundamental diagrams show the differences in the distance values of the corresponding points are formed, and also a density diagram of the comparative fundamental diagrams as a reference density diagram is determined for which the sum of the difference amounts is minimal or smaller than a predetermined threshold. Method according to claim 7, with the further steps: d) automatic determination of the points of a density diagram for the points of the characteristic diagram corresponding to the reference density diagram, wherein a density value assigned to a grid point from the sum of the distances between the grid point of the grid section and the points of a subset of the points of the characteristic diagram multiplied by the corresponding distance measure Traffic parameter value results, e) automatic determination of a density value of the density diagram of the characteristic diagram which corresponds to a point of the fundamental diagram. The method of claim 8, wherein step c) comprises the steps of: Assigning the points of the comparison fundamental diagram corresponding to the reference density diagram in each case to an undersaturated or an oversaturated state in accordance with a predetermined criterion, in order to obtain two partial foundation diagrams, Determining the points of a density diagram for the characteristic diagrams corresponding to a partial foundation diagram, Assigning the points of the fundamental diagram to an undersaturated or an oversaturated state according to a predetermined criterion, and wherein in step e) the density diagram from the two density diagrams of the characteristic diagram is used, which corresponds to the state of the point of the fundamental diagram. The method of claim 9, wherein for mapping the points of the fundamental diagram the one corresponding to a point on the fundamental diagram Density value of the reference density diagram for the undersaturated State with the density value of the reference density diagram corresponding to the point for the supersaturated state according to a predetermined Criterion is compared. A method according to claim 9 or 10, wherein for assigning the points of the fundamental diagram, the differences of two in time following points of the fundamental diagram with the differences of two successive points in time of the reference density diagram corresponding comparative fundamental diagrams for the undersaturated and the supersaturated state according to a predetermined Criterion to be compared. The method of claim 1, wherein in step c) for each point of the fundamental diagram a closest point in the comparison fundamental diagram is determined according to a predetermined distance measure, and with the further steps: Determining a comparative fundamental diagram as a reference fundamental diagram for which the sum of the distances of the points of the fundamental diagram from the respectively closest points in the comparative fundamental diagram is minimal or less than a predetermined threshold value, Determine a traffic parameter value that corresponds to a point of the parameter diagram corresponding to the reference fundamental diagram. The method of claim 1, wherein in step c) the distances of each point of the fundamental diagram to all points of the at least one comparison fundamental diagram are determined according to a predetermined distance measure, and with the further steps: Determining a comparative fundamental diagram as a reference fundamental diagram for which the sum of the distances of all points of the fundamental diagram from all points of the comparative fundamental diagram is minimal or less than a predetermined threshold value, Determine a traffic parameter value that corresponds to a point of the parameter diagram corresponding to the reference fundamental diagram. Method according to one of the preceding claims, wherein in step c) the points of the fundamental diagram and the comparative fundamental diagrams transformed to common blocking and / or passage times become. Method according to one of the preceding claims, wherein in step c) comparing only for points within a predetermined traffic range is carried out. Method according to one of the preceding claims, wherein the comparison fundamental diagrams through simulations, through real measurements and / or manually, especially taking into account geometric or operational characteristics of the operator station and its inflow, to be provided. Method according to one of the preceding claims, wherein step c) and / or step d) and / or step e) with the aid of matrices become. Procedure for calibrating a bulk model with the steps: Provision of a bulk model with initial parameters for an operator station, Determining traffic parameters at the operating station according to the method according to one of the preceding claims, Adjusting the parameters of the bulk model depending on the specific traffic parameters. Method for controlling an operator station for the handling of individually moving units with the steps: Determining traffic parameters at the operating station according to the method according to any one of claims 1-17, Control of the operator station depending on the specific traffic quality. Computer program product comprising at least one computer readable Medium with which the steps of the method according to one of the preceding Claims are executed.

Images