A method for detecting a crack growth in a component of a plant or a vehicle and a corresponding system

Abstract

The present invention relates to a method for detecting crack growth in a component (3) of a plant or a vehicle during operation, wherein the method comprises: detecting continuously noise present in the component and converting the detected operating noise into a sensor signal; calculating an indicator function K(t) for each sensor signal comprising the following steps: calculating a spectrum, wherein each spectrum comprises amplitudes in a predefined frequency range; determining, for each frequency value (fi) of the predefined frequency range, a difference between the amplitude A(t,fi) of the frequency value and a mean amplitude, and calculating a normed difference based on the difference; and calculating an indicator function, wherein the indicator function is calculated based on the mean of the normed difference about all frequencies within the predefined frequency range; and comparing each indicator function with a threshold value, and signalling an overpassing of the threshold value.

Description

[0001] A method for detecting a crack growth in a component of a plant or a vehicle and a corresponding system

[0002] The present invention concerns a method for detecting crack growth in a component of a plant or a vehicle.

[0003] Further, the present invention relates to a system for detecting crack growth within a component of a plant or vehicle.

[0004] CN 110824007 discloses a method for determining a crack in a tubular pile. For that purpose, a load is applied at an excitation point of the pile top. Then a response of the excitation is recorded.

[0005] WO2017136692 A1 discloses a method for determining cracks in a tube using an electromagnetic acoustic transducer transceiver.

[0006] However, such systems cannot be used during the operation of a power plant, chemical or petro-chemical plant or in otherwise noisy environment like in a vehicle.

[0007] EP 1476733 B1 discloses an algorithm for detecting a loose part in a turbine. It uses the same algorithm as above, without the use of the quantile.

[0008] Object of the invention is to detect crack growth with a better precision, in particular during operation of a plant, for example a power plant, in particular a nuclear power plant, chemical or petro-chemical plant or a vehicle.

[0009] According to one aspect, a method is provided for detecting crack growth in a component of a plant or a vehicle during operation of the plant or the vehicle, wherein the method comprises: detecting continuously, with at least one acoustic sensor coupled to the component, noise present in the component and converting the detected operating noise into a sensor signal; calculating an indicator function K(t) for each sensor signal comprising the following steps: calculating a spectrum with a sliding first time window from the respective sensor signal, wherein each spectrum comprises amplitudes in a predefined frequency range; determining, for each frequency value of the predefined frequency range, a difference between the amplitude of the frequency value and a mean amplitude, and calculating a normed difference based on the difference between the amplitude A(t,fi) of the frequency value and a mean amplitude; and calculating an indicator function, wherein the indicator function is calculated based on the mean of the normed difference about all frequencies within the predefined frequency range; and comparing each indicator function with a threshold value, and signalling an overpassing of the threshold value as an indication of a possible crack growth.

[0010] Further embodiments may relate to one or more of the following features, which may be combined in any technical feasible combination:

[0011] - the sliding first time window for transforming the sensor signals into a spectrum is between 0,5 and 20 ms, in particular around 3 ms;

[0012] - the predefined frequency range corresponds to the ultrasound range, wherein in particular the ultrasound range is between 20 kHz and 1000 kHz, in particular between 50 kHz and 1000kHz, for example between 100 kHz and 400 kHz;

[0013] - the mean amplitude is an arithmetic mean of preceding amplitudes at the respective frequency (1) or is estimated by a recursive approach;

[0014] - the mean amplitude is calculated by the mean of an a-quantile and a 1 - a quantile of the preceding amplitudes within a second time window, wherein for example the a is between 10 and 30%, in particular between 20% and 30 %;

[0015] - a mean of the normed difference in a predefined frequency band around the respective frequencies is calculated, wherein the mean of the normed difference is used as basis for calculating the indicator function, wherein the respective frequencies comprise at least each frequency value of the predefined frequency range;

[0016] - at least two acoustic sensors are used, which coupled to the component at spaced apart coupling locations, in particular in longitudinal direction of the component, wherein crack growth is localized using the time difference between the detected crack growth from each of the acoustic sensors and the distance between the coupling locations;

[0017] - the normed difference is calculated by dividing the difference between the amplitude A(t,fi) of the frequency value and a mean amplitude by a deviation in particular the standard deviation;

[0018] - the component is a tube, a piping or a vessel, for example of a primary cooling circuit or a secondary cooling circuit of a nuclear power plant or wherein the component is a structural component, in particular a shaft, an axle or a frame;

[0019] - the component is made of metal, a metal alloy, a fibre-reinforced material or concrete;

[0020] - the plant is a power plant, in particular a nuclear power plant, a chemical plant or a petro chemical plant; and / or

[0021] - the vehicle is an aircraft, a train or a car. According to another aspect, a system is provided for detecting crack growth in a component of a plant or a vehicle during operation of the plant or vehicle, the system comprising: a least one acoustic sensors adapted to be coupled to the component (3) and to convert detected operating noise into a sensor signal, and a processor coupled to the at least one acoustic sensor, wherein the processor is adapted to calculate an indicator function K(t) for each sensor signal comprising the following steps: calculating (1010) a spectrum with a sliding first time window from the respective sensor signal, wherein each spectrum comprises amplitudes at a predefined frequency range; determining (1020), for each frequency value (fi) of the predefined frequency range, a difference between the amplitude A(t,fi) of the frequency value and a mean amplitude, and calculating a normed difference based on the difference between the amplitude A(t,fi) of the frequency value and a mean amplitude; and calculating (1030) an indicator function, wherein the indicator function is calculated based on the mean of the normed difference about all frequencies within the predefined frequency range; and compare (1040) each indicator function with a threshold value, and signalling an overpassing of the threshold value as an indication of a possible crack growth Further embodiments may relate to one or more of the following features, which may be combined in any technical feasible combination:

[0022] -the system comprises at least two acoustic sensors, wherein the processor is further adapted to determine the location of the crack by calculating the time difference between the trigger signals and based on the distance between coupling locations of the acoustic sensors on the component; and / or

[0023] - each acoustic sensor is adapted to be coupled to the component of the plant or the vehicle, in particular with an acoustic coupling device. According to a further aspect, a computer program product is provided comprising commands for executing the method according an embodiment disclosed herein, when loaded and executed on a processor.

[0024] According to an embodiment, a computer program product may be provided on a physical software product, for example a hard disc, a solid state disc, a CD-ROM, a DVD, comprising the program.

[0025] According to another aspect, a data carrier signal carrying the computer program product according to an embodiment disclosed herein is provided. Embodiments are also directed to the system for carrying out the disclosed methods steps and in particular including apparatus parts and / or devices for performing described method steps.

[0026] The method steps may be performed by way of hardware components, firmware, software, a computer programmed by appropriate software, by any combination thereof or in any other manner.

[0027] In particular, for electronic and / or software means, each of the above listed terms means and encompasses electronic circuits or parts thereof, as well as stored, embedded or running software codes and / or routines, algorithms, or complete programs, suitably designed for achieving the technical result and / or the functional performances for which such means are devised.

[0028] According to other aspects, the present invention relates to computer-readable nonvolatile storage medium, for example a hard disc, a solid state device, a CD-ROM, a DVD, storing a program containing commands for executing the method according an embodiment disclosed herein, when loaded and executed on a processor.

[0029] Further advantages, features, aspects and details are evident from the dependent claims, the description and the drawings.

[0030] The accompanying drawings relate to embodiments of the invention and are described in the following:

[0031] Fig. 1 shows schematically a system for detecting crack growth in a tube of a power plant according to an embodiment;

[0032] Fig. 2 shows signals received by the sensors; and

[0033] Fig. 3 shows schematically a flow chart of a method for detecting crack growth in a tube of a power plant according to an embodiment.

[0034] Fig. 1 shows schematically a system 1 for detecting crack growth in a component 3 of a plant or a vehicle according to an embodiment during operation of the plant or the vehicle. For example, during operation of the plant or the vehicle the sound emitted by the component 3 due to a crack growth is within the noise or slightly above the noise level, so that the signal to noise ratio is very low.

[0035] In some embodiments, the component 3 is a component of a plant, in particular a nuclear power plant, a chemical or petro-chemical plant, an airplane, a train or a car. For example, the component 3 may be a structural component, for example a tube, a piping, a shaft, an axle, a frame, a vessel or the like.

[0036] For example, the component 3 is a piping of a primary or secondary circuit of a nuclear power plant. The system 1 can be also used for detecting crack growth in another component of a plant or vehicle, in particular a component, which are prone to crack initiation and crack growth. This can be for example due to a welding.

[0037] For example, the component 3 is a metal component, a metal alloy component, a fiber-reinforced component, or a concrete component.

[0038] In other embodiments, the component 3 can be a part of an airplane like the support structure of a wing, which can be monitored by sensors located on the wing box, a part of a rotating machine like its shaft, which can be monitored by sensors located on the bearings in case of journal bearings, loaded parts of trains or cars, which can be monitored by sensors located close to these loaded parts, pressurized vessels in the chemical and petro-chemical plants, which can be monitored by sensors located on the vessels.

[0039] The system comprises at least one acoustic sensor 5a, 5b, which is coupled respectively with acoustic coupling devices 7a, 7b to the component 3 at coupling locations or is pressed or glued directly to the monitored component. For example, the acoustic coupling devices 7a, 7b are wave guides. Alternatively, the sensor can be pressed or glued directly to the monitored component. In case of a glue, the acoustic coupling device is provided by the glue. For example, the system comprises at least two, three, four or more acoustic sensors 5a, 5b. The number of acoustic sensors 5a, 5b is not limited. In Figure 1 N acoustic sensors 5a, 5b are shown, which are numbered Si to SN. Each acoustic sensor 5a, 5b is provided for a respective measuring channel. Thus, the system may comprise N measuring channels.

[0040] In an embodiment, the sensors 5a, 5b may be microphones or other types of acoustic transducers. The sensors 5a, 5b convert acoustic waves into charge signals.

[0041] In some embodiments, the sensors 5a, 5b may have a sensitivity range in the ultrasound range.

[0042] According to embodiments in the present invention, the ultrasound range is between 20 kHz and 1000 kHz, in particular between 50 kHz and 1000kHz, for example between 100 kHz and 400 kHz. According to embodiments, the ultrasound range is suitable for detecting crack growth in the monitored components.

[0043] Further, the system 1 comprises for each sensor 5a, 5b and / or each measuring channel a respective amplifier 10a, 10b (Vi to VN), which converts the charge signals of the sensors into voltage sensor signals, which enables the transmission of the analogue sensor signals for a long distance.

[0044] In addition, the system 1 includes an analogue digital converter (ADC) 12a, 12b (ADCi to ADCN) for each sensor and / or each measuring channel for converting the (amplified) analogue sensor signal into a digital sensor signal. For example, in an embodiment, each ADC can convert data with a sampling rate up to 10 MHz.

[0045] Figure 2 a) b) and c) show acquired data in the time domain for three measuring channels 1 to 3.

[0046] The system further includes at least one processor 14 for treating the digital sensor signals. In some embodiments, which may be combined with other embodiments disclosed herein, the ADC 12a, 12b for each measuring channel is included in the at least one processor 14.

[0047] The at least one processor 14 is adapted to use from the acquired data a time sequence in the time domain (Xi(t) to XN(t) in Figure 1 , with t being the time) for each acoustic sensor 5a, 5b or each digital sensor signal. The recorded time sequence in the time domain for each digital sensor signal are recorded for the respective measuring channel 1 to N.

[0048] According to embodiments, the at least one processor 14 is adapted to perform a Fourier transformation of each time sequence Xi(t) to XN(t) for each measuring channel. A first time window for calculating the spectrum is used. For example, the first time window is between 0,5 and 20 ms, in particular around 3 ms.

[0049] The following calculations apply, according to embodiments, for each of the measuring channels.

[0050] For example, the at least one processor 14 may comprise a ring memory for storing a number T of Fourier transforms for each measuring channel. In other embodiments another type of memory can be used to store the number T of Fourier transforms. In an embodiment, this transformation can be, for example, a windowed fast Fourier transformation FFT, where the window is sliding. Each spectrum has the amplitude values A(t,fi), where t is the time, fi is the frequency, and A is the intensity or amplitude. According to embodiments, 1 are discrete frequencies.

[0051] After each calculation of a spectrum a new spectrum is calculated for a further time step. In other words, the (first) time windows of the subsequent calculated spectrums overlap and the first time window is displaced by a time step for each subsequent sliding first time window.

[0052] In an embodiment, the spectrum is not calculated for each sample time. For example, the spectrum is calculated for each Z samples, where Z is between 1 / 8 and % of the length of the window for FFT (in samples), namely the first time window. For example, if the first time window has 1024 samples, the FFT is calculated every 128 or every 256 samples or a value therein between. Therefore, the next FFT is computed in time t + Z * At where At is the sampling period. A number M of discrete frequencies fi of the calculated spectrum is for example between 65 and 3200.

[0053] According to embodiments, the spectrum comprises at least a predefined frequency range, which corresponds to the ultrasound range.

[0054] For example, in an embodiment the amplitudes in the predefined frequency range are retained, in particular the frequencies in the ultrasound range.

[0055] Then, based on each time sequence for each measuring channel an indicator function Ki(t) to KN(t) dependent on the time is calculated. For that purpose, for each time t, at which the spectra are calculated a mean amplitude A(t,fi) and deviation t / , / )) is estimated from n preceding spectra-amplitudes A(tj,fi) at the respective frequency fi, where i is the index of the frequency and has for example a value between 1 and M and tj is the time with time index j, at which the spectra are computed, which means a time t + j * Z * At. For example, to is the current time t and t-i is the time of the last previous computed spectra. The n preceding amplitudes represent the second time window. For example, n represents the prior spectra within 0.2s and 4s. Thus, it allows to adapt to slow changes in the operating noise. This is for example done for each of the plurality of frequency fi of the spectrum.

[0056] For the values of the mean amplitude A (t,fi) and the deviation the standard equations for arithmetic mean and standard deviation may be used: equation (1 ) equation (2)

[0057] Alternatively, recursive formula may be applied to estimate and s(t, / )): equation (3) equation (4) where krecis a parameter depending on the length of the second time window, for example krec= 1 -1 / n. In an embodiment krgc= 0,999.

[0058] Alternatively, quantiles may be used to estimate the mean amplitude / (t^ ^ and the deviation An a quantile Qa(t,fi) of spectra amplitudes A(t, fi) is the value, under which in the observed second time window a% of all values of the frequency fi are below this value. Then, the sliding mean value and the standard deviation is calculated as follows, with an example where a is 25%:

[0059] Also other values for a may be taken. In more general terms, the equations are as follows: where qi-ais the (1 -a) quantile of the normalized Gaussian distribution or normal distribution. The term a is typically 25%; in this case q?5 = 0,67.

[0060] This quantile method can therefore be used to calculate the mean and deviation of a data record without taking account of the values that are located outside the ranges defined by a and (1 - a). As a result, substantially higher amplitudes, such as can occur in an additionally amplified fashion from a superposed burst signal, are not taken into account.

[0061] Then, a normed difference DCt, / )) is calculated: equation (9)

[0062] The normed difference is computed for each of the plurality of frequencies fi of the spectrum, and in particular for each measuring channel and the mean value and the deviation are continuously updated every time step. In another embodiment, the mean values and standard deviations ft) can be determined once for each frequency fi in a calibration run and stored as frequency specific constant values.

[0063] Optionally, the normed difference D ft) can be in addition averaged around the frequency fi in a frequency band comprising 2L+1 frequencies, i.e. comprising the frequencies fi-L,fi-L+i, ... , fi+i_: equation (10)

[0064] For example, L may be between 5 and 15 and may represent for example a frequency range smaller than 50 kHz, for example 6 kHz. For example, the average or mean around the frequency fi is calculated at least for each frequency value of the predefined frequency range.

[0065] This additional computing step leads to a reduction in the level and breadth of fluctuation of the normalized deviation in regions, in which only background signals are present.

[0066] The indicator function K(t) (in particular for each measuring channel 1 to N or each sensor signal) is calculated, for example by the at least one processor 14, by the squaring and summing the normed difference about a selected band of discrete frequencies 1 and then, optionally, by calculating the square root. According to an embodiment, alternatively S(t) is used as indicator function. The calculations are performed for example in the box g an index corresponding to a minimal frequency and Mmax being an index corresponding to a maximal frequency, wherein the frequency range corresponds to the predefined frequency range, in particular in which the crack growth can be detected with the best sensitivity.

[0067] In another embodiment, it is also possible for the indicator function to be formed as the difference between K(t) and a sliding time mean X(t) of K(t): Then

[0068] B(t) is the indicator function.

[0069] Figure 2 d), e), f) show the indicator function Ki(t), K3(t) and K3(t) for the signals Xi (t), X2 ) and X3(t) of the measuring channels 1 to 3. The indicator functions are plotted against the time. As it can be seen, the curves 2 d) and e) and f) exhibit a significant pulse, which corresponds to crack growth. In signal 1 and 2, the pulse can be seen also in the raw time signal Xi (t) and X2(t), but in channel 3 the pulse is not recognizable in the raw time signal X3(t), but only in indicator function K3(t). As indicated above, alternatively the functions S(t) or B(t) may be used instead of K(t).

[0070] The indicator function K(t) is compared, for example in a comparator, with a prescribed threshold value Ko. Overshooting of the threshold value Koserves as an index for the presence of a pulse-type signal component caused by transient mechanical effects like crack growth and generates a corresponding trigger signal TS. The choice for the threshold value depends on the application (material, geometry, environmental conditions etc.).

[0071] The trigger signal TS is fed to a transient recorder 16 in which the received data are recorded for a predefined time window. The data stored by the transient recorder 16 are then analysed by an analysing module 18. According to embodiments, also the spectra and the acquired data are stored by the transient recorder 16.

[0072] For example, if two or more acoustic transducers are used, the location of crack growth can be determined, for example if crack is located between the two or more acoustic transducers, by calculating the time difference between the trigger signals and based on the distance between the coupling locations on the component 3. In an embodiment, this is performed by the analysing module 18. For example, for the calculation of the location of the crack growth, other parameters can be used compared than to the calculation of the trigger signal. For example, the first sliding window, the second sliding window, and / or Mmin and Mmax for calculating the indicator function could be different. For that purpose, the analysing module 18 may use each time sequence Xi (t) to XN(t) for each measuring channel prior and perform the different steps discussed above with other parameters. For example, the first FFT window length can be reduced from 3.2ms to 0.8ms and quantiles can be used instead of standard deviation and mean.

[0073] Figure 3 shows a flow chart of a method according to an embodiment of the invention.

[0074] In step 1000 the noise in a component is detected continuously for example with the at least one acoustic sensor 5a, 5b coupled to the component 3. The detected operating noise is converted into a sensor signal, for example an analogue and / or digital sensor signal.

[0075] In step 1010, for each sensor signal, a spectrum is calculated with a sliding first time window from the respective sensor signal. The spectrum comprises amplitudes in a predefined frequency range. The sliding windows for transforming the sensor signals into a spectrum is between 0,5 and 20 ms, in particular around 3 ms. The predefined frequency range corresponds to the ultrasound range as indicated above.

[0076] In box 1020 for each frequency value (1), a difference between the amplitude A(t,fi) of the frequency value and a mean amplitude is determined, and a normed difference based on the difference is calculated.. The mean amplitude is an arithmetic mean of preceding amplitudes at the respective frequency (fi) or its estimation (e.g. by the recursive formula). Alternatively, the mean amplitude is calculated by the mean of an a- quantile and a 1 -a quantile of the preceding amplitudes within a second time window, wherein for example the a is between 10 and 30%, in particular between 20% and 30 %. In an embodiment, a mean of the normed difference in a predefined frequency band around the respective frequencies is calculated, wherein the mean of the normed difference is used as basis for calculating the indicator function, wherein the respective frequencies comprise at least each frequency value of the predefined frequency range. The frequency band is for example smaller than the frequency band of the predefined frequencies, for example, it is smaller than 50 kHz, for example 6 kHz.

[0077] In step 1030 an indicator function is calculated, wherein the indicator function is calculated based on the mean of the normed difference about all frequencies within the predefined frequency range. For example, the indicator function is the square route of the mean of the normalized difference about all frequencies. In step 1040 the indicator function is compared, for example in by the analysing module 18, with a threshold value, and signalling an overpassing of the threshold value as an indication of possible crack growth.

Claims

CLAIMS1 . A method for detecting crack growth in a component (3) of a plant or a vehicle during operation of the plant or the vehicle, characterized in that the method comprises: detecting (1000) continuously, with at least one acoustic sensor (5a, 5b) coupled to the component (3), noise present in the component and converting the detected operating noise into a sensor signal; calculating an indicator function K(t) for each sensor signal comprising the following steps: calculating (1010) a spectrum with a sliding first time window from the respective sensor signal, wherein each spectrum comprises amplitudes in a predefined frequency range; determining (1020), for each frequency value (fi) of the predefined frequency range, a difference between the amplitude A(t,fi) of the frequency value and a mean amplitude, and calculating a normed difference based on the difference between the amplitude A(t,fi) of the frequency value and a mean amplitude; and calculating (1030) an indicator function, wherein the indicator function is calculated based on the mean of the normed difference about all frequencies within the predefined frequency range; and comparing (1040) each indicator function with a threshold value, and signalling an overpassing of the threshold value as an indication of a possible crack growth.

2. The method according to claim 1 , wherein the sliding first time window for transforming the sensor signals into a spectrum is between 0,5 and 20 ms, in particular around 3 ms.

3. The method according to claim 1 or 2, wherein the predefined frequency range corresponds to the ultrasound range, wherein in particular the ultrasound range is between 20 kHz and 1000 kHz, in particular between 50 kHz and 1000kHz, for example between 100 kHz and 400 kHz.

4. The method according to one of the preceding claims, wherein the mean amplitude is an arithmetic mean of preceding amplitudes at the respective frequency (fi) or is estimated by a recursive approach.

5. The method according to one of the preceding claims 1 to 3, wherein the mean amplitude is calculated by the mean of an a-quantile and a 1 - a quantile of the precedingamplitudes within a second time window, wherein for example the a is between 10 and 30%, in particular between 20% and 30 %.

6. The method according to one of the preceding claims, wherein a mean of the normed difference in a predefined frequency band around the respective frequencies is calculated, wherein the mean of the normed difference is used as basis for calculating the indicator function, wherein the respective frequencies comprise at least each frequency value of the predefined frequency range.

7. The method according to one of the preceding claims, wherein at least two acoustic sensors (5a, 5b) are used, which coupled to the component at spaced apart coupling locations, in particular in longitudinal direction of the component (3), wherein crack growth is localized using the time difference between the detected crack growth from each of the acoustic sensors and the distance between the coupling locations.

8. The method according to one of the preceding claims, wherein the normed difference is calculated by dividing the difference between the amplitude A(t,fi) of the frequency value and a mean amplitude by a deviationin particular the standard deviation.

9. The method according to one of the preceding claims, wherein the component (3) is a tube, a piping or a vessel, for example of a primary cooling circuit or a secondary cooling circuit of a nuclear power plant or wherein the component is a structural component, in particular a shaft, an axle or a frame.

10. The method according to one of the preceding claims, wherein the component is made of metal, a metal alloy, a fibre-reinforced material or concrete.11 . The method according to one of the preceding claims, wherein the plant is a power plant, in particular a nuclear power plant, a chemical plant or a petro chemical plant.

12. The method according to one of the preceding claims, wherein the vehicle is an aircraft, a train or a car.

13. A computer program product is provided comprising commands for executing the method according an embodiment disclosed herein, when loaded and executed on a processor.

14. A system for detecting crack growth in a component (3) of a plant or a vehicle during operation of the plant or vehicle, the system comprising: a least one acoustic sensors adapted to be coupled to the component (3) and to convert detected operating noise into a sensor signal, and a processor coupled to the at least one acoustic sensor, wherein the processor is adapted to calculate an indicator function K(t) for each sensor signal comprising the following steps: calculating (1010) a spectrum with a sliding first time window from the respective sensor signal, wherein each spectrum comprises amplitudes at a predefined frequency range; determining (1020), for each frequency value (fi) of the predefined frequency range, a difference between the amplitude A(t,fi) of the frequency value and a mean amplitude, and calculating a normed difference based on the difference between the amplitude A(t,fi) of the frequency value and a mean amplitude; and calculating (1030) an indicator function, wherein the indicator function is calculated based on the mean of the normed difference about all frequencies within the predefined frequency range; and compare (1040) each indicator function with a threshold value, and signalling an overpassing of the threshold value as an indication of a possible crack growth.

15. A system according to claim 14, comprising at least two acoustic sensors (5a, 5b), wherein the processor is further adapted to determine the location of the crack by calculating the time difference between the trigger signals and based on the distance between coupling locations of the acoustic sensors (5a, 5b) on the component (3).

16. The system according to claim 14 or 15, wherein each acoustic sensor (5a, 5b) is adapted to be coupled to the component of the plant or the vehicle, in particular with an acoustic coupling device (7a, 7b).

17. The system according to any one of the claims 14 to 16, wherein the system is adapted to perform the method according to any one of the claims 1 to 12.

Images