The task is considered related to the creation of monitoring systems in the industrial and environmental technologies that allow to conduct continuous analysis of dust in the industrial premises and quality control of the products made of powders, charges, and other dispersed and composite materials. The proposed methodology is based on the acoustic spectral-timbre control. It allows to make an operational analysis of the physical and chemical properties of the dispersed systems. Using a spectrum analyzer, the acoustic signal is decomposed into a Fourier spectrum consisting of the fundamental and timbre harmonics. The higher harmonics of the spectrum provide for more complete information about the amplitude-frequency parameters of the dispersed system and allow to distinguish by classes of the dispersed composition.
The graphical representation of the division of the information space into classes with optimal characteristics allows to interpolate and visually evaluate the belonging of the analyzed spectra to the family with the parameters corresponding to these dispersed systems. When dividing the information space into standard classes, the boundaries between them are approximated by Hermite polynomials with the preservation of the first terms. This makes it possible to divide the modes of dust flows into concentration classes, as well as formally separate them from each other by constructing separating boundaries. After calculating the coefficients of the separating functions, the separating boundaries are determined, which, by virtue of the accepted agreement on the basis functions, will be piecewise linear. The coefficients of the separating functions are calculated according to the developed program of differential diagnostics by the method of potential functions, which implements this method from the theory of pattern recognition.
Patterns to be recognized and classified using an automatic recognition system must have a set of measuring statistical characteristics of the spectrum of acoustic emission signals of the dust flow. Pattern recognition systems allow to configure the measurement system to recognize the required concentration classes based on the collected information.
2. Huang C.-Y., Cai X.-S., Zhou W. Particle size measurement in liquid-solid system based on acoustic emission technique. Journal of University of Shanghai for Science and Technology. 2012. Vol. 34. Iss. 5. pp. 466–470.
3. Gubaidullin D.A., Nikiforov A.A., Utkina E.A. Acoustic waves in two-fraction mixtures of gas with vapor, droplets and solid particles of different materials and sizes in the presence of phase transitions. Fluid Dynamics. 2011. Vol. 46. pp. 72–79.
4. Nasir V., Cool J. Characterization, optimization, and acoustic emission monitoring of airborne dust emission during wood sawing. The International Journal of Advanced Manufacturing Technology. 2020. Vol. 109. Iss. 9–12. pp. 2365–2375.
5. Kuts V.P., Slobodyan S.M. Analysis Method for Disperse Composition of Aerosols, Dusts and Powders. Izvestiya Altayskogo gosudarstvennogo universiteta = Izvestiya of Altai State University Journal. 2014. № 1-1 (81). pp. 248–251. (In Russ.).
6. Lepikhova V.A., Pikina E.V. Spectral-timbre methodology of the coal dust dispersed analysis in the ventilation systems. Aktualnye problemy geologii, gornogo i neftegazovogo dela: sb. nauch. tr. (Current problems of geology, mining and oil and gas business: collection of the scientific papers). Novocherkassk: YuRGPU (NPI), 2017. pp. 108–113. (In Russ.).
7. Zyryanov B.A. Recognition in the diagnosis of NPU. Teplofizika yadernykh energeticheskikh ustanovok: mezhvuz. sb. nauch. tr. (Thermal physics of nuclear power units: interuniversity collection of the scientific papers). Sverdlovsk: UPI, 1982. pp. 14–19. (In Russ.).
8. Tu D., Gonsales R. Principles of pattern recognition. Moscow: Izd-vo «Mir», 1978. 411 p. (In Russ.).
9. Zyryanov B.A. Coherence of thermophysical characteristics of NPU. Teplofizika yadernykh energeticheskikh ustanovok: mezhvuz. sb. nauchn. tr. (Thermal physics of nuclear power units: interuniversity collection of the scientific papers). Sverdlovsk: UPI, 1983. pp. 15–20. (In Russ.).
10. Fu K. Sequential methods in the pattern recognition and machine learning. Moscow: Nauka, 1971. 255 p. (In Russ.).
11. Zhuravlev V.P., Uchitel G.S., Toropov O.A., Puresev A.I., Malykh E.A., Lepikhova V.A. Method for processing the signals from a shock pulse sensor for solid particles of the aerosol flow. Patent RF № 2097738. Applied: April 18, 1996. Published: November 27, 1997. Bulletin № 33. (In Russ.).
12. Zhuravlev V.P., Uchitel G.S., Toropov O.A., Mukhanov V.V., Puresev A.I., Lepikhova V.A., Malykh E.A. Method for determining the concentration of the solid phase of the dust and gas flow. Patent RF № 2105302. Applied: April 18, 1996. Published: February 20, 1998. Bulletin № 15. (In Russ.)