EP4155162A1 - Method and device with axle counter for operating a railway crossing - Google Patents

Abstract

The invention relates to a method for operating a level crossing (BU), in which an axle counter (AZ1 . . . AZ2) records measurement data for rail vehicles (FZ) approaching a level crossing (BU) while the rail vehicle (FZ) is crossing. The speed of the rail vehicle (FZ) and the acceleration of the rail vehicle (FZ) are determined with the aid of a computer from the measurement data as properties of the rail vehicle (FZ). A closing time is determined for triggering a closing of the level crossing (BU) as a function of the determined properties of the rail vehicle. The speed and the acceleration are determined from the measurement data of the axle counter (AZ1 ... AZ2). The closing time of the level crossing is determined taking into account the speed determined when the rail vehicle (FZ) exits from the axle counter (AZ1 ... AZ2) and the acceleration determined when the rail vehicle (FZ) exits from the axle counter (AZ1 ... AZ2). The invention also includes a device for operating a level crossing (BU), a computer program product and a provision device for the latter.

Description

• ein Achszähler während der Überfahrt des Schienenfahrzeugs Messdaten erfasst,
• rechnergestützt als Eigenschaften des Schienenfahrzeugs (FZ) die Geschwindigkeit des Schienenfahrzeugs und die Beschleunigung des Schienenfahrzeugs aus den Messdaten ermittelt werden,
• eine Schließzeit zum Auslösen eines Schließens des Bahnübergangs in Abhängigkeit der ermittelten Eigenschaften des Schienenfahrzeugs bestimmt wird.

The invention relates to a method for operating a level crossing, in which rail vehicles approaching a level crossing

• an axle counter records measurement data during the crossing of the rail vehicle,
• the speed of the rail vehicle and the acceleration of the rail vehicle are determined computer-aided as properties of the rail vehicle (FZ) from the measurement data,
• a closing time for triggering a closing of the level crossing is determined as a function of the determined properties of the rail vehicle.

• mindestens einen Achszähler zum Erfassen von Messdaten bei Überfahrt der Schienenfahrzeuge,
• einen Computer (CP), der eingerichtet ist, rechnergestützt als Eigenschaften des Schienenfahrzeugs (FZ), die Geschwindigkeit des Schienenfahrzeugs und die Beschleunigung des Schienenfahrzeugs aus den Messdaten zu ermitteln,

und den Bahnübergang für sich einem Bahnübergang nähernde Schienenfahrzeuge zu schließen.The invention also relates to a device for operating a level crossing, comprising

• at least one axle counter for recording measurement data when the rail vehicles pass,
• a computer (CP) which is set up to determine the speed of the rail vehicle and the acceleration of the rail vehicle from the measurement data in a computer-aided manner as properties of the rail vehicle (FZ),

and to close the level crossing for rail vehicles approaching a level crossing.

Finally, the invention relates to a computer program product and a provision device for this computer program product, the computer program product being equipped with program instructions for carrying out this method.

document EP 2 718 168 B1 relates to a method for operating a railway safety system with at least one route facility, taking into account when the rail vehicle enters the switch-on section

Railway safety system detected speed measurement variable. When the rail vehicle enters the switch-on section, the measured speed variable is used to check whether a correction time for the forwarding of a message from the one route facility to an associated railway safety system must be set according to the measured speed variable. A set correction time is then checked to determine whether it should remain effective as a function of at least one other influencing variable of the rail vehicle that determines the travel time.

In the post-release European application with the file number 20202457.6 it is described that axle counters can be used to recognize patterns of rail vehicles, to the effect that train types can be recognized on the basis of determined wheelbases. Among other things, a level crossing can also be operated with an optimized closing time if the train types are known, whereby the closing times for freight trains, for example, can be shifted to a later point in time.

The object of the invention is to detect characteristics of the train with sufficient certainty, if possible without additional sensors to be installed, so that a level crossing can be operated with an optimized closing time. For this purpose, a method and a device suitable for using the method are to be specified. In addition, the object of the invention consists in specifying a computer program product and a provision device for this computer program product, with which the aforementioned method can be carried out.

• der ermittelten Geschwindigkeit bei Ausfahrt des Schienenfahrzeugs aus dem Achszähler und
• der ermittelten Beschleunigung bei Ausfahrt des Schienenfahrzeugs aus dem Achszähler

bestimmt wird.This object is achieved according to the invention with the subject matter of the claim (method) specified at the outset in that the speed and the acceleration are determined from the measurement data of the axle counter and the closing time of the level crossing is taken into account

• the speed determined when the rail vehicle exits the axle counter and
• the acceleration determined when the rail vehicle exits the axle counter

is determined.

Because the axle counter is used to determine the speed and acceleration of the rail vehicle, it is advantageously possible for the axle counter, which can also be designed as a double axle counter, to record several types of properties of the rail vehicle for the closing time of the level crossing are relevant can be used. The axle counter alone is preferably used to determine measured values for all properties that are to be used to determine the closing time of the level crossing.

As a result, the outlay on hardware that is required for detecting the properties of the rail vehicle is advantageously reduced. This also means that the reliability when carrying out the method can be increased. Axle counters are highly reliable components, avoiding the use of other sensors that would reduce the overall reliability of the method.

Another advantage is that the exact location of the axle counter is known and the measured values can therefore be assigned directly to this location. The use of locating methods, from which parameters such as speed and acceleration can also be calculated, involves additional measurement uncertainties that can be avoided when carrying out the method according to the invention.

The rail vehicle can consist of individual vehicles or of a train, the latter occurring significantly more frequently in the approach area of level crossings.

Closing times within the meaning of the invention are understood to be the time at which the level crossing is closed. The Closing times could thus also be referred to as the closing time. On the other hand, there is talk of a closing period when it comes to the period in which the level crossing remains closed, starting with the closing time. At the end of the closing period is the opening time, when the level crossing is opened again.

The great advantage of using the method for level crossings is that the closing times of the level crossing can advantageously be set individually depending on the properties of the approaching train. At least if the characteristics of the train can be reliably identified, in many cases the closing time can be postponed to a later point in time without lowering the safety standards in the operation of the level crossing. The crossing traffic, which in many cases does not have to wait as long at the level crossing, benefits.

• Fahrtrichtung des Zuges: durch Vergleich der Beeinflussung bei Doppelsensoren (z.B. durch Auswertung des Zeitversatzes bei der Signalgenerierung)
• Geschwindigkeit des Zuges bei Überfahrt einer Achse: aus dem Abstand der Doppelsensoren z. B. durch Auswertung des Zeitversatzes bei der Signalgenerierung oder dem zeitlichen Abstand des Passierens der geschätzten Radmittelpunkte bei bekannten Achsabständen.
• durchschnittliche Geschwindigkeit bei der Überfahrt und/oder die Beschleunigung bei der Überfahrt: aus der Mittelung über verschiedene Räder bzw. numerische Ableitung der Geschwindigkeit.

The following measuring principles of axle counters can be used as an example to determine the properties.

• Direction of travel of the train: by comparing the influence of double sensors (e.g. by evaluating the time delay when generating the signal)
• Speed of the train when crossing an axle: from the distance between the double sensors, e.g. B. by evaluating the time offset in the signal generation or the time interval between passing the estimated wheel centers with known wheelbases.
• Average speed when crossing and/or acceleration when crossing: from the averaging over different wheels or numerical derivation of the speed.

In the context of the invention, “computer-aided” or “computer-implemented” can be understood to mean an implementation of the method in which at least one computer or processor executes at least one method step of the method.

The term "calculator" or "computer" covers any electronic device with data processing properties. Computers can be, for example, personal computers, servers, handheld computers, mobile phones and other communication devices that process computer-aided data, processors and other electronic devices for data processing, which can preferably also be combined to form a network.

In connection with the invention, a “processor” can be understood to mean, for example, a converter, a sensor for generating measurement signals, or an electronic circuit. A processor can in particular be a main processor (Central Processing Unit, CPU), a microprocessor, a microcontroller, or a digital signal processor, possibly in combination with a memory unit for storing program instructions, etc. A processor can also be understood to mean a virtualized processor or a soft CPU.

In the context of the invention, a “memory unit” can be understood to mean, for example, a computer-readable memory in the form of a random-access memory (RAM) or data memory (hard disk or data carrier).

The "interfaces" can be realized in terms of hardware, for example wired or as a radio connection, and/or software, for example as an interaction between individual program modules or program parts of one or more computer programs.

"Cloud" is to be understood as an environment for "cloud computing" (German computer cloud or data cloud). What is meant is an IT infrastructure that is made available via interfaces of a network such as the Internet. It usually includes storage space, computing power or software as a service, without these having to be installed on the local computer using the cloud. The services offered as part of cloud computing cover the entire spectrum of information technology and include, among other things, infrastructure, platforms and software.

"Program modules" should be understood to mean individual functional units that enable the program flow according to the invention. These functional units can be implemented in a single computer program or in several computer programs that communicate with one another. The interfaces implemented here can be implemented in terms of software within a single processor or in terms of hardware if multiple processors are used.

Unless otherwise indicated in the following description, the terms "create", "determine", "compute", "generate", "configure", "modify" and the like preferably refer to processes that create and/or data change and/or convert the data into other data. The data are available in particular as physical quantities, for example as electrical impulses or also as measured values. The necessary instructions program commands are combined in a computer program as software. Furthermore, the terms "send", "receive", "read in", "read out", "transmit" and the like relate to the interaction of individual hardware components and/or software components via interfaces.

According to one embodiment of the invention, it is provided that when the closing time is determined, a standard closing time is shifted at least once, in particular to a later closing time.

The advantage of using the standard closing time is that, on the one hand, safe operation of the level crossing can be ensured without exception and, on the other hand, the closing times can be flexibly adjusted if the characteristics of the approaching train can be determined with sufficient reliability.

According to one embodiment of the invention, it is provided that the determined property of the speed is taken into account in this way that, taking into account the distance between the axle counter and the level crossing, it is determined whether the rail vehicle has reached the level crossing.

If the distance between the axle counter and the level crossing is known, the arrival time of the rail vehicle at the level crossing can be calculated directly. Depending on how long the level crossing should be closed before the rail vehicle arrives, the closing time can be calculated directly. If a specific distance between the rail vehicle and the level crossing is required as a trigger criterion, the closing time can also be calculated from this. However, this calculation does not take into account any acceleration values determined by the axle counter (more on this below).

According to one embodiment of the invention, the determined property of the acceleration is taken into account in such a way that, taking into account the distance of the axle counter from the level crossing and the property of the acceleration, it is determined whether the rail vehicle has reached the level crossing.

The acceleration provides information on how the probable point in time when the rail vehicle passes the level crossing is influenced by the acceleration state of the train. A closing time that results from the sole consideration of the speed of the rail vehicle must therefore be corrected accordingly.

Advantageously, it can be assumed that the most unfavorable case is that the rail vehicle is (further) accelerated until it reaches the level crossing. If the vehicle is accelerated negatively when leaving the axle counter (i.e. braked), this can also be taken into account, since this shifts the time of reaching the level crossing (compared to a constant speed) to a later point in time. Here, too, the most unfavorable case can be taken into account, that the operating state can change from braking to accelerating after exiting the axle counter.

• rechnergestützt anhand der ermittelten Achsabstände eine erste weitere Eigenschaft ermittelt wird, nämlich, ob es sich bei dem Schienenfahrzeug um einen Personenzug oder einen Güterzug handelt,
• bei der Ermittlung der ersten weiteren Eigenschaft in einem ersten Prüfungsschritt überprüft wird, ob zumindest in einem überwiegenden Teil der Folge der Achsabstände identische oder ähnliche Muster festgestellt werden können, und
• wenn kein Muster festgestellt werden konnte, dem Schienenfahrzeug als erste weitere Eigenschaft die Eigenschaft eines Güterzuges zugeordnet wird, oder
• wenn ein Muster festgestellt wurde, dem Schienenfahrzeug als erste weitere Eigenschaft die Eigenschaft eines Personenzuges zugeordnet und/oder ein weiterer Prüfungsschritt durchgeführt wird.

According to one embodiment of the invention, it is provided that

• a first further property is determined with the aid of a computer on the basis of the wheelbases determined, namely whether the rail vehicle is a passenger train or a freight train,
• When determining the first further property, a check is carried out in a first test step to determine whether identical or similar patterns can be determined at least in a predominant part of the sequence of wheelbases, and
• if no model could be determined, the rail vehicle is assigned the property of a freight train as the first additional property, or
• if a sample has been identified, the property of a passenger train is assigned to the rail vehicle as the first additional property and/or a further test step is carried out.

In other words, it is advantageously provided to estimate the distances between the axles of the train set from the raw data of the axle counter during the crossing. In particular, passenger trains such as ICE or regional trains consist of fixed units that usually stay together, from which no cars are decoupled. Therefore, there are patterns that can be measured several times in a row by coupling these units and that are similar to each other. Passenger trains have a fixed "fingerprint", so to speak, which is only changed by measurement errors, etc.

On the other hand, there are other, usually variable and therefore not similar or identical patterns for freight trains, depending on which and how many units are coupled together. This allows the data to be represented as multidimensional vectors whose components estimate the distances between the axes, i.e. H. Axle 1 to axle 2 up to axle n-1 to axle n (if the train has n axles, in reality up to 250) .

The terms identical (identical) and similar (similar) should be understood in their meaning in terms of pattern recognition. This means that a comparison of patterns can result in them being identified as identical or similar (or not as identical and not similar, i.e. not related). This assessment is preferably computer-aided.

Patterns are considered to be identical if all test criteria in the pattern comparison lead to the result that the test criteria match. Since the test criteria are based on measured values, a tolerance interval can be defined for the measurement, within which the test criterion can lie in order to be understood as identical.

Patterns are understood to be similar if an evaluation of the test criteria shows that they at least largely correspond to each other. It should be noted here that similarity also exists if the samples are identical.

When the question can be answered in the affirmative, that the criteria at least largely correspond to each other, must be determined for the implementation of the pattern recognition. In general, the following relationship applies to the recognition of the first further property mentioned (freight train or passenger train). The stricter the criteria for the recognition of similarity, the greater the probability that the recognized similar patterns are actually always used to recognize people. However, the probability that passenger trains will not be recognized also increases. The less stringent the criteria for detecting similarity, the higher the probability that all passenger trains will be recognized. However, the likelihood of freight trains being mistakenly recognized as passenger trains also increases.

Regardless of the strictness of the criteria, the method according to the invention works in the technical sense. In operation, however, it must be determined where there is an optimum in terms of the strictness of the criteria in relation to safe operation. A predominant part of the wheelbase sequence is present when the repeating pattern can be identified for more than 50% of the wheelbases. It can also preferably be defined that the limit at which a predominant part is assumed is to be more than 60%, particularly preferably more than 70%, 80% or 90%.

The advantage of using pattern recognition in train operation is that parameters of train operation, such as the closing time of the level crossing, can be flexibly adapted to the properties of the vehicles. In this way, for example, a greater route utilization can be achieved.

In particular, the maximum acceleration capability of the detected train type can be taken into account (see above for the consideration of speed and acceleration measurements). This means, for example, that a passenger train can accelerate faster than a freight train. If a freight train is detected, its lower acceleration capacity can be taken into account so that the level crossing can be closed later. This saves the crossing traffic unnecessary waiting times that would arise if closing times were calculated for the freight train in question, which take into account the high acceleration capacity of passenger trains.

According to one embodiment of the invention, it is provided that in the second checking step in the sequence of the center distances, a number of center distances at the beginning of the series and/or a number of center distances at the end of the series remain unconsidered.

By ignoring a number of wheelbases at the beginning or at the end of the sequence of wheelbases, it can be advantageously achieved that locomotives or power cars, which form the beginning or end of the train, for example in passenger trains, are not checked for the presence of patterns in the sequence of wheelbases become. Both the locomotives and often also the power cars have different wheelbases (which are therefore also characterized by different patterns) than the vehicles in the middle of the train, which are usually identical in a passenger train and therefore form similar or identical patterns. The first test step can therefore advantageously be carried out faster and more reliably if it is limited in particular to the central part of the train.

How many axles should not be taken into account depends on the train operation to be monitored. If the traction engines to be used are known, the sequence of wheelbases that are not to be taken into account corresponds to those of the locomotives or power cars that are to be used. However, a general value can also be assumed for unknown locomotives and power cars. This can be four, six or eight axes, for example.

According to one embodiment of the invention, it is provided that in the second or a further test step the amount of the wheelbase is determined, with the rail vehicle being assigned the property of a passenger train as the first additional property, as long as the amount of the largest wheelbase occurring in the sample exceeds a specified limit value .

The limit value that allows reliable conclusions to be drawn about passenger trains also depends not least on the specifics of the respective train operation that is to be monitored. This limit value can thus be determined as a function of the route if it is known which passenger trains are running on the relevant route. It is important that the longest occurring wheelbase of the relevant towing vehicle is taken into account. However, in the event that train cars with different longest wheelbases are used, the shortest of the longest distances between the different train cars running on the route must be taken into account as a limit value.

According to a particularly advantageous alternative of the invention, this limit value can also differ from typical center distances of freight cars, so that the center distance can be used as a particularly reliable criterion for distinguishing between freight cars. However, since the center distance is used as an additional criterion for the patterns to be recognized (i.e. in addition), compliance with this difference is not obligatory.

According to one embodiment of the invention, it is provided that in the second or a further test step, the determined patterns of the wheelbases are compared with reference patterns of wheelbases and, in the event that the pattern matches a reference pattern, the rail vehicle is given a type of train linked to the reference pattern as a second additional one property is assigned.

The reference patterns can be stored in a memory device, for example. A server can make the reference patterns available so that a comparison with the determined patterns is made possible. Another possibility is that the reference patterns are stored in a storage device that forms part of the axle counter. This creates the possibility of technically modifying the axle counters with a certain intelligence, in other words as autonomous or partially autonomous units.

The advantage of storing reference patterns in a storage device is that they are available at all times and can be called up without a time delay if required. The memory devices can also store the various reference patterns in a route-specific manner, so that only specific reference patterns are made available to specific axle counters on specific route sections.

A parameter set suitable for the application is advantageously selected from a parameter set determined in this way. For example, one can omit the wheel diameter if it is more or less the same for all trains on the route. Site-specific, representative data are now being collected or measured and classified for the relevant parameters, e.g. B. Passenger train, freight train. It is a finite number of integer or real values, e.g. For example, this could be the speed and the number of axles, to give a clear two-dimensional example here. i.e. in principle one obtains a classification task, as in the following figure 5 described. Overall, the collection of further parameters in addition to the patterns to be compared makes the recognition of vehicle properties more robust against errors. Advantageously, a higher degree of reliability can be achieved in the train detection, so that the train traffic can be controlled more effectively. Which parameters are to be taken into account for a given control task for train traffic then depends on the circumstances of the individual case. They are to be selected in a suitable manner when designing the control process.

According to one embodiment of the invention, it is provided that the criteria for the second test step and/or the further test steps are evaluated using a machine learning method.

Machine learning advantageously enables the ongoing processes to be optimized, i. H. the reliable detection of the properties mentioned, in particular train types, during operation. This allows the system to automatically adapt to changing operating conditions. For example, additional patterns can be created when a new type of passenger train is deployed on a particular route segment. For this purpose, e.g. B. neural networks or other devices with artificial intelligence.

In the context of this invention, artificial intelligence (also abbreviated to KI below) is to be understood in the narrower sense as computer-assisted machine learning (machine learning, also abbreviated to ML below). It is about the statistical learning of the parameterization of algorithms, preferably for complex applications. Using ML, the system recognizes and learns patterns and regularities in the recorded process data using previously entered learning data. With the help of suitable algorithms, ML can independently find solutions to emerging problems. ML is broken down into three fields - supervised learning, unsupervised learning and reinforcement learning, with more specific applications such as regression and classification Structure recognition and prediction, data generation (sampling) or autonomous action.

In supervised learning, the system is trained through the relationship between the input and the associated output of known data, and in this way it learns approximately functional relationships. It depends on the availability of suitable and sufficient data, because if the system is trained with unsuitable (e.g. non-representative) data, it learns faulty functional relationships. With unsupervised learning, the system is also trained with sample data, but only with input data and without any connection to a known output. It learns how data groups are to be formed and expanded, what is typical for the use case and where deviations or anomalies occur. This allows use cases to be described and error states to be discovered. With reinforcement learning, the system learns through trial and error by proposing solutions to given problems and receiving a positive or negative evaluation of this suggestion via a feedback function. Depending on the reward mechanism, the AI system learns to perform corresponding functions.

• Entwicklung neuer Verbindungen
• Löschen existierender Verbindungen
• Ändern der Gewichtung (der Gewichte von Neuron j zu Neuron i)
• Anpassen der Schwellenwerte der Neuronen, sofern diese Schwellwerte besitzen
• Hinzufügen oder Löschen von Neuronen
• Modifikation von Aktivierungs-, Propagierungs- oder Ausgabefunktion

Machine learning can be carried out, for example, by artificial neural networks (hereinafter referred to as artificial neural network, ANN for short). Artificial neural networks are usually based on the networking of many neurons, such as McCulloch-Pitts neurons or slight modifications thereof. In principle, other artificial neurons can also be used in ANN, e.g. B. the high-order neuron. The topology of a network (the assignment of connections to nodes) must be determined depending on its task. After the construction of a network, the training phase follows, in which the network "learns". A network can learn, for example, using the following methods:

• development of new compounds
• Delete existing connections
• Changing the weight (the weights from neuron j to neuron i)
• Adjust the thresholds of neurons if they have thresholds
• Adding or deleting neurons
• Modification of activation, propagation or output function

In addition, the learning behavior changes when the activation function of the neurons or the learning rate of the network changes. In practical terms, an ANN learns mainly by modifying the weights of the neurons. An adaptation of the threshold value can be taken care of by an on-neuron. As a result, ANN are able to learn complicated non-linear functions using a learning algorithm that attempts to determine all parameters of the function from existing input and desired output values by means of an iterative or recursive procedure. ANN are a realization of the connectionist paradigm, since the function consists of many simple, similar parts. The behavior only becomes complex when they are added together.

According to one embodiment of the invention, it is provided that probability densities for the properties are determined from the measurement data of a large number of measurements.

Knowing the probability densities makes it possible to define classification limits for the assignment of properties. The method is advantageously very robust with regard to the classification limits, because with the comparatively low-dimensional problems according to the invention, the probability densities for the two classes can be estimated from the data (e.g. with density estimation of the measurement results) and thus also the error probabilities for an incorrect classification determine.

According to one embodiment of the invention, it is provided that a data pool is used to determine the closing time, in which closing times are linked to the properties of the rail vehicles that can be determined, in particular types of train. These closing times present in the data pool can in particular be used as standard closing times depending on the detected Type of train are used and modified according to the inventive method for the final determination of the closing time by this is shifted at least once, in particular to a later closing time.

The data pool can be determined deterministically and/or created and/or further developed using the machine learning methods already explained above during operation. As soon as the data is available in the data pool, it can advantageously be used with short access times. During operation, the data in the data pool can be further optimized so that train operations are increasingly streamlined.

According to one embodiment of the invention, it is provided that a standard closing time for the level crossing is selected in the event that the property of the rail vehicle could not be determined.

According to the invention, the standard closing time for the level crossing should be understood to be that closing time which can reliably prevent any risk to crossing vehicle and passenger traffic, regardless of the properties of the trains running on the route. Critical here are the slow-moving freight trains, which take the longest from the activation point of the track safety system to the level crossing and therefore require the longest closing time. This can thus be defined as the standard closing time.

• die Geschwindigkeit und die Beschleunigung aus den Messdaten des Achszählers zu ermitteln,
• eine Schließzeit zum Auslösen eines Schließens des Bahnübergangs in Abhängigkeit der ermittelten Eigenschaften des Schienenfahrzeugs zu bestimmen,

wobei die Schließzeit des Bahnübergangs unter Berücksichtigung

• der ermittelten Geschwindigkeit bei Ausfahrt des Schienenfahrzeugs aus dem Achszähler und
• der ermittelten Beschleunigung bei Ausfahrt des Schienenfahrzeugs aus dem Achszähler

bestimmt wird.The object is also achieved according to the invention with the device specified at the outset in that the computer is set up to

• determine the speed and acceleration from the measurement data of the axle counter,
• to determine a closing time for triggering a closing of the level crossing depending on the determined properties of the rail vehicle,

taking into account the closing time of the level crossing

• the speed determined when the rail vehicle exits the axle counter and
• the acceleration determined when the rail vehicle exits the axle counter

is determined.

The advantages that have already been explained in connection with the method described in more detail above can be achieved with the device. What has been said about the method according to the invention also applies correspondingly to the device according to the invention.

Furthermore, a computer program product with program instructions for carrying out the method according to the invention and/or its exemplary embodiments is claimed, with the method according to the invention and/or its exemplary embodiments being able to be carried out in each case by means of the computer program product.

In addition, a provision device for storing and/or providing the computer program product is claimed. The provision device is, for example, a storage unit that stores and/or provides the computer program product. Alternatively and/or additionally, the provision device is, for example, a network service, a computer system, a server system, in particular a distributed, for example cloud-based computer system and/or virtual computer system, which stores and/or provides the computer program product preferably in the form of a data stream.

The provision takes place in the form of a program data block as a file, in particular as a download file, or as a data stream, in particular as a download data stream, of the computer program product. However, this provision can also be made, for example, as a partial download consisting of several parts. Such a computer program product is read into a system, for example using the provision device, so that the method according to the invention is executed on a computer.

Further details of the invention are described below with reference to the drawing. Identical or corresponding drawing elements are each provided with the same reference symbols and are only explained several times insofar as there are differences between the individual figures.

The exemplary embodiments explained below are preferred embodiments of the invention. In the exemplary embodiments, the described components of the embodiments each represent individual features of the invention to be considered independently of one another, which also develop the invention independently of one another and are therefore also to be regarded as part of the invention individually or in a combination other than the one shown. Furthermore, the components described can also be combined with the features of the invention described above.

• Figur 1 ein Ausführungsbeispiel der erfindungsbemäßen Vorrichtung mit ihren Wirkzusammenhängen schematisch,
• Figur 2 und 3 schematisch teils identische bzw. ähnliche Muster von Achsabständen für einen Personenzug und einen Güterzug,
• Figur 4 ein Ausführungsbeispiel des erfindungsgemäßen Verfahrens als Flussdiagramm, wobei die Funktionseinheiten und Schnittstellen gemäß Figur 1 beispielhaft angedeutet sind,
• Figur 5 symbolhaft für zwei Normalverteilungen für ermittelte Messdaten, aber im Prinzip funktioniert das für alle Verteilungen.
• Figur 6 verschiedene Weg-Zeitverläufe in einem Diagramm der zurückgelegten Weges s über der Zeit t, die den querenden Zug (Schienenfahrzeug) nach Passieren der Achszähler charakterisieren könnten, je nach seinen Eigenschaften und dem aus der Achszählung gewonnenen Kenntnisstand.

Show it:

• figure 1 an exemplary embodiment of the device according to the invention with its functional relationships schematically,
• Figure 2 and 3 schematic partly identical or similar models of wheelbases for a passenger train and a freight train,
• figure 4 an embodiment of the method according to the invention as a flowchart, wherein the functional units and interfaces according to figure 1 are indicated as examples,
• figure 5 symbolic for two normal distributions for determined measurement data, but in principle this works for all distributions.
• figure 6 different path-time curves in a diagram of the covered path s over time t, which could characterize the crossing train (rail vehicle) after passing the axle counter, depending on its properties and the knowledge gained from the axle count.

In figure 1 a track system with a track GL, a control center LZ and a signal box SW is shown. A vehicle FZ in the form of a train is driving towards a level crossing BU on the track GL. A first axle counter AZ1 and a second axle counter AZ2 are installed on the track GL, which are set up in a manner known per se to count the axles of the vehicle FZ. The use of two axle counters creates redundancy in order to increase the security (safety) of the method against failure. The first axle counter AZ1 and the second axle counter AZ2 are so-called double axle counters, each with two axle counting sensors. The axle counting sensors are arranged one after the other in the direction of travel of the trains so that they generate a measurement signal in quick succession. This measurement signal can be used in a manner known per se to determine the direction of travel FR of the train and the speed v of the train. The center distances L can also be determined from the speed v (also specified below with additional capital letters). Shown in figure 1 is the wheelbase L of a bogie of the vehicle FZ shown. If the speed development from axis to axis of the crossing train is considered, an acceleration value a can also be derived from this consideration (more on this below).

The axle counter AZ1 is connected via a first interface S1 and the second axle counter AZ2 via a second interface S2 to the signal box SW, strictly speaking to a computer CP present in this signal box. In addition, the computer CP has a third interface S3 for the level crossing BU. In addition, the computer CP is connected to a storage unit SE via a sixth interface S6.

The signal box SW has a first antenna system A1, the control center LZ has a second antenna system A2 and the vehicle FZ has a third antenna system A3. This enables both the signal box SW to communicate with the control center LZ via a fourth interface S4 and the vehicle FZ to communicate with the control center LZ via a fifth interface S5. At the fourth interface S4 and the fifth In this respect, interface S5 is a radio interface. The first interface S1, the second interface S2 and the third interface S3 can represent both wired and radio interfaces, with the antenna technology that would be required for forming radio interfaces not being shown for the latter case.

If the vehicle FZ moves on the track GL towards the level crossing BU, the axles of the vehicle FZ first pass the second axle counter AZ2 and then the first axle counter AZ1. The measured values recorded can be transmitted to the computer CP via the first interface S1 and the second interface S2, the computer CP being set up to carry out the method according to the invention. The computer CP can also take over the control of the level crossing BU directly. Another possibility is that the computer CP can be connected to another computer (in figure 1 not shown) is connected, which is used via a further interface to control the level crossing BU.

In figure 2 is one on the track GL as a vehicle FZ according to figure 1 moving passenger train PZ shown. This passenger train PZ consists of a locomotive LK, several passenger cars PW and a power car TK at the end of the passenger train PZ opposite the locomotive LK.

Furthermore, the center distances between the individual axles (indicated by wheels) are shown schematically. It can be seen that in the passenger train PZ different wheelbases occur several times, so that the sequence of the wheelbases can be examined for the presence of patterns. The center distances are marked with the capital letters A to G. The sequence of center distances consists of FFEFFGABACABACABACADA.

If one disregards the locomotive LK and the power car TK, since these differ from the passenger cars PW in terms of their wheelbase, the result for the consecutive passenger cars, which are identical in construction, is a sequence of wheelbases that is repeated over and over again. These form in this respect a model MT, which is marked with a curly bracket for the passenger car PW following the locomotive LK. The sequence of the center distances in the in figure 2 illustrated pattern MT is ABAC. This sequence of wheelbases also applies to the two following passenger cars.

The situation is different in the in figure 3 illustrated freight train GZ on the track GL, which consists of a locomotive LK and a first freight car GW1, a second freight car GW2 and a third freight car GW3. These have different lengths and axle numbers, resulting in several different axle spacings, which are marked with the capital letters A to H. In figure 3 it becomes clear that no repetitive pattern can be discovered in the sequence ABACDEDFGFH shown, which allows the conclusion that it is a freight train.

In figure 4 shows how the method according to the invention can take place. First, this is started in a first step START. At this point in time, the closing time of the level crossing BU is set to the value of a standard closing time SZS. A separate memory area is reserved for this in the memory device, to which a controller (e.g. the computer CP or a Figures 1 to 3 further computer not shown) of the level crossing can access to call up the currently stored closing time.

A measuring step MS follows by the relevant axle counters AZ1, AZ2 (cf. figure 1 ). In this measuring step, the axle counting sensors of the two axle counters are used to determine, among other things, the speed of the passing wheels (representing the axle), distances between the axles, the speed of the axles and the speed development from axle to axle, and thus the acceleration, with the Acceleration can be constant or can undergo an evolution.

This measurement step is followed by a first test step PS1, this first test step consists of an evaluation step for the speed ELVv and an evaluation step for the Acceleration ELVa, in which the effects of the speed v of the train and the acceleration state a of the train during the crossing via the axle counters are taken into account.

The slower the train is traveling, the slower it will reach the level crossing BU. This means that the closing time of the level crossing can be postponed to a later date. Specifically, this means that the standard closing time SZS, which takes into account the worst case, can be corrected and a later closing time SZ1 can be output to the storage device SE.

When evaluating the acceleration, it must be taken into account that the slower the acceleration while crossing the axle counter, the less the train can accelerate due to its inertia on the way from the axle counter to the level crossing BU. By assuming the worst case, that the train is accelerated to the maximum for a specific train type after exiting the axle counter, the closing time as the second closing time SZ2 will in most cases have to be corrected to earlier closing times. However, if the train is accelerated negatively, i.e. decelerated, the braking effect must first be canceled before a positive acceleration can take place. This can also be taken into account by specifying a later closing time, which is output to the storage device SE as the second closing time SZ2.

Then follows a second test step PS2, in which the sequence of the center distances (as in Figure 2 and Figure 3 described) can be determined and checked. It is either possible to recognize patterns MT in the sequence of the center distances or not. In a subsequent query step GZ, PZ? It is checked whether the sequence of wheelbases (by finding samples) can be used to conclude that it is a GZ freight train or a passenger train. If this is not the case, the second closing time SZ2 remains.

If the first property, i.e. the question of whether it is a freight train GZ or a passenger train PZ, can be determined, a further query step GZ? follows in the computer CP, whether it is is a freight train GZ. If this is the case, a third modified closing time SZ3 is transferred to the storage unit SE (replaces a previously stored closing time). If it is not a freight train or if there is no clear result, a second test step PS2 is carried out in the computer CP.

The third test step PS3 is used to determine the amounts of the center distances. In a subsequent test step |A| <GW can therefore be asked whether the determined amounts of the axle distances are smaller than a typical limit value for freight wagons GW. If this is the case, it is a freight train GZ, so that the third modified closing time SZ3 can be transferred to the storage unit SE (replaces a previously stored closing time). If this is not the case, a fourth test step PS4 is initiated in the computer CP.

In the fourth test step PS4, reference patterns RMT are loaded from the memory unit SE. The center distances or their amounts are now compared with the reference samples, whereby in a test step MT=RMT? it can be checked whether the determined pattern MT corresponds to a reference pattern RMT. If this is not the case, a fourth modified closing time SZ4 is transferred to the storage unit SE (replaces a previously stored closing time). However, if a pattern MT has been recognized, a fifth modified closing time SZ5 can be transferred to the memory unit SE (replaces a previously stored closing time), which individually matches the reference pattern RMT.

The aforementioned modified closing times SZ3, SZ4, SZ5, SZ6 are closing times that are calculated taking into account the closing times determined in the first test step and are modified in this sense. In other words, the insights gained in the first test step with regard to the speed and acceleration of the approaching train are also included in the modified closing times, in which the type of train is also taken into account. In other words, they are refinements typical of a train, each one enable more and more precise determination of the closing time, whereby the safety (safety) of the operation of the level crossing is not impaired due to the additional knowledge.

Should the query step MT=RMT? turn out to be negative, the determined new pattern MT can also be transferred to the control center LZ via the interface S4. At the same time, driving data FD from the vehicle FZ can also be transferred to the control center LZ via the fifth interface S5. Using the data available in the control center LZ, a new sixth modified closing time SZ6 adapted to the determined train type can then be determined in a modification step MOD, and transmitted to the storage unit SE via an output step OUT. This sixth modified closing time SZ6 can then be used as an individual closing time for the level crossing BU (replaces a previously saved closing time). At the same time, an output can be made in the memory unit SE such that the sixth modified closing time SZ6 is written in the memory unit SE as a supplement to the database together with the newly determined reference pattern RMT, which belongs to the vehicle FZ just analyzed.

In the storage unit SE there is now a closing time for the level crossing BU in the separate storage area. Depending on the course of the procedure, this can be the standard closing time SZS, the first closing time SZ1, the second closing time SZ2, the third modified closing time SZ3 or the fourth modified closing time SZ4, the fifth modified closing time SZ5, the sixth modified closing time SZ6 (or further closing times, which are not according to the example figure 4 are described) act.

This closing time is now available in the separate memory area of the memory device SE in order to be able to be activated by controlling the level crossing BU (cf. figure 1 ), So the computer CP or another controller of the level crossing BU (not shown) to be handed over. In this respect, the level crossing BU can be operated with an individually determined closing time.

In figure 5 two parameters measured or determinable by the axle counter according to the invention are shown abstractly in a plane which could also be referred to as the xy plane and on which the measured value distribution MV of the measured values becomes clear. Accordingly, the speed GSW would be shown on the x-axis and the center distances A...H on the y-axis. The z-axis is used to represent the (e.g. estimated) probability densities

In this example, site-specific, representative data is collected or measured and classified for the relevant parameters, e.g. B. Passenger train as normal distribution NV2 and freight train as normal distribution NV1, as already described above. It is a finite number of integer or real-valued measurement data from the axle counters, e.g. For example, this could be the speed and the center distance, to give a clear two-dimensional example here. That is, in principle one obtains a classification task as in figure 5 shown schematically.

If representative data is available, it is known how such pattern recognition problems can be solved using machine learning methods, e.g. B. Neural networks solves. In this application with axle counters, there is a great deal of leeway as to how one sets the classification limit, because with such low-dimensional problems one can also estimate the probability densities for the two classes from the data (e.g. with density estimation). This can be used to determine the error probabilities for an incorrect classification (cf. e.g. B. Duda et al.: Pattern Classification, Wiley, 2001 ), figure 5 shows this symbolically for a first normal distribution NV1 and a second normal distribution NV2, but in principle this also works for other distributions than normal distributions.

If you assume in the example that the small ellipse is the first classification limit KG1 for freight trains and the large ellipse is the classification limit KG2 for passenger trains, then you could use the estimated distributions to determine the error probabilities calculate. If the misclassification probability for freight trains were too high, the classification limits would be changed. In the example according to figure 5 a smaller ellipse would then be obtained for the first classification limit KG1. However, there can also be applications where the classification errors are asymmetrical, ie the errors do not have the same meaning. For example, in the case of timed switching on of a level crossing, it would be irrelevant from a safety perspective if a slow freight train were classified as a fast passenger train, while in the case of a tunnel crossing ban this would be dangerous. i.e. In any case, the safety aspect must be taken into account when evaluating the types and probabilities of errors.

In figure 6 shows the distance traveled by trains in a distance-time diagram. The time t is shown on the x-axis and the distance traveled by the trains s on the y-axis. The position of the first axle counter AZ1 and the level crossing BU is also shown on the y-axis, because the implementation of the method is about covering this stretch and estimating the time required for this.

The curves K1 to K4 show different profiles of a route progression of a train under consideration. Closing times SZS and SZ1 to SZ4 are calculated using these curves. These are plotted in the path-time diagram and have in the embodiment according to figure 6 always a constant time interval ZA to the planned arrival time of the vehicle at the level crossing BU. The arrival time is represented by dash-dotted vertical lines that intersect the x-axis at the respective arrival times.

The zero point of the time axis is at the point in time at which the rail vehicle leaves the relevant axle counter AZ1. Up to this point in time, the speed and acceleration were recorded during the crossing based on the axle count events, so that curve K0 is a real measured curve. Without taking this curve into account, the standard closing time SZS is provided, which is the worst-case scenario for the approach of the rail vehicle to the level crossing. This is the constant for the maximum speed permitted on the stretch of road, which is why curve K1 in the distance-time diagram is linear.

How to figure 4 explained, various specifications can be taken into account by determining the speed and acceleration in the steps ELVv and ELVk. As in figure 6 shown, the rail vehicle was braked before crossing over the first axle counter AZ1 (cf. curve K0, which has a right-hand curve). Therefore, curve K2 can be assumed for the distance-time behavior, which takes into account the maximum acceleration of the fastest rail vehicle running on the route. This must first build up the speed (ie reverse the negative acceleration into a positive one), which is why, in comparison to curve K1, less critical scenarios SZ1, SZ2 with a later closing time can be assumed.

Curve K3 occurs when a freight train is detected in the second test step PS2 (cf. figure 4 ). This can only, as in figure 6 is shown, accelerate much more slowly, which is why the closing time SZ3 can be shifted to later points in time due to the third curve K3 compared to the second curve K2.

However, if a passenger train is detected, a comparison with a reference model RMT can be carried out in test step PS4 in order to determine the train type more precisely. Although it is in accordance with the example figure 6 If it is a passenger train, the fourth curve K4 can be assigned to it if it turns out that its acceleration capacity does not correspond to that of the fastest train running on the route (cf. curve K2). In comparison to the second curve K2, the closing point in time SZ4 can also be shifted to a later point in time when the fourth curve K4 is taken into account.

Reference List

GL

Track

car

rail vehicle

BU

Railroad Crossing

LZ

control center

SW

signal box

A1...A3

antenna

AZ1 ... AZ2

axle counter

S1...S6

interface

CP

computer

SE

storage unit

v

speed

a

acceleration

PZ

passenger train

LK

locomotive

pw

passenger car

TC

power car

GZ

freight train

GW1 ... GW3

Freight wagons

A...H,L

center distance

MT

Pattern

RMT

reference pattern

GW

limit

MS

measuring step

PS1 ... PS3

test step

SZS

Default closing time

SZ1 ... SZ6

Closing time (calculated)

FD

driving data

IN

input step

MOD

modification step

OUT

output step

GZ?

Query step freight train?

MV

measured value distribution

GSW

speed

KG1 ... KG2

classification limit

NV1 ... NV2

normal distribution

s

Distance (covered by the rail vehicle)

t

Time

K1...K4

displacement-time curves

Claims (14)

Method for operating a level crossing (BU), in which rail vehicles (FZ) approaching a level crossing (BU) • an axle counter (AZ1 ... AZ2) records measurement data while the rail vehicle (FZ) is crossing, • the speed (v) of the rail vehicle and the acceleration (a) of the rail vehicle are determined computer-aided as properties of the rail vehicle (FZ) from the measurement data, • a closing time (SZ1 ... SZ4) for triggering a closing of the level crossing (BU) is determined depending on the determined properties of the rail vehicle, characterized,
that the speed and acceleration are determined from the measurement data of the axle counter and the closing time (SZ1 ... SZ4) of the level crossing is taken into account • the determined speed (v) when the rail vehicle (FZ) exits from the axle counter (AZ1 ... AZ2) and • the determined acceleration (a) when the rail vehicle (FZ) exits the axle counter (AZ1 ... AZ2) is determined. Method according to claim 1,
characterized,
that when determining the closing time (SZ1 ... SZ4), a standard closing time (SZS) is modified by it being shifted at least once, in particular to a later closing time. Method according to one of the preceding claims,
characterized,
that the determined property of the speed is taken into account in such a way that, taking into account the distance of the axle counter (AZ1 ... AZ2) to the level crossing (BU), the reaching of the level crossing by the rail vehicle (FZ) is determined. Method according to one of the preceding claims,
characterized,
that the determined property of the acceleration is taken into account in such a way that, taking into account the distance of the axle counter (AZ1 ... AZ2) to the level crossing (BU) and the property of the acceleration, reaching the level crossing by the rail vehicle (FZ) is determined. Method according to one of the preceding claims,
characterized in that • a first additional property is determined with the aid of the computer based on the wheelbases (A ... H) determined, namely whether the rail vehicle (FZ) is a passenger train (PZ) or a freight train (GZ), • when determining the first further property, in a first test step (PS1) it is checked whether identical or similar patterns (MT) can be determined at least in a predominant part of the sequence of the center distances (A ... H), and • if no model (MT) could be determined, the rail vehicle (FZ) is assigned the property of a freight train (GZ) as the first additional property, or • if a sample (MT) has been determined, the rail vehicle (FZ) is assigned the property of a passenger train (PZ) as the first additional property and/or a further test step (PS2, PS3) is carried out. Method according to claim 5,
characterized,
that in the second test step (PS2) in the sequence of center distances (A ... H) a number of center distances (A ... H) at the beginning of the sequence and/or a number of center distances (A ... H) at the be disregarded at the end of the episode. Method according to one of claims 5 or 6,
characterized,
that in the second (PS2) or a further test step (PS3) the amount of the center distances (A ... H) is determined, with the rail vehicle (FZ) being assigned the property of a passenger train (PZ) as the first further property, as long as the The amount of the largest center distance (A ... H) occurring in the pattern (MT) exceeds a specified limit value (GW). Method according to one of claims 5 to 7,
characterized, that in the second (PS2) or a further test step (PS3) the pattern (MT) of the center distances (A ... H) with Reference patterns (RMT) of wheelbases (A ... H) are compared and in the case of a recognized match of the pattern (MT) with a reference pattern (RMT) the rail vehicle (FZ) with the reference pattern (RMT) linked type of train as a second further property is assigned. Method according to one of claims 5 to 8,
characterized,
that the criteria for the second test step (PS2) and/or the further test steps (PS3, PS4) are evaluated using a machine learning method. Method according to one of claims 5 to 9,
characterized,
that probability densities for the properties are determined from the measurement data of a large number of measurements. Method according to one of the preceding claims,
characterized,
that for determining the closing time (SZ1 ... SZ4) a data pool is used in the closing times (SZ1 ... SZ4) are linked to the identifiable properties of the rail vehicles (FZ), in particular train types. Device for operating a level crossing (BU), comprising • at least one axle counter (AZ1 ... AZ2) for recording measurement data when the rail vehicles (FZ) pass, • a computer (CP), which is set up, computer-aided as properties of the rail vehicle (FZ), the determine the speed of the rail vehicle and the acceleration of the rail vehicle from the measurement data, • and to close the level crossing for rail vehicles (FZ) approaching a level crossing (BU), characterized,
that the computer (CP) is also set up to • determine the speed (v) and the acceleration (a) from the measurement data of the axle counter, • to determine a closing time (SZ1 ... SZ4) for triggering a closing of the level crossing (BU) depending on the determined properties of the rail vehicle, taking into account the closing time of the level crossing • the speed determined when the rail vehicle (FZ) exits from the axle counter (AZ1 ... AZ2) and • the acceleration determined when the rail vehicle (FZ) exits the axle counter is determined. Computer program product with program instructions for carrying out the method according to one of Claims 1 - 11. 14. Provision device for the computer program product according to claim 13, wherein the provision device stores and/or provides the computer program product.

Images