A method for counting axles using computer-aided evaluation

Abstract

A method for counting axles, wherein a wheel passes an axle counter sensor mounted on a track, the axle counter sensor generates a measurement signal (U1 ... U2), and a curve (VL1 ... VL2) of the measurement signal (U1 ... U2) is evaluated in a computer-aided manner to distinguish the wheels. When evaluating the measurement signal (U1 ... U2), at least one maximum (M1 ... M4) of the signal amplitude is sought in the curve (VL1 ... VL2) of the measurement signal (U1 ... U2). A dynamic time normalization is performed on the curve (VL1 ... VL2) of the measurement signal (U1 ... U2) before and after the maximum value (M1 ... M4). The curve (NV1 ... NV2) of the measurement signal (U1 ... U2) normalized by amplitude normalization and time normalization is compared with a pattern (M1 ... M2) of at least one curve (VL1 ... VL2) for the measurement signal (U1 ... U2) when a wheel (RD) passes by and at least one curve (VL1 ... VL2) for the measurement signal (U1 ... U2) when an error occurs. The invention also includes a computer program product and a device for providing a computer program product.

Description

Technical Field

[0001] The invention relates to a method for counting axles, in which a wheel passes an axle counter sensor mounted on a track, the axle counter sensor generates a measurement signal, the curve of which is evaluated computer-aided, wherein the wheels are distinguished. The invention also relates to a computer program product and a device for providing the computer program product, wherein the computer program product is provided with program instructions for carrying out the method. Background Art

[0002] When counting axles with an axle counter, various natural disturbances are known to occur, from simple roars or environmental factors to cable sags on the train or the so-called swerving effect of bogies in narrow curves. It is therefore desirable to eliminate disturbances such as wheel or bogie signals in particular in order to be able to reliably identify wheel signals.

[0003] Related problems must be solved, for example, in the case of computer-assisted recognition of handwriting. For the recognition of handwriting, Claus Bahlmann et al. describe a suitable method in the article "Writer-independent online handwriting system frog on hand and cluster-generated statistical time warping" in IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINEINTELLIGENCE, 3 March 2004, Vol. 26. Despite the differences that may arise from different handwritings, the letters should still be recognized. However, this cannot be directly transferred to the axle counter, because for the axle counter, a distinction must be made between useful signals indicating the passage of a wheel and interference signals.

[0004] Furthermore, sufficient safety must be achieved here. It should be taken into account that interference signals may reach such an extent that they are misinterpreted as a wheel passing. For this purpose, the following example is given without limiting the generality.

[0005] From a measurement point of view, a bogie consists of two wheels that follow one another, that is, two maxima of the signal amplitude of the measurement signal with a certain level of flatness between the two maxima. This can lead to measurement errors, which are called skew travel. In so-called skew travel, the level is increased, so that the third wheel is incorrectly detected. Summary of the invention

[0006] The technical problem to be solved by the present invention is to provide a method for counting axles, which has a relatively high safety against erroneous recognition of the passage of wheels. In addition, the technical problem to be solved by the present invention is to provide a computer program product and a preparation device for the computer program product, by means of which the above method can be implemented.

[0007] According to the present invention, the technical problem is solved by a method for axle counting, wherein when evaluating a measurement signal, at least one maximum value of the signal amplitude is found in the curve of the measurement signal, and during amplitude normalization, the amplitude of the measurement signal is normalized (or standardized) so that the maximum value is the same as a predetermined target value, and dynamic time normalization is performed on the curve of the measurement signal before and after the maximum value, wherein the curve of the measurement signal normalized by amplitude normalization and time normalization is compared with a pattern of at least one curve of the measurement signal when a wheel passes by and with a pattern of at least one curve of the measurement signal when an error occurs.

[0008] The measurement signal is the time curve of a measured value, preferably a signal voltage, which results from the passage of a wheel of an axle, but which also has corresponding maximum values ​​due to interference factors. This means that by computer-aided evaluation of the measurement signal it is possible to identify the event to be detected, namely that a wheel has passed the axle counter sensor, but interference signals can also be incorrectly identified as such a wheel passing.

[0009] According to the invention, amplitude normalization and dynamic time normalization (also referred to as dynamic time warping (DTW)) are applied to the measurement results. This has the advantage that the measurement signal is normalized not only with respect to the amplitude of its maximum value but also with respect to the length of its time curve. This subsequently simplifies the comparison of the curve to be evaluated of the measurement signal with patterns of different curves (more on this in the following). This improves the reliability of the pattern recognition for the wheel passage and minimizes the probability of erroneous evaluation results.

[0010] Amplitude normalization is performed in such a way that the maximum value of the observed curve of the measurement signal is identical to a predetermined target value after normalization. Preferably, normalization to 1 can be performed, i.e. the target value is equal to 1. However, this is not absolutely necessary. It is important that the predetermined target value of the maximum value coincides with the maximum value contained in the pattern, with which the relevant curve of the measurement signal is to be compared.

[0011] DTW is performed in order to identify, in a time-defined portion of the entire curve of the measured signal, which extends before and after a maximum, which patterns should be compared in order to detect a wheel passing or an error occurring. Since the wheel passing generates a signal curve which rises to a maximum and then falls again, the curve identified by DTW contains a maximum.

[0012] The purpose of the time normalization is to be able to compare the relevant curve of the measurement signal with the pattern. In particular, it should be taken into account that the curve of the measurement signal depends in particular on the speed of the vehicle passing the axle counter sensor. Higher speeds produce a steeper, shorter rise to a maximum value (and a corresponding subsequent fall). In contrast, lower speeds produce a flatter, longer rise to a maximum value (and a corresponding subsequent fall).

[0013] For example, the principle of DTW is known from speech recognition (recognition of speech features during dictation): Here, individual words can be recognized from a spoken text by comparison with stored speech patterns. One problem is that words are often pronounced differently. Vowels in particular are often spoken longer or shorter. For successful pattern matching, the words should therefore be expanded or compressed accordingly, but not uniformly, but primarily on the longer or shorter pronounced vowels. Dynamic Time Warping algorithms provide this adaptive time normalization. Another application is handwriting recognition. There, pattern recognition of individual letters is performed in order to recognize letters from different handwritings.

[0014] The invention makes use of the recognition that the measurement signals of an axle counter have a relatively low complexity compared to handwriting recognition or speech recognition. On the other hand, however, there are errors which can be confused with a wheel passing and thus lead to erroneous results during the evaluation. These errors must be reliably recognized despite the relatively low complexity of the pattern. The invention proceeds here by defining patterns not only for events to be recognized for wheel passing of different vehicles, but also for typically occurring errors which are then recognized as such and cannot be confused with a wheel passing.

[0015] In other words, the object of the invention is not only to identify events that are expected and should be counted, but also to deliberately identify those events that should not occur and therefore should not be counted, but which may be erroneously identified as events to be counted. If these events are reliably identified as errors, they can be excluded as counted events, even if their evaluation as wheel-passing events to be counted is uncertain. This is the added value of increasing the reliability of the identification according to the invention.

[0016] In the context of the present invention, “computer-assisted” or “computer-implemented” may be understood as an implementation of a method, wherein at least one computer or processor performs at least one method step of the method.

[0017] The term "computer" or "computer" covers any electronic device with data processing properties. Computers may 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 may also preferably be combined to form a network.

[0018] In connection with the present invention, a "processor" may be understood as, for example, a converter, a sensor or an electronic circuit for generating a measurement signal. The processor may specifically be a central processing unit (CPU), a microprocessor, a microcontroller or a digital signal processor, possibly in combination with a memory unit for storing program instructions. The processor may also be understood as a virtualized processor or a software CPU.

[0019] In the context of the present invention, a "memory unit" is to be understood as a computer-readable memory, for example in the form of a random access memory (RAM) or a data storage (hard disk or data carrier).

[0020] An “interface” can be implemented in terms of hardware, such as a wired or wireless connection, and / or software, such as the interaction between individual program modules or program parts of one or more computer programs.

[0021] "Program modules" are to be understood as individual functional units of a program sequence capable of implementing the steps of the method according to the invention. These functional units can be implemented in a single computer program or in multiple computer programs that communicate with each other. The interfaces implemented here can be implemented in software within a single processor or in hardware if multiple processors are used.

[0022] According to one embodiment of the invention, a curve of the measurement signal normalized by amplitude normalization and time normalization is compared with a pattern of at least one curve of the measurement signal when passing a single wheel and a pattern of at least one curve of the measurement signal when passing two wheels of the bogie.

[0023] This embodiment of the invention makes use of the knowledge that the double axle of a bogie, i.e. in this case two wheels passing through the axle counter sensor, produces a characteristic pattern with two maximum values. If these two maximum values ​​are identified by DTW as belonging to the bogie, normalization can be performed with reference to this double event. Subsequently, this double event can be compared with the associated pattern. This further increases the reliability. Thus, the identified bogie is counted twice in the axle counting because it has two axles.

[0024] If the bogie cannot be identified, the two wheels can be identified as a single wheel and evaluated. Here, the same counting results are obtained, provided that both wheels are identified. This shows that by defining a pattern belonging to the bogie, additional identification possibilities can be provided, thereby increasing the reliability of the identification. This is because the pattern of the bogie provides more characteristic evaluation criteria and can therefore be identified more easily. However, if the bogie is not identified as a single wheel, the fallback position for identifying the single wheels still exists.

[0025] According to one embodiment of the invention, the comparison with the pattern of the course of the measurement signal is carried out when passing two wheels of the bogie only if the time shift of the maximum values ​​in the course of the measurement signal does not exceed a limit value predefined as a function of the speed of the vehicle passing the axle counter sensor.

[0026] This measure is based on the knowledge that when a bogie passes, the axle counter sensor records two maximum values ​​in quick succession. In other words, if no maximum value is measured within a speed-dependent time interval characteristic of the bogie, it can be ruled out that the bogie is a bogie.

[0027] In order to be able to predetermine the limit value, the speed at which the vehicle passes the axle counter sensor must be known. There are different possibilities for this. The speed can be determined, for example, by means of another sensor and entered as an input variable into the method. For example, the speed in the vehicle can be measured and transmitted by radio to a computer which performs the calculations according to the method of the invention.

[0028] Another possibility is to estimate the speed from the mutual relationship of the pattern of maximum values ​​(corresponding to the axle counter pulses). Bogies are usually installed on a vehicle of a certain length, so that the bogies produce maxima close to each other, followed by a longer stop (through the center of the vehicle) or a shorter stop (between two coupled vehicles). From the ratio of the stops, the speed can be estimated and thus the speed-related limit value can be determined.

[0029] Another possibility is to use a so-called double axle counter, in which two axle counter sensors are installed in short sequence. Since the spacing of the axle counter sensors is known, the speed can be inferred by determining the time offset of the maximum values ​​generated by the same wheel in the two axle counter sensors (more on this below).

[0030] According to one embodiment of the invention, a normalized curve of the measurement signal by amplitude normalization and time normalization is compared with at least one pattern of the curve of the measurement signal when a skew occurs when the bogie is traveling on a curve.

[0031] As mentioned above, skew travel refers to a measurable signal overshoot of the measurement signal of the axle counter sensor, which produces a maximum value between two maxima passed by the wheels of the bogie. Skew travel preferably occurs when the axle counter sensor is installed in a bend and the measurement is performed during the bend travel of the bogie.

[0032] If possible sideways travel effects are defined as patterns of errors that occur, then when sideways travel occurs during measurement by the axle counter sensor, a curve can be generated within the scope of DTW, which can be associated with this error after comparison with the existing pattern of sideways travel. If this association is unambiguous or one-to-one, the relevant curve of the measurement signal can be excluded from the association with wheel passing events. This is particularly advantageous in the case where the association with wheel passing can be extreme and, in doubtful cases, non-existent axles can be counted incorrectly.

[0033] In other words, there are situations in which the method according to the invention can be used with greater reliability when counting axles belonging to a bogie. The occurrence of erroneously counted axles can thus be excluded or the probability of such an event at least reduced.

[0034] According to one embodiment of the invention, an axle counter is used which has a first axle counter sensor and a second axle counter sensor arranged one behind the other in the direction of travel, wherein the method is carried out one after the other for the first axle counter sensor and the second axle counter sensor.

[0035] This is a so-called two-axle counter, which is very widely used. Therefore, the two installed axle counter sensors (i.e. the first axle counter sensor and the second axle counter sensor) produce the same maximum value in the time course of the measurement signal in short order, at least in the absence of interference. In this case, the maximum value corresponds to the wheel being counted. Otherwise, interference signals leading to the maximum value can also be detected.

[0036] The use of two axle counter sensors does not change the operating principle of the axle counter. The first axle counter sensor and the second axle counter sensor operate in the same way as an axle counter sensor with only a single axle counter sensor installed. Therefore, unless otherwise stated, statements made in conjunction with the present invention apply equally to the axle counter sensor or the first axle counter sensor and the second axle counter sensor.

[0037] The advantage of using a first axle counter sensor and a second axle counter sensor is that the axle counter has a higher safety against errors. In addition, as long as the first axle counter sensor and the second axle counter sensor are in operation, the sensor signals can also be used to determine the speed of the vehicle passing the axle counter. In this case, the maximum values ​​generated by the same wheel are checked in the first axle counter sensor and the second axle counter sensor with respect to their time offset, and the speed is determined taking into account the known distance between the first axle counter sensor and the second axle counter sensor.

[0038] According to one design solution of the present invention, it is provided that a maximum value in a first curve detected by the first axle counter sensor and a maximum value in a second curve detected by the second axle counter sensor are compared, and only the maximum values ​​(present in both the first curve and the second curve) of the first curve and the second curve normalized by amplitude normalization and time normalization of the measurement signal are compared with the pattern.

[0039] This embodiment of the invention makes use of the knowledge that the event of a wheel passing an axle counter sensor is reliably detected as a maximum in the curve of the measurement signal. Therefore, these maxima must also occur in both measurement curves of the measurement signal. If a maximum value only occurs in one of the two curves of the measurement signal, it is possible to conclude that this is a disturbing signal which itself should not be counted. It is therefore advantageous to exclude this maximum value a priori from the assessment of the presence of a wheel passing, thereby advantageously increasing false detection and thus the safety against errors when identifying the wheel.

[0040] According to one embodiment of the invention, a maximum value in a first curve detected by a first axle counter sensor is compared with a maximum value in a second curve detected by a second axle counter sensor, and the curve of the measurement signal before and after the maximum value to be taken into account in the dynamic time normalization is determined taking into account a time offset between a comparable maximum value of the first curve and the second curve.

[0041] If mutually corresponding maxima are found in the first and second curves, the time offsets that can be determined from these maxima can be advantageously used to obtain a speed-dependent measure of the time boundaries of the curves to be considered in the dynamic time normalization. This advantageously ensures that the change curves in the dynamic time normalization have a sufficient span to contain the characteristic to be evaluated for the subsequent comparison with the pattern.

[0042] Furthermore, protection is claimed for a computer program product having program instructions for executing the method according to the invention and / or its embodiments, wherein the method according to the invention and / or its embodiments can respectively be executed by means of the computer program product.

[0043] Furthermore, a provision device for storing and / or providing a computer program product is also claimed. The provision device is, for example, a memory unit, which stores and / or provides the computer program product. As an alternative and / or in addition, 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 a virtual computer system, which stores and / or preferably provides the computer program product in the form of a data stream.

[0044] The computer program product is provided in the form of program data blocks as a file, in particular as a download file, or as a data stream, in particular as a download data stream. However, it can also be provided, for example, as a partial download consisting of a plurality of parts. Such a computer program product is read into the system, for example using a provision device, so that the method according to the invention is executed on a computer.

[0045] The invention will be described in further detail below with reference to the drawings. Identical or corresponding elements of the drawings are provided with the same reference numerals and only the differences between the individual drawings are explained again. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] The embodiment described below is a preferred embodiment of the present invention. In this embodiment, the components of the described embodiment are the individual features of the present invention that can be viewed independently of each other, and these features also form the expanded design of the present invention independently of each other and can therefore also be regarded as components of the present invention alone or in a combination different from the combination shown. In addition, the described parts can also be supplemented by other features of the present invention that have been described.

[0047] In the attached picture:

[0048] Figure 1 As a block diagram, Figure 1Embodiment of a computer infrastructure of a device of the type described herein, wherein the individual functional units contain program modules which are each executable in one or more processors and which can thereby implement an interface in software or hardware terms.

[0049] Figure 2 A flow chart showing an exemplary embodiment of the method according to the invention is shown, wherein the individual method steps can be executed individually or in groups by program modules, and wherein according to Figure 1 The functional units and interfaces are shown by way of example. DETAILED DESCRIPTION

[0050] exist Figure 1 , a vehicle FZ is shown, which is traveling on a track GL in a travel direction FR. The vehicle FZ has bogies DG, each of which is provided with two axles. Figure 1 Indicated by wheel RD.

[0051] As soon as the wheel RD passes the axle counter AZL with the first axle counter sensor AZ1 and the second axle counter sensor AZ2, the measurement signals U1, U2 (see Figure 2 ) in the process (more on this below).

[0052] The axle counter AZL is connected to an evaluation unit AE which has a first computer CP1. The computer CP1 is connected to both the first axle counter sensor AZ1 and the second axle counter sensor AZ2 via a sixth interface S6. Instead of two axle counter sensors, a single axle counter sensor AZ can also be used, so that one of the two axle counter sensors is denoted by the reference numerals AZ and AZ1.

[0053] A first memory device SE1 connected to the first computer CP1 via a fifth interface S5 is also installed in the evaluation unit AE. The first memory device contains, for example, a program for executing the method according to the invention and a library with different modes P1, P2 (see Figure 2 ), the different modes P1, P2 are used to determine the curves VL1, VL2 to be measured, which are represented by normalized curves NV1, NV2, NV3 (see Figure 2 ).

[0054] Furthermore, the first computer CP1 is connected to a second computer CP2 in the control center LZ via a third interface S3. The second computer CP2 has a fourth interface S4 to a second memory device SE2. The control center represents a line-side device, such as a signal box or an automatic train control system.

[0055] The vehicle FZ and the control center LZ have antennas AT, so that they can communicate with each other via the second interface S2. In addition, the vehicle FZ can communicate with the satellite STL via the first interface S1. In this way, for example, the vehicle FZ can be positioned, wherein the satellite STL is a navigation satellite.

[0056] The method according to the invention has a program module which runs optionally in the first computer CP1 or in the second computer CP2. This depends on how "intelligent" the device for axle counting formed by the axle counter AZL and the evaluation unit AE is.

[0057] exist Figure 2 The method according to the invention is shown in the figure with reference to a flow chart. A schematic diagram of the signal curves is selected here in order to illustrate the individual method steps. Figure 2 The upper part of shows the curve VL1 of the first axle counter sensor AZ1 and the curve VL2 of the second axle counter sensor AZ2. For this purpose, the following diagram was selected, in which the measurement signals U1, U2 are shown in the form of output voltages over time t. Figure 2 The lower part of FIG. 5 shows the subsequent processing steps of normalization using the results of the normalized curves NV1 , NV2 , NV3 and the comparison with the patterns P1 , P2 .

[0058] exist Figure 2 In the method shown in Figure 1 An axle counter AZL is used together with a first axle counter sensor AZ1 and a second axle counter sensor AZ2. Rather, it is conceivable to use an axle counter having only one axle counter sensor AZ, wherein, for example Figure 2 , that is, the diagram of curve VL1 and the measures associated therewith (indicated by arrows) will be omitted.

[0059] From the curves VL1 and VL2 it can first be seen that the axle counter sensors AZ1, AZ2 are mounted in the track GL with a transverse offset in the direction of travel. This results in a time offset ZVM between comparable maximum values. This is Figure 2 , this is shown in that the first maximum value M1 generated due to the passage of the first wheel RD of the bogie DG is selected in the first curve VL1 and the second curve VL2.

[0060] Furthermore, it can be seen in the curves VL1, VL2 that they involve the passing of two wheels (axles) of the bogie. This can be seen because in the curves VL1, VL2, in addition to the time-shifted first maximum M1, a further second maximum M2 can also be seen, which is also shifted by the time offset ZVM and has a great similarity to the first maximum M1. The first maximum M1 and the second maximum M2 are each spaced apart from each other by a time offset ZVR between the wheel passing. This time offset ZVR corresponds exactly to the time difference between the wheel passing of the first wheel of the bogie DG and the second wheel RD of the bogie DG.

[0061] In order to generate the normalized curves NV1, NV2, NV3, the method according to the invention carries out a normalization N in a manner not shown in detail. This normalization includes normalizing the amplitudes of the measurement signals U1, U2 to a target value ZW, which is calculated according to Figure 2 In the embodiment of , it is 1. In addition, a dynamic time normalization is performed, wherein the first curve VL1 and the second curve VL2 are respectively taken into account before and after the identified maximum values ​​M1, M2, M3 to such an extent that the curves connected to the maximum values ​​M1, M2, M3 can be characterized (and can be compared with the patterns P1, P2, which will be explained in more detail below). As a result, normalized curves NV1, NV2, NV3 appear in the time windows ZF1, ZF2, ZF3, which correspond to the patterns P1, P2 in terms of their time extension.

[0062] As shown, the evaluation of the first maximum value M1 leads to the generation of a first normalized curve NV1 and the evaluation of the second maximum value M2 leads to the generation of a third normalized curve NV3. Furthermore, the third maximum value M3 can be seen in both the first curve VL1 and the second curve VL2, which leads to the generation of a second normalized curve NV2. The fourth maximum value M4 can only be determined in the second curve VL2 and is therefore excluded for normalization (indicated by X). This can be based on the fact that the fourth maximum value M4 cannot be a wheel passing, because this fourth maximum value must be visible not only in the first curve VL1 but also in the second curve VL2.

[0063] In the last step, a pattern comparison of the normalized curves NV1, NV2, NV3 is performed. Here, it is found that the first normalized curve NV1 and the third normalized curve NV3 are respectively consistent with the first pattern P1 representing the passage of the wheel. This leads to a count result of 2. The second normalized curve NV2 is identified with the help of the second pattern P2 representing the passage of the wheel. Therefore, the normalized curve NV2 is excluded from the count (indicated by X).

[0064] exist Figure 2, it is shown that the mode P1 and the second mode P2 have a confidence range indicated by shaded areas, which allow certain fluctuations in the normalized curves NV1, NV2, NV3. This leads to the situation that the measured curves VL1, VL2 are subject to certain tolerance fluctuations. In addition to the measurement tolerances, it is also necessary to consider that different vehicles generate different measurement signals, which depend on conditions such as, for example, wheel wear of the vehicle.

[0065] List of Reference Numerals

[0066] LZ Control Center

[0067] FZ Vehicles

[0068] DG Bogie

[0069] RD Wheels

[0070] FR Driving Direction

[0071] GL Track

[0072] AT Antenna

[0073] STL Satellite

[0074] AZL axle counter

[0075] AZ, AZ1, axle counter sensor

[0076] AZ2

[0077] AE Evaluation Unit

[0078] CP1…CP2 Computer

[0079] SE1…SE2 Memory Device

[0080] S1…S5 interface

[0081] VL1…VL2 curve

[0082] M1…M4 maximum value

[0083] Normalized curve of NV1…NV3

[0084] ZF1…ZF3 time window

[0085] U1…U2 Measuring signal

[0086] P1…P2 mode

[0087] ZW Target Value

[0088] ZVM Time offset between comparable maximum values

[0089] ZVR Time shift between wheel passes

[0090] 2 Counting results

[0091] X Exclude

Claims

1. A method for counting axes, wherein: The wheel (RD) passes the axle counter sensor installed on the track (GL), The axle counter sensor generates a measurement signal, The curve of the measurement signal is evaluated with the aid of a computer, wherein the wheels (RD) are identified, It is characterized in that When evaluating the measurement signal, Finding at least one maximum of the signal amplitude in the curve of the measured signal, Normalizing the amplitude of the measured signal in amplitude normalization so that the maximum value is the same as a preset target value (ZW), For the curve of the measured signal, a dynamic time normalization is performed before and after the maximum value, Here, the curve of the measured signal normalized by amplitude normalization and time normalization is a first pattern with at least one curve of the measurement signal when the wheel (RD) passes by and a second pattern with at least one curve of the measurement signal when an error occurs In comparison.

2. The method according to claim 1, It is characterized in that Normalized curve of the measured signal by amplitude normalization and time normalization A third pattern with at least one curve of the measurement signal when a single wheel (RD) passes by and a fourth mode with at least one curve of the measurement signal when two wheels (RD) of the bogie (DG) pass by In comparison.

3. The method according to claim 2, It is characterized in that A comparison with a fourth pattern of the course of the measurement signal when two wheels (RD) of a bogie (DG) pass by is performed only if the time shift of the maximum values ​​in the course of the measurement signal does not exceed a limit value predefined as a function of the speed of the vehicle (FZ) passing the axle counter sensor.

4. The method according to claim 1, It is characterized in that The normalized curve of the measurement signal by amplitude normalization and time normalization is compared with at least one fifth pattern of the curve of the measurement signal when the bogie (DG) is traveling sideways while traveling on a curve.

5. The method according to claim 1, It is characterized in that An axle counter (AZL) is used which has a first axle counter sensor (AZ1) and a second axle counter sensor (AZ2) arranged one behind the other in the direction of travel (FR), wherein the method is carried out one after the other for the first axle counter sensor (AZ1) and the second axle counter sensor (AZ2).

6. The method according to claim 5, It is characterized in that A maximum value in a first curve detected by the first axle counter sensor (AZ1) and a maximum value in a second curve detected by the second axle counter sensor (AZ2) are compared, and only the maximum values ​​in the first curve (VL1) and the second curve (VL2) normalized by amplitude normalization and time normalization of the measurement signal, which are present in both the first curve (VL1) and the second curve (VL2), are compared with the first pattern and the second pattern.

7. The method according to claim 5 or 6, It is characterized in that A maximum value in a first curve (VL1) detected by the first axle counter sensor (AZ1) and a maximum value in a second curve (VL2) detected by the second axle counter sensor (AZ2) are compared and, taking into account a time shift between comparable maxima of the first curve (VL1) and the second curve (VL2), curves of the measurement signal before and after the maximum value are determined, which curves are to be taken into account in a dynamic time normalization.

8. A computer program product having program modules for implementing the method according to any one of the preceding claims 1 to 7.

9. A device for providing a computer program product according to claim 8, wherein: The providing device stores and / or provides a computer program product.

Images