Method and apparatus for determining the absolute value of the flow velocity of a particle-transporting medium

Abstract

The invention relates to a method for determining the absolute value of the flow velocity (v) of a particle-transporting medium. At least two measurement laser beams (L_i) with linearly independent, non-orthogonal measurement directions (b_i) are emitted. The measurement laser beams (L_i) scattered at particles are detected and one measurement signal (m_i) is generated in each case for each measurement laser beam (L_i). The measurement signals (m_i) are evaluated, wherein absolute values of velocity components (v_i) are ascertained as projections of the flow velocity (v) on the respective measurement directions (b_i), wherein a solid angle region is ascertained for the prevalent direction of the flow velocity (v) and signs assigned to this solid angle region are chosen for the individual velocity components (v_i), and wherein the absolute value of the flow velocity (v) is determined using the ascertained absolute values of the velocity components (v_i) and using the chosen signs for the velocity components (v_i).

Claims

1. A method for determining the absolute value of the flow velocity of a particle-transporting medium, comprising the steps of: emitting at least two measurement laser beams with linearly independent, non-orthogonal measurement directions; detecting the measurement laser beams scattered at particles and generating one measurement signal in each case for each measurement laser beam; and evaluating the measurement signals, a. wherein absolute values of velocity components are ascertained as projections of the flow velocity on the respective measurement directions; b. wherein a solid angle region is ascertained for the prevalent direction of the flow velocity and signs assigned to this solid angle region are chosen for the individual velocity components; and c. wherein the absolute value of the flow velocity is determined using the ascertained absolute values of the velocity components and using the chosen signs for the velocity components.

2. The method according to claim 1, characterized in that at least three measurement laser beams with different measurement directions are emitted, of which at least three are non-orthogonal with respect to one another and three are linearly independent; and in that the value of the flow velocity is determined using the ascertained absolute values of the velocity components for three linearly independent measurement directions and using the chosen signs for the velocity components.

3. The method according to claim 2, characterized in that each of the three linearly independent measurement directions is respectively assigned two mutually opposing predefined solid angle regions and in that the remaining two predefined solid angle regions cover the remaining measurement surroundings.

4. The method according to claim 3, characterized a. in that the at least three absolute values of the velocity components are compared to one another, b. in that the maximum absolute value of the velocity components is ascertained and c. in that, if the difference between the maximum absolute value of the velocity components and the remaining absolute values of the velocity components exceeds a predetermined first limit value, the predetermined solid angle region that is assigned to the measurement direction with the maximum absolute value of the velocity components is determined as solid angle region for the prevalent direction of the flow velocity.

5. The method according to claim 1, characterized in that the ascertained absolute values of the velocity components are transformed into an orthogonal coordinate system using the ascertained signs for the velocity components and the flow velocity of the particle stream is calculated in the orthogonal coordinate system, at least apart from the sign thereof.

6. The method according to claim 1, characterized in that the solid angle region for the prevalent direction of the flow velocity is ascertained as one of eight predefined solid angle regions, which together cover the entire measurement surroundings, and in that respectively two predefined solid angle regions lying opposite one another are assigned the same signs for the individual velocity components.

7. The method according to claim 6, characterized in that: at least one of a contrast value of the associated particle count rates and the mean particle dwell times is calculated for each pair of measurement directions, and the solid angle region for the prevalent direction of the flow velocity is ascertained as a predefined solid angle region that is assigned to no measurement direction if at least one of the contrast values exceeds a predetermined second or third limit value.

8. The method according to claim 1, characterized in that a particle count rate is determined for every measurement direction and in that the differences between the particle count rates are taken into account when ascertaining the solid angle region for the prevalent direction of the flow velocity.

9. The method according to claim 8, characterized in that: at least one of a contrast value of the associated particle count rates and the mean particle dwell times is calculated for each pair of measurement directions, and the solid angle region for the prevalent direction of the flow velocity is ascertained as a predefined solid angle region that is assigned to no measurement direction if at least one of the contrast values exceeds a predetermined second or third limit value.

10. The method according to claim 1, characterized in that a mean particle dwell time in the region detected by the corresponding measurement laser beam is determined for each measurement direction and in that the differences between the mean particle dwell times are taken into account when ascertaining the solid angle region for the prevalent direction of the flow velocity.

11. The method according to claim 1, wherein at least four measurement laser beams are emitted in different measurement directions, characterized in that, for at least two different sets of three linearly independent and non-orthogonally aligned measurement directions, a solid angle region for the prevalent direction of the flow velocity and an absolute value of the flow velocity are ascertained as estimates in each case and in that the prevalent direction of the flow velocity and the absolute value of the flow velocity are then determined by a consistency check.

12. The method according to claim 11, characterized in that the absolute value of the at least one velocity component not belonging to the set is ascertained on the basis of the corresponding measurement signal, in that a hypothetical absolute value for the at least one velocity component not belonging to the set is determined on the basis of the estimates, ascertained for the set, of the solid angle region for the prevalent direction of the flow velocity and of the absolute value of the flow velocity, and in that the absolute value and the hypothetical absolute value for the at least one velocity component not belonging to the set are compared to one another in order to assess the accuracy of the respective estimates.

13. The method according to claim 11, wherein a solid angle region for the prevalent direction of the flow velocity is ascertained in each case for at least three different sets of three linearly independent and non-orthogonally aligned measurement directions, characterized in that, for the purposes of determining the absolute value of the flow velocity, the solid angle region that has been ascertained for the plurality of sets is used for the prevalent direction of the flow velocity.

14. The method according to claim 1, characterized in that the measurement signals are generated by self-mixing interference between the emitted measurement laser beams and the corresponding scattered measurement laser beams and in that the absolute values of the velocity components are each determined on the basis of the frequency shift between the emitted measurement laser beam and the corresponding scattered measurement laser beam.

15. Use of the method according to claim 1 for determining a particle number per unit measurement volume, wherein the measurement volume is determined on the basis of the ascertained absolute value of the flow velocity.

16. An apparatus for determining the absolute value of the flow velocity of a particle-transporting medium, comprising: an optical emitter device configured to emit at least three measurement laser beams with different measurement directions, of which at least three are non-orthogonal with respect to one another and three are linearly independent; a detector device configured to detect the measurement laser beams that have been scattered at particles, and to in each case generate a measurement signal for each measurement laser beam; and an evaluation device for evaluating the measurement signals, said evaluation device being configured a. to ascertain absolute values of velocity components as projections of the flow velocity on the respective measurement directions; b. to ascertain a solid angle region for the prevalent direction of the flow velocity and choose signs assigned to this solid angle region for the individual velocity components; and c. to determine the absolute value of the flow velocity using the ascertained absolute values of the velocity components for three linearly independent measurement directions and using the chosen signs for the velocity components.

17. The apparatus according to claim 16, characterized in that the optical emitter device comprises at least three laser diodes and in that the detector device comprises at least one photodiode for each of the three laser diodes integrated into the laser diode.

18. The apparatus according to claim 17, characterized in that the evaluation device is configured to determine the absolute values of the velocity components-on the basis of the frequency shift, ascertained by beam interference, between the emitted measurement laser beam and the corresponding scattered measurement laser beam.

19. The apparatus according to claim 16, characterized in that the evaluation device is further configured to determine one or more of a particle count rate and a mean particle dwell time for each measurement direction.

20. The apparatus according to claim 16 further comprising a first memory device for storing measurement events and measurement quantities, wherein the detection of a particle is referred to as a measurement event and one or more of the occurring measurement direction, velocity components, count rates and dwell times are stored as measurement quantities.

21. The apparatus according to claim 20, further comprising at least one further memory device, in which is stored one or more of at least one first limit value for the difference between the maximum velocity component and the remaining velocity components, and at least one second limit value for the contrast value of the particle count rates, and at least one third limit value for the contrast value of the particle dwell times.

22. The apparatus according to claim 16, wherein the evaluation device is further configured to determine the measurement volume on the basis of the ascertained absolute value of the flow velocity.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) In detail:

(2) FIG. 1 shows a schematic oblique view of an apparatus for determining the absolute value of the flow velocity of a medium according to one embodiment of the invention;

(3) FIG. 2 shows an illustration of the dependence of the error of the velocity determination as a function of a zenith angle and an azimuth angle of the flow velocity under an assumption of orthogonality;

(4) FIG. 3 shows an illustration of the dependence of the error of the velocity determination as a function of the zenith angle and the azimuth angle of the flow velocity if the signs of the velocity components are not taken into account;

(5) FIG. 4 shows an illustration of the dependence of the signature on the frequency shifts for flow velocities in the plane;

(6) FIG. 5 shows an illustration of the dependence of the signature on the zenith angle and the azimuth angle of the flow velocity;

(7) FIG. 6 shows an illustration of the dependence of the error of the velocity determination as a function of the zenith angle and the azimuth angle of the flow velocity when determining the signs on the basis of the absolute values of the velocity components;

(8) FIG. 7 shows an illustration of the selection of the signature S.sub.0 as a function of the zenith angle and the azimuth angle of the flow velocity when determining the signs on the basis of the particle count rate using a limit value;

(9) FIG. 8 shows an illustration of the selection of the signature S.sub.0 as a function of the zenith angle and the azimuth angle of the flow velocity when determining the signs on the basis of the particle count rate using a further limit value;

(10) FIG. 9 shows an illustration of the dependence of the error of the velocity determination as a function of the zenith angle and the azimuth angle of the flow velocity when determining the signs on the basis of the particle count rate using a limit value;

(11) FIG. 10 shows an illustration of the dependence of the error of the velocity determination as a function of the zenith angle and the azimuth angle of the flow velocity when determining the signs on the basis of the particle count rate using a further limit value;

(12) FIG. 11 shows an illustration of the dependence of the error of the velocity determination as a function of the zenith angle and the azimuth angle of the flow velocity when determining the signs on the basis of the absolute values of the velocity components and on the basis of the particle count rate using a limit value;

(13) FIG. 12 shows an illustration of the dependence of the error of the velocity determination as a function of the zenith angle and the azimuth angle of the flow velocity when determining the signs on the basis of the absolute values of the velocity components and on the basis of the particle count rate using a further limit value; and

(14) FIG. 13 shows a flowchart for explaining a method for determining the absolute value of the flow velocity of a medium according to one embodiment of the invention.

(15) Equivalent or functionally equivalent elements and apparatuses are denoted by the same reference signs in all figures.

DESCRIPTION OF THE EXEMPLARY EMBODIMENTS

(16) FIG. 1 shows a schematic oblique view of an apparatus 1 for determining the absolute value of the flow velocity v of a particle-transporting medium 5. The flow velocity v should be understood to mean the velocity relative to the apparatus 1. To this end, the apparatus 1 can be situated in the medium 5; however, the apparatus 1 can also be spaced apart from the medium 5, wherein the medium 5 may be, for instance, ambient air with particles, a gas with particles flowing past the apparatus 1 or a liquid with particles flowing past the apparatus 1.

(17) The apparatus 1 comprises an optical emitter device 2 with three laser diodes 21, 22, 23, which are embodied as VCSELs and which emit measurement laser beams L.sub.i, i=1, 2, 3 perpendicular to a sensor plane that is set by a substrate of the apparatus 1. A lens or any other optical system steers the measurement laser beams L.sub.1,L.sub.2, L.sub.3 in three measurement directions b.sub.1,b.sub.2,b.sub.3. The measurement directions b.sub.i of the measurement laser beams L.sub.i differ and are linearly independent of one another. Further, the measurement directions b.sub.i are pairwise non-orthogonal and preferably include acute angles in pairwise fashion, as a result of which a compact structure of the apparatus 1 can be achieved. In particular, use can be made of space-saving optical elements and individual optical units for each individual measurement laser beam L.sub.i may be economized where appropriate by using common optical units.

(18) The apparatus 1 further comprises a detector device 3 with three photodiodes 31, 32, 33, which are integrated into the laser diodes 21, 22, 23 and which detect the measurement laser beams L.sub.i that are scattered at particles. The detector device 3 generates an associated measurement signal m.sub.i for each measurement laser beam L.sub.i. Preferably, as shown in FIG. 1, the optical emitter device 2 comprises a number of laser diodes that correspond to the measurement laser beams L.sub.i, said laser diodes preferably possibly being VCSELs. The light emitted by the laser diodes can be aligned in the respective measurement directions b_i by means of optical elements such as lenses, prisms or mirrors. However, it is also possible for the number of laser diodes to be smaller than the number of measurement laser beams L.sub.i if the measurement laser beams L.sub.i are generated by way of splitting the beam of the emitted laser beams.

(19) As shown in FIG. 1, the detector device 3 preferably comprises a number of photodiodes corresponding to the laser diodes, wherein the photodiodes can preferably each be integrated into a laser diode. In this case, each photodiode respectively detects the result of the interference of the generated measurement laser beam L.sub.i with the measurement laser beam L.sub.i that has been scattered at particles in the medium, with frequency shifts arising in the process. The detector device 3 measures the size of the frequency shift, which depends on the absolute value of the velocity component v.sub.i of the particles in the measurement direction b_i, i.e., in the direction of the measurement laser beam L.sub.i. Consequently, the detector device 3 can determine the associated absolute value of the velocity component v.sub.i on the basis of the frequency shift for each measurement direction b_i and can output said absolute value as a measurement signal m.sub.i. In the context of the invention, a velocity component v.sub.i should be understood to mean the projection of the flow velocity v on the respective measurement direction b_i.

(20) Three linearly independent measurement directions form an oblique coordinate system, wherein the flow velocity v can be considered to be a vector that points away from the origin of this coordinate system. Each of the measurement directions divides the space into two half-spaces by means of a plane that passes through the origin and that is perpendicular to the measurement direction, specifically into a positive half-space, in which vectors of the flow velocity v have a positive velocity component v.sub.i, and into a negative half-space, in which vectors of the flow velocity v have a negative velocity component v.sub.i. The planes that separate the half-spaces from one another intersect for the various measurement directions, and so a total of 2.Math.2.Math.2=8 solid angle regions arise. For orthogonal measurement directions, i.e., Cartesian coordinate systems, these solid angle regions would precisely correspond to the octants known from the geometry.

(21) Consequently, one of the possible combinations of signs of the velocity components v.sub.i, which is referred to as signature, corresponds to each solid angle region. As described above, signatures that only differ by a common sign should be considered to be equivalent for the calculation of the absolute value of the flow velocity v.

(22) An evaluation device 4 of the apparatus 1 is embodied to ascertain the solid angle region into which the vector of the flow velocity v points, i.e., a solid angle region corresponding to the prevalent direction of the flow velocity v. To this end, the evaluation device 4 can compare the absolute values of the velocity components v.sub.i with one another and/or take account of a particle count rate cr.sub.i and/or a mean particle dwell time dt.sub.i in relation to the respective measurement directions. These evaluation methods are explained in more detail below with reference to FIGS. 5 to 13.

(23) The evaluation device 4 now chooses the signs for the velocity components v.sub.i that are assigned to the ascertained solid angle range. Consequently, both the absolute values of the velocity components v.sub.i and the signs of the velocity components v.sub.i are known; i.e., the evaluation device 4 ascertains the velocity components v.sub.i themselves. The evaluation device 4 can calculate the value of the flow velocity v by means of Formula (2) specified above, wherein the velocity components v.sub.i can be converted into an orthogonal system by means of the matrix M that is predetermined by the measurement directions. According to further embodiments, the evaluation device 4 can also calculate the flow velocity v itself in addition to the absolute value of the flow velocity v; i.e., said evaluation device can calculate the vectorial form apart from the sign. If the flow velocity v should be specified in an orthogonal coordinate system, then the velocity components v.sub.i, which are already known relative to the measurement directions, can be converted into the orthogonal coordinate system with the aid of the matrix M. Consequently, the evaluation device 4 can be embodied to also determine the flow velocity v itself, apart from the sign, in addition to the absolute value of the flow velocity v.

(24) FIG. 2 shows, as a function of a zenith angle θ measured in relation to the sensor plane and an azimuth angle φ of the flow velocity v, the dependence of the error of the velocity determination if the non-orthogonality of the measurement directions is simply neglected during the calculation; i.e., if the mixed terms v.sub.iv.sub.j with the unknown signs, occurring in Formula (1), are set to 0. Plotted is the ratio r of the calculated flow velocity v to the actual flow velocity v.

(25) FIG. 3 shows, as a function of the zenith angle and the azimuth angle of the flow velocity v, the dependence of the error of the velocity determination if the correct signs of the occurring mixed terms v.sub.iv.sub.j are not taken into account in the calculation according to Formula (1), for example if all are set to 1 or +.

(26) As may be gathered from FIGS. 2 and 3, significant errors may occur during the calculation; this is why the signs and/or the relative signs, i.e., the signature, are likewise ascertained according to the invention. Relative signs should be understood to mean that the product of the signs of the velocity components v.sub.i is assigned to each pair of velocity components v.sub.i as a relative sign. Expressed differently, all that matters is the difference between the signs while an absolute sign is irrelevant on account of the symmetry when calculating the value of the flow velocity.

(27) FIG. 4 shows the dependence of the signature on the frequency shifts for flow velocities v lying in a xy-plane that extends parallel to the sensor plane. The planes that separate the half-spaces from one another are straight lines in this two-dimensional case, said straight lines separating six angle regions from one another that correspond to the solid angle regions to which the signatures S.sub.1, S.sub.2 and S.sub.3 are assigned. The maximum frequency shift Δf.sub.i corresponds to a maximum absolute value of the velocity components v.sub.i. Now, for example, should the frequency shift Δf.sub.1 be at a maximum in respect of a first measurement direction b.sub.1, i.e., should said frequency shift be greater than the frequency shifts in the two other measurement directions, or should, equivalently, the absolute value of the velocity components v.sub.i be at a maximum in respect of the first measurement direction b.sub.1, then this means that the angle between the flow velocity v and the first measurement direction b.sub.1 is smaller than the angles between the flow velocity v and the two other measurement directions. For a symmetrical arrangement of three measurement directions, i.e., if the angles between respectively two measurement directions all have the same size, this is tantamount to the flow velocity v pointing into an angle region which is assigned the signature S.sub.1. Preferably, the measurement directions are consequently symmetrical with respect to one another. In the case of a non-symmetrical arrangement, the boundaries of the solid angle regions are displaced accordingly.

(28) Then, the evaluation device 4 can be embodied to ascertain the velocity component v.sub.i(max) that has the largest absolute value. Then, the evaluation device 4 chooses the solid angle region of the associated measurement direction and chooses the signature assigned to the solid angle region, i.e., chooses the combination of signs of the velocity components v.sub.i assigned to the solid angle region.

(29) FIG. 5 shows the dependence of the signature on the zenith angle and the azimuth angle of the flow velocity v. As may be seen, the signature equals S.sub.0 for zenith angles in the vicinity of the poles. The reason for this is that the flow velocity v in this case is always located in the positive half-space or always located in the negative half-space, i.e., it substantially always points in the same direction as the measurement directions or always points in the opposite direction to the measurement directions.

(30) FIG. 6 shows the error of the velocity determination as a function of the zenith angle and the azimuth angle of the flow velocity v, wherein the signs are ascertained according to the method, just described above, on the basis of the absolute values of the velocity components v.sub.i. The method supplies very good values for flow velocities v that are substantially parallel to the sensor plane. For angles that are more pronounced, the error is greater since the signature S.sub.0 is not identified by the above-described method.

(31) Therefore, the evaluation device 4 can be alternatively or additionally embodied to determine the signature on the basis of a particle count rate cr.sub.i of the particles in the measurement laser beam or on the basis of a mean particle dwell time dt.sub.i of the particles in the measurement laser beam. Should a contrast value of the particle count rate cr.sub.i or a contrast value of the particle dwell time dt.sub.i exceed a respective predetermined second or third limit value for a pair of measurement directions, the evaluation device 4 can recognize that the signature equal to S.sub.0 must be selected. Otherwise, the evaluation device 4 can select any signature S.sub.1, S.sub.2 or S.sub.3. By way of example, it is possible to always select the signature S.sub.1.

(32) The contrast rate within the scope of the determination on the basis of the particle count rate cr.sub.i depends on the particle density since the latter influences the particle count rate cr.sub.i. The decision as to whether the particle count rates cr.sub.i differ significantly (for instance, at a 95% confidence level) can be implemented by means of hypothesis testing on the basis of the underlying Poisson distribution for each measurement direction. The time duration of the measurement of the particle count rates cr.sub.i may therefore depend on the particle density and may be chosen to be longer with lower particle density. Such a dependence on the particle density does not exist for the particle dwell time dt.sub.i, and so the latter can be used for determining the signature, particularly in the case of lower densities.

(33) FIG. 7 shows the selection of the signature S.sub.0 as a function of the zenith angle and the azimuth angle of the flow velocity v on the basis of a contrast value of the particle count rates cr.sub.i using the second limit value.

(34) FIG. 8 differs from FIG. 7 in that a higher value was chosen as a second limit value.

(35) FIG. 9 shows, as a function of the zenith angle and of the azimuth angle of the flow velocity, the dependence of the error of the velocity determination. Plotted is the ratio of the calculated absolute value of the flow velocity v to the actual absolute value of the flow velocity v. Here, the second limit value was chosen like in FIG. 7.

(36) FIG. 10 shows, as a function of the zenith angle and of the azimuth angle of the flow velocity v, the dependence of the error of the velocity determination. Here, the second limit value was chosen like in FIG. 8.

(37) Further, a combination of the methods is possible. Thus, whether the signature equals S.sub.0 can be initially ascertained on the basis of a contrast value of the particle count rate cr.sub.i and/or on the basis of a contrast value of the particle dwell time dt.sub.i. If this is not the case, the evaluation device 4 can ascertain the signature S.sub.1, S.sub.2 or S.sub.3, as described above, on the basis of determining the maximum velocity component v.sub.i.

(38) Further, provision can also be made for the signature to be chosen to equal S.sub.0, even as soon as this is yielded by only one of the determinations on the basis of the particle count rate cr.sub.i and on the basis of the particle dwell time dt.sub.i. Only if both determinations on the basis of the contrast values yield that the signature is not equal to S.sub.0 does the evaluation device 4 ascertain the signature S.sub.1, S.sub.2 or S.sub.3 on the basis of determining the maximum velocity component v.sub.i.

(39) FIG. 11 shows the dependence of the error of the velocity determination on the zenith angle and on the azimuth angle of the flow velocity v in such a combined method. Here, the second limit value was chosen like in FIG. 7.

(40) FIG. 12 shows the dependence of the error of the velocity determination on the zenith angle and on the azimuth angle of the flow velocity v in such a combined method. Here, the second limit value was chosen like in FIG. 8.

(41) Furthermore, it is possible to emit more than three measurement laser beams L.sub.i. The redundancy occurring in this case can help with improving the measurement results, for instance by virtue of at least one of the above-described consistency checks being carried out by means of the evaluation device 4. In particular, provision can be made of a fourth measurement laser beam L.sub.4, which is perpendicular to the sensor plane, for example.

(42) FIG. 13 shows a flowchart of a method for determining the absolute value of the flow velocity v of a medium according to one embodiment of the invention.

(43) In a first step S1, at least two measurement laser beams L.sub.i, preferably at least three measurement laser beams L.sub.i, are emitted in linearly independent measurement directions b_i, which are not orthogonal to one another.

(44) In a second step S2, the measurement laser beams L.sub.i scattered at particles are detected. A measurement signal m.sub.i is generated for each measurement laser beam L.sub.i. The measurement signals m.sub.i are evaluated. By way of example, the measurement signals m.sub.i can comprise frequency shifts in the case of an interference between the emitted measurement laser beams L.sub.i and the scattered measurement laser beams L.sub.i.

(45) In a third step S3, the measurement signals m.sub.i are evaluated, wherein, on the basis of the frequency shifts, the absolute values of velocity components v.sub.i are ascertained as projection of the flow velocity v on the respective measurement direction b_i. Further, a solid angle region, in which the prevalent direction of the flow velocity v is located, is ascertained. This solid angle region is selected from a number of predetermined solid angle regions which, together, cover the entire measurement surroundings of the measurement arrangement. A signature is assigned to each of these predetermined solid angle regions, and so signs can be assigned to the respective absolute values of the velocity components v.sub.i by way of ascertaining the solid angle region in which the prevalent direction of the flow velocity is located. Consequently, it is then also possible to determine the absolute value of the flow velocity v.

(46) According to further embodiments, provision can be made for at least four measurement laser beams L.sub.i to be emitted. Respectively one absolute value of the flow velocity v is calculated as an estimate according to the method described for different sets of three measurement directions in each case, said measurement directions being linearly independent and non-orthogonal. The estimated absolute values of the flow velocity v are compared to one another and a final value for the absolute value of the flow velocity v is ascertained within the scope of a consistency check.

(47) By way of example, the sign combination can be ascertained for one of the sets such that the flow velocity v is calculated in vectorial form using the absolute values of the velocity components v.sub.i. Then, the flow velocity v is projected on a measurement direction that is not part of the set and the absolute value is formed so as to determine a hypothetical absolute value of the velocity component v.sub.i for this measurement direction. This value is compared to the absolute value of the velocity component v.sub.i determined on the basis of the interference measurements. If the deviation is below a predetermined threshold value, the calculation of the signs is assessed as being correct and the absolute value of the flow velocity v determined on the basis of the set is output. Additionally, the flow velocity v itself can be output in vectorial form.

(48) The absolute value of the flow velocity v can also be calculated for at least three different sets of measurement directions. The sign combination occurring most frequently in these calculations or the solid angle region occurring most frequently is used to calculate the absolute value of the flow velocity v.