Empirical Mode Decomposition Technique Research Articles

Characteristics of 2020 stock market crash: The COVID-19 induced extreme event.

A sudden fall of stock prices happens during a pandemic due to the panic sell-off by the investors. Such a sell-off may continue for more than a day, leading to a significant crash in the stock price or, more specifically, an extreme event (EE). In this paper, Hilbert-Huang transformation and a structural break analysis (SBA) have been applied to identify and characterize an EE in the stock market due to the COVID-19 pandemic. The Hilbert spectrum shows a maximum energy concentration at the time of an EE, and hence, it is useful to identify such an event. The EE's significant energy concentration is more than four times the standard deviation above the mean energy of the normal fluctuation of stock prices. A statistical significance test for the intrinsic mode functions is applied, and the test found that the signal is not noisy. The degree of nonstationarity test shows that the indices and stock prices are nonstationary. We identify the time of influence of the EE on the stock price by using SBA. Furthermore, we have identified the time scale ( τ) of the shock and recovery of the stock price during the EE using the intrinsic mode function obtained from the empirical mode decomposition technique. The quality stocks with V-shape recovery during the COVID-19 pandemic have definite τ of shock and recovery, whereas the stressed stocks with L-shape recovery have no definite τ. The identification of τ of shock and recovery during an EE will help investors to differentiate between quality and stressed stocks. These studies will help investors to make appropriate investment decisions.

Identification of the typical frequency range associated with subsurface gas zones: a study using Hilbert-Huang transform and wavelet analysis

Frequency attributes embedded in nonlinear geophysical signals such as well-log data or seismic data can provide vital clues in the effective characterization of the subsurface reservoir. Therefore, it is always important to recognize suitable mathematical techniques that need to be employed to extract such vital information from the nonlinear geophysical data. Several nonlinear mathematical techniques include wavelet analysis, fractal and multifractal analyses, and the fully data-adaptive Hilbert-Huang transform (HHT) technique. In the present study, we employ the HHT and the wavelet analysis techniques to seismic data of the Dutch sector of southern North Sea. We use a small portion of the seismic data, consisting of inline section 133 and time slice section 520 ms of the F3 block, to extract the instantaneous frequency associated with the subsurface gas zone. HHT constitutes two independent techniques, namely, the empirical mode decomposition (EMD) technique and the Hilbert spectral analysis (HSA). While the EMD technique facilitates to decompose the data into different signals of varied frequencies, called intrinsic mode functions (IMFs), the HSA helps to determine the instantaneous amplitudes and frequencies of the IMFs. EMD technique, together with HSA, has delineated a low-frequency spectral signature, believed to be a subsurface gas zone in the chosen data, corroborating the results of earlier studies of the same data sets. The instantaneous frequency range of the gas zone estimated by HSA was validated with the two-dimensional continuous wavelet transform (CWT) of the data sets using a suitable frequency range of Morlet wavelet. Results of the combined HHT and CWT analyses confirmed that the subsurface gas zone could be well identified in the frequency range 14–18 Hz. Results have direct implications in situ, in the sense that, by performing CWT analysis of seismic data using a Morlet wavelet in the frequency range 14–18 Hz, one can easily identify the location of subsurface gas zones.

Simulating a Stochastic Signal of Urban Water Demand by a Novel Combination of Data Analytic and Machine Learning Techniques

In this research, a new methodology is presented to forecast the stochastic component of urban water demand for Baghdad City from 2003 to 2014. The methodology contains data preprocessing to analyse raw time series of water via Empirical Mode Decomposition (EMD) technique and select the best scenario of independent variables by a stepwise regression method. Artificial neural network (ANN) is integrated by Backtracking Search Algorithm (BSA) to find the best factors of the ANN model. The outcomes reveal that data pre-processing can detect the stochastic signal of water data and choice the best model input’s scenario. BSA successfully determines the parameters of the ANN model. The methodology accurately simulated the stochastic signal of water time series depend on different statistical criteria such as coefficient of determination and mean absolute relative error equal to 0.99 and 0.0208, respectively.

A hybrid approach for El Niño prediction based on Empirical Mode Decomposition and convolutional LSTM Encoder-Decoder

El Niño can affect climate patterns, causing extreme weather events, such as floods and droughts, around the world. Accurate forecasting of El Niño events allows preparation for El Niño-related disasters. However, the performance of current methods for predicting El Niño events one year in advance is not effective. This study proposes a hybrid approach to predicting the El Niño-related Oceanic Niño Index (ONI) and El Niño events with a lead time of 12 months. The proposed approach combines the convolutional Long Short-Term Memory (LSTM) Encoder-Decoder model with the Empirical Mode Decomposition (EMD) technique. The EMD technique can decompose time series data into a set of Intrinsic Mode Functions and a residue. Subsequently, the convolutional LSTM Encoder-Decoder model is employed to make an independent prediction for each component. Finally, the predicted data from each model can be reconstructed to obtain the forecasting results. The proposed approach is applied to the monthly ONI dataset for 1950–2019. The prediction model is trained and validated on the historical ONI values from 1950 to 2007 and forecasts El Niño events over a period of 12 years (2008–2019) with a lead time of 12 months. The results demonstrate that the proposed approach can successfully forecast that, for this period, 2009–2010, 2015–2016, and 2018–2019 are El Niño years. The performance of the proposed approach is then assessed by comparing it with the standalone convolutional LSTM Encoder-Decoder model, the LSTM-based models, and machine learning algorithms. The evaluated results indicate that the proposed approach outperforms these models in ONI predictions and El Niño event forecasts for a lead time of 12 months.

Automated detection of hypertension using wavelet transform and nonlinear techniques with ballistocardiogram signals

Arterial hypertension (HT) is a common cardiovascular disease and, if not treated at an early stage, can lead to serious complications. It is difficult to precisely describe because it is a dynamic physiological state. Monitoring of HT is subjective and prone to mistakes. Therefore, various computer-assisted diagnostic methods have been designed. The proposed work is based on ballistocardiography (BCG) signals for the diagnosis of healthy control (HC) and HT subjects. The identification of HT based on BCG signals of 30-s duration using empirical mode decomposition (EMD) and wavelet transform (WT) with nonlinear techniques is proposed in this work. The BCG signal is decomposed into five sub-bands (SBs) using WT and five-level of intrinsic mode functions (IMFs) using EMD. Then, the various non-linear features are calculated for all five-level wavelet decomposition and IMFs. The non-linear features are extracted from the SBs of WT and IMFs of EMD. These features are fed to ensemble gentleboost (EGB) classifier with 10-fold cross-validation strategy for automated detection of HC and HT groups. In this study, the performance of WT and EMD techniques are compared. The proposed HT identification model accomplished the highest average classification accuracy of 89% using WT method. In future, we plan to extend this work using more BCG data.

Improving Purity of Blasting Vibration Signals Using Advanced Empirical Mode Decomposition and Wavelet Packet Technique

Improving Purity of Blasting Vibration Signals Using Advanced Empirical Mode Decomposition and Wavelet Packet Technique

Analysis and Forecasting COVID-19 Outbreak in Pakistan Using Decomposition and Ensemble Model

COVID-19 has caused severe health complications and produced a substantial adverse economic impact around the world. Forecasting the trend of COVID-19 infections could help in executing policies to effectively reduce the number of new cases. In this study, we apply the decomposition and ensemble model to forecast COVID-19 confirmed cases, deaths, and recoveries in Pakistan for the upcoming month until the end of July. For the decomposition of data, the Ensemble Empirical Mode Decomposition (EEMD) technique is applied. EEMD decomposes the data into small components, called Intrinsic Mode Functions (IMFs). For individual IMFs modelling, we use the Autoregressive Integrated Moving Average (ARIMA) model. The data used in this study is obtained from the official website of Pakistan that is publicly available and designated for COVID-19 outbreak with daily updates. Our analyses reveal that the number of recoveries, new cases, and deaths are increasing in Pakistan exponentially. Based on the selected EEMD-ARIMA model, the new confirmed cases are expected to rise from 213,470 to 311,454 by 31 July 2020, which is an increase of almost 1.46 times with a 95% prediction interval of 246,529 to 376,379. The 95% prediction interval for recovery is 162,414 to 224,579, with an increase of almost two times in total from 100802 to 193495 by 31 July 2020. On the other hand, the deaths are expected to increase from 4395 to 6751, which is almost 1.54 times, with a 95% prediction interval of 5617 to 7885. Thus, the COVID-19 forecasting results of Pakistan are alarming for the next month until 31 July 2020. They also confirm that the EEMD-ARIMA model is useful for the short-term forecasting of COVID-19, and that it is capable of keeping track of the real COVID-19 data in nearly all scenarios. The decomposition and ensemble strategy can be useful to help decision-makers in developing short-term strategies about the current number of disease occurrences until an appropriate vaccine is developed.

A Smart Sensor-Based cEMD Technique for Rotor Bar Fault Detection in Induction Motors

Induction motors (IMs) are commonly used in industrial and domestic applications. Reliable IM fault detection can improve machinery production quality and operation safety and prevent unexpected failures. However, reliable motor fault diagnosis is still a challenging research and development task, especially in real machinery monitoring applications, due to the limitations in data acquisition (DAQ) systems and fault detection techniques. The first objective of this work is to develop a wireless smart sensor DAQ system for the current signal measurement. The second objective is to propose a new correlation empirical mode decomposition (cEMD) technique to detect broken rotor bar faults. The proposed cEMD technique can differentiate the line current and its sideband components from other frequencies and accentuate fault features over intrinsic mode function sidebands so as to improve fault detection accuracy. The effectiveness of the developed smart sensor DAQ system and the cEMD technique is verified experimentally under different motor conditions.

Dispatch Planning of a Wide-Area Wind Power-Energy Storage Scheme Based on Ensemble Empirical Mode Decomposition Technique

This article addresses the challenging task of developing a procedure for the day-ahead dispatch planning of wind power which emanates from a wide geographical area. Using the complete ensemble empirical mode decomposition technique, it is shown that the low-frequency components of the area aggregated wind power account for the largest proportions of the perturbing energy harnessed from the wind. By taking advantage of the slow-varying characteristics of the low-frequency components, accurate forecast of these components is readily obtained and incorporated into the developed dispatch planning procedure. The dispatchability of the wide-area wind generation is facilitated by the buffering actions offered by a centralized power dispatch energy storage system, operating under a proposed power flows control strategy. The efficacy of the developed procedure is illustrated using a pumped hydroelectric system as the dispatch energy storage medium.

Optimal EEG channels selection for alcoholism screening using EMD domain statistical features and harmony search algorithm

Alcoholism can be analyzed by Electroencephalogram (EEG) data. Finding an optimal subset of EEG channels for alcoholism detection is a challenging task. The paper reports a new methodology for the detection of optimal channels for alcoholism analysis using EEG data. The proposed technique employs the Empirical Mode Decomposition (EMD) technique to extract the amplitude and frequency modulated bandwidth features from the Intrinsic Mode Function (IMF) and ensemble subspace K-NN as a classifier to classify alcoholics and normal. The optimum channels are selected, using a harmony search algorithm. The fitness value of discrete binary harmony search (DBHS) optimization algorithms is calculated using accuracy and sensitivity achieved by the ensemble subspace K-Nearest Neighbor classifier. Experimental outcomes indicate that the optimal channel selected by the harmony search algorithm has biological inference related to the alcoholic subject. The proposed approach reports a classification accuracy of 93.87%, with only 12 detected EEG channels.

DC Arc-Fault Detection Based on Empirical Mode Decomposition of Arc Signatures and Support Vector Machine

Protection devices are extensively utilized in direct current (DC) systems to ensure their normal operation and safety. However, series arc faults that establish current paths in the air between conductors introduce arc impedance to the system. Consequently, they can result in a decrease of current, and thus conventional protection devices may not be triggered. Undetected series arc faults can cause malfunctions and even lead to fire hazards. Therefore, a series arc-fault detection system is essential to DC systems to operate reliably and efficiently. In this paper, a series arc-fault detection system based on arc time-frequency signatures extracted by a modified empirical mode decomposition (EMD) technique and using a support vector machine (SVM) algorithm in decision making is proposed for DC systems. The oscillatory frequencies from the arc current are decomposed by the EMD with an analysis of the Hurst exponent ( ${H}$ ) to reject interference from the power electronics noise. ${H}$ analyzes the trend of a signal and the intrinsic oscillations of the signal are those with values of ${H}$ larger than 1/2. Comparing to traditional filters or wavelet transforms, this method does not require knowledge of the frequency range of the interference which varies from system to system. The capability and applicability of the proposed technique are validated in a photovoltaic system. The effectiveness of arc-fault detection is significantly improved by this technique because it can acquire sufficient and accurate arc signatures and it does not need to predefine various thresholds.

Adaptive denoising of ECG using EMD, EEMD and CEEMDAN signal processing techniques

Using adaptive signal processing techniques denoising of ECG signal is performed which is obtained from physionet database. In this paper, the baseline wandering noise is removed using different adaptive techniques such as Empirical Mode Decomposition (EMD), Ensemble Empirical Mode Decomposition (EEMD) and Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN). All these algorithms are effectively used to decompose the noisy ECG signal into different Intrinsic Mode Functions (IMFs) and further these IMFs are filtered using low pass filtering method to extract the low frequency baseline component. The high frequency noise present in the reconstructed signal is reduced by further decomposing into IMFs using all the three methods. These IMFs are soft thresholded to remove the high frequency noise. The results obtained from the CEEMDAN outperform EMD and EEMD in extracting signal from noise. Further, distinct parameters such as skewnesscrest factor, RMS value and kurtosis are estimated for the reconstructed signal to analyse their behaviour.

Arrhythmia Classification Using Radial Basis Function Network With Selective Features From Empirical Mode Decomposition

In this piece of work, the authors have attempted to classify four types of long duration arrhythmia electrocardiograms (ECG) using radial basis function network (RBFN). The data is taken from Massachusetts Institute of Technology-Beth Israel Hospital (MIT-BIH) arrhythmia database, and features are extracted using empirical mode decomposition (EMD) technique. For most informative contents average power (AP) and coefficient of dispersion (CD) are evaluated from six intrinsic mode function (IMFs) of EMD. Principal component analysis (PCA) is used for feature reduction for effective classification using RBFN. The performance is shown in the result section, and it is found that the classification accuracy is 95.98%.

Probability Density Forecasting of Wind Speed Based on Quantile Regression and Kernel Density Estimation

Based on quantile regression (QR) and kernel density estimation (KDE), a framework for probability density forecasting of short-term wind speed is proposed in this study. The empirical mode decomposition (EMD) technique is implemented to reduce the noise of raw wind speed series. Both linear QR (LQR) and nonlinear QR (NQR, including quantile regression neural network (QRNN), quantile regression random forest (QRRF), and quantile regression support vector machine (QRSVM)) models are, respectively, utilized to study the de-noised wind speed series. An ensemble of conditional quantiles is obtained and then used for point and interval predictions of wind speed accordingly. After various experiments and comparisons on the real wind speed data at four wind observation stations of China, it is found that the EMD-LQR-KDE and EMD-QRNN-KDE generally have the best performance and robustness in both point and interval predictions. By taking conditional quantiles obtained by the EMD-QRNN-KDE model as the input, probability density functions (PDFs) of wind speed at different times are obtained by the KDE method, whose bandwidth is optimally determined according to the normal reference criterion. It is found that most actual wind speeds lie near the peak of predicted PDF curves, indicating that the probabilistic density prediction by EMD-QRNN-KDE is believable. Compared with the PDF curves of the 90% confidence level, the PDF curves of the 80% confidence level usually have narrower wind speed ranges and higher peak values. The PDF curves also vary with time. At some times, they might be biased, bimodal, or even multi-modal distributions. Based on the EMD-QRNN-KDE model, one can not only obtain the specific PDF curves of future wind speeds, but also understand the dynamic variation of density distributions with time. Compared with the traditional point and interval prediction models, the proposed QR-KDE models could acquire more information about the randomness and uncertainty of the actual wind speed, and thus provide more powerful support for the decision-making work.

Comparison of wavelet and empirical mode decomposition hybrid models in drought prediction

Drought is a major area of interest within the field of water resources management, agriculture, energy resources and community health. Recently researchers have examined not only the mathematical expression of drought indices but also statistical predictions. Accordingly, high accuracy results were obtained using stand-alone machine learning techniques such as artificial neural networks (ANN) and support vector machine (SVM). However, lately, hybrid models have been introduced, which are created by integrating different time series decomposition techniques into standalone models, since the accuracy of stand-alone models used in the drought prediction being low particularly for mid-term and long-term drought predictions. In this study, self-calibrated Palmer Drought Severity Index (sc-PDSI) values were predicted by using three different standalone models and six hybrid models which are performed by two different decomposition techniques, such as Empirical mode decomposition (EMD) and Wavelet decomposition (WD). The main purpose of this study is to evaluate the effect of using EMD and WD for decomposing time series into their sub-bands on drought prediction. sc-PDSI time series were used to achieve 1, 3 and 6-month lead time predictions for Adana and Antalya cities located in the southern part of Turkey. Model performance indicators such as mean square error (MSE), Nash-Sutcliffe efficiency coefficient (NSE) and determination coefficient (R2) were employed to compare the proposed models. The results revealed that the accuracy of the stand-alone models, particularly in mid-term sc-PDSI predictions, was unsatisfactory. However, the prediction accuracy has been increased significantly with the introduction of EMD and WD techniques. Considering the Adana region, the hybrid wavelet models outperformed the models hybridized by EMD for not only 1-month lead time (NSEWD-ANFIS = 0.981 and NSEEMD-M5 = 0.890), but also for 3-month (NSEWD-SVM = 0.878 and NSEEMD-ANFIS = 0.811) and 6-month (NSEWD-ANFIS = 0.857 and NSEEMD-ANFIS = 0.783) lead times. According to the obtained results for Antalya region, similar findings were also observed among hybrid models. Thus, it is concluded that the predictions made using WD have higher accuracy than EMD, and the correct wavelet type selection has a significant effect on the results.

A Computerized Method for Automatic Detection of Schizophrenia Using EEG Signals.

Diagnosis of schizophrenia (SZ) is traditionally performed through patient's interviews by a skilled psychiatrist. This process is time-consuming, burdensome, subject to error and bias. Hence the aim of this study is to develop an automatic SZ identification scheme using electroencephalogram (EEG) signals that can eradicate the aforementioned problems and support clinicians and researchers. This study introduces a methodology design involving empirical mode decomposition (EMD) technique for diagnosis of SZ from EEG signals to perfectly handle the behavior of non-stationary and nonlinear EEG signals. In this study, each EEG signal is decomposed into intrinsic mode functions (IMFs) by the EMD algorithm and then twenty-two statistical characteristics/features are calculated from these IMFs. Among them, five features are selected as significant feature applying Kruskal Wallis test. The performance of the obtained feature set is tested through several renowned classifierson a SZ EEG dataset. Among the considered classifiers, theensemble bagged tree performed as the best classifier producing 93.21% correct classification rate for SZ, with an overall accuracy of 89.59% for IMF 2. These results indicate that EEG signals discriminate SZ patients from healthy control (HC) subjects efficiently and have the potential to become a tool for the psychiatrist to support the positive diagnosis of SZ.

Comparison of time-frequency-analysis techniques applied in building energy data noise cancellation for building load forecasting: A real-building case study

Time-frequency analysis that disaggregates a signal in both time and frequency domain is an important supporting technique for building energy analysis such as noise cancellation in data-driven building load forecasting. There is a gap in the literature related to comparing various time–frequency-analysis techniques, especially discrete wavelet transform (DWT) and empirical mode decomposition (EMD), to guide the selection and tuning of time–frequency-analysis techniques in data-driven building load forecasting. This article provides a framework to conduct a comprehensive comparison among thirteen DWT/EMD techniques with various parameters in a load forecasting modeling task. A real campus building is used as a case study for illustration. The DWT and EMD techniques are also compared under various data-driven modeling algorithms for building load forecasting. The results in the case study show that the load forecasting models trained with noise-cancelled energy data have increased their accuracy to 9.6% on average tested under unseen data. This study also shows that the effectiveness of DWT/EMD techniques depends on the data-driven algorithms used for load forecasting modeling and the training data. Hence, DWT/EMD-based noise cancellation needs customized selection and tuning to optimize their performance for data-driven building load forecasting modeling.

A New Approach for Reconstruction of IMFs of Decomposition and Ensemble Model for Forecasting Crude Oil Prices

Accurate forecasting for the crude oil price is important for government agencies, investors, and researchers. To cope with this issue, in this paper, a new paradigm is designed for the reconstruction of intrinsic mode functions (IMFs) of decomposition and ensemble models to reduce the complexity in computation and to enhance the forecasting accuracy. Decomposition and ensemble methodologies significantly enhance the forecasting accuracy under the framework of “divide and conquer” with the proposed reconstruction of IMFs method. The proposed approach used the autocorrelation at lag 1 of all IMFs for the reconstruction. The ensemble empirical mode decomposition (EEMD) technique is employed to decompose the data into different IMFs. Models that utilized the decomposed data relatively perform well, as compared to its application to the undecomposed data. However, sometimes, the decomposition may produce poor results due to the error accumulation at the end. Thus, in this study, the reconstruction of IMFs is proposed for minimizing the aforementioned error, thereby increasing the forecasting accuracy. The Brent and West Texas Intermediate (WTI) datasets (daily and weekly) are exploited to compare the forecasting performance of autoregressive integrated moving average (ARIMA) along with artificial neural network (ANN) models with the decomposed data. The results have proven that the new paradigm of reconstruction of IMFs through autocorrelation was a better and simple strategy that significantly improved the performance of single models including ARIMA and ANN. Hence, it is concluded that the proposed model takes less computational time and achieved higher forecasting accuracy with the reconstruction of IMFs as opposed to using all IMFs.

Classification of arc fault between broken conductor and high‐impedance surface: an empirical mode decomposition and Stockwell transform‐based approach

The high-impedance arc fault (HIAF) is a result of the interaction between the current-carrying conductor and the high-impedance surface. This study presents a detailed and comparative time-frequency based analysis for the classification of different arcs, generated by the interaction of a broken conductor and different arcing surfaces. The real-time arcing voltage signals are considered as the basis of the whole time-frequency based analysis from a medium voltage distribution line. In this proposed approach, empirical mode decomposition (EMD) is used to study the real-time arc voltage signals of various surfaces. The intrinsic mode functions obtained by the application of the EMD technique is used as the input to various machine learning techniques, which later successfully classifies different HIAFs from their harmonic footprints. The Stockwell transform (ST) is additionally applied to the same test cases, and the classification of the arc is performed using similar learning algorithms to get a quantitative view of both outcomes of EMD and ST towards the selection of appropriate signal processing technique and machine learning algorithm for the same.

EMD and Singular Value Difference Spectrum Based Bearing Fault Characteristics Extraction

Aiming at extracting rolling bearing defect characteristic effectively under strong background noise, a defect diagnosis technique based on empirical mode decomposition (EMD) and singular value difference spectrum is put forward. Firstly, the non-stationary original bearing vibration experimental data is decomposed into several intrinsic mode functions (IMFs) using EMD technique to eliminate the noise influence. Secondly, the effective intrinsic mode functions are selected by using correlation functions to reconstruct the bearing transient vibration signal. Thirdly, the singular value decomposition of the reconstructed signal is carried out to calculate the singular value difference spectrum (SVDS), and the principal components are selected according to the SVDS. Finally, the envelope spectrum of the reconstructed principal components data is calculated, and rolling bearing defect characteristic is picked up according to the envelope spectrum technique. The results of the experiment exhibit that the proposed technique can be effectively applied to rolling bearing defect characteristics extraction.