A prediction of the frequency of non-periodic signals based on convolutional neural networks
The problem on creation of mathematical support for construction of forecast models based on convolutional neural networks is solved in the work. A method is proposed for using convolutional neural networks to predict the frequency of non-periodic signals. To determine the frequency of the signal, i...
Збережено в:
| Дата: | 2018 |
|---|---|
| Автори: | , , |
| Формат: | Стаття |
| Мова: | Ukrainian |
| Опубліковано: |
Інститут проблем реєстрації інформації НАН України
2018
|
| Теми: | |
| Онлайн доступ: | http://drsp.ipri.kiev.ua/article/view/158515 |
| Теги: |
Додати тег
Немає тегів, Будьте першим, хто поставить тег для цього запису!
|
| Назва журналу: | Data Recording, Storage & Processing |
Репозитарії
Data Recording, Storage & Processing| id |
drspiprikievua-article-158515 |
|---|---|
| record_format |
ojs |
| spelling |
drspiprikievua-article-1585152019-12-27T01:23:42Z A prediction of the frequency of non-periodic signals based on convolutional neural networks Прогнозування частоти неперіодичних сигналів на основі згорткових нейронних мереж Subbotin, S. A. Korniienko, O. V. Drokin, I. V. forecasting signal training neural network convolution error gradient прогнозування сигнал навчання нейронна мережа згортка помилка градієнт The problem on creation of mathematical support for construction of forecast models based on convolutional neural networks is solved in the work. A method is proposed for using convolutional neural networks to predict the frequency of non-periodic signals. To determine the frequency of the signal, it was divided into parts, after which a fast Fourier transform was used for each part. The spectrograms obtained after the transform are used as inputs to the learning of the neural network. The output value depends on the presence or absence of a frequency that is above the critical value on the predicted interval. The first layer of the neural network uses a three-dimensional convolution, and on the next layers - a onedimensional convolution. Between the convolutional layers, there are subsampling layers used to accelerate learning and prevent retraining. The neural network contains two output neurons which determine the presence of a frequency that exceeds the critical value. The practical task of predicting the frequency of vibration of aircraft engines during their tests is solved. The construction of different neural network models, their training and testing on the data that were collected from vibration sensors during the testing of the aircraft engine has been performed. To increase the amount of data, augmentation is used. To do this, several copies of the signal with changed frequencies are added. The models constructed differ in the amount of data used and in the forecasting time. Comparison of the test results of all the models has been performed. The maximum forecasting time that can be achieved with the proposed method is determined. This time is enough for the pilot to react and change the flight mode or to land the helicopter. Вирішено завдання створення математичного забезпечення для побудови прогнозних моделей на основі згорткових нейронних мереж. Запропоновано метод використання згорткових нейронних мереж для прогнозування частоти неперіодичних сигналів. Вирішено практичне завдання прогнозування частоти вібрацій авіаційних двигунів при проведені їхніх випробувань. Виконано побудову нейромережевих моделей, їхнє навчання та тестування на даних, які було зібрано з датчиків вібрацій при проведені випробувань авіадвигуна. Порівняно результати тестування всіх побудованих моделей. Інститут проблем реєстрації інформації НАН України 2018-09-18 Article Article application/pdf http://drsp.ipri.kiev.ua/article/view/158515 10.35681/1560-9189.2018.20.3.158515 Data Recording, Storage & Processing; Vol. 20 No. 3 (2018); 29–36 Регистрация, хранение и обработка данных; Том 20 № 3 (2018); 29–36 Реєстрація, зберігання і обробка даних; Том 20 № 3 (2018); 29–36 1560-9189 uk http://drsp.ipri.kiev.ua/article/view/158515/157870 Авторське право (c) 2021 Реєстрація, зберігання і обробка даних |
| institution |
Data Recording, Storage & Processing |
| baseUrl_str |
|
| datestamp_date |
2019-12-27T01:23:42Z |
| collection |
OJS |
| language |
Ukrainian |
| topic |
forecasting signal training neural network convolution error gradient |
| spellingShingle |
forecasting signal training neural network convolution error gradient Subbotin, S. A. Korniienko, O. V. Drokin, I. V. A prediction of the frequency of non-periodic signals based on convolutional neural networks |
| topic_facet |
forecasting signal training neural network convolution error gradient прогнозування сигнал навчання нейронна мережа згортка помилка градієнт |
| format |
Article |
| author |
Subbotin, S. A. Korniienko, O. V. Drokin, I. V. |
| author_facet |
Subbotin, S. A. Korniienko, O. V. Drokin, I. V. |
| author_sort |
Subbotin, S. A. |
| title |
A prediction of the frequency of non-periodic signals based on convolutional neural networks |
| title_short |
A prediction of the frequency of non-periodic signals based on convolutional neural networks |
| title_full |
A prediction of the frequency of non-periodic signals based on convolutional neural networks |
| title_fullStr |
A prediction of the frequency of non-periodic signals based on convolutional neural networks |
| title_full_unstemmed |
A prediction of the frequency of non-periodic signals based on convolutional neural networks |
| title_sort |
prediction of the frequency of non-periodic signals based on convolutional neural networks |
| title_alt |
Прогнозування частоти неперіодичних сигналів на основі згорткових нейронних мереж |
| description |
The problem on creation of mathematical support for construction of forecast models based on convolutional neural networks is solved in the work. A method is proposed for using convolutional neural networks to predict the frequency of non-periodic signals. To determine the frequency of the signal, it was divided into parts, after which a fast Fourier transform was used for each part. The spectrograms obtained after the transform are used as inputs to the learning of the neural network. The output value depends on the presence or absence of a frequency that is above the critical value on the predicted interval. The first layer of the neural network uses a three-dimensional convolution, and on the next layers - a onedimensional convolution. Between the convolutional layers, there are subsampling layers used to accelerate learning and prevent retraining. The neural network contains two output neurons which determine the presence of a frequency that exceeds the critical value. The practical task of predicting the frequency of vibration of aircraft engines during their tests is solved. The construction of different neural network models, their training and testing on the data that were collected from vibration sensors during the testing of the aircraft engine has been performed. To increase the amount of data, augmentation is used. To do this, several copies of the signal with changed frequencies are added. The models constructed differ in the amount of data used and in the forecasting time. Comparison of the test results of all the models has been performed. The maximum forecasting time that can be achieved with the proposed method is determined. This time is enough for the pilot to react and change the flight mode or to land the helicopter. |
| publisher |
Інститут проблем реєстрації інформації НАН України |
| publishDate |
2018 |
| url |
http://drsp.ipri.kiev.ua/article/view/158515 |
| work_keys_str_mv |
AT subbotinsa apredictionofthefrequencyofnonperiodicsignalsbasedonconvolutionalneuralnetworks AT korniienkoov apredictionofthefrequencyofnonperiodicsignalsbasedonconvolutionalneuralnetworks AT drokiniv apredictionofthefrequencyofnonperiodicsignalsbasedonconvolutionalneuralnetworks AT subbotinsa prognozuvannâčastotineperíodičnihsignalívnaosnovízgortkovihnejronnihmerež AT korniienkoov prognozuvannâčastotineperíodičnihsignalívnaosnovízgortkovihnejronnihmerež AT drokiniv prognozuvannâčastotineperíodičnihsignalívnaosnovízgortkovihnejronnihmerež AT subbotinsa predictionofthefrequencyofnonperiodicsignalsbasedonconvolutionalneuralnetworks AT korniienkoov predictionofthefrequencyofnonperiodicsignalsbasedonconvolutionalneuralnetworks AT drokiniv predictionofthefrequencyofnonperiodicsignalsbasedonconvolutionalneuralnetworks |
| first_indexed |
2025-07-17T10:57:19Z |
| last_indexed |
2025-07-17T10:57:19Z |
| _version_ |
1850411303336673280 |