Energy In Different Frequency Bands Research Articles

Implementation of Machine Learning and Deep Learning Techniques for the Detection of Epileptic Seizures Using Intracranial Electroencephalography

The diagnosis of epilepsy primarily relies on the visual and subjective assessment of the patient’s electroencephalographic (EEG) or intracranial electroencephalographic (iEEG) signals. Neurophysiologists, based on their experience, look for characteristic discharges such as spikes and multi-spikes. One of the main challenges in epilepsy research is developing an automated system capable of detecting epileptic seizures with high sensitivity and precision. Moreover, there is an ongoing search for universal features in iEEG signals that can be easily interpreted by neurophysiologists. This article explores the possibilities, issues, and challenges associated with utilizing artificial intelligence for seizure detection using the publicly available iEEG database. The study presents standard approaches for analyzing iEEG signals, including chaos theory, energy in different frequency bands (alpha, beta, gamma, theta, and delta), wavelet transform, empirical mode decomposition, and machine learning techniques such as support vector machines. It also discusses modern deep learning algorithms such as convolutional neural networks (CNN) and long short-term memory (LSTM) networks. Our goal was to gather and comprehensively compare various artificial intelligence techniques, including both traditional machine learning methods and deep learning techniques, which are most commonly used in the field of seizure detection. Detection results were tested on a separate dataset, demonstrating classification accuracy, sensitivity, precision, and specificity of seizure detection. The best results for seizure detection were obtained with features related to iEEG signal energy (accuracy of 0.97, precision of 0.96, sensitivity of 0.99, and specificity of 0.96), as well as features related to chaos, Lyapunov exponents, and fractal dimension (accuracy, precision, sensitivity, and specificity all equal to 0.95). The application of CNN and LSTM networks yielded significantly better results (CNN: Accuracy of 0.99, precision of 0.98, sensitivity of 1, and specificity of 0.99; LSTM: Accuracy of 0.98, precision of 0.96, sensitivity of 1, and specificity of 0.99). Additionally, the use of the gradient-weighted class activation mapping algorithm identified iEEG signal fragments that played a significant role in seizure detection.

Representation and decoding of bilateral arm motor imagery using unilateral cerebral LFP signals.

In the field of upper limb brain computer interfaces (BCIs), the research focusing on bilateral decoding mostly based on the neural signals from two cerebral hemispheres. In addition, most studies used spikes for decoding. Here we examined the representation and decoding of different laterality and regions arm motor imagery in unilateral motor cortex based on local field potentials (LFPs). The LFP signals were recorded from a 96-channel Utah microelectrode array implanted in the left primary motor cortex of a paralyzed participant. There were 7 kinds of tasks: rest, left, right and bilateral elbow and wrist flexion. We performed time-frequency analysis on the LFP signals and analyzed the representation and decoding of different tasks using the power and energy of different frequency bands. The frequency range of <8 Hz and >38 Hz showed power enhancement, whereas 8-38 Hz showed power suppression in spectrograms while performing motor imagery. There were significant differences in average energy between tasks. What's more, the movement region and laterality were represented in two dimensions by demixed principal component analysis. The 135-300 Hz band signal had the highest decoding accuracy among all frequency bands and the contralateral and bilateral signals had more similar single-channel power activation patterns and larger signal correlation than contralateral and ipsilateral signals, bilateral and ipsilateral signals. The results showed that unilateral LFP signals had different representations for bilateral motor imagery on the average energy of the full array and single-channel power levels, and different tasks could be decoded. These proved the feasibility of multilateral BCI based on the unilateral LFP signal to broaden the application of BCI technology. https://www.chictr.org.cn/showproj.aspx?proj=130829, identifier ChiCTR2100050705.

Quantifying Urban Activities Using Nodal Seismometers in a Heterogeneous Urban Space.

Earth's surface is constantly vibrating due to natural processes inside and human activities on the surface of the Earth. These vibrations form the ambient seismic fields that are measured by sensitive seismometers. Compared with natural processes, anthropogenic vibrations dominate the seismic measurements at higher frequency bands, demonstrate clear temporal and cyclic variability, and are more heterogeneous in space. Consequently, urban ambient seismic fields are a rich information source for human activity monitoring. Improving from the conventional energy-based seismic spectral analysis, we utilize advanced signal processing techniques to extract the occurrence of specific urban activities, including motor vehicle counts and runner activities, from the high-frequency ambient seismic noise. We compare the seismic energy in different frequency bands with the extracted activity intensity at different locations within a one-kilometer radius and highlight the high-resolution information in the seismic data. Our results demonstrate the intense heterogeneity in a highly developed urban space. Different sectors of urban society serve different functions and respond differently when urban life is severely disturbed by the impact of the COVID-19 pandemic in 2020. The anonymity of seismic data enabled an unprecedented spatial and temporal resolution, which potentially could be utilized by government regulators and policymakers for dynamic monitoring and urban management.

Data-Driven Seismic Impedance Inversion Based on Multi-Scale Strategy

Seismic impedance inversion is one of the most commonly used techniques for reservoir characterization. High accuracy and high resolution seismic impedance is a prerequisite for subsequent reservoir interpretation. The data-driven approach offers the opportunity for accurate impedance prediction by establishing a nonlinear mapping between seismic data and impedance. However, existing data-driven methods take the raw seismic data directly as input, making it difficult for the network to learn high frequency weak signal information and resulting in low resolution inversion results. In order to mitigate the above issues, a data-driven seismic impedance inversion method based on multi-scale strategy is proposed. The method first obtains seismic data at different scales using frequency division techniques and do normalization on the extracted multi-scale data to ensure the consistency of the seismic signal energy in different frequency bands. The multi-scale seismic data will then be fed into the network, which helps the network to learn the high frequency information features more easily, and ultimately obtain higher resolution inversion results. We use the most commonly used convolutional neural network (CNN) as an example to demonstrate that the proposed multi-scale data-driven seismic impedance inversion method can improve the resolution of the inversion results. In addition, since the above seismic impedance inversion method is executed trace-by-trace, the f-x prediction filtering technique is introduced to improve the lateral continuity of the inversion results and obtain more geologically reliable impedance profiles. The validity of the proposed method is verified in the application of synthetic model data as well as an actual data set.

Early chatter detection in robotic milling under variable robot postures and cutting parameters

Chatter in robotic milling is closely related to robot postures and cutting parameters. The chatter frequency varies over a wide range compared with that in the milling process using traditional machine tools. These features pose a significant challenge to chatter detection in robotic milling. The objective of this paper is to propose a novel approach for early and reliable chatter detection in robotic milling under variable robot postures and cutting parameters. First, the inertance frequency response function under each robot posture is used to filter the corresponding vibration signals. This minimizes the influence of robot postures on the measured vibration signals. Second, based on the filtered vibration data, a dual-tree complex wavelet packet transform is applied to extract the energy of different frequency bands, from which the fractional energy entropy is obtained to characterize the chatter state. Therefore, the chatter state can be characterized as independent of the variable cutting parameters and chatter frequencies. Third, online chatter detection is achieved using the sequential probability ratio test, with which significance tests for the chatter states are conducted. Consequently, the detection threshold does not need to be manually set. Extensive case studies demonstrate that the proposed approach can realize early and reliable chatter detection in robotic milling under variable robot postures and cutting parameters.

Research on a Rolling Bearing Fault Detection Method With Wavelet Convolution Deep Transfer Learning

Many deep learning models for fault diagnosis have not considered the prior diagnosis knowledge of the rolling bearing. Moreover, some measuring locations cannot collect adequate data to diagnose due to equipment size or installation space problems. This paper proposes a wavelet convolutional deep transfer learning model for rolling bearing fault detection on cross-measurement points. A new convolution layer includes a redesigned convolution kernel and a new energy pooling layer. The convolution kernel is designed based on wavelet construction for mining time-frequency characteristics. The energy pooling layer is proposed to extract the energy of different frequency bands. The associated location's fault information has been transferred to domain knowledge to enhance the target features. Different domain features have adaptive matches based on multiple kernel variants of maximum mean discrepancy. The experimental results demonstrate that the accuracy of fault detection can reach 99.73%, and the robustness of the proposed method is also verified.

Application of a Wavelet Packet and SOM Neural Network in Wastewater Treatment Fault

In order to solve the problem of fault diagnosis in sewage treatment, a fault diagnosis method combining wavelet packet and improved self-organization mapping (SOM) neural network was proposed. In this method, the fault of sewage treatment is decomposed into three types by wavelet packet, and then decomposed into several frequency bands to calculate the energy of different frequency bands. By using the ratio of these energy values to the energy values of normal operating frequency bands, the feature vector of fault diagnosis of sewage treatment fault is constructed to extract fault features. SOM neural network of 3 x 3 was designed, and the neural network training was carried out using the energy fault feature vector [1], so as to determine the network parameters and achieve the purpose of fault diagnosis. The simulation results show that the fault diagnosis method is effective and accurate.

A Novel Method for Chaos Detection in Heavy Noisy Environments Based on Distribution of Energy

Detecting chaos in heavy-noise environments is an important issue in many fields of science and engineering. In this paper, first, a new criterion is proposed to recognize chaos from noise based on the distribution of energy. Then, a new method based on stationary wavelet transform (SWT) is presented for chaos detection that is recommended for data that contain more than 60% noise. This method is dependent on the distribution of signal’s energy in different frequency bands based on SWT for chaos detection which is robust to noisy environments. In this method, the effect of white noise and colored noise on the chaotic system is considered. As a case study, the proposed method is applied to detect chaos in two different oscillators based on memristor and memcapacitor. The simulation results are used to display the main points of the paper.

Intelligent Fault Diagnosis of Bearings Based on Energy Levels in Frequency Bands Using Wavelet and Support Vector Machines (SVM)

In this paper, a new method was introduced for feature extraction and fault diagnosis in bearings based on wavelet packet decomposition and analysis of the energy in different frequency bands. This method decomposes a signal into different frequency bands using different types of wavelets and performs multi-resolution analysis to extract different features of the signals by choosing energy levels in different frequency bands. The support vector machines (SVM) technique was used for faults classifications. Daubechies, biorthogonal, coiflet, symlet, Meyer, and reverse Meyer wavelets were used for feature extraction. The most appropriate decomposition level and frequency band were selected by analyzing the variation in the signal’s energy level. The proposed approach was applied to the fault diagnosis of rolling bearings, and testing results showed that the proposed approach can reliably identify different fault categories and their severities. Moreover, the effectiveness of the proposed feature selection and fault diagnosis method was significant based on the similarity between the wavelet packet and the signal, and effectively reduced the influence of the signal noise on the classification results.

Defects diagnosis in laser brazing using near-infrared signals based on empirical mode decomposition

Real-time monitoring of laser welding plays a very important role in the modern automated production and online defects diagnosis is necessary to be implemented. In this study, the status of laser brazing was monitored in real time using an infrared photoelectric sensor. Four kinds of braze seams (including healthy weld, unfilled weld, hole weld and rough surface weld) along with corresponding near-infrared signals were obtained. Further, a new method called Empirical Mode Decomposition (EMD) was proposed to analyze the near-infrared signals. The results showed that the EMD method had a good performance in eliminating the noise on the near-infrared signals. And then, the correlation coefficient was developed for selecting the Intrinsic Mode Function (IMF) more sensitive to the weld defects. A more accurate signal was reconstructed with the selected IMF components. Simultaneously, the spectrum of selected IMF components was solved using fast Fourier transform, and the frequency characteristics were clearly revealed. The frequency energy of different frequency bands was computed to diagnose the defects. There was a significant difference in four types of weld defects. This approach has been proved to be an effective and efficient method for monitoring laser brazing defects.

Application of energies of optimal frequency bands for fault diagnosis based on modified distance function

Low-dimensional relevant feature sets are ideal to avoid extra data mining for classification. The current work investigates the feasibility of utilizing energies of vibration signals in optimal frequency bands as features for machine fault diagnosis application. Energies in different frequency bands were derived based on Parseval's theorem. The optimal feature sets were extracted by optimization of the related frequency bands using genetic algorithm and a Modified distance function (MDF). The frequency bands and the number of bands were optimized based on the MDF. The MDF is designed to a) maximize the distance between centers of classes, b) minimize the dispersion of features in each class separately, and c) minimize dimension of extracted feature sets. The experimental signals in two different gearboxes were used to demonstrate the efficiency of the presented technique. The results show the effectiveness of the presented technique in gear fault diagnosis application.

Application of variational mode decomposition energy distribution to bearing fault diagnosis in a wind turbine

For the unsteady characteristics of a fault vibration signal of a wind turbine’s rolling bearing, a bearing fault diagnosis method based on variational mode decomposition of the energy distribution is proposed. Firstly, variational mode decomposition is used to decompose the original vibration signal into a finite number of stationary components. Then, some components which comprise the major fault information are selected for further analysis. When a rolling bearing fault occurs, the energy in different frequency bands of the vibration acceleration signals will change. Energy characteristic parameters can then be extracted from each component as the input parameters of the classifier, based on the K nearest neighbour algorithm. This can identify the type of fault in the rolling bearing. The vibration signals from a spherical roller bearing in its normal state, with an outer race fault, with an inner race fault and with a roller fault were analyzed. The results showed that the proposed method (variational mode decomposition is used as a pre-processor to extract the energy of each frequency band as the characteristic parameter) can identify the working state and fault type of rolling bearings in a wind turbine.

Bearing Fault Diagnosis Based on Labview

This function of wavelet packet decomposition and the energy of each band to strike is achieved within the Labview module. Signal energy in different frequency bands within the change reflects a change in the operating state. Extract wavelet packet energy spectrum of each band, making it as a feature vector. Finally the fault are classified by SVM. The two Parameters, the kernel function parameters g of radial machine support vector machine and penalty factor C of the radial machine support vector machine, are optimally chosen, automatically and rapidly, by using the method of particle swarm algorithm, avoiding the blindness of artificial selection parameters. The Matalab program of support vector machines based on particle swarm optimization are made into COM components. Mixed programming, Labview call COM component, generated by the M file, is implemented, which is divorced from the MATLAB environment, making it good for expanding the function of Labview. The effectiveness, wavelet packet energy spectrum - PSOSVM model of the bearing fault diagnosis, is verified.

A Novel Technique of Black Tea Quality Prediction Using Electronic Tongue Signals

Electronic tongue (ET) system is under extensive development for automatic analysis and prediction of quality of different industrial end products. Each sensor in an ET system generates a specific electronic response in presence of different organic or inorganic compounds in the sample. The vital part of the ET system is the discrimination of the complex pattern generated by the sensor array. In this paper, a novel technique of black tea quality estimation is using the ET signals. A moving window is used to extract discrete wavelet transform coefficients from the transient response of ET. The energy in different frequency bands are used as the features of the ET signal for different positions of the window. The prediction of a new sample is performed by the highest score obtained by a particular class by testing all the patterns generated by windowing ET signal. The performance of the proposed technique is verified to estimate black tea quality using two kernel classifiers, namely support vector machine and recently proposed vector valued regularized kernel function approximation method. High prediction accuracy of both the classifiers confirms the effectiveness of the proposed technique of tea quality estimation using ET signals.

Rupture History of the 2011 M 9 Tohoku Japan Earthquake Determined from Strong-Motion and High-Rate GPS Recordings: Subevents Radiating Energy in Different Frequency Bands

Abstract Strong‐motion records from KiK‐net and K‐NET, along with 1 sample/s Global Positioning System (GPS) records from GEONET, were analyzed to determine the location, timing, and slip of subevents of the M 9 2011 Tohoku earthquake. Timing of arrivals on stations along the coast shows that the first subevent was located closer to the coast than subevent (2), which produced the largest slip. A waveform inversion of data from 0 to 0.2 Hz indicates that the first subevent primarily ruptured down‐dip and north of the hypocenter and had an M of 8.5. The areas of this subevent that generated the low ( 0.2 Hz) frequency energy are located in the same vicinity. The inversion result for the second subevent ( M 9.0) has large slip on the shallow part of the fault with peak slip of about 65 m above about 25 km depth. This slip generated the tsunami. The preferred inversion has initiation of subevent 2 on the shallow portion of the fault so that rupture proceeded down‐dip and mainly to the south. Subevent 2 started about 35 s after subevent 1, which allows for the possibility of dynamic triggering from subevent 1. The slip model predicts displacements comparable to those found from ocean‐bottom transducers near the epicenter. At frequencies that most affect tall buildings (0.1–0.5 Hz), there is a strong pulse (subevent 3) in the strong‐motion records that arrives after the near‐field ramp from subevent 2. High‐frequency subevent 3 was located down‐dip and south of the high‐slip portion of subevent 2 and was initiated as rupture from subevent 2 proceeded down‐dip. The compact pulse for subevent 3 is modeled with an M 8.0 source in a 75 by 30 km area that ruptured down‐dip and to the south with a high slip velocity, indicating high stress drop.

Fault Diagnosis using Audio and Vibration Signals in a Circulating Pump

This paper presents the use of audio and vibration signals in fault diagnosis of a circulating pump. The novelty of this paper is the use of audio signals acquired by microphones. The objective of this paper is to determine if audio signals are capable to distinguish between normal and different abnormal conditions in a circulating pump. In order to compare results, vibration signals are also acquired and analysed. Wavelet package is used to obtain the energies in different frequency bands from the audio and vibration signals. Neural networks are used to evaluate the discrimination ability of the extracted features between normal and fault conditions. The results show that information from sound signals can distinguish between normal and different faulty conditions with a success rate of 83.33%, 98% and 91.33% for each microphone respectively. These success rates are similar and even higher that those obtained from accelerometers (68%, 90.67% and 71.33% for each accelerometer respectively). Success rates also show that the position of microphones and accelerometers affects on the final results.

Performance Degradation Assessment of a Beam Structure by Using Wavelet Packet Energy

The evaluation on damage event of a beam structure is carried out in this study through an integration of knowledge in damage detection and signal process. A damage detection approach based on wavelet packet transform (WPT) is developed by decomposing and reconstructing the structural dynamic responses to obtain the signal energy distribution at different levels. The WPT coefficients are reconstructed to extract the signals in various frequency ranges and determine the signal energy in different frequency bands. A WPT component energy index is proposed to establish the slope vector as well as curvature vector and then the judge on the slope vector and curvature vector can indicate the damage event. Parametric study is carried out to investigate the relationship between the index amplitude and the damage extent. The results from this study demonstrate that the detection approach proposed can accurately identify the damage event in the structure.

A roller bearing fault diagnosis method based on EMD energy entropy and ANN

According to the non-stationary characteristics of roller bearing fault vibration signals, a roller bearing fault diagnosis method based on empirical mode decomposition (EMD) energy entropy is put forward in this paper. Firstly, original acceleration vibration signals are decomposed into a finite number of stationary intrinsic mode functions (IMFs), then the concept of EMD energy entropy is proposed. The analysis results from EMD energy entropy of different vibration signals show that the energy of vibration signal will change in different frequency bands when bearing fault occurs. Therefore, to identify roller bearing fault patterns, energy feature extracted from a number of IMFs that contained the most dominant fault information could serve as input vectors of artificial neural network. The analysis results from roller bearing signals with inner-race and out-race faults show that the diagnosis approach based on neural network by using EMD to extract the energy of different frequency bands as features can identify roller bearing fault patterns accurately and effectively and is superior to that based on wavelet packet decomposition and reconstruction.

Responses of auditory-cortex neurons to structural features of natural sounds.

Sound-processing strategies that use the highly non-random structure of natural sounds may confer evolutionary advantage to many species. Auditory processing of natural sounds has been studied almost exclusively in the context of species-specific vocalizations, although these form only a small part of the acoustic biotope. To study the relationships between properties of natural soundscapes and neuronal processing mechanisms in the auditory system, we analysed sound from a range of different environments. Here we show that for many non-animal sounds and background mixtures of animal sounds, energy in different frequency bands is coherently modulated. Co-modulation of different frequency bands in background noise facilitates the detection of tones in noise by humans, a phenomenon known as co-modulation masking release (CMR). We show that co-modulation also improves the ability of auditory-cortex neurons to detect tones in noise, and we propose that this property of auditory neurons may underlie behavioural CMR. This correspondence may represent an adaptation of the auditory system for the use of an attribute of natural sounds to facilitate real-world processing tasks.