METHOD FOR GENERATING AN EXCITER SIGNAL AND FOR ACOUSTIC MEASURING IN TECHNICAL HOLLOW SPACES

Abstract

The invention relates to a method for acoustic measuring in technical hollow spaces, for example, for measuring reflection points along long pipelines. The method begins with establishing a broadband exciter frequency range. This is followed by establishing interference frequencies which should not be in the exciter signal, and generating a precursor signal over the frequency range with the omission of the interference frequencies. Then, the precursor signal is coupled into the pipeline and a precursor reflection signal reflected out of the pipeline is received. The precursor signal is compared with the precursor reflection signal and damping frequencies are determined at which the precursor signal on the measuring path along the pipeline is damped much more than the average damping of the entire precursor signal. Then, an exciter signal is generated over the exciter frequency range with the omission of the interference frequencies and the damping frequencies. This is followed by coupling the exciter signal into the pipeline and receiving a measuring signal reflected out of the pipeline, as well as evaluating the reflected measuring signal using suitable evaluation methods.

The invention also relates to a method for generating an exciter signal and a device for carrying out this method.

Claims

1. A method for acoustic measuring in or along a pipeline, comprising the steps of: establishing a broadband exciter frequency range in which an exciter signal is to be generated; establishing interference frequencies which should not be in the exciter signal; generating a precursor signal over the exciter frequency range with the omission of the interference frequencies; coupling the signal precursor into the pipeline and receiving a precursor reflection signal reflected from the pipeline; comparing the precursor signal with the precursor reflection signal and determining damping frequencies at which the precursor signal on the measuring path along the pipeline is damped much more than the average damping of the entire precursor signal; generating an exciter signal over the exciter frequency range with the omission of the interference frequencies and the damping frequencies; coupling the exciter signal into the pipeline and receiving a measuring signal reflected out of the pipeline; evaluating the reflected measuring signal using suitable evaluation methods, wherein the exciter signal is generated as a multi-sine burst by the following steps: 11. Establishing a burst duration TB; 12. Establishing a frequency mask of the multi-sine test signal; 13. Determining the time signal by means of inverse Fourier transform; 14. Adding zeros to the desired signal length T0 to get a burst signal; 15. Calculating the complex amplitude spectrum of the burst by means of Fourier transform; 16. Modifying the spectrum by setting the amplitudes of the spectral components as established in step 12 with retaining the values of the phase angles as they resulted in the spectrum according to step 15; 17. Determining the time signal by means of inverse Fourier transform; 18. Restoring the pulse pause by setting all related signal components to zero; 19. Jumping back to step 15 and repeating the steps until the spectral mask and the pulse pause are complied with an accuracy to be specified.

2. The method according to claim 1, wherein also the precursor signal is generated as a multi-sine burst.

3. The method according to claim 1, wherein the exciter frequency range is established in the range of 1 Hz to 1 kHz, preferably ranging from a lower cut-off frequency f.sub.u=5 Hz up to an upper cut-off frequency f.sub.o =300 Hz.

4. The method according to claim 1, wherein measuring acoustic interference signals occurring at the pipeline without coupling of the precursor or the exciter signal, is carried out for establishing interference frequencies, wherein the frequencies of the measured interference signals are established above a predetermined interference signal threshold as interference frequencies.

5. The method according to claim 1, wherein those frequencies are determined as the damping frequencies for which the damping is at least greater by a factor of 2 than the average damping of the precursor signal.

6. The method according to claim 1, wherein a test signal is generated which has a frequency within the exciter frequency range which frequency correlates with periodic reflection points in the pipeline, and in that the test signal is coupled into the pipeline in order to determine the speed of propagation taking into account a known distance between the periodic reflection points.

7. A device for acoustic measuring in or along a pipeline, comprising: a unit for generating an exciter signal; a sound generator for coupling the exciter signal generated into the hollow space; a sensor for detecting a measuring signal reflected from the hollow space; an analysis unit for evaluating the measuring signal; wherein the analysis unit is configured to carry out a method according to claim 1.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Further advantages and details of the invention will become apparent from the following description of a preferred embodiment, with reference to the drawing.

[0034] FIG. 1 shows steps of the method according to the invention as a flowchart and related signal profiles;

[0035] FIG. 2 shows a spectrum of a multi-sine signal with any desired spectral components;

[0036] FIG. 3 shows the time course of a multi-sine signal consisting of three sine components;

[0037] FIG. 4 shows the time course and spectrum of a sine burst;

[0038] FIG. 5 shows the time course and spectrum of a multi-sine burst;

[0039] FIG. 6 shows the time course and spectrum of a multi-sine burst in which the sine component f.sub.2 has been shifted by 180°;

[0040] FIG. 7 shows a flowchart for optimizing a multi-sine burst;

[0041] FIG. 8 shows exemplary signal profiles occurring in the steps according to FIG. 6;

[0042] FIG. 9 shows the time course and spectrum of the optimized multi-sine burst;

[0043] FIGS. 10 A-C show a comparison of a classic pulse technique, method with changing frequency (chirp) and multi-sine burst with a plurality of frequency bands.

DETAILED DESCRIPTION OF THE INVENTION

[0044] The preferred embodiment described hereinafter in exemplary fashion is based on the use of a broadband multi-sine burst as the exciter signal, said multi-sine burst having a spectrum optimized according to the invention in order to selectively exclude interference frequencies and possibly damping frequencies.

[0045] A multi-sine is a broadband periodic signal in which the spectral power of the individual frequency components can be easily set.

[0046] FIG. 1 shows essential steps of the method according to the invention as a flowchart and the associated signal profiles in a simplified form. The method for acoustic measuring, in particular of reflection points, in or along long pipelines begins at step 01 with the establishing of a broadband exciter frequency range in which a desired exciter signal is to be formed. In step 02, interference frequencies are established or defined which should not be in the exciter signal. In step 03, a precursor signal is generated over the exciter frequency range with the omission of the previously established interference frequencies. In step 04, coupling the generated precursor signal into a pipeline to be examined takes place. In step 05, a precursor reflection signal reflected from the pipeline is received. In step 06, the actual desired exciter signal is generated, for which purpose a comparison of the precursor signal with the precursor reflection signal and a determining of damping frequencies at which the precursor signal on the measuring path along the pipeline is damped much more than the average damping of the entire precursor signal are carried out and wherein generating the exciter signal over the exciter frequency range is carried out with the omission of the interference frequencies and the damping frequencies. The details of generating the exciter signal are described in greater detail below with reference to FIG. 7. In step 07, the exciter signal generated is coupled into the pipeline. In step 08, a measuring signal reflected from the pipeline is received. And finally, in step 09, the reflected measuring signal is evaluated using appropriate evaluation methods.

[0047] FIG. 2 shows the procedure using the spectrum of a multi-sine signal with any desired spectral components. Assuming a period duration of the multi-sine signal of T.sub.0, multiple sine signals can be accommodated within that time period if the integral multiple of the period duration T.sub.0 of the sine signals corresponds to the period duration T.sub.0 of the multi-sine signal. Therefore, the following must apply:

kT.sub.s=T.sub.0 or T.sub.0f.sub.s=k

[0048] In this case, f.sub.s is the frequency of the corresponding sine component. Therefore, any number of sine components can be accommodated in the entire signal as long as their frequency is an integer multiple of Δf=1/T.sub.0. In FIG. 1, the permitted frequencies are labelled by small circular areas. In the example shown, only 3 of the possible frequencies are actually used. The strength of each individual sine component can be predetermined by the height of the spectral lines. In the example, two sine signals have been selected to have the same strength, a third has been selected to be somewhat weaker.

[0049] FIG. 3 shows the time signal which results from the superposition of the three sine components selected in FIG. 1. It is shown over three periods. In the case shown, the individual spectral lines are independent of one another, so that any desired performance spectrum can be specified very flexibly. By inverse Fourier transform the time signal can be calculated therefrom which finally is converted into a voltage signal with the predetermined spectrum by means of a DA converter.

[0050] FIG. 4 shows the time course and the spectrum of a sine burst. In the case of a burst signal, the aforementioned independence of the spectral lines is no longer given. A sine signal of frequency f.sub.B is sent out only during the time interval T.sub.B. The spectrum of this signal now consists not only of the individual spectral line of the frequency f.sub.B (dashed line) but scatters over a certain range around this frequency. The smaller T.sub.B is selected, the stronger the scattering.

[0051] FIG. 5 shows the time course and the spectrum of a multi-sine burst, wherein the burst signal consists of the same sine components as shown in FIG. 1. The intended sine components are marked by dotted lines. A number of spectral components arise that were not originally intended. These arise from the superposition of the individual spectra (see lower diagram in FIG. 3) of the sine bursts at different frequencies. The resulting overall spectrum is therefore difficult to predict.

[0052] Changing the phasing of the three sine components relative to one another may result in another overall spectrum because spectral components can superimpose constructively or destructively.

[0053] FIG. 6 shows the time course and the spectrum of a multi-sine burst, wherein the sine component f.sub.2 has been shifted by 180°.In the example shown, the same sine components as in FIG. 4 have been used. Only the phase of the frequency component f.sub.2 was rotated by 180° . As can be seen, the change in the phasing has led to a considerable modification of the overall spectrum. It can be deduced that by a suitable choice of the phases of the sine components, for burst signals also it is possible to comply with a predetermined frequency mask at least approximately. As a result of the superposition of many spectral lines, the determination of suitable phase values is a complex optimization problem. One possibility for an approximate solution to this optimization problem is an iterative process, as described below.

[0054] FIG. 7 shows a flowchart for the approximate solution of the above optimization problem with the illustration of typical signal profiles and spectra. The sequence consists of the following steps:

[0055] 11. Establishing the burst duration T.sub.B.

[0056] 12. Establishing the frequency mask of the multi-sine test signal. In the example, the signal should only have spectral components within the areas B1 and B2.

[0057] 13. Determining the time signal by means of inverse Fourier transform. The phases of the individual spectral components can be chosen randomly.

[0058] 14. Adding zeros to the desired signal length T.sub.0 to get a burst signal.

[0059] 15. Calculating the complex amplitude spectrum of the burst by means of Fourier transform.

[0060] 16. Since the spectrum deviates from the intended shape, the amplitudes of the spectral components must be set again as established in step 12. However, those values are to be retained as the phase angle as they resulted in the spectrum according to step 15.

[0061] 17. Determining the time signal by means of inverse Fourier transform.

[0062] 18. Due to the changes in the amplitude spectrum, the signal no longer has a pronounced pulse pause. To restore such pulse pause, all related signal components are set to zero.

[0063] 19. Jumping back to step 15 with this signal and repeating the process until the spectral mask and the pulse pause are complied with an accuracy to be specified.

[0064] FIG. 8 shows exemplary signal profiles as they occur in the steps described above.

[0065] FIG. 9 shows the time course and the spectrum of the optimized multi-sine burst, thus the result of the optimization according to FIG. 6. With the exception of minor deviations, the test signal complies with both the spectral and the temporal mask.

[0066] For the final clarification of the differences between the present invention and the well-known methods, FIG. 10 shows in each case the signal profile in comparison to the classic pulse technique (FIG. 10A), methods with changing frequency (chirp from f.sub.u to f.sub.o) (FIG. 10B), and the method according to the invention multi-sine burst with a plurality of frequency bands (FIG. 10C).

[0067] As the person skilled in the art, can find detailed instructions for the construction of individual units of a suitable device, as well as mathematical and metrological procedures in the signal analysis, for example from the above cited EP 2 169 179 B1, the repetition of these well-known aspects is dispensed with for the most part.