OPTIMIZATION OF THE APPLICATION OF LOCAL APPROXIMATION FOR THE SMOOTHING PROCESS

UDC 621.3.011.712

 

Hryniuk Dzmitry Anatol’yevich – PhD (Engineering), Associate Professor, Assistant Professor, the Department of Automation Production Processes and Electrical Engineering. Belarusian State Technological University (13a, Sverdlova str., 220006, Minsk, Republic of Belarus). E-mail: hryniukda@gmail.com

Oliferovich Nadezhda Mikhaylovna – Senior Lecturer, the Department of Automation of Production Processes and Electrical Engineering. Belarusian State Technological University (13a, Sverdlova str., 220006, Minsk, Republic of Belarus). E-mail: oliferovich@belstu.by

Suhorukova Irina Gennad’yevna – Senior Lecturer, the Department of Software Engineering. Belarusian State Technological University (13a, Sverdlova str., 220006, Minsk, Republic of Belarus). E-mail: irina_x@rambler.ru

 

DOI: https://doi.org/ 10.52065/2520-6141-2024-284-9.

Key words: filtering, smoothing of measurement trends, local approximation.

For citation: Hryniuk D. A., Oliferovich N. M., Suhorukova I. G. Optimization of the application of local approximation for the smoothing process. Proceedings of BSTU, issue 3, Physics and Mathematics. Informatics, 2024, no. 2 (284), pp. 58–69 (In Russian). DOI: 10.52065/2520-6141-2024-284-9.

Abstract

The article analyzes methods for smoothing measurement trends and highlights their strengths and weaknesses. It is concluded that the local approximation method has a number of advantages. To effectively use this method, a simulation of the operation of the selected filter was carried out for periodic signals with one and several harmonics, at a constant sampling time and at different frequencies. The influence of local approximation on the measuring signal was assessed by the change in amplitude, total harmonic distortion factor and the average difference between the original and smoothed signal. When modeling, polynomials of the first, second and third orders were used for local approximation. For the smoothing process itself, the approximation window and time operator were varied. Simulation has shown that nonlinear filtering distortions are very small if you don’t go beyond the passband of this filter. The results of the filter for a first- and second-order polynomial have a significant difference, while for the second and third they coincide if the time operator is equal to half the time of the window width. When varying the time operator from the center, the smoothing results for secondand third-order polynomials seem to be very different. The possibilities of varying the time operator for a second-order polynomial, from the point of view of distortion of the original information, are wider than for the third. Based on the change in amplitude for a harmonic signal and the total harmonic distortion factor, formulas are obtained for calculating the window width by frequency or signal speed. These results are proposed to be used for the adaptation process.

Download

References

  1. Shrivastava P., Singh U. P. Noise removal using first order neighborhood mean filter. Conference on IT in Business, Industry and Government (CSIBIG). Indore, 2014, pp. 1–6. DOI: 10.1109/CSIBIG.2014.7057004.
  2. Fu B., Xiong X., Sun G. An efficient mean filter algorithm. The 2011 IEEE/ICME International Conference on Complex Medical Engineering. Harbin, 2011, pp. 466–470. DOI: 10.1109/ICCME.2011.5876785.
  3. Zhang H., Peng Z., Tang J., Dong M., Wang K., Li W. A multi-layer extreme learning machine refined by sparrow search algorithm and weighted mean filter for short-term multi-step wind speed forecasting. Sustain Energy Technol, 2022, vol. 50, article 101698. DOI: 10.1016/j.seta.2021.101698.
  4. Sinha P. K., Hong Q. H. An improved median filter. IEEE Transactions on Medical Imaging, 1990, vol. 9, no. 3, pp. 345–346. DOI: 10.1109/42.57773.
  5. Gallagher N. C. Median filters: a tutorial. IEEE International Symposium on Circuits and Systems. Espoo, 1988, vol. 2, pp. 1737–1744. DOI: 10.1109/ISCAS.1988.15271.
  6. Sun Y., Tan W., Chen T. A method to remove chattering alarms using median filters. ISA Transactions, 2018, vol. 73, pp. 201–207. DOI: 10.1016/j.isatra.2017.12.012.
  7. Savitzky A., Golay M. J. E. Smoothing and Differentiation of Data by Simplified Least Squares Procedures. Analytical Chemistry, 1964, vol. 36, no. 8, pp. 1627–1639.
  8. Magazine S. P., Schafer R. W. What is a savitzky-golay filter? IEEE Signal processing magazine. IEEE, 2011, vol. 28, issue 4, pp. 111–117. DOI: 10.1109/MSP.2011.941097.
  9. Kalman R. E. A new approach to linear filtering and prediction problems. Transactions of the ASME – Journal of Basic Engineering, 1960, vol. 82, series D, pp. 35–45.
  10. Pham S. Dinh A. Using the Kalman algorithm to correct data errors of a 24-bit visible spectrometer. Sensors, 2017, vol. 17, issue 12, pp. 1–20. DOI: 10.3390/s17122939.
  11. Deng G., Cahill L.W. An adaptive Gaussian filter for noise reduction and edge detection. IEEE Conference Record Nuclear Science Symposium and Medical Imaging Conference. San Francisco, CA, USA, 1993, pp. 1615–1619. DOI: 10.1109/NSSMIC.1993.373563.
  12. Kliewer A. J., Fletcher S. J., Jones A. S., Forsythe J. M. Comparison of Gaussian, logarithmic transform and mixed Gaussian-log-normal distribution based 1DVAR microwave temperature-water-vapour mixing ratio retrievals. Quarterly Journal of the Royal Meteorological Society, 2015, vol. 142 (694), pp. 274– 286. DOI: 10.1002/qj.2651.
  13. Baszczynska A. One value of smoothing parameter vs interval of smoothing parameter values in kernel density estimation. Acta Universitatis Lodziensis. Folia Oeconomica. Lodz University Press, 2017, vol. 6, no. 332, pp. 73–86. DOI: 10.18778/0208-6018.332.05.
  14. Rosenblatt M. Remarks on some nonparametric estimates of a density function. The Annals of Mathematical Statistics, 1956, vol. 27, issue 3, pp. 832–837. DOI: 10.1214/aoms/1177728190.
  15. Kowalski P., Smyk R. Review and comparison of smoothing algorithms for one-dimensional data noise reduction. International Interdisciplinary PhD Workshop (IIPhDW). Świnouście, 2018, pp. 277–281. DOI: 10.1109/IIPHDW.2018.8388373. 16. Hryniuk D., Suhorukova I., Oliferovich N. Adaptive smoothing and filtering in transducers. Electrical, Electronic and Information Sciences (eStream): Open Conference. Vilnius, 2016, pp. 1–4. DOI: 10.1109/eStream39242.2016.7485917.
  16. Hryniuk D. A., Oliferovich N. M., Suhorukova I. G. The use of algorithmic approaches for smoothing of measurement information. Trudy BGTU [Proceedings of BSTU], issue 3, Physics and Mathematics. Informatics, 2017, no. 2 (200), pp. 82–87 (In Russian).
  17. Hvozdzeu M., Karpovich M. Dynamic signals filtration in high level noise condition. Mokslas – Lietuvos Ateitis / Science – Future of Lithuania. Vilnius, 2020, vol. 12, pp. 1–3. DOI: 10.3846/mla.2020.11487.
  18. Hryniuk D. A., Suhorukova I. G., Alekseev V. L., Shilin A. M. Study of ski vibrations under the action of a pneumatic cylinder. Khimicheskaya tekhnologiya i tekhnika: materialy 85-y nauchno-tekhnicheskoy konferentsii professorsko-prepodovatel'skogo sostava, nauchnykh sotrudnikov i aspirantov (s mezhdunarodnym uchastiyem) [Chemical Technology and Engineering: Proceedings of the 85th Scientific and Technical Conference of Faculty, Researchers and Postgraduates (with International Participation)]. Minsk, 2021, pp. 228–230 (In Russian).
  19. Katkovnik V. Ya. Neparametricheskaya identifikatsija i sglazhivaniye dannykh: metod lokal'noy approksimatsii [Nonparametric identification and smoothing of data: local approximation method]. Moscow, Science Publ., 1985, 336 p. (In Russian).

25.04.2024