# Method of measuring level of material in reservoir

FIELD: measurement technology.

SUBSTANCE: method can be used for measuring level of liquid and loose materials as well as for measuring distances. Difference frequency spectrum is calculated and its shape is compared with shape of reference spectrum. Reference spectrum consists of permanent and variable parts which are achieved during calibration at working place. Parameters of permanent part of reference spectrum are defined by design of reservoir and kept permanent during process of measurement. Parameters of variable part of reference spectrum are subject to change till achieving minimum of difference measure for measured and reference spectra and are used for calculating distances. Reference spectrum can be formed by means of two ways. According to the first way reference spectrum is calculated from formed reference signal which has to be the sum of partial signals with parameters corresponding to reflections from members of constructions and from measured signal level. Parameters of signals that correspond to reflection from measured level are changed when conducting measurements till achieving minimum of difference measure. According to the second way during process of calibration the permanent and variable parts of spectrum of measured signal are memorized and the measured signal is used as a reference one. Value of measured frequency is the frequency at which difference measure is minimal. For both variants of forming reference spectrum the permanent and variable parts of reference spectrum do not come into interaction.

EFFECT: improved precision of measurement at the presence of background.

8 cl, 4 dwg

The invention relates to the field of measuring equipment and can be used for level measurement of liquid or bulk materials.

Known radar method of level measurement [1], including the measurement of propagation times of the radio waves radiated in the direction of the surface of the medium and reflected from it, and calculating from the measured propagation time of radio waves range to the surface environment. This method does not allow to measure with sufficient accuracy in the presence of reflections caused by the design of the tank with liquid or loose material, and the reflection from discontinuities antenna / waveguide path and cowling (cover) antenna when at her deposition vapor content of the tank or dust particles.

The known method of measuring the distance implemented in the device [2], namely, that emit frequency-modulated signal in the direction of the tank contents, take some time later, spread the reflected signal and mix it with a part of the emitted signal to obtain a signal of the difference frequency (NAC). The phase of this signal is used for measuring the distance to the surface of the controlled environment, under the condition of maintaining constant the difference frequency, by controlling the period of modulati is. During this phase of the difference frequency signal when the measuring distance is continuously changed within 2πN+ϕ proportional to distance. Here N is an integer number of periods NAC contained in the modulation period, ϕ - the number that corresponds to the remaining part of the period, i.e. the initial phase of the NAC. Thus, the determination of the distance is reduced to counting the number N, the dimension of the phase ϕ and calculating the distance.

The disadvantage of this method is the inability to measure the level with a given accuracy in the presence of reflections caused by tank construction elements, and reflections caused by the heterogeneity of the antenna / waveguide path and the vapor deposition of liquid or dust particles on the cowling (cover) antenna.

Also known a method of measuring distance [3], which consists in the fact that radiation is frequency-modulated radio signal towards the surface of the controlled environment, the reception after the propagation time of the reflected signal, mixing it with a portion of the emitted signal to obtain NAC, determine the moments of the appearance of the characteristic points in the NAC (e.g., extrema or zero), the accumulation values of the weighting function corresponding to the moments of appearance of the characteristic points of the signal, and calculate distances on the amount of accumulated values.

The disadvantage with the person is also the impossibility of level measurement with high accuracy in the presence of reflections.

The closest technical solution in essential characteristics (prototype) is a method of measuring the level [4] radar, oriented to the maximum of the echo signal from the measured surface, including radiation sequence of microwave signals, discrete angular frequencies which are evenly distributed on the scanned frequency range. The signal reflected from the measured surface, mixed with a part of the emitted signal to obtain two signals of the differential frequency shifted relative to each other in phase by the angle π/2 (signal quadratures), which, after analog-to-digital conversion are fed to the microprocessor. Processing the sequence of samples of quadratures includes direct and inverse Fourier transform for localization of the most powerful sources of radiation method high resolution MUZIC, selection of the echo signals in the time domain and in the end obtain the distance to the measured level.

The disadvantage of this method is not sufficiently high accuracy level measurement in the presence of reflections caused by tank construction elements and inhomogeneities antenna / waveguide tract due to their mutual influence on the useful signal. Analysis of the descriptions and claims of the patent suggests that the method of high resolution is the relationship between type of MUZIC or other methods, giving improved resolution in frequency, will improve the accuracy of distance measurement. However, the main reason for the lack of accuracy in the presence of reflections is the mutual influence of the spectra of useful and interfering signals, and the measurement error will depend on the ratio of the amplitudes of the useful and interfering signals.

The purpose of the present invention is the reduction of measurement error level in the presence of reflections from the tank construction elements and reflections from discontinuities antenna / waveguide path.

This objective is achieved in that in the method of measuring the level, including radiation of a frequency-modulated signal in the direction of the tank contents, the reception after the propagation time of the reflected signal, mixing it with a portion of the emitted signal to obtain a signal of the difference frequency (NAC) and the calculation of the spectrum of NAC output from the analog-to-digital Converter (ADC) receiver, additionally, the computation of the reference spectrum obtained during calibration at the workplace, and measures the differences between the measured and reference spectrum. Then you change the parameters of the reference spectrum to achieve the minimum specified measure of the difference spectra. The parameters of the reference spectrum, in which the detected minimum measures differences, use PR is the calculation of the measured distance.

The reference spectrum is represented as a sum of two components, one of which is constant and corresponds to the reflection from interfering reflectors, and the second is variable and corresponds to the reflection from the measured material. The parameters of the reference spectrum are determined during calibration. Calibration is done in the workplace with this tank filling level when the signal reflect all interfering reflectors and there is no mutual influence of side lobes spectra NAC appropriate to prevent the reflectors, and the spectrum of NAC corresponding to the reflection from the material.

There are two methods of calculating the reference spectrum.

In the first embodiment, the reference spectrum is calculated by reference signal containing two parts, which are formed during calibration in the workplace. The first part of the reference signal corresponds to the constant part of the reference spectrum and is the sum of the signals corresponding to interfering reflections and the second reference signal corresponds to the reflection from the material. Amplitude, phase and frequency of the reference signal are chosen so that the measure of the difference spectrum calculated for the reference signal and the signal measured during calibration, was minimal. Selection is carried out in two stages. The first step is a rough estimate of these parameters on the spectrum taken the wow signal. The second step is the specification of values for these parameters by small variations using as a minimum criterion measures the differences of the spectra of the measured signal and the generated signal.

The second option of calculating the reference spectrum based on the summation of two spectra obtained during calibration, one of which is a range of NAC corresponding to the reflections from interfering reflectors, and the second range NAC corresponding to the reflection from the material. These spectra are reference measurements and stored in memory.

The inventive method of measuring the level of material in the tank has a set of features known from the prior art for methods such purposes, which allows to conclude that the criterion of "novelty",

For the proof of inventive step it is necessary to consider that well-known and widely used way of measuring distance range, for example, is based on finding the maximum spectral component and further calculating the distance [5]. To increase the accuracy achieved by adjustment of the generator maximum value of the amplitude of the spectral component at the measured frequency (or minimum level of side lobes). Thus it is assumed the level of reflections, the minimum and not vlue is on the measurement accuracy. However, the appearance of reflections leads to a shift of the maximum spectral component and the emergence of additional measurement errors.

The claimed method does not have this drawback, since the difference frequency corresponding to the distance to the material, is determined by shape analysis of the spectra of the received signal and the reference minimum measures differences between the shape of the spectra. This minimum will be achieved when you become close spectral components of the reference spectrum and the spectrum obtained by measurement. Since the mutual influence of the spectra of the desired signal and clutter and accordingly the variable and constant parts of the spectrum will be the same, it will have no impact on the measurement accuracy. These differences do not follow explicitly from the available scientific and technical sources, which allows to make a conclusion on the compliance of the claimed technical solution the criteria of the invention is an inventive step.

These differences lead to the emergence of qualitatively new properties of the claimed method is the possibility of measuring the level of material in the presence of reflections. This new property allows to increase the accuracy of the measurements.

These differences do not follow explicitly from the available scientific and technical sources, which allows to make a conclusion on compliance with ablanovo technical solutions the criteria of the invention " Inventive step"

A method of measuring the level of material in the tank is as follows. Form the excitation RF signal with a periodic frequency modulation and emit the generated signal in the direction of the sensed material. Take some time later, the spread of the reflected signal is mixed with a part of the irradiated signal and produce a signal of the difference frequency. Calculated using the Fourier transform for discrete frequencies ω_{i}spectrum of the signal S_{from}jω_{i}), the spectrum of the reference signal S_{op}jω_{i}) and a measure of dissimilarity between the Z-spectrum reference signal from the measured data. As a measure of differences Z may be any mathematical metric used to assess differences between the two functions. For example [6]:

where j is the imaginary unit, i is the number of discrete frequencies, N is the number of calculated spectral components (N=K+N_{1}To , the number of signal samples from the ADC output; n_{1}- the number of zero counts, add to interpolate the shape of the spectrum and, consequently, to increase the accuracy of the level measurement).

In this case, the measured spectrum can be represented in the form

where M is the number of reflections in the tank; S_{m}[j(ω_{i}-ω_{m})] - range NAC with frequency ω_{m}(s) the relevant range up to m-th reflector;
S_{d}[j(ω_{i}-ω_{d})]-range NAC with frequency ω_{d}corresponding to the distance to the measured level.

For calculating the reference spectrum there are two options.

In the first embodiment, the reference spectrum is calculated in accordance with the calibration reference signal, containing two parts: u_{op}(t)=u_{post}(t)+u_{lane}(t). The first part of the reference signal u_{post}(t) is constant and represents the sum of the signals corresponding to reflections from interfering reflector:

where u_{mo}that ω_{mo}and ϕ_{mo}respectively the amplitude, frequency and phase of the NAC, the corresponding m-th interfering reflector, M is the total number of interfering reflectors in this tank. The second part of the reference signal u_{lane}(t) is variable and corresponds to the reflection from the material:

where U_{to}and ϕ_{to}respectively the amplitude and the phase NAC corresponding to the reflection from the material;

ω(t) - a well-known law of modulation of the carrier frequency of the microwave signal;

t_{C}the time distribution of the signal corresponding to the simulated distance to the material.

When we change the value of t_{C}the variable part of the spectrum will move along the frequency axis.

The parameters of both parts reference signal definition is considered in the calibration of the calculated spectrum NAC by rough estimates of these quantities and then Refine them with small variations. Variations are made to the detection of the minimum measures of the differences between the measured spectrum and the spectrum calculated by reference signal.

During the measurement, when the measured material begins to close m-th interfering reflector, the amplitude of the corresponding summand reference signal U_{mo}begins to decrease. The nature of changes in the amplitude depends on the electrodynamic properties of the material.

Because the design of the tanks is known and the known distance to the structural elements that cause clutter, the measurement frequency can be eliminated, and the frequency and phase of the signals corresponding to impede the reflectors can be estimated.

The second option of calculating the reference spectrum based on the summation of two spectra obtained during calibration:

S_{op}jω_{i})=S_{post}jω_{i})+S_{lane}jω_{i}).

Continuous part of the spectrum is the spectrum of the signals reflected from interfering reflections and can be represented as follows:

where S_{mo}[j(ω_{i}-ω_{mo})] - range NAC with frequency ω_{mo}corresponding to the distance to m-th interfering reflector.

The variable part of the spectrum is the spectrum of NAC with frequency ω_{x}thus estoya the distance to the material level in the tank: S_{
lane}jω_{i})=S_{to}[j(ω_{i}-ω_{x})]. Discrete samples of these spectra are stored and used when calculating the reference spectrum.

In that case, if the characteristics of the environment (temperature, humidity) during the formation of the reference range during calibration correspond to the characteristics of the environment when carrying out measurements will satisfy the condition S_{mo}[j(ω_{i}-ω_{mo})]=S_{m}[j(ω_{i}-ω_{m})]. The shape of the spectrum S_{d}[j(ω_{i}-ω_{d})] is identical to the form of the spectrum of S_{to}[j(ω_{i}-ω_{x})].

Frequency ω_{x}in the measurement process varies from ω_{n}to ω_{in,}define respectively the lower and upper levels of the material in the tank, with a step δωset desired measurement accuracy. The frequency change ω_{x}is to achieve the minimum measures differences Z these spectra. With the aim of finding the global minimum of the function Z this process is repeated by varying the phase spectrum of the variable partby changing the values of the auxiliary variable ϕ. By the received frequency reference signal ω_{Hmin}=ω_{n}+kδωcorresponding to the best match Spectro is,
calculate the distance R from the antenna to the measured material:

where C is the speed of propagation of radio waves in free space; T_{mod}- period of the frequency modulation transmitter; Δω - adjustment range frequency transmitter; ω_{n}- the initial value of the frequency of beating, corresponding to the maximum level of the material in the reservoir; δω - step change frequency ω_{x}then there is the step of shifting the spectrum of S_{to}[j(ω_{i}-ω_{x})] along the frequency axis at the measurement level; k is the number of steps in the shift of the spectrum on frequency ω_{n}to the frequency at which the magnitude of Z takes the minimum value.

In the proposed method of level measurement is used periodic frequency modulation. Measurement level is when using NAC received in the interval

The implementation of the declared method is illustrated with drawings, shown in figure 1-4.

1 shows a device for level measurement in the presence of interfering reflectors in the tank.

Figure 2 shows the flowchart of the program signal processing in measurement mode level.

Figure 3 and 4 shows the flowchart of the program signal processing for calibration.

Devices is to measure the level contains shaper (f) 1 signal, the output of which is connected to the input of the amplifier UHF (USWC) 2, a directional coupler (BUT) 3, and the output of the microwave amplifier 2 is connected to the input 3, the circulator (C) 4, an input connected to the first output of the directional coupler 3, the antenna (A) 5, is connected to the first output of the circulator 4, a mixer (Cm) 6, the inputs of which are connected with the second output of the directional coupler 3 and the circulator 4, and the output is connected through serially connected amplifier (I) 7 and an analog-to-digital Converter (ADC) with 8 first input processor (cont'd) 9. The second output driver 1 is connected with the second input of the processor 9, the first output of the processor 9 is connected with the second ADC input 8, and the second output of the processor is the output device.

The imaging unit 1 generates a signal with a given periodic law frequency modulation (for example, a symmetric triangular). This signal, after amplification in the amplifier SHF 2 is fed through a directional coupler 3 and the circulator 4 to the antenna 5 and is emitted in the direction of the controlled material. After the time of distribution of the reflected signal is received by the antenna 5 and the second output of the circulator 4 is supplied to the first input of the mixer 6. To the second input of the mixer 6 receives a portion of the radiated signal from the second output of the directional coupler 3. NAC output of the mixer via the amplifier 7 receives the and the ADC input 8. To the second input of ADC 8 receives control pulses from the first output of the processor 9. The NAC counts in digital form are received at the first input of the processor 9. To the second input of the processor 9 receives the pulses corresponding to half of the modulation period from the second output driver 1. The measurement of material level in the tank is supplied to the second output of the processor 9.

The block diagram of the processor in the measurement mode level of the material is shown in figure 2. Block 10 is used to input digital samples NAC during the time of one half-cycle of the modulation. Beginning and duration of an interval of input samples NAC is set by the signal supplied to the second input of the unit 9 with the driver 1. In block 11, the computation times of the spectrum obtained in the NAC using the Fourier transform. Unit 12 sets the initial value of the average frequency of the variable part of the reference spectrum corresponding to the lower boundary of the expected range. At the first measurement, it corresponds to the lowest level of the material. When subsequent measurements to accelerate the procedure the initial value of the average frequency of the variable part of the reference range is selected on the basis of the measurement results in the previous step and the width of the signal spectrum. In block 13 calculates the reference range for the set value of the mean is often the s variable part. In block 14 calculates the first value measures the difference spectra:

where k_{1}and k_{2}numbers calculated spectral components, specifies the frequencyi=k_{1}, k_{2}inside difference frequency corresponding to the measured distance.

Unit 15 compares the current value of the Central frequency of the variable part of the reference spectrum with its final value. At the first measurement end value corresponds to the maximum level of the material and subsequent measurement is selected based on the results of previous measurement and the spectrum width of the signal. If the final value is reached, the switch is made to the block 16, which is the change in the value of the δωset desired measurement accuracy, and the transition to the block 13 for a new calculation of the reference spectrum. After calculation of the spectrum of the reference signal determined by the second value Z(2). The number of values computed measures the differences of the spectra of Z is determined by the required accuracy of the frequency measurement. Once you have calculated all the values of the measures differences in unit 17 searches the minimum value measures the differences and identifying the appropriate center frequency of the variable part of the reference is pectra.

Next, in block 18 is calculating the distance according to the frequency value corresponding to the minimum measures differences. Then the output of the calculation block 19 and the return to the unit 10 for entering new array counts NAC and so on, cyclically repeats the measurement procedure.

The calibration process in accordance with the first variant form the spectrum produced by the block diagram of the program shown in figure 3. In block 20 is used to input a set level material providing no mutual influence of constant and variable parts of the spectrum. In block 21 is input numeric values in the NAC in block 22 calculates the spectrum of the received signal in block 23 on the obtained spectrum is determined by the number of interfering reflectors and runs rough estimate of the amplitude, frequency and phase of the signals constant and variable parts of the reference signal. Then in block 24 on the obtained values of the amplitudes, frequencies and phases are forming samples of the reference signal in the block 25 - calculation of the spectrum and in block 26 the calculation of the measure of the difference Z of the measured spectrum and the spectrum of the reference signal. Next, in block 27 is compared with the previous value and the transition to the block 28, if not detected minimum measures differences (and the first step is always). In block 28 is a variation of the PA is amerov reference signal in accordance with the selected algorithm for finding minimum and returns to block 24 to form a new array of counts of the reference signal. When there is a minimum measure of the difference spectra, the transition occurs to the unit 29 for storing the received parameters of the reference signal and the block 30 exit calibration.

The calibration process in accordance with a second embodiment of forming the reference spectrum is performed in accordance with the flowchart of the program shown in figure 4. In block 31 is filled with the value specified level material. In block 32 is used to input digital samples NAC. Then in block 33 calculates the spectrum, in block 34 - allocation of fixed and variable parts of the spectrum, in block 35 - storing samples of the spectrum and in block 36 to exit the calibration mode.

Modeling of process level measurement showed high efficiency of the proposed method of measuring the level of material in the tank. So without using the proposed method of measuring the level of material in the tank, the measurement error reaches twice the value of the discrete errorThe use of the proposed method of measuring the level of material in the tank provides a measurement errorwhere N is the number of samples of the signal added null samples used in the calculation of the spectrum.

SOURCES of INFORMATION

1. Theoretical OS is Ovi radar. Edited by Shirman AD M, Owls. Radio, 1970.

2. Marfin V.P., Kuznetsov, F. W. the microwave transmitter. // Devices and systems management. 1979, No. 11. Pp.28-29.

3. Japan's bid No. 30-1591, MKI G 01 S 13/34.

4. U.S. patent 5 504 430. MCI G 01 S 13/08.

5. U.S. patent No. 5546088, G 01 S 13/18,13.08.1996.

6. Gorelik A., Skripkin V.A. Methods of recognition. Uch. the manual for high schools.: M, The Higher. School., 1977, 208 S.

1. A method of measuring the level of material in the tank, including the formation of the excitation RF signal with a periodic frequency modulation, the radiation-generated signal in the direction of the sensed material, after receiving the propagation time of the reflected signal, mixing it with a portion of the irradiated signal, the selection signal of the difference frequency, the calculation of the spectrum of these signals, characterized in that it further calculate the reference range and a measure of dissimilarity between the reference spectrum from the measured change of the parameters of the reference spectrum to achieve the minimum measures differences between these spectra and the parameters of the reference spectrum corresponding to the best matching spectra, calculate the distance.

2. A method of measuring the level of material in the tank according to claim 1, characterized in that the reference range is calculated as the sum of a constant part and a variable part.

3. A method of measuring the level of material in the tank according to claim 2, featuring the the action scene, the Central frequency of the variable part of the reference spectrum at the first measurement change in the limits corresponding to the minimum and maximum levels of material in the tank, and for each subsequent measurement is within the limits defined by details on the level of material in the tank from the previous measurement.

4. A method of measuring the level of material in the tank according to claim 2, characterized in that calibrate in the workplace at a certain level of the measured material, which determine the parameters constant and variable parts of the spectrum required for calculating the reference spectrum.

5. A method of measuring the level of material in the tank according to claim 4, characterized in that the material level in the calibration are chosen so that the signal reflected all interfering reflectors and was excluded the influence of side lobes of the fixed portion of the reference spectrum on the variable part of the spectrum.

6. A method of measuring the level of material in the tank according to claim 2, characterized in that the continuous part of the spectrum calculated by the fixed portion of the reference signal comprising the sum of the difference frequency signals corresponding to the distance of reflections in the tank, and the variable part of the spectrum calculated by the variable part of the reference signal represents a signal is l difference frequency, corresponding to the distance to the material.

7. A method of measuring the level of material in the tank according to claim 4, characterized in that during calibration select amplitude, phase and frequency constant and variable parts of the reference signals so as to measure the difference spectrum of this amount and the spectrum measured during calibration, was minimal and remember these settings.

8. A method of measuring the level of material in the tank according to claim 4, characterized in that the calculated spectrum of the signal of the difference frequency, separated from him spectra corresponding to impede the reflectors, i.e. continuous part of the spectrum, and the material, i.e. the variable part of the reference spectrum, and remember discrete samples of these spectra.

**Same patents:**

FIELD: measurement technology.

SUBSTANCE: method can be used for measuring level of liquid and loose materials as well as for measuring distances. Difference frequency spectrum is calculated and its shape is compared with shape of reference spectrum. Reference spectrum consists of permanent and variable parts which are achieved during calibration at working place. Parameters of permanent part of reference spectrum are defined by design of reservoir and kept permanent during process of measurement. Parameters of variable part of reference spectrum are subject to change till achieving minimum of difference measure for measured and reference spectra and are used for calculating distances. Reference spectrum can be formed by means of two ways. According to the first way reference spectrum is calculated from formed reference signal which has to be the sum of partial signals with parameters corresponding to reflections from members of constructions and from measured signal level. Parameters of signals that correspond to reflection from measured level are changed when conducting measurements till achieving minimum of difference measure. According to the second way during process of calibration the permanent and variable parts of spectrum of measured signal are memorized and the measured signal is used as a reference one. Value of measured frequency is the frequency at which difference measure is minimal. For both variants of forming reference spectrum the permanent and variable parts of reference spectrum do not come into interaction.

EFFECT: improved precision of measurement at the presence of background.

8 cl, 4 dwg