EP1453714B1 - Method and system for the detection of objects along a track - Google Patents

Abstract

To determine if a track section (1) is covered by a wagon or foreign object (6) detectors (D1-D5) are arranged as freely selectable intervals along the track. The detectors use an NMR principle with transmitted pulse packets detected by the detectors and response signals converted into the frequency domain using a Fourier transform. An object within range of a detector alters the magnetic field such that largely no response is generated. <??>After Fourier transformation of the response signals, the response signals are assigned to the detectors. Thus the absence of a response signal can be used to indicate the precise location of a possible obstacle on the track. <??>An Independent claim is also included for a corresponding system.

Description

The present invention relates to a method for the detection of objects along a track according to the preamble of patent claim 1.

• i) Streckenabschnitte werden mit sogenannten Achszählern überwacht. Sowohl beim Eintritt eines Zuges in einen Streckenabschnitt wie auch beim Verlassen des betreffenden Abschnittes werden die Achsen gezählt und bei Gleichheit erfolgt eine Freimeldung dieses Streckenabschnittes.
• ii) Speziell kürzere Gleisabschnitte wie z.B. Weichen werden mit sogenannten Gleisstromkreisen gesichert, in dem eine Schiene gegenüber der anderen Schiene elektrisch isoliert ist. Durch das Befahren eines solchen Abschnittes wird durch die Räder einer Achse ein Stromkreis geschlossen und dadurch kann ein Belegtzustand abgeleitet werden.

Evidence of whether a section of track or a section of track is occupied by a railway carriage can be made, for example, by the following means:

• i) Track sections are monitored with so-called axle counters. Both when entering a train in a section of the route as well as when leaving the relevant section, the axes are counted and in case of equality is a free message of this section.
• ii) Specially shorter track sections such as points are secured with so-called track circuits in which one rail is electrically insulated from the other rail. By driving on such a section, a circuit is closed by the wheels of an axle and thereby a busy condition can be derived.

The above techniques work reliably, but they are limited to occupying a track section by railway cars and locomotives. A car that has been stuck eg on a railroad crossing, can not be determined by the means mentioned above. The procedure with the track circuits is in winter service with salt use eg in the area of stations or level crossings subject to interference, since the saltwater located on the route has a relatively low resistance and forms a two-line electrical conduction. This effect of misdetections occurs both in a DC circuit and when using a signal of several kHz. Repeating a measurement within a few seconds does not provide new state information.

The document WO 98/58829 entitled "Vehicle Presence System" discloses a system in which passive magnetic detectors are arranged along a railway track in the track bed in order to determine the track occupancy in the vicinity of a railroad crossing. The detection takes place via a comparison of signals which are stored in advance in the absence of objects and signals which are acquired at a time determined by the detection and with the stored signals. Although this "Vehicle Presence System" is very practical, but associated with a high circuit complexity and is greatly limited in terms of the number of usable induction coils.

Document US 5,868,360 also discloses a method and system for detecting objects along a track.

The present invention is therefore based on the object of specifying a method for the detection of objects along a track, which can be realized with a low circuit complexity, is independent of direct and indirect weather influences and allows monitoring of a route including the immediate environment with scalable resolution.

This object is achieved by the measures specified in claim 1. Advantageous embodiments of the invention are specified in further claims.

• A Antwortsignale auf ein an Detektoren ausgesendetes Pulspaket werden erfasst;
• B die erfassten Antwortsignale werden in ein Signal im Frequenzraum transformiert»;
• C die in den Frequenzraum transformierten Antwortsignale werden mit vorgängig gespeicherten Antwortsignalen verglichen, die bei gesicherter Abwesenheit eines Objektes gemäss den Verfahrensschritten A und B gespeichert wurden, wobei eine festgestellte Ungleichheit als Vorhandensein eines Objektes bewertet wird;

ist ein Verfahren geschaffen, dass unabhängig von den Witterungseinflüssen ist und auf robuste Weise eine Erkennung von Objekten längs und in unmittelbarer Umgebung eines Geleises erlaubt. Das erfindungsgemässe Verfahren ist insoweit ein statisches Verfahren, als zweifelhaften Resultaten eine Prüfung des Geleises auf das Vorhandensein von Objekten mehrmals wiederholt werden kann, ohne dass eine Besichtigung vor Ort vorgenommen werden muss.By the inventive solution after what
«The detectors work according to the method Nuclear Magnetic Resonance (NMR) and through the process steps

• A response signals to a pulse packet sent to detectors are detected;
• B the detected response signals are transformed into a signal in frequency space »;
• C the response signals transformed into the frequency domain are compared with previously stored response signals stored in the assured absence of an object according to the method steps A and B, wherein a detected inequality is evaluated as the presence of an object;

a method is provided which is independent of the weather influences and allows a robust detection of objects along and in the immediate vicinity of a track. The inventive method is insofar a static method, as dubious results, a test of the track on the presence of objects can be repeated several times without a visit must be made on site.

• i) Dadurch dass
  die Detektoren längs des Geleises in frei wählbarer Sequenz positionierbar sind;
  kann das Verfahren optimal auf die mutmasslichen Gefahrenstellen beschränkt werden (Patentanspruch 2).
• ii) Dadurch dass
  die Detektoren neben dem Geleise positionierbar sind; kann ein System zur Durchführung des erfindungsgemässen Verfahrens installiert und gewartet werden, ohne dass der Eisenbahnbetrieb deswegen unterbrochen werden muss (Patentanspruch 3).
• iii) Dadurch dass
  im Verfahrensschritt C die in den Frequenzraum transformierten Antwortsignale den Detektoren zuordenbar ist;
  kann ein durch einen Detektor festgestelltes Objekt genau lokalisiert werden und es kann ein Interventionsdetachement genau an die betreffende Stelle der Eisenbahnstrecke dirigiert werden (Patentanspruch 6).
• iv) Dadurch dass
  dass im Verfahrensschritt C eine Auswahl von in den Frequenzraum transformierten Antwortsignalen auf die in einem festern Zeitraster ausgesandten Pulspakte gespeichert wird;
  lässt sich eine langsam eintretende Veränderung des Systems zur Durchführung des erfindungsgemässen Verfahrens besonders gut erkennen. Dadurch ist einerseits möglich, die ursprünglich gespeicherten Antwortsignale im Frequenzraum mittels eines statistischen Verfahrens zu korrigieren und andererseits kann exakt ein Logbuch über den zustand des genannten System geführt werden (Patentanspruch 8).

This may result in the following additional benefits:

• i) By that
  the detectors can be positioned along the track in a freely selectable sequence;
  the process can be optimally limited to the presumed danger spots (claim 2).
• ii) By that
  the detectors are positionable next to the track; For example, a system for carrying out the method according to the invention can be installed and maintained without the railway operation having to be interrupted for that reason (claim 3).
• iii) By that
  in method step C, the response signals transformed into the frequency domain can be assigned to the detectors;
  For example, an object detected by a detector can be precisely located and an intervention detective element can be precisely directed to the relevant point of the railway track (claim 6).
• iv) By that
  in method step C, a selection of response signals transformed into the frequency domain is stored on the pulse packets transmitted in a fixed time grid;
  a slowly occurring change of the system for carrying out the method according to the invention can be recognized particularly well. This makes it possible, on the one hand, to correct the originally stored response signals in the frequency domain by means of a statistical method, and, on the other hand, exactly one logbook can be maintained about the state of said system (claim 8).

Figur 1a

Mit Detektoren versehener Streckenabschnitt zur Durchführung des erfindungsgemässen Verfahrens

Figur 1b

Zuordnung von Signalen nach erfolgter Transformation in den Frequenzraum;

Figur 2

Aufbau eines Detektors;

Figur 3

Darstellung von ausgesandtem Pulspaket, Signalverlauf für Gradientenspule und Antwortsignal.

The invention will be explained in more detail with reference to the drawing, for example. Showing:

FIG. 1a

Provided with detectors track section for performing the inventive method

FIG. 1b

Assignment of signals after transformation into frequency space;

FIG. 2

Structure of a detector;

FIG. 3

Representation of emitted pulse packet, gradient coil signal and response signal.

FIG. 1a shows a section of the route in the overview. Parallel to a track 1 at a distance a from the track axis 2 metal detectors D ₁ , D ₂ , .., in freely selectable spaces, .., d ₃₄ , d ₄₅ , .. are arranged (not fully shown in Fig. 1a and ) enumerated. These spaces result from the operational requirements of a railway administration; typical sizes for these spaces are 5 m .. 200 m, for the distance a to the track axis 3 to 6 m. The metal detectors - hereinafter referred to as detectors D ₁ , D ₂ , .., are connected to a transmitting / receiving unit 5, wherein a detector D _{n is} geometrically the last detector. The index n stands for the maximum number of detectors, with typical values for n being in the order of 10... 100.

FIG. 2 shows the structure of a detector D ₁ operating according to the method NMR (Nuclear Magnetic Resonance), which has a permanent magnet 10, a gradient coil 11 and a transmitting / receiving coil 14 contains. The permanent magnet 10 is horseshoe-shaped, in the air gap, a water sample 12 is arranged, which is surrounded by both the gradient coil 11 and the transmitting / receiving coil 14. It is to be provided structurally that the B field generated in the transmitting / receiving coil 14 is orthogonal to the B field of the permanent magnet 10. The water sample 12 consists of a glass body filled with distilled water, a typical internal volume (also called measuring volume) of the glass body is of the order of magnitude of 0.5 cm ³ . The configuration of the measurement volume must be such that, on the one hand, the highest possible signal intensity is achieved - this as an advantage of a relatively large measurement volume - and, on the other hand, a large homogeneity of the permanent magnetic field must be ensured via the corresponding measurement volume. The H nuclei of the water molecules H ₂ O are particularly suitable for the NMR method. The so-called Larmor frequency of the H-core is 42MHz / Tesla. The signal intensity of the response signal S _R is significantly determined by the homogeneity of the magnetic field in the air gap. Relatively large metallic objects in the immediate vicinity of the permanent magnet affect the homogeneity of the magnetic field in the air gap, which leads to a line broadening and thus almost or entirely to an extinction of the NMR signal. The structure and mode of action of such a detector D ₁ is referred to by the skilled person as a so-called NMR detector. The individual detectors D ₁ , D ₂ , etc. are individually characterized by their own number of turns N ₁ , N ₂ , .. the gradient coil 11. The gradient coils 11 of the individual detectors D ₁ , D ₂ , etc. are connected in parallel with each other and connected to a gradient amplifier (not shown in FIG. ₁ a). The gradient amplifier can be part of the transmitting / receiving unit 5. The individual transmitting / receiving coils 14 are preferably connected in parallel to the transmitting / receiving unit 5; It is also a serial connection possible.

FIG. 3 below shows a pulse packet S _P (also referred to as transmit pulse) which is emitted by the transmitting / receiving unit 5 to the detectors D ₁ , D ₂ , etc. Immediately before, a rectangular signal S _{A is applied} to the gradient coils 11 by the gradient amplifier. The duration t _{A of} this square wave signal S _A is, depending on the acquisition time or observation time called t _A , at about 2 s. Together with the individual number of turns of the gradient coil 11 can be superimposed for each detector D ₁ , D ₂ , etc. an individual to the permanent magnetic field additional field. In this case, this additional field is matched to the local position of the relevant detector D _k . The tuning can also be carried out in the evaluation unit with correction values. Due to the individual winding number N ₁ , N _{2 of} the gradient coil of the detectors D ₁ , D ₂ arranged along a track 1, a detector-specific frequency band for the response signal S _{R is also produced} on an emitted pulse packet S _P. The pulse width t _P must be chosen so that all hydrogen atoms in the different detectors D ₁ , D ₂ , etc. can be excited simultaneously. This means that the bandwidth of the pulse S _P must be significantly greater than the frequency deviation between the lowest and highest measurement frequency. Typical values for the pulse clock S _p are given below: Carrier frequency: 21 MHz (at B ₀ = 0.5 Tesla) Pulse width: 5 μs

The response signal S _{R of} a detector D to an emitted pulse packet S _P is also shown in FIG. 3 in the time domain. The typical observation time (acquisition time) is 1 s, the amplitude in the range of mV. In the transmitting / receiving unit 5 or in an associated evaluation unit, the response signal S _{R is} mapped into the frequency domain by means of a Fourier transformation. The result of this mapping of the response signals S _R can be seen with a simplified curve S _Rf of FIG. 1b. In a broken scale - but not shown as such - due to the individual number of turns N ₁ , N ₂ , etc. various response signals P ₁ , P ₂ , etc. are shown. It is assumed that in the vicinity of the detector D3 is a metallic. Object 6 is located so that the magnetic field in the water sample so affected that on a transmitted pulse packet in the frequency space no or according to the scale no representable response signal P3 arises. As a result, the information can be derived that the route along the track 1 is occupied and therefore no adjustment of a signal for a passage of a train may be made.

After the explanation of the infrastructure provides an implementation of the inventive method provides that after installation of the aforementioned system along a TRACK IS 1 on emitted pulse packets S _P into the frequency domain transformed the detectors D _1, D ₂ attributable response signals P _1, P ₂ in either the transmitting / receiving unit 5 or stored in an associated evaluation unit. These stored response signals P ₁ , P ₂ , P ₃ , etc. can be assigned a flat rate for all the same or individual tolerance bands. For carrying out the method according to the invention, the previously mentioned rectangular signal for acquisition and the pulse packet S _{P are} emitted in a fixed time frame or as required. The incoming response signals S _R are transformed into the frequency domain and compared with the stored response signal P ₁ , P ₂ , P ₃ , and so on. An inequality or an inequality outside of an aforementioned tolerance band leads to the generation of a signal indicating an occupancy of the relevant track 1 or track section. By emitting in a fixed time frame of, for example, 5 s, it can be provided, in particular, to store a selection of the resulting response signals, for example every 10th or every 50th signal progression S _RF . This storage allows in particular to detect a slow change and either to generate a warning signal regarding the functionality of the system and / or by a statistical method to correct the stored response signals. It should be noted that outliers are not included in this so-called auto-calibration.

The above-described embodiment of the method according to the invention was primarily aimed at the detection of objects 6 outside the railway. However, the proposed method can also be applied to the railway operation itself and the tracking of a train along a track section.

The inventive system and method is not limited to the embodiment described above, possible further evaluation methods, in particular by means of a two-dimensional Fourier transform to eliminate runtime shifts can. This represents the more general case of the embodiment described above, in which the transit time influence is negligible. The method explained above is not limited to the railway technology, but can also be used, for example, in a matrix arrangement of detectors at a distance of, for example, 0.5 m for the detection of metallic objects such as mines.

List of reference numbers and abbreviations used

1

track

2

track axis

5

Transmit / receive unit

6

Object, metallic

10

permanent magnet

11

gradient coil

12

water sample

14

Transmitter / receiver coil

a

Distance of the detectors D ₁ , D ₂ from the track axis

D, D ₁ , D ₂ , D _k , D _n

detectors

d ₃₄

Distance between the detectors D ₃ and D ₄

d ₄₅

Distance between the detectors D ₄ and D ₅

N ₁ , N ₂

Number of turns of the gradient coil of the detector D ₁ , D ₂

NMR

Nuclear Magnetic Resonance

P ₁ , P ₂ ,

Response signal of the relevant detector D ₁ , D ₂ in the frequency domain

S _A

Square wave signal for acquisition

S _P

pulse packet

S _R

Response signal of a detector D in the time domain

S _RF

Waveform in the frequency domain of the totality of the transformed response signals S _R

Claims (8)

1. Method for detecting objects (6) along a track (1) using detectors (D₁, D₂,..) positioned along the track (1), wherein signals from said detectors (D₁, D₂, ..) have been stored previously in the absence of objects and signals from said detectors (D₁, D₂, ..) are recorded at a point in time determined by the detection and compared with the stored signals,
   characterised in that
   the detectors operate according to the Nuclear Magnetic Resonance (NMR) method and by means of the method steps

   A response signals (S_R) to a pulse packet (Sp) sent out to detectors (D₁, D₂, ..) are recorded;

   B the recorded response signals (S_R) are transformed into a signal (S_RF, P₁, P₂, ..) in the frequency chamber;

   C the response signals (S_RF, P₁, P₂, ..) transformed in the frequency chamber are compared with previously stored response signals which were stored in the clear absence of an object (6) according to the method steps A and B, a determined difference being judged as the presence of an object (6).
2. Method according to claim 1,
   characterised in that
   the detectors (D₁, D₂, .. ) can be positioned in an entirely optional sequence (d₃₄, d₄₅, ..) along the track (1).
3. Method according to claim 1 or 2,
   characterised in that
   the detectors (D₁, D₂, ..) can be positioned next to the track (1).
4. Method according to one of claims 1 to 3,
   characterised in that
   each detector (D₁, D₂, .) has a gradient coil (11) surrounding a permanent magnet (10), a transceiver coil (14) and a water sample (12).
5. Method according to claim 4,
   characterised in that
   each detector (D₁, D₂, ..) is characterised by an individual number of windings (N₁, N₂, ..) of the gradient coil (11).
6. Method according to one of claims 1 to 5,
   characterised in that
   in method step C the response signals (S_RF, P₁, P₂..) transformed in the frequency chamber can be assigned to the detectors (D₁, D₂, ..).
7. Method according to one of claims 1 to 6,
   characterised in that
   in method step A pulse packets (Sp) are sent out in a fixed time-frame in order to detect a slowly occurring change.
8. Method according to claim 7,
   characterised in that
   in method step C a selection of the response signals {S_RF, P₁, P₂) transformed in the frequency chamber which are in response to the pulse packets (Sp) sent out in a fixed time-frame is stored.

Images