EP1154389A1 - Method to determine the traffic situation in a road network - Google Patents

Abstract

Traffic data is captured through associated vehicles moving around in traffic. Indicative traffic data is captured for journey times on edges of routes through these associated vehicles. For each route's edge these journey times determine average vehicle numbers in a queue, average vehicle numbers, average vehicle speed outside the queues, average queuing time and/or average vehicle density outside the queues.

Description

The invention relates to a method for determining the traffic situation on the basis of traffic data that is in itself in the Moving registration vehicles are won for a Traffic network with traffic-regulated network nodes and connecting them Track edges.

Procedure for determining the current as well as the future Traffic conditions to be expected are mainly for road networks known in many forms and win because of the steadily growing volume of traffic is becoming increasingly important. Common traffic forecasting methods can be roughly divide into two types, namely historical progression forecasts and dynamic traffic forecasts. The former are based on previously obtained traffic situation data, from which an archive of so-called Hydrographs is created, based on which then from current Traffic situation data through a so-called matching process, where a best fitting gait line is selected is concluded on the future development of the traffic situation becomes. The dynamic traffic forecast is based on detection traffic objects or traffic conditions, such as free Traffic, synchronized traffic and traffic jam, from current traffic measurements and on the dynamic tracking of these individualized Traffic conditions. You can also use both forecasting methods can be used in combination. Such historical and dynamic traffic forecasts are e.g. in the patent DE 195 26 148 C2, the published documents DE 196 47 127 A1 and DE 197 53 034 A1 and the older German patent application 198 35 979.9. A necessary requirement of any traffic forecasting process is the determination of the current traffic situation at the time of the forecast and if necessary at earlier times.

Most common methods for traffic situation determination are geared towards transport networks where the dynamics of the Traffic flow essentially through the traffic conditions on the different track edges, i.e. the path connections between two network nodes, is determined itself, i.e. by the dynamics of the various identifiable traffic Objects and phase transitions between them. Such conditions are given for example for expressways.

On the other hand, others apply to traffic networks in metropolitan areas Relationships. There the traffic flow is mostly through the traffic regulation measures at the network nodes, e.g. in the form of Traffic lights at intersections, determined and hardly by the traffic dynamic effects on the often relatively short track edges between the nodes. It is known that in these Cases a queuing theory can be applied at which is the length of the queue before the respective traffic regulated Network nodes, the duration of free phases, during which the traffic at the relevant network node is released, and by Interruption phases during which the traffic at the network node the speed of the vehicles is stopped outside the typical queues in front of the network nodes, the inflows to the queue and the length of the route edges for traffic dynamics are important, see e.g. the publications S. Miyata et al., "STREAM", Proc. of the 2nd Word Congress on Intelligent Transport Systems, Yokohama, volume 1, page 289, 1995 and B. Ran and D. Boyce, "Modeling Dynamic Transportation Networks ", Springer-Verlag, Berlin, 1996.

In the unpublished, older German patent application 199 40 957.9 describes a traffic forecasting method which is especially for metropolitan transport networks is suitable. This traffic forecasting method is based on an acquisition of current, through the free and interruption phases the traffic-regulated network node time-discretized traffic state parameters based on how the current vehicle drain a queue, the current inflow of vehicles into the queue and the current number of vehicles in the queue. From the current, time-discretized traffic condition parameters become effective continuous traffic condition parameters determined, including at least one effective continuous vehicle drain from a queue and / or an effective continuous flow of vehicles into the Queue, based on which one or more traffic parameters based on dynamic macroscopic modeling traffic forecast, e.g. the at a forecast time expected travel time for a certain route and / or the expected traffic situation at least with regard to the number of queues or outside thereof moving vehicles and / or the expected Length of the respective queue. The content of this older one Registration is especially important with regard to there too explanations and definitions of relevant information Terms and physical quantities here in full Scope added by reference.

In a parallel German patent application by the applicant (our file: P033150 / DE / 1) is a procedure for the extraction of Traffic data from reporting vehicles moving in traffic, i.e. to obtain so-called FCD (floating car data), described, which is also specifically for transport networks of Conurbations, i.e. for transport networks where the Traffic through the traffic regulation measures at the network nodes is dominated. This process involves special extraction from FCD, i.e. of dynamic individual or registration vehicle data, which contain timestamp information, each of which designate a reporting time that is not earlier than that Time of leaving a relevant track edge and no later than the time at which the registration vehicle a section of a track edge traversed thereafter in front of a next considered network node reached. From this timestamp information can the routes of the reporting or FCD vehicles tracked and the expected travel times for the respective track edge can be determined, if necessary individually for each of several directional track sets of the same. The term "directional track quantity" denotes the Amount of the different direction traces of a track edge, which can each include one or more lanes and are defined by the one or more lanes a respective directional track set equal to the Vehicles can be used to connect the network node to continue in one or more assigned target directions happen. As for the present traffic situation determination procedure this FCD traffic data extraction method as a preferred one Basis for determining travel times for each Track edge can serve, as used here the content of this earlier application is also described herein in fully incorporated by reference.

The invention is a technical problem of providing based on a method of the type mentioned at the beginning, with which one or more traffic parameters indicative of the traffic situation comparatively using FCD information can be determined well, especially for transport networks of metropolitan areas.

The invention solves this problem by providing a Traffic situation determination method with the features of the claim 1. According to this procedure, be in traffic moving registration vehicles for the travel times on the track edges indicative traffic data, i.e. for determining travel times suitable FCD, obtained and based on this traffic data the travel times for the route edges determined. Based on then determined route edge-specific travel times determines one or more traffic situation parameters, namely the average number of vehicles in a queue of each Route edge in front of a traffic-regulated network node, the average number of vehicles in total on the track edge, the average vehicle speed on the track edge before a possible queue, i.e. between the Beginning of the route edge to the upstream end of the queue, the mean waiting time in the respective queue and / or the average vehicle density on the track edge the queue.

With this procedure it is possible to be based on suitable won FCD the current traffic situation especially for traffic networks in metropolitan areas where traffic dynamics dominated by the traffic control measures at the network nodes is determined with sufficient accuracy, i.e. based on the FCD to reconstruct. Other recorded traffic data, e.g. of stationary detectors, can also be taken into account however, this is not mandatory. The determined or reconstructed current traffic situation can then in turn as Basis for building a gangue database and further for route-based and / or for dynamic traffic forecasts serve. For such traffic forecasts about the expected Traffic situation on a metropolitan traffic network is the knowledge of the time-dependent queue lengths at the traffic-regulated network nodes and the time-dependent number of Vehicles on the respective track edge important by the method according to the invention can be obtained.

In a development of the invention according to claim 2, the Travel times and the traffic situation parameter (s) specifically one for each of possibly several directional track sets respective route edge determined separately. With that you can significantly improve the accuracy of the traffic situation determination, because it takes into account that on a track edge before a traffic-regulated network node generally different long queues for different directional lane quantities form and / or the traffic regulation at the network node mostly is also specific to directional lanes, i.e. different Holding and passage times, also free or interruption phases called, for the different direction trace amounts includes.

In a development of the invention according to claim 3 the determined current traffic information in the form of one one or more, specific to the edge and preferred specially determined traffic situation parameters specific to the directional lane quantity for a continuous generation of historical Graphs related to the average number of vehicles in the respective Queue, the queue length, the middle Waiting time in the respective queue and / or the middle one Number of vehicles used on the respective track edge.

In a further development of the invention as claimed in Another determined traffic situation parameter is the direction-specific quantity Vehicle turn rate at the respective network node taken into account, i.e. it determines how many vehicles at the respective point in time on average of a respective one Direction trace quantity of a terminating in an associated network node Line edge over the network node in a respective Directional track quantity of a line edge continuing from the network node drive in. This can be done by collecting appropriately Determine FCD by e.g. the recorded FCD information about the direction of travel or change of direction selected at the network node contain.

In a development of the method according to claim 5 is a distinguishing detection of the state of undersaturation on the one hand and the supersaturation, on the other hand, using an appropriate one Travel time criterion provided, in which the determined Travel time is compared with a threshold, which among other things the length of the track edge, a typical free vehicle speed on the same as the holding and the The duration of the traffic regulation depends on the network node.

In a further embodiment of the invention according to claim 6 is a Determination of traffic parameters taken into account according to the process according to different systems of equations for the two Cases of undersaturation or oversaturation are provided.

A method developed according to claim 7 allows one special, advantageous determination of the number of vehicles on one Track edge and the effective continuous inflow of vehicles to the track edge and also to a queue of the same, if suitable traffic data of two or there are more corresponding FCD vehicles that the relevant Drive through the edge of the track at intervals.

A development of the method according to claim 8 enables the detection of the state of total overfilling of a track edge, i.e. of a state where the queue is over the entire route edge and possibly still upstream further across the network node there into others Line edges extends into it.

A method developed according to claim 9 is taken into account Inflow and outflow sources of vehicles, such as those e.g. in downtown areas are formed by parking garages and parking lots.

In a method developed according to claim 10, a "thinned out" traffic network in terms of traffic situation determination considered that only part of all of the vehicles contains drivable route edges of an overall traffic network, e.g. only track edges of certain track types, such as Main roads. The remaining line edges are and sources of runoff from vehicles.

Fig. 1

ein Flussdiagramm eines Verfahrens zur Verkehrslagebestimmung für ein Verkehrsnetz mit verkehrsgeregelten Netzknoten auf der Basis von FCD,

Fig. 2

eine idealisierte Darstellung eines Netzknotens zur Erläuterung der vorliegend verwendeten streckenbezogenen Begriffe und

Fig. 3

eine schematische Darstellung eines Verkehrsnetzbereichs mit zwei benachbarten Netzknoten zur Veranschaulichung einer vorteilhaften FCD-Gewinnung.

A preferred embodiment of the invention is illustrated in the drawings and is described below. Here show:

Fig. 1

1 shows a flowchart of a method for determining the traffic situation for a traffic network with traffic-regulated network nodes based on FCD,

Fig. 2

an idealized representation of a network node to explain the route-related terms used here and

Fig. 3

is a schematic representation of a traffic network area with two adjacent network nodes to illustrate an advantageous FCD extraction.

The method according to the invention is described below in an advantageous manner Realization based on the process flow illustrated in FIG. 1 explained in detail. The procedure is suitable to determine or reconstruct the traffic situation in one Traffic network with traffic-regulated network nodes, especially in a road network in a metropolitan area. That took into account Traffic network can correspond to an entire traffic network, all the track edges that can be driven by the associated vehicles of a certain area, or in one "Thinned out" form only part of the route edges of the overall traffic network included, e.g. only roads from a certain one Street type minimum size, such as major roads. The The process begins with the acquisition of traffic data reporting vehicles moving in traffic (step 1), i.e. of FCD (floating car data). This FCD is preferably obtained by the in the above-mentioned, parallel German patent application (our file: P033150 / DE / 1) described procedures, which can be referred to for further details. The FCD can thereby via permanently installed terminal devices on the vehicle, however also e.g. recorded via mobile phones carried by the vehicle or forwarded.

To better understand this FCD extraction process and is also the route-related terms used here 2 shows an idealized network node, in the four Line edges j = 1, ..., 4 merge and from the four line edges i = 1, ..., 4 open out. Without loss of generality it is assumed that the junction edges j two each different direction trace amounts k = 1.2 and the opening Line edges i also two different directional track sets have m = 1.2. Each directional track set k, m can consist of one or more lanes exist from vehicles can be used equally on the network node to continue in one or more specific directions. For example, which a directional track set of a confluent Track edge include one or more lanes, from continue straight ahead via the network node or can be turned to the right while the other direction set may include one or more lanes which can be turned left.

The said FCD extraction process according to the parallel German The patent application is characterized in that data acquisition processes at least for successive network nodes not before leaving one of the respective network nodes confluent route edge j are triggered and in the respective Data acquisition process as FCD a timestamp information is obtained, one on the relevant network node related reporting time, which is not earlier than the time leaving the relevant line edge j and not later than the time at which the registration vehicle arrives Section of a route edge i traversed thereafter in front of a reached the next considered network node or in a queue of the next track edge i taken into account.

In a metropolitan area traffic network, as mentioned, the traffic dynamics or the behavior of traffic disruptions is mostly dominated by the traffic regulation at the network nodes. A queue is often formed at the end of a line edge leading into an associated network node. 3 schematically shows an exemplary snapshot from the area of a network node K, into which, among other things, a route edge St opens, at the end of which a queue W with an associated number N _q of vehicles has formed in front of the network node K. The downstream queue end lies on a terminating line or stop line An, which represents the boundary line of the line edge St at the junction with the network node K. Vehicles with a traffic flow q _in, q _enter queue W, and vehicles with a traffic flow q _out enter and into network node K, from there to enter one of the opening route edges. Three FCD vehicles FCD1, FCD2, FCD3 are illustrated by way of example, which have left the queue W of the route edge St in question and continue in different directions via the network node K. Specifically, a first FCD vehicle FCD1 continues straight ahead, a second FCD vehicle FCD2 is turned to the right, and a third FCD vehicle FCD3 is turned to the left. Boundary lines En1, En2, En3 are drawn in, at which the further route edges begin.

As described in detail in the aforementioned parallel German patent application, the FCD obtained in this way and containing node-related reporting time information are particularly well suited, among other things, for them to use them to calculate the currently expected travel time t _tr ^{(j, k)} for the respective route edge j according to their directional track quantities k to investigate. This is explained in more detail there and therefore does not require repeated explanation here. The determination of the travel times t _tr ^{(j, k)} for the one or more directional track sets k of the respective route edge j is the next step (2) in the course of the present method and can be carried out in accordance with the procedure described in the parallel German patent application. Alternatively, these currently expected travel times t _tr ^{(j, k)} can also be determined using any other conventional algorithm based on the FCD obtained for this purpose, if and to the extent that such is known to the person skilled in the art. In other words, the present method is independent of the manner in which the travel times t _tr ^{(j, k)} for the different route edges j of the traffic network are determined on the basis of recorded FCDs.

The determined current travel times t _tr ^{(j, k)} for the directional track quantities k of the route edges j of the traffic network are then used to determine whether a condition of undersaturation or oversaturation exists for the respective route edge j, possibly differentiated according to its different direction track amounts k (step 3). The state of undersaturation is defined here by the fact that the queue which arises during a hold or interruption phase, for example a red phase of a light signal system, at the end of the line edge is completely resolved by the subsequent pass or free phase, for example the green phase of a light signal system, what can be seen as behavior analogous to the state of free traffic on expressways. The state of supersaturation is defined by the fact that the queue that arises during an interruption phase is no longer completely resolved by the subsequent free phase, which can be regarded as behavior analogous to the state of heavy traffic on expressways. The more free phases a vehicle has to wait until it passes the traffic-regulated network node in front of it, the more this behavior of dense traffic increases in each directional amount of the relevant route edge of the metropolitan area traffic network.

To determine whether there is under- or oversaturation, the determined travel time t _tr ^{(j, k) is} compared with a threshold value t _s ^{(j, k)} , which is determined by the relationship t s (j, k) = L (j, k) / v free (j, k) (ρ (j, k) ) + β (j, k) (T R (j, k) -γ (j, k) T G (j, k) T R (j, k) / T (j, k) ) is defined, in each case specifically for the directional track quantity k of the route edge j with L the entire route length, with T _R the duration of the interruption or red phases, with T _G the duration of the free or green phases, with T = T _G + T _{R is} the associated traffic regulation period and is designated by β an appropriately predetermined constant and γ by the relationship γ (j, k) = q sat (j, k) b (j, k) / [n (j, k) v free (j, k) (ρ (j, k) )] is defined, the boundary condition γ ^{(j, k) being} kept less than one in each case and again in each case specifically for the directional track quantity k of the route edge j with q _sat a predetermined saturation outflow from the queue, with b an average vehicle distance in queues, ie an average queue vehicle periodicity length, and n denotes the number of lanes. With ρ is the mean vehicle density from outside the queue, ie between the beginning of the line and the beginning of the queue, moving vehicles and with v _free (ρ) the average vehicle speed outside the queue, which is dependent on the vehicle density ρ. In many cases, the average vehicle speed v _free outside the queue can be approximated by a constant v _eff , which corresponds to a typical value of v _free that is predetermined independently of the density. The constant β is greater than zero and less than one and is usually at or near the value 0.5. The quantities q _sat , T _G , T _R and thus T are predefined parameters or functions of the other traffic-indicative quantities. Furthermore, all traffic-related variables mentioned here are mostly time-dependent functions, as is understood by the person skilled in the art and which is therefore also not explicitly stated in the size descriptions for the sake of clarity.

In the case of road traffic, the parameters b and q _sat depend on the type of vehicle, in particular on the relative proportions of vehicles of different lengths on average, such as passenger cars and trucks. In this case, the parameters b and q _sat each result as the sum of the corresponding relative contributions of the different types, which in turn result as the product of the relative share of the type in question in the total number of vehicles multiplied by the associated type-specific average vehicle distance or saturation outflow. Insofar as the parameters b and q _sat in the form of their product q _sat · b are included in the above equation (2) and in the equations below, it should be mentioned that this product q _sat · b for each directional track quantity even in the presence of vehicles of different lengths regardless of the relative proportions of which remain approximately constant if the vehicle density in free traffic outside of traffic control queues can be assumed to be small compared to the queue vehicle density, which is fulfilled in a good approximation in most practical cases.

If the determined travel time t _tr ^{(j, k) is} less than the threshold value t _s ^{(j, k)} defined in this way, the state of under-saturation is inferred, while the transition to the state of oversaturation is assumed if the determined travel time t _tr ^{(j , k )} lies above this threshold t _s ^{(j, k)} .

The method then continues with the determination of traffic situation parameters describing the traffic situation on the basis of the determined travel times t _tr ^{(j, k)} for the directional track quantities k of the route edges j (step 4), the traffic situation parameters for the two states under- and oversaturation are calculated according to different, suitable systems of equations in order to then reconstruct or determine the current traffic situation. This preferably includes, in each case specifically for each directional track quantity k of the respective route edge j, the calculation of the mean total number N of vehicles, the mean number N _q of vehicles in the queue, the mean vehicle density ρ of the vehicles traveling outside the queue and from this the mean speed V _{free of} the vehicles outside the queue, the mean queue length L _q and the mean waiting time t _q in the queue.

L (j,k) q = b (j,k) N (j,k) q /n ( j,k )

In the case of undersaturation, this is done according to the following system of equations:

L ( j, k ) q = b (j, k) N ( j, k ) q / n ( j, k )

It was taken into account that the determined average travel time t _tr ^{(j, k) is} the sum of the waiting time t _q ^{(j, k)} in the queue and the average travel time t _free ^{(j, k)} for the route from the beginning of the route edge to the beginning of the queue , ie up to the upstream end of the queue, the latter resulting from the relationship t free (j, k) = (L (j, k) -L q (j, k) ) / v free (j, k) (ρ (j, k) ) results. Furthermore, the travel time t _tr, since the queue length L _q is not less can be regarded as zero, a minimum travel time t _{tr, min} = L / v _free ⁺ βT _R ² / T does not fall below for driving the completely vehicle-free section edge. This is checked in all calculations in the event of undersaturation and, if necessary, the travel time t _{tr is} kept limited to the minimum value t _{tr, min} . The relationship applies to the total number N of vehicles on the directional track quantity k of the route edge j N (j, k) = N q (j, k) t tr (j, k) / t q (j, k) , where the quotient q _{in, q} ^{(j, k)} = N _q ^{(j, k)} / t _q ^{(j, k)} indicates the mean inflow into the queue.

t (j,k) q = N (j,k) q T (j,k)/(T (j,k) G q (j,k) sat ) . In the case of oversaturation, equations 3 and 6 above continue to apply for the mean vehicle density ρ outside the queue and the mean queue length L _q , while for the system of equations then valid, equations 4, 5 and 7 apply to the mean total number N of vehicles, the average number N _q of vehicles in the queue and the average waiting time t _q in the queue, based in each case on the directional track quantity k of the route edge j, are replaced by the following relationships: N ( j, k ) = t ( j, k ) tr q ( j, k ) sat T ( j, k ) G / T ( j, k )

t ( j, k ) q = N ( j, k ) q T ( j, k ) / ( T ( j, k ) G q ( j, k ) sat ).

Here, γ _{1 is} defined by γ ₁ ^{(j, k)} = γ ^{(j, k)} T _G ^{(j, k)} / T ^{(j, k)} , with the parameter γ defined in equation 2 above, where again the formal boundary condition γ ₁ <1 applies. The natural boundary condition L≥L _q = bN _q / n also applies to the case of oversaturation, since the queue belonging to a route edge cannot become longer than the route itself. Furthermore, the trivial boundary condition applies to the total number of vehicles N that it cannot become greater than the maximum number N _max = nL / b of possible vehicles on the route length L. Correspondingly, the route edge travel time t _tr cannot be greater than the maximum waiting time t _{q, max} = N _max T / (T _G q _sat ) in a queue which extends over the entire route edge. In the event of oversaturation, all calculations are therefore checked to determine whether the travel time t _{tr is} below the maximum value t _{q, max} , otherwise it is kept limited to the same.

By solving the respective coupled system of equations, the essential parameters determining vehicle traffic ρ, average number of vehicles N, average number N _q of vehicles in the queue, average queue length L _q and average waiting time t _q in can thus be used for both undersaturation and oversaturation determine the queue for each directional lane quantity k of each route edge j of the traffic network on the basis of the FCD-based average travel times t _tr ^{(j, k)} , that is to say the current traffic situation can be reconstructed on the basis of suitably recorded FCDs which represent traffic data recorded on a sample basis .

In most cases, it is justified both for the case of undersaturation and for oversaturation to simplify the vehicle speed- _dependent average vehicle speed v _free ^{(j, k)} (ρ ^{(j, k)} ) to an effective speed _value v _eff ^{(j, k )} to be set, which is specified for the respective directional track quantity k of the track edge j constant regardless of the vehicle density ρ.

To determine the traffic ^{situation parameters, number of} vehicles N ^{(j, k)} on the relevant directional track quantity k of the route edge j and effective continuous inflow q _in ^{(j, k)} into the relevant directional track amount k of the route edge j and effective continuous inflow q _{in, q} ^{(j, k )} In the queue in question, a procedure can optionally be applied in which the difference Δt _tr ^{(j, k) of} the travel times t _tr ^{(j, k) is used} by at least two FCD vehicles which have the same directional track quantity k of the route edge j in one Drive through sufficient time interval Δt ^{(j, k)} . This time interval .DELTA.t ^{(j, k)} must be equal to or greater than the traffic regulation period T ^{(j, k)} , and the average travel time t _tr ^{(j, k)} for this case is made up of individual travel time values over the queue period T ^{(j, k)} averaged. More precisely, the time interval .DELTA.t ^{(j, k) is} the time difference between the times at which the FCD vehicles concerned enter the same directional track quantity k of the route edge j.

Specifically, the path edge inflow q _{in can be} specific for the respective directional track quantity k of the path edge j through the relationship q in (j, k) = (1 + Δt tr (j, k) / Δt (j, k) ) q sat (j, k) T G (j, k) / T (j, k) are described using the approximation Δt _free ^{(j, k)} «Δt ^{(j, k)} , which is usually well justified in metropolitan areas, i.e. the difference Δt _free ^{(j, k) of} the travel times from the beginning of the route edge to the beginning of the queue for two Successive FCD vehicles that enter the relevant directional track quantity k of the route edge j at a time interval Δt ^{(j, k)} is significantly smaller than the difference Δt _q ^{(j, k) of} the waiting times of the FCD vehicles in the queue. This relationship also contains the prerequisite that there are no sources and sinks of the vehicle flow on the relevant directional track quantity k of the route edge j.

Such sources and sinks can be formed, for example, in inner-city areas of parking garages and parking lots. In this case, a corresponding inflow q _Q ^{(j, k)} and outflow q _s ^{(j, k)} of vehicles result for the respective directional track quantity k of the route edge j. This can be taken into account, among other things, in the above equation 12 for the mean line edge inflow in that on the left side of the equation the quantity q _in ^{(j, k)} by the expression q _in ^{(j, k)} -q _s ^{(j, k)} + q _Q ^{(j, k) is} replaced. In an analogous manner, such sources and sinks of the vehicle flow can also be taken into account as a corresponding traffic flow correction when determining the other parameters relevant to the traffic situation, as described above. If the traffic network considered is a "thinned out" traffic network as mentioned above, the line edges and associated network nodes that are not taken into account are treated as further sources and sinks of the vehicle flow.

Modern traffic signal systems and similar traffic control devices at network nodes are often traffic-volume-controlled, i.e. the free-phase and interruption-phase durations vary depending on the traffic volume, so that, for example, for a direction-lane volume on which a relatively long queue has already formed, the free-phase duration is increased compared to its normal value to shorten the excessively long queue again. In other words, the interruption phase duration T _R , the free phase duration T _G and thus the round trip time T defined by the sum of these two time periods are functions that depend not only on the route edge j, the directional track quantity k and the time, but also on one or more traffic situation indicators Variables such as the flow of vehicles etc. In order to enable more global statements about the traffic situation to be made that are independent of such local fluctuations in traffic control measures, it is advisable in these cases to average the free and interruption periods and the round trip times, i.e. the traffic control period to be used, which are obtained by averaging over time intervals which are much larger than a typical round trip time without the influence of traffic volume.

L (j) q = b (j) N (j) sq ,

mit t_q ^(j,k) aus der obigen Gleichung 12 des Übersättigungsfalles, K^(j) als der Anzahl an Richtungsspurmengen der Streckenkante j und b^(j) als mittlerer Fahrzeuglänge. Wenn q_sat ^(j,k) und T^(j,k) für alle Richtungsspurmengen k einer Streckenkante j jeweils gleiche Werte haben, vereinfacht sich die obige Gleichung 19 entsprechend.Although in general a determination of the directional trace quantity specific to the various variables above according to the Indek k used is preferred, these variables can of course also only be determined specifically for the track edge without further differentiation into the individual directional trace quantities. In particular, from the above-mentioned, direction-track-quantity and route-edge-specific variables, summing consideration of all direction-track quantities of a respective route edge associated, only route-edge-specific variables can be derived. Thus, an average number N ^(j) of vehicles on the route edge j, an average number N _q ^(j) of vehicles in all queues of the route edge j, from this an average number of vehicles N _s ^(j) per lane and an average number of queue vehicles N _sq ^(j) per lane and a mean queue length L _q ^(j) that is purely specific to the edge of the route and an average waiting time t _q ^{(j) that is} also purely specific to the path edge are derived from the following relationships:

L ( j ) q = b ( j ) N ( j ) sq ,

with t _q ^{(j, k)} from equation 12 of the supersaturation case above, K ^(j) as the number of directional trace amounts of the track edge j and b ^(j) as the average vehicle length. If q _sat ^{(j, k)} and T ^{(j, k)} have the same values for all directional track sets k of a line edge j, the above equation 19 is simplified accordingly.

Furthermore, the present method makes it possible to determine whether there is a total overfill of the respective directional track quantity k of the plug edge j with the vehicles in the queue. This is the case when the queue length L _q ^{(j, k)} corresponding to the route length L ^{(j, k),} ie if the relationship b (j, k) N q (j, k) / n (j, k) = L (j, k) is satisfied, where N _q ^{(j, k) is} to be determined according to equation 11 above for the case of oversaturation. The travel time t _{tr, crit} ^{(j, k)} for which this criterion (equation 14) is fulfilled is referred to as the critical travel time. The criterion that an overfilled directional track quantity k of a route edge j of a metropolitan area traffic network blocks one or more upstream plug edges over one or more corresponding network nodes can then be that in this case the difference tt ₂ ^{(j, k)} between the current point in time t and the time t ₂ ^{(j, k)} at which the FCD vehicle in question entered the directional track _quantity k of the route _edge ^j is greater than this critical travel time t _{tr, crit} ^{(j, k)} .

It goes without saying that instead of the traffic situation parameters explicitly stated above depending on the application, only a part of this Parameters and / or additional traffic situation parameters the basis of the FCD-supported, track edge-specific and included preferably averaged direction-specific amount Travel times can be determined. For example, as more Traffic situation parameters the current turn rates at the respective Network nodes considered and determined in the form of a matrix whose matrix elements indicate the rates at which vehicles from a respective directional trace amount of a confluent Line edge over the relevant network node into a respective one Directional track amount of an opening route edge drive in.

The determination of the traffic situation parameters as explained here and thus the traffic situation can be as desired for use corresponding other applications. In particular can the data determined according to the procedure for the average number of vehicles in the respective queue, the queue length, the mean waiting time in the queue and the average number of vehicles on the respective directional lane quantity a track edge and continuously over current turn rates for the generation of historical curve lines over the concerned sizes relevant to the transport warehouse are used. In order to can create a hydrograph database and a corresponding hydrographic database Traffic forecasting system, e.g. to Travel time forecast. A traffic control center with a Memory must be equipped in which the corresponding information about traffic control measures at the network nodes and travel times for all the edges of a metropolitan road network based on a digital road map are saved. A processing unit in the Traffic control center can provide current information about the traffic regulation periods or the free-phase and interruption phases for the traffic-regulated intersections as well via the current FCD-based route edge-specific Receive travel times. Based on this data then a processing unit of the traffic center will be able to automatically Travel time forecasts for any trips on the transport network through a curve-based and / or dynamic Determine traffic forecast (step 5).

A dynamic forecast of traffic development is, for example with the older German patent application cited above No. 199 40 957 possible methods described. The predicted traffic data can then be compared with currently available Traffic data are compared, resulting in an error correction for the forecasting process can be derived by the determined current values e.g. for the turn rates and other parameters relevant to the traffic situation and / or the corresponding ones Values of the historical curves depending on any deviations found in the comparison Getting corrected.

Claims (10)

Method for determining the traffic situation on the basis of traffic data obtained by reporting vehicles moving in traffic for a traffic network with traffic-controlled network nodes and route edges connecting them,
characterized in that for the travel times (t _tr ^{(j, k)} ) on the route edges (j, k), indicative traffic data are obtained by reporting vehicles moving in traffic, the travel times for the route edges are determined on the basis of the traffic data obtained and one or more of the following traffic situation parameters are determined on the basis of the determined travel edge-specific travel times: (i) the mean number (N _q ^{(j, k)} ) of vehicles in a queue of the respective route edge (j, k) in front of an associated traffic-regulated network node, (ii) the average number (N ^{(j, k)} ) of vehicles on the respective track edge (j, k), (iii) the average speed (v _free ^{(j, k)} ) of the vehicles on the respective route edge (j, k) between the beginning of the route edge and the beginning of the queue, (iv) the mean waiting time (t _q ^{(j, k)} ) in a network queue of the respective route edge (j, k) and / or (v) the mean density (ρ ^{(j, k)} ) of vehicles on the respective Line edge (j, k) between the line edge start and the queue start. The method of claim 1, further
characterized in that
the travel times (t _tr ^{(j, k)} ) and the traffic situation parameter (s) are determined specifically for each directional track quantity (k) of the respective route edge (j). The method of claim 1 or 2, further
characterized in that
the traffic situation parameter value (s) determined on the basis of the determined route edge-specific travel times continuously for the generation of historical corridors with regard to the average number of vehicles in a respective queue, the length of the queue, the average waiting time in the queue and / or the average number of vehicles on the respective route edge (j, k) can be used. Method according to one of claims 1 to 3, further
characterized in that
as further traffic situation parameters, turn rates, which are determined on the basis of the route edge-specific travel times determined, and which each specify the rate of vehicles traveling from a confluence of directional traces across the network node into an outflow of directional traces. Method according to one of claims 1 to 4, further
characterized in that
a threshold value (t _s ^{(j, k)} ) according to the relationship for discriminatingly recognizing an undersaturation state on the one hand and an oversaturation state on the other hand t s (j, k) = L (j, k) / v free (j, k) (ρ (j, k) ) + β (j, k) (T R (j, k) -γ (j, k) T G (j, k) T R (j, k) / T (j, k) ) predefined and for the respective route edge ^{(j, k} ) it is concluded that there is under-saturation if the determined travel time (t _tr ^{(j, k)} ) is less than the threshold value (t _s ^{(j, k)} ), and it is concluded that there is oversaturation, if the determined travel time is greater than the threshold value, where L ^{(j, k) is} the route length of the route edge (j, k), T _R ^{(j, k) is} the traffic control interruption phase duration, T _G ^{(j, k) is} the traffic control free phase duration , T ^{(j, k)} = T _G ^{(j, k)} + T _R ^{(j, k)} the traffic regulation period, v _free ^{(j, k)} (ρ ^{(j, k)} ) the vehicle density-dependent mean vehicle speed in the area outside the queue and β ^{(j, k)} denote a predeterminable constant greater than zero and less than one and γ (j, k) = q sat (j, k) b (j, k) / [n (j, k) v free (j, k) (ρ (j, k) )] where q _sat ^{(j, k) is} the queue saturation outflow of the respective track edge (j, k), b ^{(j, k) is} the mean vehicle distance in the queue and n ^{(j, k) is} the number of lanes. Method according to one of claims 1 to 5, further
characterized in that
the route edge-specific traffic situation parameters mean vehicle density (ρ ^{(j, k)} ) outside the queue, mean number of vehicles (N ^{(j, k)} ), mean queue vehicle number (N _q ^{(j, k)} ), queue length (L _q ^{(j, k )} ) and waiting time (t _q ^{(j, k)} ) in the queue for the state of undersaturation can be determined by the following system of equations: ρ ( j , k ) = N ( j, k ) - N ( j, k ) q n ( j, k ) ( L (j , k) - L ( j, k ) q )

L ( j, k ) q = b ( j , k ) N ( j, k ) q / n ( j , k )

and for the state of supersaturation can be determined by the following system of equations:

N ( j , k ) = t ( j, k ) tr q ( j, k ) sat T ( j, k ) G / T ( j , k )

L ( j, k ) q = b ( j , k ) N ( j, k ) q / n ( j, k )

with γ ^{(j, k)} = q _sat ^{(j, k)} b ^{(j, k)} / [n ^{(j, k)} v _free ^{(j, k)} (ρ ^{(j, k)} )]
and γ ₁ ^{(j, k)} = γ ^{(j, k)} T _G ^{(j, k)} / T ^{(j, k)} , where in each case specifically for the directional track quantity k of the route edge j with L the entire route length, with T _R the Duration of the interruption or red phases, with T _G the duration of the free or green phases, with T = T _G + T _R the associated traffic regulation period, with q _sat a predefined saturation outflow from the queue, with b an average vehicle distance in Queues, with n the number of lanes, with v _free the vehicle speed-dependent average vehicle speed outside the queue, and with β a suitably predefined constant. Method according to one of claims 1 to 6, further
characterized in that the traffic situation parameters mean number of vehicles (N ^{(j, k)} ), effective continuous track edge inflow (q _in ^{(j, k)} ) and effective continuous queue inflow (q _{in, q} ^{(j, k)} ) based on traffic data from at least two registration vehicles that are in a time interval (Δt ^{(j, k)} ) greater than or equal to the traffic regulation period (T ^{(j, k)} ) travel the same route edge (j, k), using the difference (Δt _tr ^{(j, k)} ) of the determined travel times of these registration vehicles are determined and the relationship for the determination of the effective continuous line edge inflow (q _in ^{(j, k)} ) q in (j, k) = (1 + Δt tr (j, k) / Δt (j, k) ) q sat (j, k) T G (j, k) / T (j, k) and the approximation .DELTA.t _free ^{(j, k)} << .DELTA.t ^{(j, k)} is used, where .DELTA.t _free denotes the travel time difference from the beginning of the route edge to the beginning of the queue. Method according to one of claims 1 to 7, further
characterized in that
an overcrowded route edge is concluded if a reporting vehicle has been on the relevant route _edge ^{(j, k)} for a period longer than a critical travel time (t _{tr, crit} ^{(j, k)} ), the critical travel time being the travel time determined, which is the implied relationship b (j, k) N q (j, k) / n (j, k) = L (j, k) fulfilled, the mean queue vehicle number (N _q ^{(j, k)} ) being that for the oversaturation case. Method according to one of claims 1 to 8, further
characterized in that
Sources and sinking of the vehicle flow on the traffic network are taken into account when determining the traffic situation parameters by corresponding inflows (q _Q ^{(j, k)} ) and outflows (q _s ^{(j, k)} ) to and from the respective route edge (j, k) . The method of claim 9, further
characterized in that
the traffic network taken into account for determining the traffic situation only forms a predeterminable part of all route edges and network nodes of an overall traffic network and the route edges and network nodes not considered here are treated as sources and sinks of the vehicle flow on the traffic network taken into account.

Images