FEATURE ANALYSIS OF VISUAL-GRAPHIC REPRESENTATIONS OF INFORMATION CHARACTERIZING BATCHING EQUIPMENT OPERATION MODES

The purpose of the paper is the analysis of the properties of time-frequency representation of the information on one-dimensional flow signals characterizing the operation modes of batching equipment. The properties of multidimensional (2D/3D-) representations in the time-frequency space are analyzed for one-dimensional flow signals when describing the batching process performed by continuous and discrete-type devices as components of a mixture-producing unit for the production of finely dispersed dry compositions. The properties of the visual-graphic representation of information on operation modes of the batching equipment are given and described. One-dimensional flow signals at the output unit of batching devices are displayed in a wavelet environment in the two-dimensional/three-dimensional space. Such transformation transfers one-dimensional measurement signals into a multi-dimensional one-point environment where one-dimensional material-flow signals in each technological point of the batch preparation equipment are represented by their multidimensional presentations (energy distributions) in the time-frequency space. Consideration is given to four analysis problems of Wigner time-frequency distribution, which represent specific modes of a batching process. Batching modes are characterized by mono- and multi-component stationary and non-stationary flow signals generated by continuous-type batching devices. In this case time-frequency presentations of flow rate signals are presented in the form of multidimensional Wigner and Choi-Williams distributions in three-dimensional (in the first three problems) and two-dimensional (in the fourth problem) environments. The first task describes a practical example of calculation and analysis of a multidimensional presentations of stationary and non-stationary time-frequency-dependent (chirp) signals with varied parameters of the latter. The calculated discreteness of Wigner distributions is explained. Other problems demonstrate the effect of damping parasitic elements in the Choi-Williams distribution. In particular, the second problem deals with the representation variant of a two-component harmonic signal. The third task presents the analysis of a batching signal with two chirp components having mutually reverse frequencies. The fourth problem also analyzes a two-component signal, one of which is a chirp signal with a linearly increasing frequency, another - a complex chirp-signal with a falling sine-modulated instantaneous frequency. The last task also presents a practical example of obtaining a multidimensional presentation of a multi-component signal from the block of two or more batchers. The semantic transparency of representations is noted under monitoring of the process of multi-component batching. The corrected Wigner distribution (Choi-Williams one) is more appropriate for the purposes of computer-aided control (in the form of visual-graphic monitoring) and automatic control in the regimes of stabilization and tracking the dynamics of non-stationary processes of various nature including the batching processes in different branches of industry and agro-industrial complex.

batching modes, continuous-type and discrete batchers, multi-component flow rate signals, virtual distribution elements, chirp-signals, Wigner and Choi-Williams representations

1. Cohen L. Time-frequency distributions - A Review // Proceedings of the IEEE, 1989, July. Vol. 77. No. 7. P. 94-981.

2. Mallat St. A wavelet tour of signal processing, Academic Press, 2nd Ed., Ecole Politechnique, Paris; Courant Institute, New York University, Library of Congress Catalog Card Number: 99-65087. 1999, Reprinted 2001. 637 p.

3. Boashash B., Touati S., Auger F., Flandrin P., Chassande-Mottin E., et al. Measures, performance assessment, and enhancement TFDs // Time-frequency signal analysis and processing: a comprehensive reference, Academic Press. 2016, January. P. 387-452.

4. Auger F., Chassande-Mottin, E. Quadratic time-frequency analysis I: Cohen's class / in "Time-frequency analysis: concepts and methods", ISTE. 2008 (January). P. 131-163.

5. Debnath L. Recent development in the Wigner-Ville distribution and time-frequency signal analysis, PINSA, 68, A, No. 1. 2002. January. P. 35-56.

6. Debnath L. Wavelet transforms and their applications, Birkhauser, Boston, 2002.

7. Mallat S., Zhang Z. Matching Pursuit With Time- Frequency Dictionaries, IEEE Transactions on Signal Processing. 1993. Vol. 41. No. 12. P. 3397-3414.

8. Choi H.L. and Williams W. J. IEEE Trans. Acoust., Speech, Signal Processing. 1989. Vol. 37. P. 862.

9. Мартуганова Е.Р. Модель web-сервиса по специализированной обработке данных на основе жадных алгоритмов. М.: МГУ им. М.В. Ломоносова. 2014. 86 p.

10. Townsend S., Lee B., Jr. Sparse Approximation and Atomic Decomposition: Considering Atom Interactions in Evaluating and Building Signal Representations, Dissertation, March 2009. 260 p.

11. Ferrando S.E., Doolittle E.J., Bernal A.J., Bernal L.J. Probabilistic matching pursuit with Gabor dictionaries, Signal Processing. 2000. Vol. 80. P. 2099-2120.

12. Davis G.M., Mallat S.G., Zhang Z. Adaptive time-frequency decomposition with matching pursuit, Proc. SPIE 2242, Wavelet Applications. 402. 1994.

13. Gribonval R., Depalle P., Rodet X., Bacry E., and Mallat S. Sound signals decomposition using a high resolution matching pursuit. In Proc. Int. Computer Music Conf. (ICMC'96), August 1996. P. 293-296.

14. Tropp J.A., Gilbert A.C. Signal recovery from random measurements via orthogonal matching pursuit, IEEE Transactions on information theory. 2007. Vol. 53. No. 12. P. 4655-4666.