WO2015169277A1

WO2015169277A1 - Method for carrying out spatial and/or time resolution of a first underwater sound source from a second underwater sound source and associated device and watercraft

Info

Publication number: WO2015169277A1
Application number: PCT/DE2015/100098
Authority: WO
Inventors: Jörg HURKA; Kevin Brinkmann
Original assignee: Atlas Elektronik Gmbh
Priority date: 2014-05-07
Filing date: 2015-03-11
Publication date: 2015-11-12
Also published as: DE102014106351A1

Abstract

The invention relates to a method for carrying out spatial and/or time resolution of a first underwater sound source from a second underwater sound source, and the first underwater sound source and the second underwater sound source each emit and/or reflect an underwater sound signal and the method comprises picking up the reflected and/or emitted underwater sound signals in a frequency range, dividing the frequency range into at least two frequency bands and determining a signal amplification profile with spatial and/or time resolution for each frequency band.

Description

Method for spatially and / or temporally dissolving a first underwater sound source from a second underwater sound source and associated apparatus and

water craft

The invention relates to a method for location and / or time resolution of a first underwater ^{sound source} from a second underwater ^{sound source} , wherein the first underwater sound source and the second

Underwater sound source each emit and / or reflect an underwater sound signal and a device, which performs the method and an associated watercraft, which at least one hydrophone or a Hydrophonanordnung and a computer and has.

[02] Target tracking with active or passive sonars attempts to locate individual vessels or underwater vehicles and record the movement of these underwater vehicles.

[03] To estimate the price, the signals are monitored over time and displayed accordingly. In the ideal case, a continuous graph is obtained, which corresponds to the course of the object to be tracked.

However, if several vessels are pursued in parallel, which are in particular very close to each other or whose courses intersect, this can sometimes no longer be resolved with the known active and passive sonars. Also it can happen that in one Overlapping area several meaningful courses, for passive sonar bearings, for a vehicle to be displayed. This is rather unsatisfactory because it is not always clear what the true heading / bearing of the vessel is.

[05] The task is to improve the state of the art.

[06] The problem is solved by a method for location and / or time resolution of a first

Underwater sound source from a second

Underwater sound source, the first

Underwater sound source and the second

Underwater sound source each emit and / or reflect an underwater sound signal, the method having the following steps

Recording the reflected and / or emitted underwater sound signals in a frequency range,

- Dividing the frequency range in at least two frequency bands and

Determining a spatially and / or time-resolved signal strength curve for each frequency band.

By subdividing the frequency range into at least two frequency bands, two underwater sound sources, in particular in the frequency band with the higher frequencies, can be separated from one another, which are not resolved in an overall representation in the (total) frequency range. [08] In particular, by evaluating the frequency bands, the course of the individual underwater sound source can be tracked clearly longer even with intersecting courses of the underwater sound sources.

[09] The following is explained:

[10] An "underwater sound source" can be any natural or artificial object which

Underwater sound waves emitted or reflected. In particular, these may include ships, buoys, submarines, AUV ^ s

(Autonomous Underwater Vehicles) or ROVs (Remotely Operating Vehicles). Stationary structures such as oil platforms or buoys are also included. In particular, one or both of the underwater sound sources can move in or on the water.

[11] The underwater sound signal is in particular a signal emitted by the underwater sound source, such as a screw sound. Also reflected signals, such as an active sonar, which are sent in the direction of the underwater sound source and sent back from the underwater sound source are included. The underwater sound signal is in particular by a hydrophone or a hydrophone arrangement

(Hydrophone array) recorded. Such a hydrophone, for example, for this purpose, electrical elements (eg piezo-ceramics) have, which oscillate according to the underwater sound signals and generate an electronic signal, which is converted accordingly. [12] It is also possible to use a hydrophone array or several hydrophones instead of just one hydrophone, so that a direction determination can be realized.

[13] "Spatial resolution" is understood in particular to mean that, for example, two underwater sound signals can be separated over an angular range.

[14] A "time resolution" is to be understood in particular that the time-varying positions, which correspond to a course or a bearing course (variation of the bearing), the individual

Underwater sound sources can be assigned.

[15] "Recording of the reflected and / or emitted underwater sound signals" is to be understood in particular as meaning the measurement by means of a hydrophone or a hydrophone arrangement and the electronic evaluation by means of a corresponding electronic system and an optionally downstream computer.

[16] The "frequency range" is in particular the frequency range, which is measurable or measured with the hydrophone, which basically has a lower limit and an upper limit, depending on the type of hydrophone the frequency range is for example between a few Hz and several kHz.

[17] When "dividing the frequency range", the frequency range is divided in particular on the basis of the frequency or the wavelength. [18] The "signal strength curve", for example, the (sound) energy (-einhüllende) of the respective

Frequency band over the place (for present text also synonymous for bearing) determined. Especially with an all-round display (360 °), the sound intensity is determined by the angle.

[19] The "time-resolved signal strength curve" can be used to determine the course of the underwater sound source for a detected underwater sound source, which can be determined individually for each frequency band the other frequency band are determined.

[20] In another embodiment, dividing the frequency band range into three, four, five, six, seven, eight, nine, ten, eleven, twelve, thirteen, fourteen, fifteen, sixteen, seventeen, eighteen, nineteen, twenty or more frequency bands provided. Thus, advantageously, the frequency range can be finely divided.

[21] In particular, by the effect that

Underwater noise is frequency dependent attenuated by the water, so can be done on the individual frequency bands a more detailed time resolution and / or local resolution of the underwater sound sources from each other.

[22] In order to achieve a higher sound energy contribution or to leave certain frequency ranges unimportant, For example, dividing the frequency range may be such that adjacent frequency bands form a frequency band gap or frequency band overlap. Thus, in turn, an improved method for time resolution and / or spatial resolution of two underwater sound signal sources can be provided from each other.

[23] If an antenna with a (total) frequency bandwidth of 20 Hz to 200 kHz is used and twenty frequency bands are determined, each frequency band may comprise a bandwidth of 20 kHz. In this case there is a (very low) "frequency band overlap".

[24] For an antenna with one (total

) Frequency bandwidth from 20 Hz to 200 kHz twenty frequency bands each with 5 kHz can be provided, each having no overlaps but each have a frequency band spacing. It is also possible that partially a frequency band overlap and partly a frequency band spacing is used.

[25] In another embodiment, two, three or more frequency bands are grouped into a first frequency band group, the grouped frequency bands being adjacent to one another in particular.

[26] This can be particularly advantageous if the sound energy which is represented in a frequency band by means of the signal strength curve is not clearly suitable for time resolution and / or spatial resolution due to the unfavorable signal-to-noise ratio. By the Combining multiple frequency bands can provide a better signal-to-noise ratio, which then allows for better location and / or time resolution.

[27] For the signal-to-noise ratio of several

To improve frequency bands, a second frequency band group, a third frequency band group and / or other frequency band groups may be formed by the grouping.

[28] In "grouping", for example, the respective signal strengths can be added depending on the position (for example, angle-dependent).

[29] Frequently higher frequencies allow a better local resolution of two underwater sound signals, one of the frequency band groups can be successively formed, in particular from high frequencies to low frequencies.

For example, the frequency bands with the two highest frequency bandwidths can become one

Frequency band group are summarized. In the present case, if the signal-to-noise ratio is not yet sufficient, the frequency band with the third highest frequency bandwidth can be added to the frequency band group. If this is still not enough, this can be done with the fourth highest frequency bandwidth and so on.

[31] In another embodiment, the signal strength characteristic of one of the frequency bands and / or one of the frequency band groups is the first Underwater sound source and the second

Underwater sound source determined. By this determination, the underwater sound sources can be separated from each other.

In order to be able to provide a particularly simple method for determining the first underwater sound source and the second underwater sound source, the first underwater sound source and the second underwater sound source can be determined by means of a maximum determination and / or by means of a mathematical adaptation of two basic functions to the signal strength profile.

In the case of the "maximum determination", in particular two local maxima, which are adjacent to one another, can each be assigned to an underwater sound source, which represents a particularly simple method.

[34] In cases where this can not be done, for example, mathematical fitting can be done using two basic functions. For example, two Gaussian curves, whose maxima are close to each other, overlappingly form part of the signal strength curve, which does not form separate, local maxima. This overlap function only has a maximum. If this total function is adapted mathematically (fitting is also called "fitting"), for example, two assumed Gaussian functions can form the overlapping function descriptively or mixtures of the two, eg Voigtfunktionen, which arise from a convolution of the Gaussian curves and the Lorentz curve, the overlap function form.

[35] In order for a human operator to track the time-resolved and / or spatially-resolved subsonic sound signals, one of the signal strength curves as a function of location or the two subsonic sound signals can be displayed as a function of location and time. As a display element, for example, in the present case serve a display or a screen.

In a further embodiment, the first underwater sound source and / or the second

Underwater sound source, in particular by means of a

Tracking algorithm, tracked.

[37] This can be used to conclude a future course or the actual course of one of the underwater sound sources.

[38] Tracking can be done by extrapolation. Thus, an unresolved and, for example, interrupted signal with high probability can be assigned. For example, frayed signal end lanes are compared with frayed signal start lanes, and the most probable lane is selected by extrapolating the ends and checking for an overlap. Tracing is also called "tracking".

[39] To determine the most probable course (for tracking the entire text synonymously with the bearing course), the tracking can be carried out for individual frequency bands or for all frequency bands and / or for individual frequencies

Frequency band groups or for all frequency band groups.

Thus, in the high-frequency frequency bands, the track can be analyzed by the well-resolved maxima of the underwater sound signals, while at lower frequencies, frequency bands are combined into frequency band groups and the course determination

(Beilungsverlaufermittlung) additionally for this takes place. By comparing and averaging the results, an optimized result can be obtained.

[41] In the inventive method, the tracking (SHT process) and / or by means of a multi ^¬ hypothesis tracker process (MHT method) can be performed by a single hypothesis tracker method.

[42] These tracking methods (SHT, MHT) are well-suited methods that can be used for individual frequency bands as well as for frequency band groups. In particular, the tracking methods and their adaptation are disclosed in EP 2 267 475 A1, the relevant content of which is part of the present description.

[43] In one embodiment, a track fusion is performed in which different tracking algorithms, in particular for different frequency bands and / or frequency band groups, interactively determine a track. [44] Interaction is understood in particular to mean that different tracking algorithms and / or different tracking methods for different frequency bands or frequency band groups are looking for an optimal solution depending on each other.

In particular, so frays of each track tracks can be analyzed and

Track track results are eliminated and / or optimally combined.

[46] For example, a tracking algorithm can be used for each frequency band, so that a track track is determined for each frequency band. This can lead to fraying in an overall representation of the track tracks. Thus, tracking results, which are definitely exclude due to their location, can be omitted, so that, for example, only half of the tracking results are used. These in turn can be averaged, for example. Also strongly divergent track tracks can be detected and omitted.

[47] In a further aspect of the invention, the object is achieved by a device having a hydrophone and a computer, wherein the device is set up such that a previously described method can be carried out.

[48] Such a computer has in particular working memory, fixed memory, input and output means and corresponding calculation algorithms. The hydrophone naturally has an electronic evaluation unit on, which processes signals in such a way that they can be evaluated by the computer, for example.

[49] In an additional aspect, the object is achieved by a watercraft having the device described above.

[50] In the following, the invention will be explained with reference to exemplary embodiments. Show it

Figure 1 is a schematic representation of a

Geometry of a linear antenna with a ramped plane wave,

Figure 2a-e is a schematic representation of a

Converting an underwater sound signal obtained from an antenna to an energy envelope,

Figure 3 is a schematic representation of several

Frequency bands,

Figure 4a is a schematic representation of a

Energy envelopes with associated goals,

Figure 4b is a schematic representation of

Bearing marks (courses) of the targets from FIG. 4a,

Figure 5 is a schematic representation of two crossing targets according to the prior art and Figure 6 is a schematic representation of the intersecting courses of Fig. 5 according to the invention.

[51] In the first place, the theory of the present invention, which is the basis of the present technique, will be explained briefly, clearly with associated advantages.

[52] In addition to the energy envelope generated in the broadband signal processing per time cycle in a time-lapse representation, which represents the total energy of the sound intensity of a particular frequency band (depends on the selected antenna) depending on the direction of incidence, in particular in the location representation (bearing) per Time clock several

Generates energy envelope, in each case the total energy of the sound intensity of a particular frequency band of the entire frequency range is determined in dependence on an incident direction.

[53] Due to the wavelength or

Frequency dependence of a main lobe width of

Directional characteristic of a transducer arrangement

(Hydrophone array), two targets that are close together in a line of sight can be resolved better at higher frequencies than at lower frequencies.

[54] Especially in the area of target crossings, the consideration of the energy envelopes of higher-frequency frequency bands is therefore advantageous. [55] Two targets can be detected up to a smaller bearing distance than individual local maxima in the envelope. If the location of the local maxima is regarded as the detection of a bearing to an existing target, the failure of detections of a target can be shortened in time. The time sequence of the detections of a specific destination can be summarized to the time history of a destination track. To accomplish this, an automatic tracking system is used.

[56] The task of a tracking system is to extract as many true target lanes as possible from the available detections and to estimate their bearing course and additionally to prevent the extraction of false target lanes as well as possible. The already mentioned target intersections represent a challenge for such a system due to the lack of detections. If a track intersection is extended over a very long time, an assignment of target track pieces before and after such a crossing according to the prior art is not possible.

[57] To provide a solution for this, a reduction in the time gaps in the target lanes can be improved. This is done in particular by an improved spatial resolution.

[58] The basics to the actual

Waveform processing and processing can be found in Richard Nielsen's Basic Book, Sonar Signal Processing, Artech House, 1991 (ISBN 0-89006-453-9). be removed. In particular, reference is made here to Chapter 2 and especially to Chapter 2.5, the content of which is part of the present document.

In Figure 2 essential aspects of the generation of the Engergieeinhüllenden for a linear antenna are shown in an illustrative manner.

[60] With the help of beamforming, a number of discrete beams are formed for the linear antenna at the bow of a submarine 311. The total sound intensity (energy) contained in a beam depends on whether there is a loud target ship 313, 315 or not in the appropriate direction. The targets 313, 315 in the north (0 °) and east (90 °) lead to much stronger intensity (larger bars in the illustration) in the partially overlapping beam sectors 321, which point in the direction of the targets 313, 315.

By folding the beam signals in the circle 331 special representation can be generated. By cutting the circle 331 can be a linear representation

(Fig. 2d.) Of the energy distribution of the horizon can be achieved.

[62] By interpolation, the discrete beam signals can be converted into a continuous distribution (waveform, Fig. 2e.).

[63] In Figure 3, a selection of spatially resolved waveforms (energy envelopes) for linear

Hydrophone arrangement of a real lake scenario shown. In this case, the axis is plotted on the abscissa and the energy is plotted on the ordinate (proportional to the sound intensity). The signal curves are plotted for different frequency bands (Nos. 5, 8, 11, 14, 17, 20). The higher the frequency band number, the higher the evaluated sound intensities.

[64] Higher frequency bands have a positive effect on the target separation power in envelopes: It can be seen that two local maxima 461, 463 are visible in the higher bands at about 130 °, whereas in the lower bands (numbers 5 and no 8) these local maxima are inseparable. That is, if the temporal evolution of such energy envelopes is considered, approximate targets in higher frequency bands longer than two targets are resolvable in the bearing. This is of great importance in tracking goals.

FIG. 4a shows a spatially resolved signal curve 351 with three maxima 561, 563, 565. On the abscissa is again the bearing and on the ordinate the energy is plotted. From the spatially resolved

Waveform representation from FIG. 4 a, a time-resolved representation (FIG. 4 b) can be determined.

[66] In Fig. 4b, a so-called waterfall diagram or Bearing Time Record (BTR) is shown, which represents the time course of the target detections (maxima 561, 563, 565). It is therefore a time-resolved representation. The bearing is plotted on the abscissa and the time on the ordinate. [67] N _bancl envelopes are generated per time _cycle . For the BTR, the local maxima 561, 563, 565 are extracted from each of these envelopes. For each time clock, the local maxima and the associated bearing are determined. Due to the time course, the course

(Bearing history) of a target (maximum).

FIG. 5 shows a waterfall diagram according to the prior art, in which the total frequency width was used to determine the maxima. The abscissa has a bearing Θ and the ordinate the time t. In addition, the course 661, 663 of a first target 461 and a second target 463 is shown. In the intersection area 671 of the two courses 661, 663, for the weaker destination 663 there is an opening 673 of the course, so that the course can not be clearly predicted.

[69] The flared courses are now analyzed for the three highest frequency bands using a multi-hypothesis tracking method. In addition, all low-frequency frequency bands are combined to form a frequency band group. For this frequency band group, the multi-hypothesis tracking method is also performed. Thus, there are four (most likely) courses determined by MTH procedures.

[70] These are averaged, so that then there is a combined total price. Alternatively, the four courses are given a weighting, where the rate of the highest frequency band is a factor of 4, followed by a lower rate the factor 3, then the subsequent course receives the factor 2 and the course from the frequency band group the factor 1.

[71] In Fig. 6, to determine the targets, the frequency bands 17, 18, 19, and 20 are grouped into a frequency band group after the fanning 673 is detected. The maxima are determined for the signal course of this frequency band group and the courses 661, 663 are determined. Fray 673 is omitted and rates 661, 663 are uniquely determinable. Possibly. the course is extrapolated 663.

LIST OF REFERENCE NUMBERS

100 linear antenna

103 plane wave

311 submarine

313 target ship

315 target ship

321 Beam sector

331 circle

333 Slicing the circle 331

341 linear representation

343 own course

351 waveform

461 first local maximum

463 second local maximum

561 local maximum and destination

563 local maximum and destination

565 local maximum and destination

661 Course Goal 1

663 Course Goal 2 671 Intersection 673 Fanning the course

Claims

Claims:

1. A method for location and / or time resolution of a first

Underwater sound source from a second

Underwater sound source, the first

Underwater sound source and the second

Underwater sound source one each

Send and / or reflect underwater sound signal, the method has the following steps

- Dividing the frequency range in at least two frequency bands and

2. Method according to claim 1, characterized in that when dividing the frequency range three, four, five, six, seven, eight, nine, ten, eleven, twelve, thirteen, fourteen, fifteen, sixteen, seventeen, eighteen, nineteen, twenty or more frequency bands are provided.

3. The method according to any one of the preceding claims, characterized in that the division of the frequency range is performed such that adjacent frequency bands form a frequency band gap or a frequency band overlap.

4. The method according to any one of claims 2 or 3, characterized in that two, three or more frequency bands are grouped into a first frequency band group, wherein the grouped frequency bands are adjacent to each other in particular. A method according to claim 4, characterized in that a second frequency band group, a third

Frequency band group and / or other frequency band groups are formed by the grouping.

Method according to one of Claims 4 or 5, characterized in that one of the frequency band groups is formed successively, in particular from high frequencies to low frequencies.

Method according to one of the preceding claims, characterized in that by means of the signal strength curve of one of the frequency bands and / or one of the frequency band group, the first underwater sound source and the second underwater sound source is determined.

A method according to claim 7, characterized in that the determination of the first underwater sound source and the second underwater sound source by means of a

Maximum determination and / or by means of a mathematical adaptation of two basic functions to the

Signal strength course takes place.

Method according to one of the preceding claims, characterized in that one of the signal strength curves depending on the location or the two

Underwater sound sources as a function of location and time is displayed.

Method according to one of claims 7 to 9, characterized in that the first underwater sound source and / or the second underwater sound source, in particular by means of a tracking algorithm, be tracked.

A method according to claim 10, characterized in that the tracking for individual frequency bands or for all frequency bands and / or for individual Frequency band groups or for all

Frequency band groups takes place.

Method according to one of claims 10 or 11, characterized in that the tracking is carried out by means of a single hypothesis tracker method and / or by means of a multi-hypothesis tracker method.

Method according to one of Claims 10 to 12, characterized in that a track fusion is carried out in which different tracking algorithms, in particular for different frequency bands and / or frequency band groups, interactively determine a course or course or bearing course.

Device with a hydrophone and a computer, wherein the device is set up such that a method according to one of the preceding claims is feasible.

A watercraft having a device according to claim 14.