System for determining wheel-rail adhesion values ​​for rail vehicles

Abstract

A system for determining an adhesion value between a wheel (W) and a track (R) of a rail vehicle (RV) is described, comprising: an optical information acquisition device (3) arranged to acquire optical information from the track; a database (5) arranged to store the acquired optical information at least during a learning phase and to associate the acquired optical information with a corresponding actual wheel-track adhesion measured when the optical information was acquired; a control unit (7) arranged to determine a current wheel-track adhesion value based on a comparison between currently acquired optical information and optical information previously stored in the database (5). The control unit (7) determines that the current wheel-track adhesion value associated with the currently acquired optical information corresponds to the adhesion value associated with the optical information stored in the database (5) and having the highest similarity to the currently acquired optical information.

Description

Technical Field

[0001] The present invention generally relates to the field of rail vehicles, and in particular to a system for determining a wheel-rail adhesion value of a rail vehicle. Background Art

[0002] Existing technologies for measuring wheel-track adhesion are essentially "indirect" solutions based on vision or reflected light sensors, or "direct" solutions implementing braking or traction control specifically for estimating adhesion. The second category includes, for example, techniques that apply a known braking force to one or more axles and detect their speed response. Using the braking force and the angular acceleration, it is easy to infer the adhesion available in the wheel-track contact.

[0003] Disadvantageously, however, optical solutions based on vision or reflection are generally not used because they must be properly pre-calibrated through complex calibration processes, which are however susceptible to the ambient conditions of visibility and lighting in which the system operates.

[0004] On the other hand, direct adhesion measurement, while reliable, can only be performed when a braking torque is applied to the axle. Therefore, the measurement can only be performed during the braking phase unless a dedicated brake control is implemented, for example, by periodically applying a known braking force and detecting the angular acceleration response of the axle. The disadvantage is that specialized control solutions allow adhesion measurement with the required temporal and spatial resolution, but require that a braking force be periodically applied to at least one axle of the rail vehicle, which is not necessary for the vehicle to move. Summary of the invention

[0005] The object of the present invention is to provide a system for determining wheel-track adhesion values ​​of a rail vehicle, which system is capable of self-calibration during use, for example based on artificial intelligence techniques. The system for determining wheel-track adhesion values ​​of a rail vehicle is based on the results of calibrating an optical sensor on the basis of direct adhesion measurements, and thanks to the above system, the interpretation of indirect measurements based on optical sensors will become increasingly reliable, thanks to the self-calibration process of the system, which will accumulate a series of events, namely direct adhesion measurements. The reliability of the indirect measurements, obtained by a properly calibrated system, will allow reliable and continuous measurement of wheel-track adhesion, while at the same time will allow reducing or eliminating the need for braking solely for measuring adhesion.

[0006] According to one aspect of the invention, the above and other objects and advantages are achieved by a system for determining wheel-rail adhesion values ​​of a rail vehicle having the features defined in claim 1. Preferred embodiments of the invention are defined in the dependent claims, the contents of which are to be understood as an integral part of the present description. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] The functions and structural features of some preferred embodiments of the system for determining the wheel-rail adhesion value of a rail vehicle according to the present invention will now be described with reference to the accompanying drawings, wherein:

[0008] Figure 1 An embodiment of a system for determining a wheel-rail adhesion value of a rail vehicle is shown; and

[0009] Figure 2 Another embodiment of a system for determining a wheel-rail adhesion value of a rail vehicle is shown.

[0010] Figure 3A and 3B An example of direct adhesion measurement by applying a braking force to the axle is shown respectively. DETAILED DESCRIPTION

[0011] Before describing in detail a number of embodiments of the present invention, it should be clear that the application of the present invention is not limited to the construction details and configurations of the components presented in the following description or shown in the accompanying drawings. The present invention can adopt other embodiments and can be implemented or constructed in practice in different ways. It should also be understood that the words and terms have descriptive purposes and should not be interpreted as limiting. The use of "including" and "comprising" and their variations should be understood to include the elements listed below and their equivalents, as well as additional elements and their equivalents.

[0012] First reference Figure 1 , a system for determining adhesion values ​​between wheels W and a rail R of a rail vehicle RV is shown.

[0013] The system comprises an optical information acquisition device 3 for acquiring optical information from a track R and a database 5 for storing the optical information at least during a learning phase and associating the optical information with a corresponding actual adhesion value between the wheel W and the track R, the actual adhesion value being actually measured when the optical information acquisition device 3 acquires the optical information. Obviously, the database 5 may be configured to store the optical information and associate the optical information with a corresponding actual adhesion value between the wheel W and the track R, the actual adhesion value being actually measured when the optical information acquisition device 3 acquires the optical information, even during use after the initial learning phase, so as to continue training the system even during operational use of the system.

[0014] In other words, in the database 5 , the optical information may be a function of the corresponding actual adhesion value between the wheel W and the rail R.

[0015] The actual adhesion value between the wheel W and the rail W can be measured by braking the axles of the rail vehicle RV and analyzing the behavior of the axles. The situation where only one axle is braked does not affect the behavior of the rail vehicle, because a single braked axle is not sufficient to influence the dynamics of the movement of the rail vehicle. In this way, the actual wheel-rail adhesion value can be measured without negatively affecting the speed of the rail vehicle and the comfort of the passengers on the rail vehicle. Preferably, the axles of the rail vehicle are braked until the axle begins to slip or reaches a predetermined limit value.

[0016] like Figure 3A and Figure 3B , two examples of direct adhesion measurement are shown by applying a (dedicated) braking force to the axle.

[0017] In this figure, 300 represents the maximum braking force for an axle that may be required for the exemplary rail vehicle in question. Typically, the maximum braking force is the braking force applied during emergency braking.

[0018] The process of direct adhesion measurement includes applying a braking force 302 to one or more axles in a gradually increasing manner. Figure 3A As shown in the figure, if the maximum braking force of 300 is reached, or if Figure 3B As shown, if the axle in question begins to slip, ie the tangential velocity 304 of the axle deviates from the vehicle's travel velocity 306, the gradual increase is interrupted, returning the braking force to zero.

[0019] exist Figure 3A It is not possible to quantify wheel-track adhesion in the case of , but it can be said that the wheel-track adhesion is sufficient to apply maximum braking force, such as emergency braking. Typically, the adhesion for emergency braking is about 0.15. Therefore, in Figure 3A In this case, it can be said that the wheel-track adhesion is greater than, for example, 0.15.

[0020] In any case, the adhesion measurement is not particularly important in cases where it exceeds the adhesion required for emergency braking.

[0021] Or, in Figure 3B In the case of a braking force, the wheel-track adhesion is insufficient to apply the maximum braking force and once a certain braking force value is reached as shown by the dashed line, the axle starts to slip. At this point, the braking force is immediately reset to zero and the axle resumes the vehicle's speed.

[0022] Adhesion can be calculated using the following relationship:

[0023]

[0024] in:

[0025] μ is the wheel-track adhesion;

[0026] F brake is the braking force applied to the axle;

[0027] R is the wheel radius;

[0028] J is the inertia of the axle;

[0029] is the angular acceleration of the axle;

[0030] M is the weight on the axle;

[0031] G is the acceleration due to gravity.

[0032] like Figure 1 As shown, the database 5 can be arranged on the rail vehicle, or as Figure 2 As shown, the database 5 may be remotely arranged relative to the rail vehicle. If the database 5 is remotely arranged, the system will be able to communicate with the database via appropriate wireless communications.

[0033] The system further comprises a control unit 7 configured to determine a current wheel-rail adhesion value based on a comparison between currently acquired optical information and optical information previously stored in the database 5 .

[0034] Advantageously, based on the currently determined wheel-track adhesion value, a recommendation can be provided to the operator via a suitable human-machine interface arranged in the driver's cab. Alternatively or additionally, based on the currently determined wheel-track adhesion value, a braking system of the rail vehicle and / or a WSP system of the rail vehicle and / or an acceleration level of the rail vehicle can be automatically controlled.

[0035] The control unit 7 may be, for example, a PLC, a microprocessor, a microcontroller or an FPGA.

[0036] As shown in the figure, the optical information acquisition device 3 may be arranged in the rail vehicle to acquire optical information of the rail located in front of the first axle A1 of the rail vehicle according to the traveling direction V of the rail vehicle.

[0037] In this way, the optical information obtained is not disturbed by the wheels of the rail vehicle passing over the track, which will facilitate cleaning the track or in any case modifying the condition of the track.

[0038] The control unit 7 is configured to determine that the current wheel-rail adhesion value associated with the current optical information acquired by the optical information acquisition device 3 corresponds to an adhesion value associated with optical information stored in the database 5 and having a higher similarity to the current optical information.

[0039] Obviously, the similarity can be determined by the control unit 7 through a deep learning algorithm.

[0040] "Deep learning" refers to a field of study in machine learning and artificial intelligence based on different levels of representation. Deep learning is a set of techniques based on artificial neural networks organized in different layers, where each layer calculates the value of the subsequent layers, making the information processing more and more complete.

[0041] Image recognition is a branch of computer science called "computer vision".

[0042] Algorithms for image recognition can be used to define similarity. In a broad sense, automatically recognizing images means using algorithms that receive an input image and are able to extract various information from it. The information extracted can be divided into several levels: low-level (e.g. statistics of the presence of various shades of gray or color, sudden changes in brightness, etc.); mid-level (features associated with image regions and relationships between regions), or high-level (determination of semantically meaningful objects). Two images can be identified as similar even based only on low-level, mid-level, or high-level features.

[0043] Preferably, the optical information acquisition device 3 is a visual or image acquisition sensor, such as a video camera or a still camera.

[0044] Alternatively, the optical information acquisition device 3 may be a reflection-based optical sensor. In this case, the reflection-based optical sensor is arranged to send light signals to the track and detect the quantity and / or distribution of light signals reflected from the track to indicate the condition of the surface of the track.

[0045] In order to calibrate the system, the database 5 may store in advance a predetermined amount of known optical information, which is associated with corresponding known adhesion values, the correctness of which has been verified in advance. The stored information will then be enhanced and updated during normal operation of the rail vehicle. For example, the optical information acquisition device 3 may acquire optical information at determined time intervals or when adhesion degradation is detected.

[0046] Now, to give a practical example, in a first stage of filling the database, a determined amount of optical information is acquired by the optical information acquisition device 3, which is associated with a corresponding actual wheel-track adhesion value, which is actually measured when the optical information acquisition device 3 acquires the corresponding optical information. In a second operation step, the optical information acquisition device 3 will pick up one or more additional items of optical information, by comparing these additional items with the optical information previously stored in the database 5, to determine the current wheel-track adhesion value. In particular, an adhesion value corresponding to the current wheel-track adhesion value is determined, which adhesion value is associated with the optical information stored in the database 5 and has a higher similarity with the additional current optical information.

[0047] The advantage achieved is thus that a system for determining wheel-rail adhesion values ​​of a rail vehicle is provided which is capable of self-calibration in an increasingly reliable manner, since the self-calibration of the system will accumulate a series of past events. Thanks to a properly calibrated system, the interpretation of indirect measurements based on optical sensors will be increasingly reliable, thus allowing the need for brakes dedicated only to measuring adhesion to be reduced or eliminated.

[0048] Having described various aspects and embodiments of the system for determining the wheel-track adhesion value of a rail vehicle according to the present invention, it should be understood that each embodiment can be combined with any other embodiment. In addition, the present invention is not limited to the described embodiments, but can be varied within the scope defined by the appended claims.

Claims

1. A system for determining the adhesion value between a wheel (W) and a rail (R) of a railway vehicle (RV), comprising: An optical information acquisition device (3) is configured to acquire optical information from the track (R); a database (5) configured to store the acquired optical information at least during a learning phase, and to associate the acquired optical information with corresponding actual wheel-track adhesion values, the actual wheel-track adhesion values ​​being actually measured when the optical information acquisition device (3) acquires the optical information; a control unit (7) configured to determine a current wheel-track adhesion value based on a comparison between currently acquired optical information and optical information previously stored in the database (5); The control unit (7) is configured to: determine that the current wheel-track adhesion value associated with the current optical information acquired by the optical information acquisition device (3) corresponds to an adhesion value associated with optical information stored in the database (5) and having the highest similarity to the currently acquired optical information; and The control unit (7) is configured to determine the optical information stored in the database (5) and having the highest similarity with the current optical information through an image recognition algorithm based on artificial intelligence and / or machine learning.

2. The system according to claim 1, wherein: The optical information acquisition device (3) is a visual sensor.

3. The system according to claim 2, wherein: The visual sensor includes a video camera or a still camera.

4. The system according to claim 1, wherein: The optical information acquisition device (3) is a reflection-based optical sensor.

5. The system according to claim 4, wherein: The reflection based optical sensor is arranged to transmit a light signal to the track (R) and detect the amount and / or distribution of the light signal reflected by the track to indicate the condition of the surface of the track.

6. A system according to any one of the preceding claims, wherein: The database (5) is arranged remotely from the rail vehicle.

7. The system according to any one of claims 1 to 5, wherein: The optical information acquisition device (3) is arranged in the rail vehicle (RV) to acquire optical information of a rail (R) located in front of a first axis (A1) of the rail vehicle according to a traveling direction (V) of the rail vehicle.

8. The system according to any one of claims 1 to 5, wherein: The database (5) is configured to store a predetermined amount of known optical information, the predetermined amount of known optical information being associated with corresponding known adhesion values, the correctness of the known adhesion values ​​having been verified in advance.

9. The system according to any one of claims 1 to 5, wherein: The actual wheel-rail adhesion value is measured by analyzing the behavior of the axle of the rail vehicle during a braking phase.

10. The system according to claim 9, characterized in that The axle of the rail vehicle is braked until the axle begins to slip or reaches a predetermined limit value.

Images