Papers by Keyword: IMF

Abstract: Lung sound analysis plays an important role in the assessment and diagnosis of respiratory conditions and diseases. It can provide valuable information about the functioning of the respiratory system, including the airways, lungs, and associated structures. By analyzing the characteristics of lung sounds, healthcare professionals can gain insights into the presence of abnormalities, such as airway obstructions, lung diseases, and respiratory infections. In this paper, a multiple channel model for processing and classifying abnormalities in lung sound is proposed, which utilize the characteristics of Mel spectrogram and the Empirical Mode Decomposition (EMD). Unlike previous research which directly convert the lung sound into scalogram or spectrogram, the pre-processing of the original audio signal is considered and focused in this paper. This pre-processing step includes denoising, resampling, padding and augmentation, which incredibly increase the quality of the input signal. Finally, the multiple channel is put into the VGG16 deep learning model to classify the abnormalities in lung sound, including wheezes, crackles, and both. The model is trained and tested on the benchmark ICBHI dataset. The proposed model has shown better performance when compared with the state-of-the-art researches.

Abstract: In the article it is solving problems of effective functioning of the financial system in the direction of economic development of regions of Ukraine.

Abstract: In the process of imaging, digitalization and transmission, images are generally contaminated by Gaussian noise and salt & pepper noise, which cannot be eliminated completely at the same time only by Mean filter or Median filter. Aiming at solving this problem, an improved hybrid median-mean filter algorithm based on the Improved Median Filtering (IMF) algorithm is proposed in this paper. The experimental results show that the new algorithm shows better performance than either Median filtering algorithm or Mean filtering algorithm, which can not only get rid of Gaussian noise and salt & pepper noise simultaneously, but also minimize the contradictions between noise erasing and image details protecting effectively.

Abstract: A new method of pitch detection is proposed based on Hilbert-Huang Transform (HHT). Firstly noisy speech signal is filtered by morphological filtering to remove the noise, and then HHT is employed to get Hilbert-Huang spectrum and to calculate instantaneous energy and its derivative. Distinguish the unvoiced and voiced using mutation of instantaneous energy and track pitch.

Abstract: Health monitoring of the bridge structure has gradually become one of the hot topics. The signal decomposition technology is the key technique of the bridge structural health monitoring. The traditional data analysis and processing methods, which can only be applied to stationary or linear signal processing, have significant limitations. However, the structural response signals tested are mostly non-stationary and nonlinear. So methods that can effectively analyze non-stationary and nonlinear signal are urgently needed. Based on the summarization and analysis of the shortage of wavelet analysis method, the application of local wave method for data processing and analysis in structural health monitoring is put forward. The feasibility and superiority of local wave method is discussed. Experimental simulation results show that the application of local wave method in bridge health monitoring signal decomposition is feasible.

Abstract: In ultrasonic testing of coarse-grain materials, signal to noise ratio (SNR) is so poor because of the serious structure noise, and reflected wave from defects is difficult to be identified. In order to improve SNR and increase the reliability of ultrasonic testing for coarse grain materials, Hilbert-Huang Transform (HHT) is introduced to analyze and process the testing signal here. Firstly, detected signals from the coarse grain material can be collected by using ultrasonic test system; And then many Intrinsic Mode Function (IMF) can be obtained according to Empirical Mode Decomposition (EMD), and marginal spectrum of different mode can be gotten by Hilbert transform; And finally, the noise should be removed after analyzing the time-frequency information, and SNR is able to be enhanced and the reflection wave from defect is being more obvious. It was shown from the experimental result that the ineffective structure noise could be removed after HHT, and SNR could be improved and the defect reflection is more outstanding.

Abstract: Aiming at the problem that Current Transformer (CT) affects differential protection, this paper proposes a new method to identify CT saturation based on Empirical Mode Decomposition (EMD). Differential current can be decomposed into few Intrinsic Mode Functions (IMFs) by EMD. When CT is linear transfer, the current waveform is nearly a sine wave, and it only contains one dominant IMF. However, CT saturation leads to the distortion of secondary current which contains at least two dominant IMFs. From the defined projection area on t-axis of each IMF and the specific gravity coefficient, the number of dominant IMF can be got. Thus CT saturation can be identified. Theoretical analysis and simulation results show that this method can identify CT saturation of different degree. It is convenient to achieve and hardly to be affect by aperiodic component.

Abstract: It can obtain intrinsic mode function (IMF) of signal wave with empirical mode decomposition (EMD) in harmonic analysis of power system. Harmonic frequency, amplitude and duration can be obtained through analysis of IMFs. Through EMD analysis on distortion waveform of single-phase AC inverter output as an example, combined with applied scene of distorted voltage, cause of distortion waveform can be deduced. The result shows that EMD analysis on non-stationary signals is of good performance, and a new substitute method of FFT transform in harmonic analysis.

Abstract: In order to solve the endpoint effect and modal aliasing phenomenon in EMD and EEMD,Improved EEMD is put forward, and the application in signal singularity detection is researched in this paper. The improved EEMD will signal drops down into a series of different IMF to highlight the different local characteristics of original data, and then calculate Hilbert marginal spectrum and time-frequency spectrum to determine the frequency of these mutations and mutations position. To compared with FT, STFT, WVD,WT, EMD and EEMD etc, No cross-terms and no false IMF components are produced in the Hilbert time-frequency spectrum of the improved EEMD. Different frequency components and frequency mutations position are clearly distinguished at the same time. The Hilbert time-frequency spectrum of the improved EEMD has more superior detection signal singularity ability.

Abstract: Condition monitoring of reciprocating machines through the analysis of their vibrations has been recognized to be a difficult issue, essentially because of the strong nonlinearity of the vibration signals. A new method of multi-component singular entropy is put forward to resolve this problem. Local Wave method is combined with Singular Entropy to extract features from the IMF of the vibration signals of reciprocating machines. And the features will be used as the input of ANFIS to classify and recognize the fault mode. The results are classified correctly. The conclusion shows that this method is feasible.