阅读说明：本技术 用于对声源进行空间定位的装置、系统和方法 (Apparatus, system and method for spatially localizing a sound source ) 是由 T·里滕朔贝尔 于 2019-03-19 设计创作，主要内容包括：本发明描述了一种装置,其包括至少一个可动地布置的第一传声器(10)、至少一个第二固定式传声器(11)和至少一个传感器(16)。所述传声器可以检测声源所发出的声波,所述传感器可以检测所述第一传声器的空间坐标。此外,还描述了一种对应的方法和一种具有上述装置的系统。(The invention describes a device comprising at least one movably arranged first microphone (10), at least one second stationary microphone (11) and at least one sensor (16). The microphone may detect sound waves emitted from a sound source, and the sensor may detect spatial coordinates of the first microphone. Furthermore, a corresponding method and a system having the above-described device are described.)

1. An apparatus, comprising:

at least one movably arranged first microphone (10) adapted to detect a sound source in a measurement scene,

a second stationary microphone (11) adapted to detect a sound source in said measurement scenario, and

at least one position sensor (16) adapted to detect a position of the first microphone (10).

2. The apparatus of claim 1, wherein the first and second electrodes are disposed on opposite sides of the housing,

wherein the device is supported in such a way that it can be rotated about a rotational axis (A), wherein the second microphone (11) is arranged on the rotational axis (A) in such a way that it does not change its position during the rotation.

3. The apparatus of claim 2, wherein the first and second electrodes are disposed in a common plane,

wherein the device has the shape of an aerodynamically shaped wing.

4. The apparatus of any of claims 1 to 3, further comprising:

an electronics unit (40) adapted to process sensor signals provided by the first microphone (10), the second microphone (11) and the position sensor (16).

5. The apparatus of claim 2, the apparatus further comprising:

a balancing weight (17) arranged on the device in such a way that no imbalance of the device occurs in a rotational movement about the rotational axis (A).

6. The device of any one of claims 1 to 5,

wherein the device comprises a first part and a second part, wherein the first part is movably supported relative to the second part, wherein the first microphone (10) is arranged on the first part and the second microphone (11) is arranged on the second part.

7. A system, comprising:

a support;

the device of any one of claims 1 to 6 mounted on the support and

a device (20) fixable on the support, the device being adapted to receive measurement data from the apparatus based on the sensor signal and to enable visualization of a measurement result based on the measurement data.

8. The system of claim 7, further comprising:

a computing unit contained in the device (20) or coupled with the device by a communication link.

9. The system of claim 8, wherein the first and second sensors are arranged in a single package,

wherein the computing unit is comprised of a cloud computing service.

10. A method, comprising:

-moving the first microphone (10) relative to a fixed reference position, the second microphone (11) being arranged at said reference position;

-detecting sensor signals provided by the first microphone (10) and the second microphone (11) and simultaneously detecting a position signal indicating the position of the first microphone (10) relative to the reference position;

-calculating a sound source intensity at one or more reconstruction points based on sensor signals of the first (10) and second (11) microphones and further based on the measured position signals.

11. The method of claim 10, the method further comprising:

acquiring an optical image, the optical image comprising one or more sound sources;

superimposing the calculated sound source intensity for one or more reconstruction points with an optically acquired image.

12. The method according to any one of claims 8 or 9, the method further comprising:

the collected measurement data is transmitted over a communication link to a computing unit, in particular a cloud computing service,

wherein the sound source intensity at the one or more reconstruction points is calculated by the calculation unit.

13. A method, comprising:

at least one sensor signal (p) is detected by means of at least one first microphone (10) which moves along a defined path₁(t)) and simultaneously measuring a relatively fixed reference position (M) of the first microphone (10)₂) Position (M)₁(t))；

By being located at said reference position (M)₂) A second microphone (11) in (b) detects a second sensor signal (p)₂(t))；

Based on the first sensor signal (p)₁(t)), the second sensor signal (p)₂(t)) and the measured position of the second microphone (10) relative to the fixed reference position (M)₂) Is virtually located in a plurality of predefinable reconstruction points (R) with respect to the second sensor signal (p) measured₂(t)) of the signal power.

14. The method of claim 13, the method further comprising:

-taking an image by means of a camera having a fixed position with respect to said first microphone (11);

-associating said reconstruction point (R) with a corresponding pixel of said image; and

-showing the image, wherein the pixels corresponding to the reconstruction points (R) are rendered based on the components calculated for each reconstruction point (R) of the virtual sound sources located in said reconstruction point (R).

15. The method according to claim 13 or 14, wherein for each reconstruction point (R), the component determination of a virtual sound source comprises:

-comparing the first sensor signal (p)₁(t)) into a converted first sensor signal

calculating the second sensor signal (p)₂(t)) with the converted first sensor signalCoherency between

for a predefinable frequency band ([ F ]₁,F₂]) Calculating a second sensor signal (p) relative to the measured virtual sound source located in each reconstruction point (R)₂(t)) of the signal power.

16. Method according to claim 15, wherein the first sensor signal (p) is sensed₁(t)) includes:

-transmitting the first sensor signal (p) with a time span related to the distance between the first microphone (10) and the reconstruction point (R) at any starting point in time of the measurement ₁(t)) back-propagating into said reconstruction points (R);

-applying the counter-propagating first sensor signal (p) with a time-varying fractional delay filter in a time span representing the time required for an acoustic signal to travel from the reconstruction point to the first microphone (10)₁(t)) filtering;

filtering the filtered signal with a time span related to the distance between the first microphone (10) and the reconstruction point (R) at any starting point in time of the measurementSensor signal counter-propagating into the reconstruction point (R), wherein the resulting counter-propagating filtered sensor signalRepresenting acoustic signals emitted by virtual sound sources located in respective reconstruction points (R);

-filtering the counter-propagating filtered sensor signal with a time span related to the distance between the reference position and the reconstruction point (R)

17. Method according to claim 15 or 16, wherein the calculation of the components of virtual sound sources located in said reconstruction points (R) comprises:

in the predefinable frequency band ([ F ]₁,F₂]) In-range pair coherence

18. A device for spatially localizing sound sources on arbitrary surfaces by recording measured variables of the sound field by means of two microphones (10, 11),

It is characterized in that the preparation method is characterized in that,

the first microphone (10) moves on a path, the second microphone (11) is fixed,

wherein the relevant acoustic signal is transmitted to a unit moving together with the first motion microphone (10) and consisting of:

a sensor system for detecting the spatial coordinates of the first motion microphone (10),

a data acquisition device for the data of the two microphones (10, 11),

a sensor (16) for detecting the spatial coordinates of the first moving sensor (10) and

a power supply device (49),

and forwarding the relevant acoustic signals to a data processing device (20) comprising a display and an operating face for controlling the measurements and representing the results in the form of a superposition of the image of the measurement scene acquired by means of the camera (21) and a reconstruction map of the acoustic source, and

wherein the first motion microphone (10), the second stationary microphone (11) and the unit moving together with the first motion microphone are integrated into a frame structure (100, 200).

19. The apparatus of claim 18,

the frame structure (100, 200) is rotatably supported about a fixed axis (54), whereby controlled rotation of the frame structure (100, 200) occurs.

20. The apparatus of claim 19,

the frame structure (100, 200) is coupled to a drive (30) such that an automatic rotation of the frame structure (100, 200) about the fixed axis (54) is achieved.

21. The apparatus of claim 19,

from an array of microphones distributed along the frame structure (100, 200), the first motion microphone (10) is selected by an electronic multiplexer (45) controlled by the data acquisition device depending on the time and/or angular position of the frame structure (100, 200), whereby an approximately helical motion of the first motion microphone (10) can be achieved.

22. The apparatus of claim 19,

the first motion microphone (10) slides on a radially oriented track embedded in the frame structure (200), the radial motion of the microphone being synchronized with the motion of the frame structure (200) about the fixed axis (54), thereby causing a helical motion of the first motion microphone (10) upon motion of the frame structure (200).

23. The apparatus of claim 18,

the first motion microphone (10) and the second stationary microphone (11) are essentially acoustically decoupled from the frame structure (100, 200) with respect to mechanical vibrations, so that the drive noise, the motion noise and the wind noise have only a minimal effect on the measurement dynamics.

24. The apparatus of claim 18,

the first mobile microphone (10) and the second stationary microphone (11) are assigned wind-shielding means in order to suppress the aeroacoustic noise of the moving frame structure (100, 200) and to influence the measurement dynamics only to a minimal extent.

25. The apparatus of claim 18,

the surface of the frame structure (100, 200) is shaped aerodynamically in the direction of movement in order to suppress the aeroacoustic noise of the moving frame structure and to influence the measurement dynamics only to a minimal extent.

26. The apparatus of claim 18,

the sensor system for detecting the spatial coordinates of the first motion microphone (10) is realized by a rotation angle sensor (16) for measuring the rotation angle relative to the rigid rotation axis (54).

27. The apparatus of claim 18,

the sensor system for detecting the spatial coordinates of the first motion microphone (10) is realized by an angular velocity sensor and a three-axis acceleration sensor, which are at least uniaxial and oriented collinear with the rotational axis (54) of the frame structure (100, 200).

28. The apparatus of claim 18,

the sensor system for detecting the spatial coordinates of the first motion microphone (10) is realized by a motion tracking system.

29. The apparatus of claim 18,

the distance of the first motion microphone (10) from the centre of rotation of the frame structure (200) can be varied in order to improve the spatial resolution, especially in the case of sound sources having a lower frequency.

30. The apparatus of claim 18,

the data of the data acquisition devices integrated into the frame structure (100, 200) and moving together are wirelessly transmitted to a personal computer, laptop or mobile device for further processing or visualization.

31. A method for reconstructing the intensity of a sound source in an arbitrary spatial point by estimating the coherence between the temporal signal of a first stationary acoustic sensor and the transformation of the temporal signal of a second moving sensor,

converting the time signal of the second motion sound sensor to the reconstruction point by time shifting taking into account the propagation velocity in the observed medium,

the resulting signal is used as excitation of a monopole sound source in the reconstruction point,

Time-shifted by means of a time delay relative to the time dependence between the reconstruction point and the position of the second motion-acoustic sensor with an inverse time delay profile of the direct-current component, and

finally mapped onto the position of the first stationary acoustic sensor with a constant time shift,

wherein the coherence function measured in the frequency domain in this way defines the value of the component of the source radiating from the reconstruction point with respect to the level measured at the position of the stationary sensor.

32. The method of claim 31,

the coherence estimate is analyzed in the frequency domain at a specific frequency or integrated in a defined frequency band in order to analyze the sound source distribution in the frequency domain.

33. The method of claim 32,

the sound source intensity is reconstructed at a plurality of spatial points in order to achieve a mapping of the sound source intensity on an arbitrarily shaped surface.

34. The method of claim 33,

and superposing the mapping of the sound source intensity and the image of the measurement scene acquired optically so as to realize spatial correspondence.

Technical Field

The present invention relates to a device, a system and a method for spatially localizing a sound source, and thus belongs to the technical field of acoustic measurements.

Background

Spatial localization of acoustic sound sources is an important task for quantifying the noise and vibration characteristics of products in the automotive and aerospace fields, as well as in the consumer electronics and industrial equipment fields. During product development, this task is called Noise, Vibration, harshness (nvh) Testing. NVH testing is essential in order to develop products that meet regulatory and customer-related requirements in terms of noise limits and sound perception, and therefore the products must be measured with corresponding reliability and accuracy. The localization of sound sources is not limited to the above-mentioned products of the industrial field, but also in other fields, for example for measuring ambient noise in workplaces, public spaces to identify sources of acoustic interference or for making sound-proof evaluations in architectural acoustics. Users of systems for spatially localizing sound sources include, but are not limited to, manufacturers of products in the aforementioned industrial fields, engineering services providers, construction phonologists, construction industry companies and public institutions.

In particular, the required regulatory and desired sound quality is checked relatively late in the product development process. Here, product developers need easy-to-use and intuitive tools that help them analyze NVH issues and make decisions on meeting product specifications. Similar problems exist in the field of architectural acoustics, in field inspection of construction projects, quality monitoring during manufacturing of products, and condition monitoring of machines and processes.

So-called acoustic cameras for visualizing sound sources are known from the prior art. Such acoustic cameras usually have a microphone array with a large number of microphones arranged on a disc-shaped surface. The construction of such acoustic cameras is often complex, and in particular it is often necessary to connect a large number of microphones together with an efficient system for parallel data acquisition and processing.

The inventors set themselves the object of providing a device for spatially localizing and visualizing a sound source which offers inherent technical advantages over the prior art, for example in terms of image quality, in particular in terms of contrast range, spatial resolution and maximum representable frequency of the sound source, and which, moreover, is particularly easy to operate and, owing to a reduced technical complexity, is less expensive to manufacture.

Disclosure of Invention

The solution of the invention to achieve the above object consists in a device according to claims 1 and 18, a system according to claim 7 and a method according to claims 10, 13 and 31. Various embodiments and further aspects are the subject matter of the dependent claims.

A device is described, which comprises at least one movably arranged first microphone, at least one second stationary microphone and at least one sensor. The microphone may detect sound waves emitted from a sound source, and the sensor may detect spatial coordinates of the first microphone.

Furthermore, a system with a corresponding device is described, which comprises a stationary data processing apparatus on which the device is mounted so as to be rotatable about a rotational axis. The data processing device may receive measurement data from the apparatus and show the sound source intensity of the object to be measured.

The method according to one embodiment comprises the steps of: providing a rotatably supported device having at least one movably arranged first microphone and at least one second stationary microphone; rotating the device about an axis of rotation, wherein the first microphone rotates together and the second microphone remains stationary; detecting sound waves emitted by an object to be measured through the first microphone and the second microphone, and detecting the space coordinate of the first microphone; the source intensity of the sound producing object is calculated and imaged based on the measured measurement data.

The invention also describes a method and apparatus for imaging an acoustic emission object, comprising: recording (i) acoustic variables of the sound field, (ii) path coordinates of the moving sensor and (iii) an optical image of the scene to be measured by means of the moving and stationary sensors, recording the data in an electronic unit moving together with the sensing mechanism and forwarding these data to the terminal device for processing the data and representing the result as a superposition of a color-coded sound image and an optically acquired image of the object to be measured.

The device described herein can be used for a variety of applications, in particular for checking the regulatory requirements and the desired sound quality of products in the automotive, aerospace, consumer electronics and industrial equipment fields, but also in the construction acoustics field, quality monitoring during the manufacture of products and condition monitoring of machines.

The concept described herein for imaging acoustic emitting objects can be used for different applications, for example because the measuring instrument is less expensive compared to known systems with comparable performance and image quality based on sensor arrays and beam forming techniques. The associated cost reduction makes this technology available to more users. The substantial reduction in size and weight makes it easy to transport, the reduced complexity reduces the preparation time for the measurements and improves operational reliability.

In certain applications, the concepts described herein may be advantageously employed because the amount of data that must be processed is reduced by one to two orders of magnitude compared to known sensor arrays and beamforming techniques, thus greatly reducing the requirements on the hardware to be used for data acquisition and processing and allowing results to be computed significantly faster in the form of acoustic source images.

Drawings

The invention will be described in detail hereinafter with reference to an embodiment shown in the drawings. The drawings are not necessarily to scale and the invention is not limited to the illustrated aspects. Rather, the emphasis is placed upon illustrating the principles upon which the invention is based. Wherein:

FIG. 1 illustrates, in connection with a block diagram, an example of a method of calculating an image of an acoustic emitting object based on measurements of acoustic variables and characteristics of measured sensor motion.

Fig. 2 is a front view (left side, block diagram including electronic unit) and a cross-sectional view (right side) of a first embodiment of the device.

Fig. 3 is a front view (left side, block diagram including electronics unit), a cross-sectional view (middle), and a back view (right side, without housing cover and driver) of an alternative embodiment of the device.

Detailed Description

The present invention addresses the problem of spatial localization of sound sources, referred to in the literature as the "Direction of arrival" (DoA) problem. It is noted that the DoA problem and related solutions may be relevant in almost all cases where the physics on which they are based are characterized by wave propagation, such as radar, sonar, seismology, wireless communication, radio astronomy, imaging in medical technology, acoustics, etc. In terms of acoustics, the aim is to reconstruct the direction or position of a sound source with respect to the position and orientation of the measurement system under observation, based on measurements of physical variables of the sound field (e.g. by means of microphones).

In a commercially available solution, the surveying instrument and the program described herein are combined with a camera, a data logger connected to the sensor and a computer for executing algorithms for reconstructing a sound image, see for example the system description in publication WO2004/068085 a 2.

In commercially available solutions for sound source localization, the following physical variables and associated sensor technologies are taken into account: pressure (non-directional pressure microphone); sound velocity (hot wire microphone), sound intensity (dual microphone technique); air refractive index (laser doppler vibrometer); lateral velocity of the reflecting surface (laser doppler vibrometer). The measurement of the transverse velocity of an optically reflective sound-transmitting membrane by means of a laser doppler vibrometer for spatial localization of sound sources is also described in the scientific literature. The physical variables measured by means of different sensor technologies are usually converted electrically, acquired digitally by means of corresponding single-channel or multi-channel data acquisition systems, preprocessed and finally fed to the actual measurement program.

For further observation, this method is advantageously divided into a first group with a plurality of stationary time-synchronized sensors (e.g. microphones), which are referred to in the literature as sensor arrays; and a second group of methods with separate or multiple moving and time-synchronized sensors. Depending on the application, the sensor array may have a one-dimensional, two-dimensional or three-dimensional geometrical arrangement, for example, linear, circular, cross-shaped, spiral, spherical, regular or randomly distributed in a plane or volume. A variant in the form of a wheel with laterally offset spokes is described in publication US 7,098,865B 2 (original publication DK 174558B 1). In all of the following methods, the configuration of the array in terms of the distance between the sensors or the spatial extent (aperture) of the array plays a decisive role in the achievable spatial resolution, the suppression of artifacts and the maximum detectable frequency content of the acoustic source to be measured.

Beamforming refers to a signal processing technique that corrects in a first step the time offset due to different transit times from the sound source to the various sensors in the array with respect to the observed focal point and adds all the time-corrected signals in a second step. Thus, the output signal of the beamforming algorithm increases in its amplitude when the acoustic signal comes from the direction of the observed focus, and attenuates when the acoustic signal comes from other directions. Thus, beamforming corresponds to spatial filtering of the acoustic signal. The algorithmic correlation of the sensor signals implemented in the manner described is also referred to as the "delayed Sum" or "controlled Response Power" (SRP) algorithm.

The first step of the signal processing of the sensor signal, i.e. the time correction with respect to the focus, can be replaced by a filter in order to optimize the ratio of received signal power to signal noise. Related methods are referred to as "super directional beamforming" and "adaptive beamforming". This method can also be applied to moving sound sources, such as trains, cars or rotating objects in mechanical manufacture that are driven by the side.

Acoustic Holography (NAH-Near field Acoustic Holography) in the Near field of Acoustic emitting objects produces an image of the Acoustic source on the surface based on measurements with a checkerboard-like arrangement of sensor arrays in the passive area, where this array is transverse to the direction of propagation of the sound and should cover most of the radiating surface. In the case of a measurement with a pressure microphone, the sound pressure in the source plane is determined by means of a function which describes the phase shift from the hologram plane to an arbitrary parallel and passive plane by means of the inverse fourier transform of the (two-dimensional) spatial fourier transform of the sound pressure distribution in the measurement plane (hologram plane). The calculation method can reconstruct the sound velocity and the sound intensity through an Euler equation. The so-called Patch-NAH method (local near-field acoustic holography) is directed to the measurement problem of large areas and large structures, where the area to be measured is divided into smaller areas (local). An extension of the NAH method is based on the boundary Element method (ibem (inversebound Element method) based NAH) of NAH, where helmholtz integration theory is applied in order to calculate the influence of sound sources on the envelope on points inside the observed volume. In the literature, the approximation of the above method is referred to as the "helmholtz equation, Least Squares" (HELS) method, in which the sound field is approximated by an allowed basis function with a mass function with respect to the minimum squared error.

The Time Difference of Arrival (TDOA) describes a two-stage method in which, in a first step, the Time delay between the sensor signals of spatially adjacent sensor pairs is determined for a sensor array by means of cross-correlation. Curves or planes are generated by knowing the time delays of the pairs and the spatial positions of the sensors and the intersection of the curves or planes resulting from the optimization problem is determined as the estimated position of the sound source.

The Multiple Signal Classifier (MUSIC) algorithm belongs to the group of Subspace (Subspace) methods and is capable of identifying Multiple harmonic sound sources (narrowband sources) at higher spatial resolutions. The spatial spectrum of the source distribution is calculated by decomposing the cross-correlation matrix of the sensor signals into mathematical subspaces and eigenvalues on the basis of a linear mapping, column by column, consisting of so-called "steering" vectors, which map the signals of a plurality of a priori defined spatial sources at a particular point in time onto the measurement signals of the sensor array at a particular point in time. The peaks in the spectrum correspond to the direction of sound arrival of the parameterized source in the model. ESPRIT (Estimation of Signal Parameters using a rotation invariant Technique to estimate Signal Parameters) is an adaptation of the above subspace approach in order to react less sensitively to changes in position, gain and phase errors of the sensor.

The time-synchronized spatial scanning of the sensor signals of the sensor array in the form of a circle causes a doppler effect in the composite signal. A time-discrete Teager-Kaiser operator can be used to analyze the modulation and, in the case of a plane, the direction of the arriving sound. The localization of multiple sources and the calculation of the spatial spectrum of the source distribution is achieved by determining a probability density function based on an analysis of the respective time segments. Alternatively, a center of gravity algorithm may be used in order to reduce noise sensitivity and to reduce the limitations in the frequency domain that is actually available.

Publication US 9,357,293B 2 proposes a frequency domain method based on scanning a microphone signal from a virtually moving sub-aperture of a linear array with a large number of sensors. By moving the sub-apertures along the array at different speeds, different signal superpositions are produced frequency-dependently according to the position of the sound source. Information about the different mixes and signal energies in the spectrum during different motions of the sub-apertures may be used in order to finally infer the sound source location at a particular frequency.

In so-called "Light refractive tomography", the refractive index change of a medium in which a sound field propagates along a laser beam, which is reflected on a rigid reflector, is calculated by means of a laser scanning vibrometer. In this case, the laser beam is directed transverse to the direction of propagation of the acoustic field. As an extension, a computer tomography method can be used in order to reconstruct a tomographic image of the sound field in the plane in which the refractive index variations are calculated.

A sensor with single or multiple motions is discussed below in a method that utilizes the doppler effect.

Individual sound sources located in space with constant frequency and amplitude are localized by a phase-locked loop (PLL) demodulation of the phase modulation of the observed signal through a priori known motion of the acoustic sensor on a circular path.

According to publication US 2,405,281 a, the direction of arrival of the main sound or electromagnetic wave can be identified by means of a sensor moving on a circular path based on the doppler effect. In this case, the measured bandwidth of the doppler history of the harmonic source is greatest when the plane of rotation of the sensor is orthogonal to the receive direction of the source. EP 2948787 a1 and EP 2304969B 1 are other publications which identify the individual main sound sources by the movement of one or more microphones, using the doppler effect.

The limitation of using beamforming to spatially filter sound waves arriving at a sensor array (with uniformly distributed sensors) is that the maximum detectable (critical) frequency of a sound source is defined by the distance between two adjacent sensors of the array. Attempts to localize sound sources outside this frequency result in so-called "ghosts", also referred to in the literature as spatial aliasing artifacts, in the reconstructed sound image. The use of a fixedly arranged sensor array with sensors above a critical frequency and the associated suppression of spatial aliasing artifacts can be achieved by spatial motion of the sensor array. The point spread function (point spread function) of an imaging system comprising a sensor array moving with a constant angular velocity over a circular path can be simulated by means of a delay-and-sum algorithm and the suppression of aliasing artifacts can be experimentally checked by means of a single harmonic sound source.

As already mentioned in the beginning of this section, it is also worth mentioning methods for locating sources or reflecting structures which are not applied directly to acoustics, but whose physics is based on is characterized by the propagation of waves. Thus, Synthetic Aperture Radar (SAR) may provide imaging of a surface by scanning an electromagnetic field generated by a moving transmitter having a known position. The movement of this transmitter relative to the surface causes a doppler effect in the collocated receivers. By correlating the measured doppler history, i.e. the time course of the doppler frequency of the object echo from the initial positive value to the negative value, an image of the scanned surface is reconstructed with a higher resolution by means of a so-called replica, i.e. based on knowledge of the movement of the doppler history produced by the transmitter from the point reflector for each distance.

The basic idea of the new concept described herein is to reduce the measuring instrument expenditure for acquiring sound field information to a minimum while improving the quality of the sound source spatial imaging, in particular in terms of contrast range, spatial resolution and maximum representable frequency. An inherent disadvantage of the measurement methods requiring sensor arrays is the use of a large number of sensors in order to obtain a usable spatial resolution with an acceptable contrast range and a reasonable upper frequency limit.

In both beam forming and acoustic holography, spatial resolution is described by the size of the aperture of the array and the maximum representable frequency, and the field of view is described by the distance between the sensors. The increase in spatial resolution therefore determines an increase in the number of sensors on the measuring area.

The TDOA method only allows the identification of the most dominant source, while the MUSIC/ESPRIT method has a limitation on the number of sources that can be identified, with the upper limit depending on the number of sensors used. When using TDOA and MUSIC/ESPRIT, measurements of the sound field with an unknown number of sound sources may cause misinterpretations. The DREAM (Discrete reconstruction array modeling) method requires multiple runs in order to characterize the signal spectrum of the acoustic source in a first step and finally in a second step to determine therefrom the size and velocity of the sub-apertures of the linear array. Due to the sequential measures used to determine the configuration, this method cannot be used to measure individual acoustic events without a priori information.

The photorefractive tomography approach described above presents a challenge to measurement technicians in creating a vibration-free environment for laser scanning vibrometers and rigid reflectors, since the medium refractive index variation according to the acoustic field is typically many orders of magnitude lower than the vibration level of the induced source. As with beam forming and acoustic holography, an array of laser doppler vibrometers along the perimeter of the tomographic image to be reconstructed is required in order to achieve a usable spatial resolution. A technical difficulty with measuring instruments for implementing arrays with laser doppler vibrometers is that the complexity of the measuring instruments used is several orders of magnitude higher than when arrays with pressure or sound velocity sensors are employed.

The method of demodulation of the phase modulation of the observed signal using a sensor of circular motion is generally only used to identify harmonic sources or mainly fixed sources with constant amplitude and frequency.

Fig. 1 shows in a block diagram an example of a signal processing method which can be used in the embodiments described herein to determine an image of an acoustic emitting object (so-called acoustic image, in which the acoustic pressure is, for example, color-coded). The embodiments which will be described in more detail below all use a fixed position M which is known per se₂And one or more sensors moving along a circular path. In the present example, a moving microphone is observed, for example by means of a time-dependent position M₁(t) to describe the circular or spiral motion of this microphone. But trace M₁(t) does not necessarily have to travel along a circular path or a helical path. However, in the case of circular or spiral trajectories, the instantaneous position M₁(t) can be measured relatively easily by means of an angle sensor (rotation angle encoder).

The following variables are considered as input variables for the method: (i) the point of reconstruction R, described by spatial coordinates relative to a coordinate system, has its origin at the position M of the stationary sensor ₂At least one of (1) and (b); (ii) sensor signal of a moving first sensor (microphone signal p)₁(t)) and the sensor signal of the stationary second sensor (microphone signal p)₂(t)); (iii) movement M of a sensor by a first movement described with respect to the spatial coordinates of a coordinate system₁(t) with its origin at the position of the stationary sensor; (iv) observation time T, which is determined by a complete scan of the apertureDefining (e.g. one revolution of the sensor steered on a circular path) and determining a time window for the time signal converted into the frequency domain; and (v) the observed frequency band [ F ]₁,F₂]。

In a first step, with a value τ₁＝d(R,M₁)_t＝0C time signal p of sensor for movement with time t equal to 0₁(t) backward propagation in time (backward shift in time) into reconstruction point R, where d (R, M)₁) Represents the distance between the reconstruction point R and the position of the motion sensor, and c represents the propagation speed (speed of sound) of the disturbance in the observed medium. In a second step, by having a variable time delay of 2 τ₁-₁To the time signal p1(t + τ)₁) A time shift is performed. Variable time delay₁＝τ₁+τ(M₁) Representing the position M of an acoustic disturbance in the observed medium from the reconstruction point R to the moving sensor ₁The time difference required. τ is the position M of the moving microphone₁As a function of (c). In a third step, as in the first step, the time signal is propagated back in time to the reconstruction point R (at the value τ)₁). Generated time signalRepresenting the acoustic signal emanating from the virtual sound source (acoustic monopole) in the reconstruction point R, and then finally, in a fourth step, this time signal

With time delay τ₂＝d(R,M₂) C forward propagation (forward in time) to position M of stationary sensor₂Where d (R, M)₂) Representing the distance between the reconstruction point R and the position of the stationary sensor. For the actual radiation source in the reconstruction point R, the movement M for the first sensor₁The resulting temporally variable doppler shift is compensated for. Thus, for the radiation source in the reconstruction point R, a time signal is generated

Is the time signal p of a moving sensor₁(t) position M to a stationary sensor₂Is mapped. The time signal component from the actual radiation source away from the reconstruction point R may experience an additional temporally variable doppler shift.

By means of a time signal p₂(t) andbased on the spectral power density function P_p2p2(f) Andand spectral cross power density function Determining coherence estimates in the frequency domainWherein

The estimated coherence can be adjustedAnd the spectral power density function P_p2p2(f) Multiplication so as to pass the observed lower frequency F₁And a higher frequency F₂Integration of the defined frequency bands, the estimated coherence is analyzed according to:

the value Q defined in this way corresponds to the slave band [ F ]₁,F₂]Relative to the position M of the stationary sensor at the source of the reconstruction point R radiation in (1)₂The component of the measured signal power. This component is related to the reference point M₂The signal power at (the position of the fixed microphone) is correlated and can be expressed in dB.

Fig. 1 shows the reconstruction of the sound source intensity at reconstruction point R in space, and this concept can be extended to surfaces of any shape by spatially discretizing the surface into a plurality of reconstruction points and calculating the sound source intensity in each of the discretized reconstruction points. Finally, the calculated mapping of the source intensities in the reconstruction points can be superimposed with the optically acquired image of the measurement scene (for example by means of a camera) in order to achieve a spatial correspondence. In one embodiment, the surface on which the reconstruction point is located may be an image plane, which is, for example, at right angles to a rotational axis about which the moving microphone is rotated and which has a defined, predefinable distance from the stationary microphone.

Fig. 2 shows a system for locating and imaging sound sources and their intensities in a measurement scene with one or more sound sources. This system comprises an apparatus constructed as a movable frame structure 100 and a stationary data processing device connected thereto, which may be a mobile device 20 (e.g. a smartphone or a tablet) with a camera 21. Other types of data processing devices capable of receiving and transmitting data and mapping images via wired or wireless communication links may also be used.

According to one embodiment, the frame structure 100 is rotatably supported about a (fixed) shaft 54 (the shaft 54 having the rotation axis a). The stationary microphone 11 (position M) is arranged in the center of the rotational movement, i.e. on the rotational axis a₂) Whereas a plurality of (e.g. electronically multiplexed) microphones 10 are arranged along the longitudinal axis of the frame structure 100. The aforementioned longitudinal axis of the frame structure 100 is at right angles to the axis of rotation a, so that the microphone 10 moves in a circular path about the axis of rotation a as the frame structure 100 rotates about the axis 54. An electronics unit 40 can also be arranged in the frame structure 100, to which the microphones 10 and 11 are connected. Instead of a rigid shaft 54, a rotatable shaft with an axis of rotation a may also be used.

For power supply, the electronic unit 40 may have a battery 49, which is electronicThe remaining components of the cell supply the supply voltage. The charging unit 44 is used to charge the battery 49. But other forms of power supply may be used. According to the example shown in fig. 2, the electronic unit 40 further comprises (i) a (e.g. magnetic or optical) rotation angle encoder 16 for determining the angular position of the frame structure 100 with respect to the rotation axis a, (ii) sensor signals p for the stationary microphone 10 and the moving microphone 11₂(t) and p₁(t) a microphone amplifier 41 for analog pre-amplification, (iii) data acquisition means (analog-to-digital converter 42, memory 46) for digitizing and storing the sensor signals of the microphones 10, 11 and of the rotation angle encoder 16, (iv) an electronic multiplexer 45 for selecting the moving microphone 10 connected to the data acquisition means, and (v) a module 47 for wireless transmission of the acquired data to the mobile device 20 for further processing of the measurement data or for analysis by the user. The microcontroller 43 controls the multiplexer 45, the analog-to-digital converter 42, the data flow and the data transmission. Other sensor systems, such as a system comprising an angular velocity sensor and a three-axis acceleration sensor or a motion tracking system for direct detection of motion, may also be used for the spatial coordinates M of a microphone for detecting motion ₁(t)。

The mobile device 20 with the integrated camera 21 receives digitized measurement data from the electronic unit 40 with its optical axis parallel to the rotation axis a in a wireless path and transmits these measurement data, for example, over a wireless network connection to a cloud computing service for computing the sound image according to the method described above in connection with fig. 1 and superimposes the result (of the cloud computing service) on the optical image of the measurement scene acquired by the camera 21 with the sound source contained therein. These computations are outsourced to cloud computing services in order to reduce the power consumption of mobile devices, to store data continuously and persistently with continuous measurements, to simplify the accessibility of data for other users with corresponding access rights, and to enable Web-based integration of sensors in arbitrary measurement and regulation technology systems. It is also not necessary to outsource computing power to a cloud computing service, rather, the computation may be performed by any other computer (e.g., a workstation) connected to the mobile device 20.

The frame structure 100 can be rotated by a manual or electric drive 30 (motor). Alternatively, the drive can also be effected by a mechanical spring mechanism. In the example shown, the force transmission of the drive 30 is effected via a drive wheel 31 which is rigidly connected to the frame structure 100 and is rotatably mounted on the shaft 54. The shaft 54 and the driver 30 are fixed and mounted, for example, on the bracket 22. The drive wheel 31 is driven, for example, by a pinion 32 on the motor shaft of the electric motor. A controller for controlling the motor 30, which comprises a battery or a connector for external power supply, is integrated into the holder 22. The support 22 can be arranged, for example, on a tripod (as is also the case, for example, for a camera).

In one embodiment, the shell 13 (i.e. the outer sheath) of the frame structure 100 may have a cross-section (perpendicular to the longitudinal axis of the frame structure 100) having the shape of the aerodynamic wing profile 12. This has the advantage that the rotational movement of the frame structure 100 does not cause any airborne sound sources or at least substantially reduces the generation of these. In the present embodiment, four microphones 10 are arranged on the frame structure 100, and these microphones are rotated together with the frame structure 100. As described above, the microphone 11 is located at a position on the rotation axis, and therefore, the position of this microphone does not change. The microphones 10, 11 are arranged along the longitudinal axis of the frame structure 100, wherein the distance between two adjacent microphones may (but need not) be the same. In the example shown, the moving microphone 10 is arranged on one side of the frame structure 100, while the electronics unit 40 and the balancing weight 17 are arranged on the other side of the frame structure 100. The balancing weight 17 may be sized and positioned in such a way that the inherently asymmetric frame structure 100 does not have any imbalance when rotated about the rotational axis a. Even when a symmetrical design of the frame structure is used, the balancing weights 17 are generally required, since the mass of the components fixed to the frame structure 100 is asymmetrical with respect to the axis of rotation.

According to the example shown in fig. 2, the microphones 10, 11 may be fixed to the frame structure 100 by means of a resilient support structure 15. The resilient support structure 15 may help mechanically decouple the microphones 10, 11 from the frame structure 100 and prevent vibrations caused by, for example, the driver 30 or the pivot bearing 50 from being transmitted to the microphones. In other words, the elastic damping support of the microphones 10, 11 causes the mechanical vibration path to the microphones to be interrupted. In the example shown, wind-deflecting elements 14 can be provided on the housing 13 of the frame structure 100, which wind-deflecting elements cover the microphones 10, 11 in order to suppress the coupling of wind noise and signals of other air-sound sources into the sensor signal in dependence on the movement of the frame structure 100. These wind-deflecting elements 14 are optional depending on the particular application.

In summary, the function of the embodiment shown in fig. 2 can be described as follows: when the frame structure 100 is rotated about the axis a, the fixed microphone 11 does not change its position, while the other microphones 10 follow a circular path. The microphones 10, 11 detect sound waves emitted by the sound source in the form of sound pressure, while the rotation angle encoder 16 detects the spatial coordinates of the moving microphone 10. These spatial coordinates are defined by the angular position of the frame structure 100 and the (fixed) position of the microphone 10 relative to the frame structure 100. The obtained sensor signals are received by the electronic unit 40, digitized and sent to the mobile device 20. As described above, the mobile device itself can calculate the source intensity of the sound source located on the measured object from the received measurement data or can outsource this calculation to an external calculation unit. The camera 21 acquires an optical image of the object (or objects) to be measured, which can be superimposed with the calculated source intensity in order to obtain a graphical representation and correspondence of the sound source and its source intensity with respect to the optical camera image. This optical image may be recorded in black and white, for example, and the source intensities in the image may be color coded.

Fig. 3 shows another embodiment of a system for locating and imaging a sound source on an object to be measured. The system shown in fig. 3 differs from the system shown in fig. 2 only in the design of the frame structure 200. In the present example, instead of four microphones 10 which are movable along a circular path, only one movable microphone 10 is provided, which is supported on the frame structure 200 in a radially movable manner (along its longitudinal axis). Thus, the distance between the microphone 10 and the axis of rotation (and the radius of the circular movement of the microphone 10) can be varied. Other arrangements of microphones, in particular having a plurality of movably supported microphones 10, may also be used.

By varying (enlarging or reducing) the radial distance between the microphone 10 and the rotation axis a during rotation of the frame structure 200, the microphone 10 effectively performs a spiral movement around the rotation axis a. Adjustability of the distance between the microphone 10 and the axis of rotation a (i.e. the position of the microphone 10 relative to the frame structure 200) may be achieved, for example, by a pull cord arrangement 60. In this case, the cable 61 can be connected to a microphone holder 65 which is supported in a linearly displaceable manner on the frame structure 200. For this purpose, the frame structure 200 may, for example, have two guide rods 62, which are substantially parallel to the longitudinal axis of the frame structure and along which the microphone carrier 65 can slide. Thus, the guide rods 62 together with the microphone holder form a linear guide for the microphone 10. In the example shown, the cable 61 is guided around a plurality of deflection rollers 63 and a disc 66 rigidly connected with the shaft 54. In the example shown, this rope portion passes through one of the (hollow) guide rods 62.

When the frame structure 200 is rotated about the fixed axis 54, the cord 61 is unwound on the circumference of the disc 66, which displaces the microphone holder 65, which in turn causes an approximately helical movement of the microphone 10, wherein the radial position of the moving microphone 10 can be unambiguously assigned to the measured angular position of the frame structure; the microphone will move a distance corresponding to the circumference of the disc 65 for each complete rotation of the frame structure. As in the previous example, the frame structure has a housing 213 which encloses a pull cord arrangement 60 consisting of a cord 61, a deflecting roller 63, a disc 65, a guide rod 62 and a microphone holder 65 and has an elongated opening (slit) for the microphone 10. Otherwise, the example shown in fig. 3 is the same as the previous example shown in fig. 2.