ТЕХНОЛОГІЇ ФОРМУВАННЯ ЕКВІВАЛЕНТНИХ ЗАВАД ЗАШУМЛЕНИХ СИГНАЛІВ ТА ЇХ ЗАСТОСУВАННЯ

• Дата: 2020
• Автори: Aliev, T.A., Musaeva, N.F., Rzayeva, N.E., Mamedova, A.I.
• Формат: Стаття
• Мова: English
• Опубліковано: V.M. Glushkov Institute of Cybernetics of NAS of Ukraine 2020
• Теми: сигнал; корисний сигнал; перешкода; зашумлений сигнал; еквівалентні відліки; оцінка; похибка; кореляційна функція; об’єкт; діагностика; ідентифікація
• Онлайн доступ: https://jais.net.ua/index.php/files/article/view/469
• Назва журналу: Problems of Control and Informatics

It is shown that in controlled objects noisy signals are formed by useful signals and noises. In this case, it is usually impossible to isolate the noise from the noisy signal. For this reason, the estimates of the statistical characteristics of noisy signals calculated using traditional technologies contain significant errors. At the same time, separate processing of the useful signal and the noise makes it possible to identify moments of change in the current state of a technical object, as well as to extract the necessary diagnostic information contained in the noise characteristics. Therefore, algorithms and technologies are developed for forming, from samples of the noisy signal, samples of the useful signal and samples of the noise separately, which are called equivalent samples of the useful signal and equivalent samples of the noise. It is shown that despite the difference in the values of real and equivalent samples, the characteristics of these signals practically coincide. This allows, firstly, increasing the accuracy of estimates of the correlation and spectral characteristics of noisy signals in comparison with estimates obtained by traditional technologies. Secondly, the same characteristics of the equivalent useful signal and the equivalent noise obtained separately allow controlling the beginning of the latent period of the initiation of malfunctions of objects under investigation. In addition, it is also found that the estimates of the cross-correlation function between the useful signal and the noise contain diagnostic information. Computational experiments are carried out. For this, the useful signal and the noise with set characteristics are modeled, and the noisy signal is generated. Equivalent samples of the noise and the useful signal are determined. The characteristics of the generated useful signal and the noise, as well as the equivalent useful signal and equivalent noise, are calculated by the traditional algorithms. A comparative analysis is carried out. To this end, the relative errors of the characteristics of the generated and equivalent useful signals and noise are determined. The calculation results show that the characteristics of the generated useful signal and interference, as well as the equivalent useful signal and noise practically coincide. It is shown that the developed algorithms and technologies can be used in monitoring and control systems to improve the accuracy of the analysis of noisy signals received at the outputs of the sensors. These technologies can also be used in information measuring and other measuring complexes and systems, which will significantly improve their efficiency.