EP3437958A1 - Estimation of expected journey times of a rail vehicle - Google Patents

Abstract

A method for estimating an estimated travel time (t F) of a rail vehicle (31a, 31b, 31c) will be described. In the method, current context information (KI) is determined which relate to current driving time-relevant influencing factors for a route (FS) to be covered by the rail vehicle (30). It will compare context information (KI) with historical reference context information (RKI) to the route (FS) or a route comparable to the route from a database. In this case, the reference context information (RKI) additionally includes historical travel time information for an assigned driving route. Finally, the estimated travel time (t F) for the route to be traveled (FS) based on the comparison of the context information (KI) with the historical reference context information (RKI) is determined, the determination of the estimated travel time (t F) based on a by the reference context information (RKI) trained estimation method stored in the database (DB) and the training of the estimation method comprise a machine learning method. A travel time estimator (20) is also described. Moreover, a traffic system (30) will be described.

Description

The invention relates to a method for estimating a probable travel time of a rail vehicle. Moreover, the invention relates to a driving time estimating device. Furthermore, the invention relates to a traffic system.

When organizing rail transport, it is particularly important to be able to plan what travel time a rail vehicle needs from one location to another. On the basis of this information, timetables, for example timetables, for public transport systems can be created so that passengers can estimate the time they need to reach a particular destination.

Usually, in navigation systems distances between individual locations and information about speed limits are used to determine a probable travel time. However, there are a variety of other factors that can affect the travel time of a rail vehicle.

In DE 10 2016 215 767 A1 A method for predicting an expected train run is described. Historical conditions and driving data from a database are compared with the current general conditions and driving data, and the expected train run is determined on the basis of the historical conditions and journey data.

In DE 2005 051 077 A1 For example, a method for detecting and accounting for crosswind loads in a traveling rail vehicle is described.

In DE 10 2009 023 704 A1 For example, a method of informing drivers in a rail vehicle via a presentation means is described.

In DE 196 33 525 an expected arrival time and / or departure time of a vehicle is determined on the basis of an expected driving speed and transmitted via a communication device to a user.

It is therefore the object to provide a method and a device with which a more accurate time planning for the journey of rail vehicles is possible.

This object is achieved by a method for estimating a probable travel time of a rail vehicle according to claim 1, a driving time estimating device according to claim 5 and a traffic system according to claim 6.

In the method according to the invention for estimating a probable travel time of a rail vehicle, current context information relating to current driving time-relevant influencing factors for a travel distance to be covered by the rail vehicle is determined.

Furthermore, context information is compared with historical reference context information about the route or a route comparable to the route from a database, the reference context information additionally having historical travel time information about the assigned route. A route comparable to the current route is a route with comparable route parameters. For example, the length and the topography of the route are comparable, there are similar traffic densities, etc. It may be assumed in such a case that with a similar context, for example with a similarly fast vehicle under similar traffic conditions, a similar travel time is achieved. Finally, an estimated travel time for the route to be traveled is determined based on the comparison of the context information with the historical reference context information. For example, the estimated travel time may be determined on the basis of reference data records selected using the comparison and the current context information. In this case, an estimation method trained by the reference context information stored in the database is used, wherein the training of the estimation method comprises a machine learning method.

Due to the more precise travel time determination, railway operators can advantageously create more exact timetables. The method may also be advantageously used to update timetables in response to a changing context and thus to provide passengers with more reliable and accurate information and to fine-tune the driving experience.

The use of a machine learning method is particularly advantageous if a route section is passed through for the first time and no historical data from this route section is available yet. In this case, the travel time can be based on reference data from other routes having similar characteristics as the current route.

In addition, context data or reference data can be obtained and evaluated in real time from the currently already traveled part of the traveled section of the route. The two approaches can also be combined to provide a more accurate estimate of travel time.

Advantageously, with the aid of such an estimated estimation method, the estimated travel time can be estimated on a broad database of reference data records. so that due to statistical effects, the reliability and accuracy of the estimation result is increased.

In one embodiment of the method according to the invention for estimating a probable travel time of a rail vehicle, the training of the estimation method comprises a machine learning method.

With a machine learning process, examples can be learned and these are generalized after completion of the learning phase. That is, the examples are not simply learned by heart, but patterns and laws are identified in the learning data. This allows unknown data to be assessed afterwards.

The travel time estimation device according to the invention for a rail vehicle comprises a context determination unit for determining current context information, which relates to current driving time-relevant influencing factors for a travel distance to be traveled by the rail vehicle. Part of the driving time estimation device according to the invention is also a comparison unit, which is set up to compare the context information with historical reference context information about the route or a route comparable to the route from a database. The reference context information additionally has historical travel time information about the route to be traveled.

Finally, the driving time estimating device according to the invention also includes an estimation unit for determining the estimated travel time for the route to be traveled on the basis of the comparison of the context information with the historical reference context information using an estimation method trained by the reference context information stored in the database by a machine learning method. The driving time estimating device according to the invention shares the advantages of the method according to the invention for estimating an expected journey time of a rail vehicle.

The traffic system according to the invention comprises the time estimator according to the invention. The time estimating device according to the invention shares the advantages of the time estimating device according to the invention.

Some components of the time estimating device according to the invention can for the most part be designed in the form of software components. This concerns in particular parts of the context determination unit, the comparison unit and the estimation unit. In principle, however, these components can also be partly realized, in particular in the case of particularly fast calculations, in the form of software-supported hardware, for example FPGAs or the like. Likewise, the required interfaces, for example, if it is only about a transfer of data from other software components, be designed as software interfaces. However, they can also be configured as hardware-based interfaces, which are controlled by suitable software. The calculations can also take place outside the rail vehicle in an external central computer unit, for example as the central computer of a data network. Current context information collected by vehicle sensors may optionally be transmitted by the rail vehicle to this external central computer unit. Conversely, travel time information determined by the central computer unit can also be transmitted to individual rail vehicles.

A largely software-based implementation has the advantage that already existing in a traffic control system existing computer systems can be easily retrofitted by a software update to work in the inventive way. In this respect, the problem is also solved by a corresponding computer program product with a computer program, which directly into a memory device of such computer system is loadable, with program sections to perform all the steps of the inventive method when the computer program is executed in the computer system.

Such a computer program product may contain, in addition to the computer program, additional components, e.g. a documentation and / or additional components, also hardware components, such as e.g. Hardware keys (dongles, etc.) for using the software include

For transport to the storage device of the computer system and / or for storage on the computer system, a computer-readable medium, for example a memory stick, a hard disk or another portable or permanently installed data carrier can be used, on which the computer program readable and executable by a computer unit are stored. The computer unit may e.g. for this purpose have one or more cooperating microprocessors or the like.

The dependent claims and the following description each contain particularly advantageous embodiments and further developments of the invention. In this case, in particular the claims of a claim category can also be developed analogously to the dependent claims of another claim category and their description parts. In addition, in the context of the invention, the various features of different embodiments and claims can also be combined to form new embodiments.

• Aktualisieren der voraussichtlichen Fahrzeit auf den Bildanzeigen des Schienenfahrzeugs,
• Informieren von Passagieren und/oder anderen Zügen über die voraussichtliche Fahrzeit,
• Optimieren des Geschwindigkeitsprofils und/oder des Energieverbrauchs des Schienenfahrzeugs,
• Vorbereiten der Infrastruktur, wie z.B. Weichen, Signale, Haltepunkte, auf die Ankunft des Schienenfahrzeugs.

Advantageously, the estimated travel time determined by the method according to the invention can be used for at least one of the following measures:

• Updating the estimated travel time on the image displays of the rail vehicle,
• Informing passengers and / or other trains about the expected journey time,
• Optimizing the speed profile and / or energy consumption of the rail vehicle,
• Prepare the infrastructure, such as turnouts, signals, breakpoints, on the arrival of the rail vehicle.

The measures described improve the effectiveness of rail transport and increase passenger comfort.

In one refinement of the method according to the invention for estimating a probable travel time of a rail vehicle, a reference data record with reference context information is identified on the basis of the comparison which comes closest to the determined context information and the estimated travel time is determined on the basis of the historical travel time information assigned to the selected reference data record. Advantageously, in this variant, a historical reference data set is selected, which most closely corresponds to the current scenario. The similarity of the historical reference data record can be determined by comparing the individual historical context information associated with it with the corresponding context information of the current traffic situation. The better the historical reference data set corresponds to the current scenario, the easier it is to transfer the travel time information of the historical reference data record to the current situation.

• Wetterdaten,
• Jahreszeitdaten,
• das Datum und die Uhrzeit,
• den Fahrzeugtyp,
• den Streckentyp,
• den Streckenzustand,
• den Namen und/oder eine Identitätsnummer und/oder ein anderes eindeutiges Kennzeichen des Fahrers, welches zu dessen eindeutiger Identifizierung genutzt werden kann,
• die letzte Fahrzeit für die Fahrstrecke,
• Zeitpunkte und Orte öffentlicher Ereignisse und Veranstaltungen, wie zum Beispiel Ferienzeiten, Konzerte usw.,
• die Anzahl von Buchungen und/oder Reservierungen,
• die Anzahl von Zwischenstopps und Verbindungen mit anderen Zügen, Verspätungszeiten der einzelnen Züge und ihre Ankunftszeiten.

In a preferred embodiment of the method according to the invention for estimating a probable travel time of a rail vehicle, the context information comprises information which can improve the estimation of the travel time, preferably at least part of the following types of information:

• Weather data
• Season data
• the date and time,
• the vehicle type,
• the route type,
• the route condition,
• the name and / or identity number and / or other unique identifier of the driver which can be used to identify him or her clearly,
• the last travel time for the route,
• Times and places of public events and events, such as holidays, concerts, etc.,
• the number of bookings and / or reservations,
• the number of stops and connections with other trains, delays of each train and their arrival times.

Since the information mentioned can have an influence on the speed of a rail vehicle, it is advantageous to take into account at least a part of it in the travel time determination. If reference data sets for the same route or a route with comparable route parameters whose context information corresponds to the current context situation of the rail vehicle are found, it is possible to directly estimate a travel time for a route currently being traveled on the basis of travel times of the reference data records.

In a variant of the method according to the invention for estimating a probable travel time of a rail vehicle, in the event that no reference context information has been found which corresponds to the current situation, the determination of the probable travel time is based on a reference data set stored in the database estimation method.

FIG 1

ein Flussdiagramm, welches ein Verfahren zum Abschätzen einer voraussichtlichen Fahrzeit eines Schienenfahr-zeugs gemäß einem ersten Ausführungsbeispiel der Erfin-dung veranschaulicht,

FIG 2

ein Blockdiagramm, welches eine Zeitschätzeinrichtung gemäß einem Ausführungsbeispiel der Erfindung veran-schaulicht,

FIG 3

ein Blockdiagramm, welches ein Verkehrssystem mit einer Zeitschätzeinrichtung gemäß einem Ausführungsbeispiel der Erfindung veranschaulicht.

The invention will be explained in more detail below with reference to the accompanying figures with reference to embodiments. Show it:

FIG. 1

a flowchart, which is a method for estimating a probable travel time of a rail vehicle according to a first embodiment of the invention,

FIG. 2

3 is a block diagram illustrating a time estimator according to an embodiment of the invention;

FIG. 3

a block diagram illustrating a traffic system with a time estimator according to an embodiment of the invention.

In FIG. 1 a flow chart 100 is shown which illustrates a method for estimating a probable travel time t _{F of} a rail vehicle according to an embodiment of the invention. In step 1.I, a route FS with start and destination is first determined by a user. Alternatively, parameter data of the route can also be obtained from a database or from an on-board computer unit. Subsequently, in step 1.II, a route length SL of the travel distance FS to be traveled by the rail vehicle is determined. Furthermore, in step 1.III current context information KI, which relate to current driving time-relevant influencing factors for a route to be covered by the rail vehicle, are determined. For this purpose, current weather information, information about the current traffic situation, information about daytime, regularly occurring restrictions and information about the driving profile of the driver are determined, for example via the Internet. In step 1.IV, the context information KI and the determined route information FS, LS and additionally historical reference context information RKI and route information RFS, RLS from a database are then used to estimate an anticipated travel time t _F. There are a variety of methods available for estimating travel time. For example, monitored or unmonitored machine learning techniques, nearest neighbor comparisons, or simulation based approaches may be used.

In FIG. 2 a block diagram is shown which illustrates a time estimator 20 according to an embodiment of the invention. The time estimator 20 may, for example, be centrally located in a traffic control center. The time estimator 20 includes a communication interface 21 for communication with external units and users. Via the communication interface 21, the position determining device 20 receives, for example, information FS about a current route, in particular start and destination, and possibly additional information, such as the route length SL, from a user. The input data FS, SL are forwarded to a context determination unit 22, which searches on the basis of these data SL, FS suitable context information KI to the current route. For this purpose, corresponding context data KI, for example weather data, are searched for and acquired on the Internet via the communication interface 21. The determined context information KI and further data FS, SL characterizing the relevant route FS are transmitted to a comparison unit 23. The comparison unit 23 compares the determined context information KI with historical reference context information RKI.

The historical reference context information RKI is retrieved from an external database DB, and a reference data record NRD with reference context information RKI which is most similar to the context information KI of the current route is finally identified as a suitable reference data record NRD and transmitted to an estimation unit 24. The estimation unit 24 then determines, based on the selected reference data set NRD, an anticipated travel time t _F for the selected travel route FS. The determined travel time t _F can finally be output via the communication interface 21 to a user or a traffic control device. As already mentioned, the estimation process may alternatively be based on a plurality of reference data sets NRD, these being for training the estimation unit 24, which then estimates a probable travel time t _F on the basis of a trained model and the detected current context information KI of the current route FS.

In FIG. 3 a traffic system 30 is illustrated. The traffic system 30 comprises a plurality of rail vehicles 31a, 31b, 31c and a travel time estimation device 20 of the rail vehicles 31a, 31b, 31c data, in particular current context data KI, but also route parameters FS, LS over different routes over a communication interface 21 to the Driving time estimation device 20 transmitted. On the basis of these data KI, FS, LS and additional reference data, in particular reference context information RKI, which are obtained for the different routes FS from a database DB, the driving time estimating device 20 determines estimated travel times t _F for the individual vehicles 31a, 31b, 31c and transmits these data t _F to these vehicles 31 a, 31 b, 31 c as well as a traffic control device 32.

It is finally pointed out again that the above-described methods and devices are merely preferred embodiments of the invention and that the invention can be varied by a person skilled in the art without departing from the scope of the invention, as far as it is specified by the claims. For the sake of completeness, it is also pointed out that the use of indefinite articles does not exclude "a" or "one", that the characteristics in question can also be present multiple times. Similarly, the term "unit" does not exclude that it consists of several components, which may also be spatially distributed.

Claims (8)

Method for estimating a probable travel time (t _F ) of a rail vehicle (31a, 31b, 31c): Determining current context information (KI), which relate to current driving time-relevant influencing factors for a route (FS) to be covered by the rail vehicle (31a, 31b, 31c), Comparing the context information (KI) with historical reference context information (RKI) to the route (FS) or a route comparable to the route (FS) from a database (DB), the reference context information (RKI) additionally including historical travel time information have the assigned route, Determining the estimated travel time (t _F ) for the route to be covered (FS) on the basis of the comparison of the context information (KI) with the historical reference context information (RKI), wherein the determination of the probable travel time (t _F ) based on a by the in the database (DB) stored reference context information (RKI) trained estimation method and the training of the estimation method comprises a machine learning method. Method according to claim 1, wherein the determined estimated travel time (t _F ) is used for at least one of the following measures: Updating the estimated travel time on image displays of the rail vehicle (31a, 31b, 31c), Informing passengers and / or other trains about the expected journey time (t _F ), Optimizing the speed profile and / or the energy consumption of the rail vehicle (31a, 31b, 31c), Prepare the infrastructure for the arrival of the rail vehicle (31a, 31b, 31c). The method of claim 1 or 2, wherein based on the comparison, a reference data set (NRD) with reference context information (RKI) is identified, which comes closest to the determined context information (KI) and the determination of the probable travel time (t _F ) based on the historical reference time information associated with the selected reference data set (NRD). Method according to one of the preceding claims, wherein the context information (KI) comprises at least part of the following types of information: - weather data, - season data, - the date and time, - the vehicle type, - the route type, - the route condition, the name and / or an identity number and / or other unique identifier of the driver which can be used for his unique identification, - the last driving time for the current route (FS), - Times and places of public events and events - the number of bookings and / or reservations - the number of stops and connections with other trains - the delays of each train and their arrival times. Travel time estimating device (20) for a rail vehicle (31a, 31b, 31c), comprising: a context determination unit (21) for determining current context information (KI), which relate to current driving time-relevant influencing factors for a route (FS) to be covered by the rail vehicle (31a, 31b, 31c), - A comparison unit (22) for comparing the context information (KI) with historical reference context information (RKI) to the route (FS) or one with the route (FS) comparable route from a database (DB), the reference context information (RKI) additionally having historical travel time information about the assigned travel route, - An estimation unit (24) for determining the estimated travel time (t _F ) for the route to be covered (FS) based on the comparison of the context information (KI) with the historical reference context information (RKI) using a in the database (DB ) stored reference context information (RKI) using a machine learning method trained estimation method. A traffic system (30) comprising a travel time estimator (20) according to claim 5. A computer program product comprising a computer program directly loadable into a memory unit of a traffic system (30), having program sections for performing all the steps of a method according to any one of claims 1 to 4 when the computer program is executed in the rail vehicle. A computer-readable medium on which program sections executable by a computer unit are stored in order to execute all the steps of the method according to one of claims 1 to 4, when the program sections are executed by the computer unit.

Images