EP1528524B1 - Method for traffic prognosis based on historical data - Google Patents

Abstract

Process for traffic prognosis comprises constructing a database using an aggregation of all traffic patterns at permanently effective bottlenecks. On identification of a new, temporarily effective bottleneck an aggregated traffic pattern is determined and the database is updated. A specific travel time is assigned to each aggregated traffic pattern. The traffic prognosis is prepared by determining and displaying travel times for all or one part of the section of the traffic network taking into account the travel times assigned to the aggregated traffic patterns of the updated database.

Description

The invention relates to a method for gang-based traffic forecasting according to the preamble of claim 1. In this method, a hydrographic database is constructed taking into account dense traffic patterns at effective bottlenecks of a considered traffic route network, and traffic predictions are made using current traffic measurements and comparison thereof created with data from the hydrograph database.

Such traffic forecasting methods are described in various earlier patent applications of the applicant on this subject, see for example the DE 100 62 856 A1 . DE 100 57 796 A1 and the older one German patent application 103 36 590.7 , Like there and for example also in the DE 199 44 075 A1 described in detail, the dynamics of dense traffic at the so-called effective bottlenecks of a considered traffic route network, eg a road network, within a considered area, by the formation and behavior of the so-called dense traffic pattern describe, in this case also called traffic pattern. Such dense traffic patterns consist of the distinguishable state phases "synchronized traffic", "compressed synchronized traffic" and "wide traffic congestion", with which traffic dynamics can be adequately described with additional consideration of the "free traffic" state. Such a procedure is also used in the EP 1 176 589 described.

In the DE 197 53 034 A1 A method is described for gangway-based traffic forecasting using a hydrograph database based on traffic data supplied by multiple data sources, such as fixed detectors and in motion mitbewegende sampling vehicles, so-called FCD vehicles. For prognosis, probabilistic selection will select a most likely hydrograph.

In most cases, the conventional gang-based forecasting methods operate on the basis of stationary vehicle counting stations and other means of measuring traffic flows, which are then converted into medium speeds and finally required travel times. Investigations show that this method usually leads to satisfactory results in free-flowing traffic, since in this case a medium speed can be unambiguously assigned to the traffic flow via the so-called fundamental diagram. On the other hand, the reliability of this method is limited when there are dense traffic conditions. Depending on the other circumstances, these may include different of the above-mentioned individualizable states "synchronized traffic", "compressed synchronized traffic" and "moving wide traffic jams" with the same traffic flow. On the other hand, the average speed and the travel time are strongly dependent on which of these dense traffic conditions exists, see the journal articles B.S. Kerner and P. Konhäuser, Phys. Rev. E, 48, p. R2335, 1993 and Phys. Rev. E, 50, p. 54, 1994 and B.S. Kerner and H. Rehborn, Phys. Rev. E, 53, p. 1297, 1996.

Although it can be considered as a remedy, instead of the measurement of traffic flows and the subsequent estimation of the average speeds and travel times to make a direct measurement of travel times regardless of the traffic flows, for example by FCD vehicles. This approach will often work well for fixed transport infrastructure, but will encounter difficulties with infrastructure changes, such as new ones construction sites and other temporary effective bottlenecks. Because a prognosis of travel time increase by a newly created construction site requires in this method, new, time-consuming travel time measurements on different days by FCD vehicles, so that in this way travel time-line forecasts can not be created within a short time after setting up a construction site.

The invention is based on the technical problem of providing a method of the type mentioned above, with the help of which traffic-based traffic forecasts can be created comparatively reliably, even with temporary effective bottlenecks of the considered traffic route network.

The invention solves this problem by providing a method having the features of claim 1. In this method, an aggregation of all traffic patterns at permanent effective bottlenecks of the considered traffic route network is performed to build the hydrographic database. In addition, when a newly occurring temporary effective bottleneck is detected, an associated aggregated traffic pattern is determined and the hydrograph database is updated with it. Each aggregated traffic pattern is assigned a certain travel time, and to generate the traffic forecasts travel times for all or a selectable part of the sections of the considered traffic route network are determined and displayed on this basis.

Thus, according to the invention, traffic patterns based on the composition of the various dense traffic dense states are the basic building blocks of an aggregated hydrograph forecast. The invention makes use of the knowledge that each such traffic pattern at an effective bottleneck can be assigned a specific travel time increase in comparison with the state of free traffic. Therefore, track-related travel times, especially for passing effective bottlenecks, directly and automatically from the traffic patterns of the hydrograph database. In the case of a newly occurring, temporary effective bottleneck, such as a construction site, it can be assigned to a specific traffic pattern relatively quickly, if the traffic flows in the area of this temporary effective bottleneck can be determined.

In one embodiment of the invention according to claim 2, the influence of a traffic pattern, which is assigned to a newly occurred, temporary effective bottleneck, is predicted and taken into account for the other traffic patterns upstream and downstream of the new temporary effective bottleneck in the context of updating the hydrograph database. This further increases the accuracy of the traffic-based traffic forecast.

In a further embodiment of the invention according to claim 3 traffic pattern and travel time data are automatically determined in an update of the hydrograph database due to a newly occurred, temporary effective bottleneck, and the automatically generated in this way hydrographs are deleted automatically when canceling the temporary effective bottleneck. Thus, the influence of temporary effective bottlenecks on the traffic situation in the gangway-based traffic forecast can be considered relatively quickly and limited to the duration of existence of this bottleneck.

A development of the invention according to claim 5 provides advantageous possibilities of displaying the predicted traffic condition for the user, in particular a driver before.

Fig. 1

eine schematische Darstellung eines ersten, verfahrensgemäß zur gangliniengestützten Verkehrsprognose und insbesondere zur Reisezeitprognose im Bereich effektiver Engstellen benutzten aggregierten Verkehrsmusters aus synchronisiertem Verkehr und gestauchtem synchronisiertem Verkehr,

Fig. 2

eine Darstellung entsprechend Fig. 1 für ein zweites aggregiertes Verkehrsmuster aus synchronisiertem Verkehr, gestauchtem synchronisiertem Verkehr und sich bewegenden breiten Staus,

Fig. 3

eine Darstellung entsprechend Fig. 1, jedoch für ein weiteres aggregiertes Verkehrsmuster in Form eines übergreifenden Musters über mehrere Engstellen aus synchronisiertem Verkehr und gestauchtem synchronisiertem Verkehr und

Fig. 4

eine Darstellung entsprechend Fig. 3, jedoch für ein viertes aggregiertes Verkehrsmuster in Form eines übergreifenden Musters über mehrere Engstellen aus synchronisiertem Verkehr, gestauchtem synchronisiertem Verkehr und sich bewegenden breiten Staus.

Advantageous embodiments of the invention are illustrated in the drawings and will be described below. Hereby show:

Fig. 1

a schematic representation of a first, according to the method of traffic-based traffic forecast and in particular for travel time forecast in the area of effective bottlenecks used aggregated traffic pattern of synchronized traffic and compressed synchronized traffic,

Fig. 2

a representation accordingly Fig. 1 for a second aggregated traffic pattern of synchronized traffic, compressed synchronized traffic and moving congestion,

Fig. 3

a representation accordingly Fig. 1 but for another aggregated traffic pattern in the form of an overlapping pattern covering several bottlenecks of synchronized traffic and compressed synchronized traffic and

Fig. 4

a representation accordingly Fig. 3 but for a fourth aggregated traffic pattern in the form of an overlapping pattern across multiple bottlenecks of synchronized traffic, compressed synchronized traffic and moving congestion.

The present method makes use of the traffic-based traffic forecasting of aggregated traffic patterns at permanent and temporary effective bottlenecks of a considered traffic route network. Some essential traffic patterns will be discussed below with reference to the figures. For carrying out the method, any suitable means are used, as are customary to the person skilled in the art for this purpose and therefore do not require any further explanation. In particular, reference may also be made to the publications cited at the beginning and the literature cited therein.

According to the method, the required input data are initially obtained for the construction of a hydrograph database and corresponding traffic measurements are obtained, for example, about the amount of traffic, travel speeds and travel times for a considered traffic route network, for example a road network in a certain area. In the area of permanent effective bottlenecks of the traffic route network, traffic patterns consisting of the individualizable dense traffic states "synchronized traffic", "compressed synchronized traffic" and "moving congestion" are determined and used as basic components of an aggregated hydrograph forecast.

For this purpose, an aggregation of all traffic patterns at each effective bottleneck in the traffic network is carried out, and each aggregated traffic pattern is assigned a travel time or travel time increase in comparison with the travel time in the state of free traffic. In addition, upon detection of a newly occurring temporary effective bottleneck, for example, upon receipt of a message about a newly constructed construction site or a prolonged accident site, that temporary effective bottleneck is assigned a corresponding associated traffic pattern and the impact of that traffic pattern on the other traffic patterns upstream and downstream predicted by the new temporary effective bottleneck.

With this information, the hydrograph database is automatically updated, ie, hydrographs belonging to the new temporary effective bottleneck for the said traffic patterns and associated travel times are automatically generated and managed in the hydrograph database. This preferably also includes predicting the impact of the traffic pattern of the new temporary effective bottleneck on the other traffic patterns in the hydrograph database of permanent or current temporary effective bottlenecks upstream and downstream of the new bottleneck. Furthermore is preferably provided that automatically generated in the manner described manner hydrographs of the particular traffic pattern of a newly occurred, temporary effective bottleneck are automatically deleted in the hydrograph database once this bottleneck no longer exists, for example, when a new site was canceled.

As a further option, it is provided to process the aggregated traffic patterns for constructing the corresponding hydrograph database or the corresponding traffic-pattern-based hydrograph forecasts together with traffic reports automatically generated, for example, by a control center, in particular those based on RDS / TMC, and such an interaction of traffic reports such as RDS / TMC messages, and aggregated traffic patterns are also used to update the hydrograph database.

The hydrograph database created on the basis of traffic patterns at effective bottlenecks then forms part of a corresponding traffic-route-based traffic prediction system in which a best-fitting hydrograph is selected from a comparison of current traffic data with the stored hydrographs according to one of the algorithms familiar to the skilled person. Optionally, a database for the mentioned traffic reports, for example in RDS / TMC format, may additionally be provided. The representative patterns of dense traffic at effective bottlenecks can likewise be determined from suitable traffic measurement data according to one of the algorithms known for this purpose and predicted in their temporal / local course.

By means of a conventional visualization system, the user is presented with the results of the traffic-based traffic forecast in the desired manner, for example as a textual and / or graphical display of current and / or future traffic disturbances on a screen. In particular, this can be used as a graphical and / or textual display of traffic information on the temporal and local course of the aggregated traffic patterns of effective bottlenecks on a digital road map, for example in the form of current and future travel loss times, which are forecast due to the pattern of dense traffic at the respective effective bottleneck. Instead of or in addition to the lost time, the respective current or future travel time for relevant sections of the route can be displayed textually and / or graphically on the digital road network map. Another representation option is to reproduce the temporal and local course of the aggregated traffic patterns in the form of current and future aggregated symbols textually and / or graphically on a map for corresponding route sections. For example, the aggregated symbols may indicate congested traffic congestion situations by "congestion" and "slow traffic" or by "high congestion", "medium congestion" and "low congestion" or other such classifications.

In the FIGS. 1 to 4 For illustration, representative types of aggregated traffic patterns are shown, which are constructed from the individualizable dense traffic states "synchronized traffic", "compressed synchronized traffic" and "moving wide traffic jams" and used for the method according to the invention, in particular for rapid gangway-based prognostic determination of Travel times on sections with permanent and / or temporary effective bottlenecks.

Specially shows Fig. 1 a first possible aggregated traffic pattern composed of an area B _S synchronized traffic and an upstream adjacent area B _GS compressed synchronized traffic and may form upstream of an effective bottleneck at a route point x _{S, F.} This waypoint x _{S, F of} the effective bottleneck forms the boundary to a downstream subsequent area B _F free traffic. The transition point between synchronized traffic and compressed synchronized traffic is denoted by x _{GS, S.} Upstream of the upstream flank x _{F, GS of} the area B _{GS of} compressed synchronized traffic, there is also a free traffic area B _F. As mentioned above and explained in detail in the literature cited above, the individual areas of free traffic B _F , synchronized traffic B _S and compressed synchronized traffic B _GS customizable traffic conditions with predictable temporal local course dar. It is also known that traffic patterns from the individualized states form dense traffic upstream of a respective effective bottleneck in a manner dependent on the conditions of the actual bottleneck and the traffic volume.

Fig. 2 shows a case in which an aggregated traffic patterns upstream of the location x _{S, F} forms an effective bottleneck which consists of a region B _S of synchronized traffic, an upstream adjoining region B _GS compressed synchronized traffic and an upstream adjoining region B _St is moving wide jam is composed, while downstream of the effective bottleneck and upstream of the aggregated traffic pattern in turn there are areas B _F free traffic. The transition points between synchronized traffic B _S and compressed synchronous traffic B _GS and between the latter and the area B _St are denoted by x _{GS, S} and x _{St, GS} respectively, the upstream transit point to free traffic is denoted by x _{F, St.}

Fig. 3 shows as a further possible case of an aggregated traffic pattern used for the present method, and in particular the hydrograph database, a so-called overarching pattern, ie, extending upstream over several effective bottlenecks. Specifically, this extends over four effective bottlenecks in the direction of travel are spaced successively in the considered section and their locations with x _{S, F1} , x _{S, F2} , x _{S, F3} and x _{S, F4 are} designated. Upstream and downstream of the overarching, aggregated traffic pattern, in turn, there is in each case an area B _{F of} free traffic. The overarching pattern in the example shown consists of a downstream area of synchronized traffic and an upstream area of compressed synchronized traffic.

Fig. 4 As another possible example of an aggregated traffic pattern at an effective bottleneck, one such as that of FIG Fig. 3 extends across four effective bottlenecks x _{S, F1} , x _{S, F 2} , x _{S, F 3} , x _{S, F 4} between areas B _F free traffic and in this case from an upstream to the most downstream effective bottleneck x _{S, F4} subsequent area of synchronized traffic, an upstream adjoining area of compressed synchronized traffic, and a wide traffic jam which adjoins this upstream area.

It can be seen that the inventive aggregation of traffic information using aggregated patterns of dense traffic at effective permanent and temporary bottlenecks in the described manner achieves reliable forecasts of current and future traffic conditions with the aid of travel times determined travel times for the individual sections of a considered traffic route network when traffic conditions are significantly affected by dense traffic patterns at effective bottlenecks.

Claims (5)

1. Method for predicting traffic on the basis of historical characteristic curves, in which

   - a database of historical characteristic curves is set up taking into account patterns of dense traffic at effective bottlenecks of a traffic route network under consideration, and

   - traffic predictions are made by using current traffic measurements and comparing them with data from the database of historical characteristic curves,

   characterized in that

   - in order to set up the database of historical characteristic curves, all of the traffic patterns at permanent effective bottlenecks of the traffic route network under consideration are aggregated,

   - if a temporary effective bottleneck which has recently occurred is detected, a respective associated aggregated traffic pattern is determined and is used to update the database of historical characteristic curves,

   - each aggregated traffic pattern in the database of historical characteristic curves is assigned a particular journey time, and

   - in order to make the traffic predictions, journey times for all of the road sections, or a selectable part of the road sections, in the traffic route network under consideration are determined and displayed taking into account the journey times assigned to the aggregated traffic patterns in the respective updated database of historical characteristic curves.
2. Method according to Claim 1, also
   characterized in that
   in order to update the database of historical characteristic curves using an aggregated traffic pattern for a temporary effective bottleneck which has recently occurred, the influence of this aggregated traffic pattern on the other aggregated traffic patterns upstream and downstream of the temporary effective bottleneck which has recently occurred is predicted.
3. Method according to Claim 1 or 2, also
   characterized in that
   new associated historical characteristic curves for traffic patterns and journey times are automatically generated and stored for the aggregated traffic pattern for a temporary effective bottleneck which has recently occurred, in order to update the database of historical characteristic curves, and, once the temporary effective bottleneck which has recently occurred has been removed, the historical characteristic curves which have recently been generated by its occurrence are automatically deleted again in the database of historical characteristic curves.
4. Method according to one of Claims 1 to 3, also
   characterized in that
   for the database of historical characteristic curves, the aggregated traffic, patterns are processed together with automatically generated traffic messages.
5. Method according to one of Claims 1 to 4, also
   characterized in that
   the prediction data obtained from the traffic prediction on the basis of historical characteristic curves with the aid of the temporal and spatial profile of the traffic patterns are textually or graphically displayed on a digital route map in the form of current and future journey times for road sections and/or in the form of current and future journey time losses caused by a respective traffic pattern and/or in the form of symbols which classify the traffic state.

Images