Method for Avoiding Phase Jumps

Abstract

A method for processing a measurement signal that is captured by a measuring device, wherein, in order to capture the measurement signal, the measuring device emits a transmission signal and receives a component of the transmission signal that is reflected by an object as a reception signal, wherein a first phase difference between a first target phase position and a first actual phase position contained in the measurement signal is determined, and wherein a second phase difference between a second target phase position and a second actual phase position contained in the measurement signal is determined, and a phase difference progression in the form of an, in particular, linear, functional relationship is determined on the basis of the first and the second phase differences, and a measured value is determined by means of the functional relationship.

Claims

1-11. (canceled)

12. A method for processing a measurement signal that is captured by a measuring device, comprising the steps of: emitting with the measuring device a transmission signal and receiving a component of the transmission signal that is reflected by an object as a reception signal; determining a first phase difference between a first target phase position and a first actual phase position contained in the measurement signal; determining a second phase difference between a second target phase position and a second actual phase position contained in the measurement signal; ascertaining a phase difference progression in the form of an, in particular, linear, functional relationship on the basis of the first and the second phase differences; and determining a measured value by means of the functional relationship.

13. The method according to claim 12, wherein: the functional relationship is ascertained as a function of a distance of the object to be determined from the measurement signal.

14. The method according to claim 12, wherein: the functional relationship is used to avoid a phase jump when determining the measured value.

15. The method according to claim 14, further comprising the step of: using a parameter of the phase difference progression—especially, a slope and/or an axial offset of the, in particular, linear functional relationship to recognize and/or eliminate a phase jump of the measured value.

16. The method according to claim 12, wherein the measuring device is a radar measuring device, and the method further comprises the steps of: converting the received signal into a digital measurement signal, especially by means of a scanning method; and determining the distance of the object with runtime information contained in the measurement signal, in order to ascertain the functional relationship.

17. The method according to claim 12, wherein: the functional relationship is ascertained with the assistance of a linear regression.

18. The method according to claim 12, wherein: the functional relationship is updated iteratively with additionally ascertained phase differences.

19. The method according to claim 12, further comprising the step of: determining the first phase difference at a first distance of the object, wherein the first target phase position is calculated for this first distance; ascertaining the first actual phase difference at this first distance; determining the second phase difference at a second distance of the object; calculating the second target phase position for this second distance; and ascertaining the second actual phase position at this second distance.

20. The method according to claim 12, wherein: the first target phase position and/or the second target phase position are/is ascertained with a target phase progression; that the target phase progression is ascertained by means of a calibration; and that the calibration is carried out by means of at least one known distance.

21. A measuring device that serves to capture and process a measurement signal, wherein the measuring device serves to emit a transmission signal, and receiving a component of the transmission signal reflected by an object as a reception signal, comprising: an electronics unit, wherein: said electronics unit serves to determine a first phase difference between a first target phase position and a first actual phase position contained in the measurement signal; said electronics unit serves to determine a second phase difference between a second target phase position and a second actual phase position contained in the measurement signal; and said electronics unit serves to ascertain a phase difference progression in the form of an, in particular, linear, functional relationship on the basis of the first and the second phase differences, thereby determining a measured value by means of the functional relationship.

22. The measuring device according to claim 21, wherein: the measuring device is a fill-level measuring device.

Description

[0038] A preferred embodiment of the invention will be explained in more detail with reference to the following figures. Illustrated are:

[0039] FIG. 1: A graphic representation of an echo function of an FMCW radar measuring device;

[0040] FIG. 2: A graphic representation of a target phase progression of a measurement signal as a function of the distance;

[0041] FIG. 3: A first phase difference progression to which a linear function was adapted;

[0042] FIG. 4a: A second phase difference progression for an FMCW radar measuring device that has a systematic error, and thus phase jumps;

[0043] FIG. 4b: A corrected phase difference progression, wherein the phase jumps of the phase difference progression shown in FIG. 4a have been eliminated;

[0044] FIG. 5: An exemplary method flow, particularly for the instance in which the linear functional relationship is adapted to the phase difference progression;

[0045] FIG. 1 shows a graphic representation of an echo function 1 of an FMCW radar measuring device, wherein the echo function 1 represents the echo amplitudes as a function of the distance X. A digitized intermediate frequency signal IF serves as a starting point for depicting the echo function 1. The digitized intermediate frequency signal IF is a real-valued signal within the time range. An echo function 1 is formed by deriving the received signal amplitude from the intermediate frequency signal IF and depicted as a function of time or distance X.

[0046] To form the echo function 1, first, a direct component of the digitized intermediate frequency signal IF is eliminated. The different frequency components contained in the intermediate frequency signal IF are retained.

[0047] A Hilbert transformation can be performed as the next step, in which the imaginary part belonging to the real-valued intermediate frequency signal IF is calculated. As the result of the Hilbert transformation, a complete, complex-valued intermediate frequency signal IF within the time range is obtained. This complex-valued signal then serves as a basis for the further processing of the signal.

[0048] Subsequently, a device-specific correction curve can be applied to the complex-valued signal. With the assistance of this device-specific correction curve, the specific characteristics of the radar measuring device can be taken into account that are caused, inter alia, by the HF module, the mode converter, and the antenna, as well as the HF cable, connecting elements, and/or plug-in connectors.

[0049] Subsequently, a fast Fourier transform (FFT) is carried out. The frequency spectrum of the intermediate frequency signal IF is obtained as result of the fast Fourier transform. Subsequently, the echo frequency component contained in this frequency spectrum is detected. In the process, each frequency peak A, B, C in the frequency spectrum is assigned an object. Of particular interest in fill-level measurement is the frequency peak A of the useful echo that is produced by the reflection of the radar signals from the surface of the liquid or medium. This relevant frequency peak A of the useful echo, which is relevant for measuring the fill level, can be identified by means of a search algorithm. A person skilled in the art is aware of many such search algorithms. The actual phase position is determined as a phase position at the position of the frequency peak A as a result of the Fourier transform.

[0050] Alternatively, a digital bandpass filter can be defined for determining the phase position, which is configured to filter out from the frequency spectrum the frequency peak produced by the surface of the medium, and to suppress other interfering frequency peaks. To accomplish this, the lower limit frequency and the upper limit frequency of the bandpass filter are set so that the frequency peak caused by the surface of the medium lies within the passband of the bandpass filter. Interfering frequency components that lie below the lower limit frequency, or above the upper limit frequency, of the bandpass filter are filtered out. As a result of bandpass filtering, a frequency spectrum is obtained that basically contains a single frequency peak, i.e., frequency peak A, which is relevant to measuring the fill level and is produced by reflection from the surface of the medium.

[0051] This bandpass-filtered frequency spectrum is then transformed back into the time domain by means of an inverse Fourier transform. For example, an inverse fast Fourier transform (IFFT) can be carried out. As a result of the inverse fast Fourier transform, a complex-valued time signal is obtained that basically contains a single frequency component, i.e., the frequency component that arose from the reflection of the radar signal from the surface of the medium. The actual phase position can then be ascertained directly from this time signal, and a determination of the measuring distance can be performed.

[0052] FIG. 2 shows a graphic representation of a target phase progression 2 of a measurement signal as a function of the distance X. When the aforementioned correction curve is being generated, a target phase progression 2 of the measurement signal at distance X can be ascertained. To ascertain or generate the device-specific correction curve, a reference measurement for an object with a known distance is performed, and the device-specific phase position at the known distance is derived from this reference measurement. The phase position can be derived from this time signal at any point in time for the known distance. A second reference measurement can be performed for a second known distance, in order to increase the reliability of the ascertainment of the target phase progression 2.

[0053] Accordingly, a device-specific target phase progression 2 can be created. The target phase progression 2 as shown in FIG. 2 has a sawtooth shape, wherein the target phase position grows linearly for each phase period 3, or for each wavelength.

[0054] In measuring mode, an unknown distance is precisely ascertained by means of this target phase progression 2. In determining the distance, a rough determination 4 is performed using the position of the frequency peak A of the useful echo in the echo function 1 as depicted in FIG. 1. Then, the actual phase position of this frequency peak A is ascertained. In FIG. 2, a point resulting from such a determination is identified with reference number 4. The measured value is corrected so that the ascertained phase position 4, or the actual phase position 4, corresponds with the target phase position 2a. The measured value is accordingly shifted within a distance range that is defined by the phase period determined by the rough determination of distance.

[0055] FIG. 3 shows a first phase difference progression 5, to which a linear function 6 has been adapted. The progression of the phase difference 5 is not influenced in this context by a systematic error; accordingly, the difference between the actual phase position 4 and the target phase position 2a is basically less than +/−0.5π over the entire measuring path X (about 27 m in this context). The adapted line has a slope that is approximately (0.5*π)/30, and the axial offset is approximately 0.25 radian. Because the phase difference 5 is always less than +/−π, the last phase period can always be clearly identified. There is, therefore, no reason to expect a phase jump in the measurement signal evaluation. A phase jump can occur only when the difference between the actual phase position and the target phase position is approximately +/−π. Only then is it no longer possible to clearly identify the last phase period.

[0056] The phase difference progression 5 depicted in FIG. 3 is from a calibrated radar measuring device. Accordingly, a reference measurement was performed as described above, so that the target phase position and actual phase position substantially correspond. The axial offset and the slope of the adapted linear function accordingly lie within the predefined value ranges W1, W2.

[0057] If this radar measuring device is used with another surge pipe, this correspondence no longer exists. FIG. 4a shows a phase difference progression 7 as an example of this. FIG. 4a shows a second phase difference progression 7 for an FMCW radar measuring device that has a systematic error, and thus phase jumps.

[0058] The systematic error in this case is, for example, an incorrectly indicated surge pipe diameter. The phase difference 7 rises linearly at a measuring distance X between about 2 and 15 m from the radar measuring device. The phase difference 7 changes over this measuring is distance X by 2*π radian. At about 15 m, the phase difference 7 is accordingly about +1*π, and the difference between target phase position 2a and actual phase position 4 jumps between +1*π and −1*π. Here, the actual phase position is 4π radian distant from a target phase position 2a in a first phase period, and π radian distant from the target phase position in a second phase period. Accordingly, there is an ambiguity as to which target phase position of the two target phase positions in the first and second phase periods should be assigned the actual phase position 4.

[0059] This ambiguity is eliminated by means of the, in particular, linear functional relationship 6. A line 6 is accordingly adapted to the phase difference progression 7, wherein this line, or this linear functional relationship 6, is ascertained while determining the phase difference progression 7. The slope and axial offset of the line are ascertained and compared to set criteria, in order to ensure that the actual phase position 4 is always assigned to a target phase position 2a that lies within a single phase period. This is explained in greater detail in conjunction with FIG. 4b.

[0060] FIG. 4b shows a corrected phase difference progression 8, wherein the phase jump 9 of the phase difference progression 7 shown in FIG. 4a is eliminated. The phase difference progression 8 is approximated in FIG. 4b with the assistance of a linear regression. During the measuring mode of the radar measuring device, at least five numerical values are ascertained and updated. The at least five numerical values are, for example:

[0061] 1) the average measuring distance of the frequency peak of the useful echo, which captures the measuring distance X from the runtime information of the measurement signal,

[0062] 2) the average phase difference 5 between the actual phase position 4 at frequency peak A of the useful echo and the target phase position 2a,

[0063] 3) the average value of the square of the measuring distance X,

[0064] 4) the average value of the product of the measuring distance X and the phase difference 5, and

[0065] 5) a number n of learning points.

[0066] The number of learning points is the number of measuring distances X that are used to adapt the linear functional relationship 6 to the phase difference progression 5, 7. This number n starts at zero and is incremented by one with each new learning point until a set limit value n.sub.max is reached.

[0067] During measuring mode, these numerical values are ascertained and saved in a memory unit of the radar measuring device. An arithmetic unit of the radar measuring device can use these numerical values for determining the slope and the axial offset of the line 6 to be adapted by means of the following formulas:

[00001] ${Slope}_{n} = \frac{\begin{matrix} {\overline{Measuring .Math. .Math. .Math. distance * Phase .Math. .Math. difference}}_{n} - \\ {\overline{Measuring .Math. .Math. distance}}_{n} * {\overline{Phase .Math. .Math. difference}}_{n} \end{matrix}}{\overline{Measuring .Math. .Math. {distance}_{n}^{2}} - {({\overline{Measuring .Math. .Math. .Math. distance}}_{n})}^{2}}$ $Axial .Math. .Math. {offset}_{n} = {\overline{Phase .Math. .Math. difference}}_{n} - {Slope}_{n} * {\overline{Measuring .Math. .Math. distance}}_{n}$

wherein the numerical values can be updated while ascertaining learning points, i.e., new measuring distances X and phase differences 5, 7, using the following formulas:

[00002] $n = {\begin{matrix} n + 1, .Math. if .Math. .Math. n < n_{\max} \\ n, .Math. if .Math. .Math. n = n_{\max} \end{matrix} .Math. .Math. {\overline{Measuring .Math. .Math. distance}}_{n} = \frac{n - 1}{n} * {\overline{Measuring .Math. .Math. .Math. distance}}_{n - 1} + \frac{1}{n} * Measuring .Math. .Math. distance .Math. .Math. {\overline{Phase .Math. .Math. difference}}_{n} = \frac{n - 1}{n} * {\overline{Phase .Math. .Math. .Math. difference}}_{n - 1} + \frac{1}{n} * Phase .Math. .Math. .Math. difference .Math. .Math. \overline{Measuring .Math. .Math. {distance}_{n}^{2}} = \frac{n - 1}{n} * \overline{Measuring .Math. .Math. {distance}_{n - 1}^{2}} + \frac{1}{n} * Measuring .Math. .Math. distance * Measuring .Math. .Math. distance .Math. .Math. .Math. {\overline{Measuring .Math. .Math. .Math. distance * Phase .Math. .Math. difference}}_{n} = .Math. \frac{n - 1}{n} * .Math. .Math. {\overline{Measuring .Math. .Math. .Math. distance * Phase .Math. .Math. difference}}_{n - 1} .Math. + .Math. \frac{1}{n} * Measuring .Math. .Math. distance * Phase .Math. .Math. .Math. difference$

[0068] The linear functional relationship 6 is accordingly adapted to the phase difference progression 5, 7. During this dynamic adaptation, the slope and axial offset of the adapted line vary correspondingly. By specifying a first value range W1 for the slope and a second value range W2 for the axial offset, the arithmetic unit can check the adaptation after each update. It is accordingly determined whether the actual phase position 4 can be clearly assigned a target phase position 2a of the target phase progression 2.

[0069] For example, the first value range W1 can be so specified for the slope that first ascertained phase difference values 7a, which are shown in FIG. 4a, b, are excluded for adapting the line. The first phase difference values 7a, which are at about 2 m in FIG. 4a, b, jump between +π and −π. A line that is adapted to these phase difference values 7a has a very steep slope. The first value range W1 can be set so that this steep slope is excluded.

[0070] In FIG. 4b, the jumping phase difference values 7b, which are at about 15 m, are also excluded. The second value range W2 for the axial offset generally spans phase difference values between +/−π. In updating the numerical values for learning points N that correspond to the measuring distance range, for example, between 16 to 20 m, a linear functional relationship 6 is ascertained that has a slope which falls within the first value range W1. However, the axial offset of the line lies outside of the second value range W2 +/−π. Accordingly, the arithmetic unit of the radar measuring device recognizes that a phase jump 9 has occurred. The phase difference values 7c in this measuring distance range are shifted upwards in this case by an addition of 2*π. The corrected phase difference progression 8 is created in this manner.

[0071] FIG. 5 shows an exemplary method flow—in particular, for the instance in which the linear functional relationship 6 is adapted to the phase difference progression 7. When the method enters step 100, a check is performed as to whether or not the numerical values calculated according to the above formulas are present. If there are no numerical values, e.g., when the radar measuring device is first turned on, the numerical values are initialized with the current measured values in step 200 as follows:

[00003] $n = 1$ ${\overline{Measuring .Math. .Math. distance}}_{1} = Measuring .Math. .Math. distance$ ${\overline{Phase .Math. .Math. difference}}_{1} = Phase .Math. .Math. difference$ $\overline{Measuring .Math. .Math. {distance}_{1}^{2}} = Measuring .Math. .Math. distance * Measuring .Math. .Math. distance$ $\overline{Measuring .Math. .Math. distance * Phase .Math. .Math. {difference}_{1}} = Measuring .Math. .Math. distance * Phase .Math. .Math. difference$

[0072] The line 6 runs flat through the first measuring point. The flat line 6 has a slope equal to zero. The phase jump is excluded at the first measuring point; consequently, a correction of the phase difference value is also unnecessary.

[0073] If numerical values already exist, the line 6, after step 100, is calculated in step 300 according to the above formulas, and the phase difference is immediately shifted from the line 6 to the value range W2 +/− *π by adding +/−2*π.

[0074] Then, the number n of learning points can be limited in step 400 by specifying a limit value n.sub.max. Upon reaching the limit value n.sub.max (i.e.: n=n.sub.max), the number n of learning points is then no longer incremented or increased stepwise with each new learning point. This ensures that the new learning points are still taken into consideration with a minimum weight of 1/n.sub.max even during a lengthy measurement operation. This ensures that the line 6 is continuously adapted to the current numerical values. The adaptation of the line 6 is accordingly sensitive to newly ascertained learning points, as well as dynamically updatable.

[0075] The numerical values or learning points are then updated in step 500, and the line 6 is updated or calculated with the added learning points. Then, the standard deviation of the average measuring distance X is evaluated in step 600 and compared with the specified limit values.

[0076] If the standard deviation of the average measuring distance X exceeds the specified limit values, the axial offset of the line 6 is checked in step 700. If the maximum amount of the axial offset of the line 6a is greater than π, or perhaps π plus a tolerance value, the axial offset is brought in step 800 to the value range W2 between +/−π by adding +/−2*π to the axial offset of the line 6a. In so doing, the numerical values or learning data must be correspondingly adapted in the memory unit of the radar measuring device.

[0077] If the standard deviation of the measuring distance does not exceed the specified limit value, or if the axial offset is less than π, or perhaps π plus a tolerance value, the slope of line 6 is investigated in step 900. If the slope lies outside of the first specified value range W1, the numerical values are re-initialized according to step 200 and inserted into a flat line 6 as a first learning point. If the slope in step 900 is within the first value range W1, or if the slope is not excessively large, +/−2*π is, in step 1000, added in turn to the phase difference until it lies between +/−π of the line 6 in the value range W2. A reliable phase difference value 8 that is without the phase jumps can accordingly be provided, to increase the precision of the radar measuring device.

LIST OF REFERENCE NUMBERS

[0078] 1 Echo function

[0079] 2 Target phase progression

[0080] 2a Target phase position

[0081] 3 Phase period

[0082] 4 Actual phase position

[0083] 5 First phase difference progression

[0084] 6 Linear functional relationship/line

[0085] 6a Line

[0086] 7 Second phase difference progression

[0087] 7a First phase difference values

[0088] 7b Jumping phase difference values

[0089] 7c Phase difference values at a measuring distance range of 16-20 m

[0090] 8 Corrected phase difference progression

[0091] 9 Phase jump

[0092] A, B, C Echo peaks/frequency peaks

[0093] X Distance/measuring distance

[0094] IF Intermediate frequency signal

[0095] W1, W2 First, second values range

Method for Avoiding Phase Jumps

Inventors

Cpc classification

Classification Explorer

G01S13/88

PHYSICS

Classification Explorer

G01S7/40

PHYSICS

Classification Explorer

G01S7/352

PHYSICS

Classification Explorer

G01S7/356

PHYSICS

Classification Explorer

G01S13/34

PHYSICS

Classification Explorer

G01S7/023

PHYSICS

Classification Explorer

G01F23/284

PHYSICS

International classification

Classification Explorer

G01S13/34

PHYSICS

Classification Explorer

G01S7/40

PHYSICS

Classification Explorer

G01F23/284

PHYSICS

Classification Explorer

G01S7/35

PHYSICS

Abstract

Claims

Description