Method and system for detecting, characterizing and assessing the quality of a spray

Abstract

The present disclosure relates to a method for the detection, characterization and assessment of the quality of a spray which is produced by atomizing liquids with nozzles of different designs and constructional forms. The spray may have an application-specific droplet size distribution ranging from fine to very coarse droplets which move at different speeds. A radar signal is directed into and reflected by the spray. The reflected radar signal is subject to a Doppler shift caused by the movement of the droplets in the spray. The transmitted and received radar signals are mixed to create a low-frequency Doppler oscillation signal which is sampled at a predetermined rate with an analog-digital converter, the output of which is stored in a data array and transformed from the time domain into the frequency domain for further processing.

Claims

1. A method for detecting, characterizing and assessing the quality of a spray of atomized liquid, comprising: transmitting a continuous wave radar signal with a frequency f.sub.O into the spray; receiving a radar signal reflected off the spray, the received radar signal having a frequency f.sub.D; generating a low frequency Doppler signal with a frequency f.sub.S=|f.sub.Of.sub.D| by mixing the transmitted radar signal and the received radar signal; sampling the low frequency Doppler signal with an analog to digital converter having a predetermined sample rate; storing the sampled data in a data array with an index n=[0, 1, 2, . . . , N]; transforming the sampled data from the time domain to the frequency domain; and analyzing the transformed data to derive the quality of the spray, wherein the spray is generated by a plurality of nozzles, wherein transforming the sampled data from the time domain to the frequency comprises calculating a radar signature, and wherein the radar signature of an individual nozzle is compared and assessed, in a superordinate processing system, with a reference signature.

2. The method as in claim 1, wherein transforming the sampled data comprises a discrete Fourier transform DFT with splitting the data array into even indices n=[0, 2, 4, . . . , N2] and odd indices n=[1, 3, 5, . . . , N1].

3. The method as in claim 2, wherein the discrete Fourier transform DFT $X\left(k\right)=\underset{n=0}{\overset{N-1}{.Math.}}x\left(n\right)e^{-\left(i\frac{2\mathrm{nk}}{N}\right)}$ is calculated as a Radix 2 Fourier Transform FFT (Fast Fourier Transform) with $X\left(k\right)=\underset{n=0}{\overset{\frac{N}{2}-1}{.Math.}}x\left(2n\right)e^{-\left(i\frac{2\left(2n\right)k}{N}\right)}+\underset{n=0}{\overset{\frac{N}{2}-1}{.Math.}}x\left(2n+1\right)e^{-\left(i\frac{2\left(2n+1\right)k}{N}\right)},X\left(k\right)=\underset{n=0}{\overset{\frac{N}{2}-1}{.Math.}}x\left(2n\right)e^{-\left(i\frac{2\mathrm{nk}}{\frac{N}{2}}\right)}+e^{-\left(i\frac{2k}{N}\right)}\underset{n=0}{\overset{\frac{N}{2}-1}{.Math.}}x\left(2n+1\right)e^{-\left(i\frac{2\mathrm{nk}}{\frac{N}{2}}\right)},\mathrm{and}$ $X\left(k\right)={\mathrm{DFT}}_{\frac{N}{2}}\left[\left[x\left(0\right),x\left(2\right),\mathrm{.Math.},x\left(N-2\right)\right]\right]+W_N^k{\mathrm{DFT}}_{\frac{N}{2}}\left[\left[x\left(1\right),x\left(3\right),\mathrm{.Math.},x\left(N-1\right)\right]\right].$

4. The method as in claim 1, wherein transforming the sampled data from the time domain to the frequency domain comprises a wavelet transformation or a filterbank.

5. The method as in claim 1, wherein the continuous wave radar signal is generated by a radar sensor having an antenna with a wide aperture angle in azimuth and elevation so that the transmitted radar signal penetrates at least 50% the spray of an associated nozzle.

6. The method as in claim 1, wherein the reference signature has been taught in advance in accordance with a supervised machine learning method.

7. The method as in claim 1, wherein the reference signature is an averaged signature of the plurality of nozzles.

8. The method as in claim 7, wherein a similarity of a first FFT data series i=[0, 1, 2, . . . , n] of the radar signature of the individual nozzle is compared with a second FFT data series of an averaged radar signature over all nozzles j=[0, 1, 2, . . . , k] of a spray system.

9. The method as in claim 1, wherein the sampled data in the frequency domain is compared with a set of reference data and wherein a Pearson correlation coefficient between the sampled data in the frequency domain and the reference data is calculated.

10. The method as in claim 9, wherein FFT bins are used in transforming the sampled data in the frequency domain and wherein the calculation of the Pearson correlation coefficient is performed with ${\mathrm{Kor}}_e\left(x,y\right)\text{:}=\frac{\frac{1}{n-1}\underset{i=1}{\overset{n}{.Math.}}\left(x_i-\stackrel{\_}{x}\right)\left(y_i-\stackrel{\_}{y}\right)}{\sqrt{\frac{1}{n-1}\underset{i=1}{\overset{n}{.Math.}}{\left(x_i-\stackrel{\_}{x}\right)}^2}.Math.\sqrt{\frac{1}{n-1}\underset{i=1}{\overset{n}{.Math.}}{\left(y_i-\stackrel{\_}{y}\right)}^2}},$ wherein $\stackrel{\_}{x}=\frac{1}{n}\underset{i=1}{\overset{n}{.Math.}}{x}_i$ is the empirical mean value of a data series of the FFT bins i=[0, 1, 2, . . . , n] of a present radar signature of the spray of an individual nozzle, wherein $\stackrel{\_}{y}=\frac{1}{n.Math.k}\underset{j=1}{\overset{k}{.Math.}}\underset{i=1}{\overset{n}{.Math.}}{y}_{\mathrm{ij}}$ is an empirical mean value of all data series of FFT bins i=[0, 1, 2, . . . , n] of all nozzles j=[0, 1, 2, . . . , k] of a spraying system having a plurality of nozzles, and wherein y.sub.i is the respective FFT-bin value i=[0, 1, 2, . . . , n] of a reference signature.

11. The method as in claim 1, wherein the transmitted continuous wave radar signal is modulated using FSK (Frequency Shift Keying) or FMCW (Frequency Modulated Continuous Wave) modulation to provide a spatial resolution and spatial allocation of a size distribution and associated speeds of droplets within the spray.

12. The method as in claim 1, wherein the transmitted continuous wave radar signal is provided by more than one focusing radar systems or by a multi-antenna MIMO (multiple in, multiple out) system to provide a spatial resolution and spatial allocation of a size distribution and associated speeds of droplets within the spray.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) Further objects, features, advantages and possible applications of the disclosed method as well as a device for monitoring the working quality of spray systems can be gathered from the following description of several exemplary embodiments with reference to the drawings.

(2) FIG. 1 shows exemplary radar signatures from a flat-jet injector nozzle with a metering aperture Teejet AIXR 110025 VP at low and at high pressure of the liquid.

(3) FIG. 2 shows the radar signature of a flat jet nozzle Teejet 8001E-SS in normal operation without blockage and in the case of partial blockade by a needle tip inserted centrally in the nozzle mouth.

(4) FIG. 3 is a side view and a rear view showing a single nozzle with spray, a radar system and its detection area.

(5) FIG. 4 shows a plant protection device spraying system with a linkage and a liquid-conducting tubes with nozzles.

DETAILED DESCRIPTION

(6) Referring to FIG. 1, two exemplary radar signatures of the same flat jet injector nozzle Teejet AIXR 110025 VP are shown. A first line 19 shows the radar signature of the spray at low pressure and a second line 20 shows the radar signature of the spray at high pressure of the liquid. The representation of the signal amplitudes of the FFT bins is log-scaled. The differences in radar signals with respect to frequencies and signal amplitudes over the frequency are clearly evident and can therefore be used in the disclosed method to determine the quality of the present spray.

(7) Referring to FIG. 2, two exemplary radar signatures of a flat jet nozzle Teejet 8001E-SS are shown. The graph shows normal operation without blockage 22 and a simulated case of partial blockade 21 which has been generated by inserting a needle tip centrally in the nozzle mouthpiece. The representation of the signal amplitudes of the FFT bins is log-scaled. The differences in radar signals with respect to frequencies and signal amplitudes over frequency are also clear and can thus also be used to in the disclosed method to determine the quality of the present spray.

(8) FIG. 3 and FIG. 4 shows a spray system 1, in this case an exemplary plant protection device, which comprises a linkage 10 with a liquid-conducting tube 11 with nozzles 12. A radar sensor 14 associated with the nozzle 12 is attached to the linkage 10 or the tube 11 by a bracket 13 in such a way that the radar detects a wide range of the respective spray 2.

(9) The spraying system 1 preferably operates in a time-slot method to avoid interferences of the radars and to reduce the overall current consumption of the system. The spray system 1 is designed for monitoring preferably up to 256 nozzles. A plurality of, for example, eight radar sensors 16 are connected to a distributor 15. In the distributor 15, substantial portions of the processing take place, such as the calculation of the FFTs.

(10) By a serial bus system, a plurality of distributors 15 are connected to one another and to an ECU (electronic control unit) 17 processing and control unit. The enumeration of the radar sensors 16 takes place during system initialization by the ECU 17.

(11) A main unit 18 is connected to the ECU 17 as a superordinated system with monitoring software and user interface via a serial bus system. The system comprising of ECU 17, distributors 15 and radar sensors 16 is thereby adaptable flexibly and can be connected to OEM systems. The monitoring software in the main unit analyzes the similarity of the detected radar signatures and the reference radar signature and warns of deviations which go beyond a user-adjustable signal threshold. The corresponding nozzle number (s) is/are displayed and optically signaled at the concerned distributor.

LIST OF REFERENCE NUMERALS

(12) 1 spraying system 2 spray 10 linkages 11 piping 12 nozzles 13 brackets 14 radar sensor 15 distributors 16 radar sensors 17 ECU 18 main unit 19 low pressure radar signature 20 high pressure radar signature 21 blocked, partial blockade radar signature 22 radar signature without blockade