Time Latency-Centric Signal Processing: A Perspective of Smart Manufacturing †

Frequency and time domain analyses of sensor signals in drilling—I. Correlation with drill wear

Phonocardiogram (PCG) signal represents recording of sounds and murmurs resulting from heart auscultation. Analysis of these PCG signals is critical in diagnosis of different heart diseases. Over the years, a variety of methods have been proposed for automatic analysis of PCG signals in time, frequency, and time-frequency domains. This paper presents a comprehensive survey of different methods proposed for automatic analysis of PCG signals with the objective to evaluate the current state-of-the-art and to determine the potential domains of effective analysis. An important aspect of our contribution is that the review is carried out as a function of domains of analysis rather than simply discussing different methods. Our method further splits analysis into pre-processing, localization, and classification, and details are presented in terms of techniques and classifiers used during these phases. Finally, results are summarized for normal heart beat, noisy heart beat, and different pathologies using metrices like accuracy and detection rate. In addition to time and frequency domain, time-frequency based methods including wavelet, empirical mode decomposition (EMD) and time-frequency representation (TFR) are selected for detailed analysis. The review concludes that the time-frequency representations like EMD and wavelets represent areas of exploration in future along with perspective of using different time-frequency techniques together.

The CNC machine tool is the passive executor of machining code. It cannot predict the machining accuracy during machining. If the error is found to be out of tolerance after processing, it will not only scrap the parts, but also greatly affect the processing efficiency. This phenomenon is very prominent when machining sculptured surface parts with five-axis machine tools. Therefore, this paper proposes a Digital Twin (DT) modeling method of five-axis machine tools for predicting Continuous Trajectory Contour Error (CTCE) caused by tracking errors and geometric errors. The DT consists of three parts: the Setpoints Trajectory (ST) model, the Actual Trajectory (AT) model considering tracking errors and geometric errors and the CTCE model. For a specific machine tool, according to the basic geometric information of the machine tool (tool length, kinematic chain information, etc.) and 41 geometric errors, the DT can be established. Inputting the Setpoints Positions (SPs) and the Linear Encoder Detection Positions (LEDPs), the DT can be used to predict the Tool-Tip Position Trajectory (TTPT) contour error and the Tool Orientation Trajectory (TOT) contour error. In order to verify the proposed method experimentally, the KMC400S U five-axis machine tool is selected to establish its DT by which the contour error of the S-shaped trajectory are predicted offline. Then, the DMU50 five-axis machine tool is selected to establish its DT to predict the contour error of the circular trajectory in real time. Combined with the deep motion mechanism, this paper proposes a DT modeling method for the vertical application scene of parts machining accuracy prediction, which is of great significance to developing the DT application theory and ensuring the machining accuracy of parts.

Heart period variability has been considered for clinical assessment of autonomic function, determining the presence of haemorrhage or disease states, and for predicting mortality from traumatic injury. However, for heart period variability to be clinically useful, a number of important methodological issues should be addressed, including the minimum number of R-R intervals (RRI) required for accurate derivation, and the reproducibility of these metrics. ECGs were recorded for > or =10 min in 18 resting, supine subjects (12 M/6 F; 19-55 years). Heart period variability analyses included 21 time, frequency and complexity domain metrics. For assessment of minimum RRIs required, measurements were made from ECG recordings of 5 min down to 30 s for time and frequency domain metrics, and from 800 RRIs down to 100 RRIs for complexity metrics, by methodical truncation of the data set. Inter-subject variability was assessed by calculating the range and co-efficient of variation (%CV) across all subjects. Two independent 30 s or 100 RRI ECG segments were used to assess intra-subject variability via calculation of %CV in each subject. Six time and frequency domain metrics were robust down to 30 s of data, while five complexity metrics were robust down to 100 RRIs. All time and frequency domain metrics (except for RRI) exhibited high inter-subject variability (CVs > or = 30.0%), while five of eleven complexity metrics displayed low inter-subject variability (CVs < or = 8.5%). In the assessment of intra-subject variability in metrics valid with 30 s or 100 RRIs of ECG, only one time domain and four complexity metrics had CVs < 10%. Metrics that are highly reproducible and require few RRIs are advantageous for patient monitoring as less time is required to assess physiological status and initiate early interventions. Based on our analyses from healthy, resting humans, we have identified a select cohort of heart period variability metrics that performed well in regards to these two criteria.

We describe the main properties of the EQUATOR airborne geophysical system and how the system can be employed to solve some engineering and environmental problems. To date, the EQUATOR system has been used for groundwater search in 2014 and 2015 in Siberia. The name “EQUATOR” originates from the intention to EQUATE, or to balance time and frequency domain approaches in airborne electromagnetics, which is why the system was originally designed. To the best of our knowledge, EQUATOR is the only system to collect electromagnetic data both in time and in frequency domains. We further explain the difference between the two methods on some numerical and field examples. Also we describe some advantages of their coupling. In particular, by virtue of our approach, a numerically more robust solution of the one-dimensional electromagnetic inverse problem may be achieved. The recorded full-time measurements are further converted to in-phase, quadrature and off-time responses, as opposed to standard frequency or time domain solutions, which have been widely adapted within the industry. When inverting the data, our main focus is within the top 100 m of the Earth, which are the most relevant for engineering and environmental studies. We emphasize that a combined use of frequency and time domain data gives more detailed information about geological structures in the near-surface. As an example, we present some case studies obtained over a platform (Siberia) and a crystalline shield (Western Africa) setting. We further suggest using these results in such near-surface studies as sediment thickness estimation, soil state analysis, groundwater search, and geological structural analysis.

The relevance of research is due to the need to establish the true causes and patterns of changes in the hydrometric characteristics of the global water balance as a whole and its components, natural environments on Earth.&#x0D; Purpose of research: to establish patterns and causes of changes in the components of the global water balance on Earth: trends in the increase in water mass reserves, dynamics of the water balance of the South Seas of Europe, the Black Sea, dynamics of levels of the Indian Ocean, Western and Central Pacific Ocean, Eastern Pacific and Atlantic Oceans, level fluctuations of the World Ocean, changes in sea ice cover, river resources, reservoirs, variability of sea levels in the Arctic Ocean according to observations at coastal stations, changes in the balance of ice masses in glacial areas, changes in interseasonal thawing in the permafrost zones of the continent.&#x0D; Objects of research: time series of heliocosmic, climatic variables, hydrometric characteristics of the components of the global water balance.&#x0D; Methods of research: developed by the author, a method for interaction of observations of a variable or variables with groups of specified heliocosmic and climatic factors in the frequency and time domains, obtained using wavelet transformations of observations with the calculation of phase-frequency and phase-time characteristics in equal time intervals; comparative analysis of changes in the obtained phase characteristics of groups of variables with the calculation of their consistency matrices and the construction of graphs in the frequency and time domains.&#x0D; Main results of research: in the observed time intervals, in the trend of periodic changes in the reserves of water masses (glaciers, groundwater, lakes, reservoirs) in the frequency domain, significant negative unidirectional influences of changes in heliocosmic and multidirectional influences of climatic variables are observed; changes in the reserves of the World Ocean are more influenced by multidirectional changes in climatic factors compared to the unidirectional influences of heliocosmic factors; the frequency bands of changes in water mass reserves are less than the frequency bands of heliocosmic factors, due to the influence of climatic factors; The variability of water mass reserves on Earth is significantly and unidirectionally consistent with changes in evaporation, evaporation and precipitation on land, and in different directions with changes in cloudiness on the continents. Similar patterns are observed in changes in the water balance of the South Seas in Europe, the Black Sea, sea ice cover, river resources, reservoirs on land, ice mass balances in glacial areas, and interseasonal thawing in the permafrost zones of the continent. In the dynamics of the levels of the Indian, Western and Central Pacific, Eastern Pacific and Atlantic, fluctuations in the level of the World Ocean, significant multidirectional changes in levels in the frequency domain are observed on the effects of heliocosmic variables with the greatest influences of solar activity and long-term solar insolation, the Ap-index of geomagnetic activity, the parameter the ratio of plasma pressure to magnetic solar wind, changes in the ozone layer in the ionosphere, changes in the speed of rotation of the Earth, the influence of climate variables in the frequency domain with the greatest influences of solar radiation, precipitation, atmospheric processes, PTV is significant. The variability of sea levels in the Arctic Ocean, as observed at coastal stations, is characterized by significant multidirectional changes in the frequency domain due to the influence of heliocosmic and climatic factors, multidirectional mutual consistency of the component levels of the ocean seas, caused by anomalous changes in the Earths magnetic field in areas of the Northern Hemisphere. In changes in the phase-time characteristics of groups of factors, lattice structures of differently and unidirectional changes in variables are observed, characterizing the cyclical nature of climate change on Earth.

In physiological signal analysis, identifying meaningful relationships and inherent patterns in signals can provide valuable information regarding subjects' physiological state and changes. Although MATLAB has been widely used in signal processing and feature analysis, Python has recently dethroned MATLAB with the rise of data science, machine learning and artificial intelligence. Hence, there is a compelling need for a Python package for physiological feature analysis and extraction to achieve compatibility with downstream models often trained in Python. Thus, we present a novel visualization and feature analysis Python toolbox, PySio, to enable rapid, efficient and user-friendly analysis of physiological signals. First, the user should import the signal-of-interest with the corresponding sampling rate. After importing, the user can either analyze the signal as it is, or can choose a specific region for more detailed analysis. PySio enables the user to (i) visualize and analyze the physiological signals (or user-selected segments of the signals) in time domain, (ii) study the signals (or user-selected segments of the signals) in frequency domain through discrete Fourier transform and spectrogram representations, and (iii) investigate and extract the most common time (energy, entropy, zero crossing rate and peaks) and frequency (spectral entropy, rolloff, centroid, spread, peaks and bandpower) domain features, all with one click. Clinical relevance- As the physiological signals originate directly from the underlying physiological events, proper analysis of the signal patterns can provide valuable information in personalized treatment and wearable technology applications.

Third order time and frequency analysis has exhibited great potential for correlation analysis of multi-sensor datasets, but is usually presented as separate time domain and frequency domain approaches. A combined framework of both frequency domain and time domain has rarely been used. This paper proposes a non-parametric third order time and frequency domain framework which used two dimensional Fourier transforms to bridge the gap between time domain and frequency domain. A unified framework offers flexibility and efficiency to apply to data. In this paper we study neural spike train data treated as stochastic point processes. In time domain direct analysis, third order cumulant densities of spike trains are applied, which need all first-, secondand third order product densities to be calculated before constructing the third order cumulant density, which brings additional challenges. The novelty in this study is that a new framework is proposed which can offer an alternative approach without calculating lower order quantities and can reveal nonlinear relationship between neural recordings. The results show that the present framework provides a novel non-parametric method to estimate both time and frequency domain measurements which is applicable to neural spike trains.

The Heisenberg-Gabor uncertainty principle defines the limits of information resolution in both time and frequency domains. The limit of resolution discloses unique properties of a time series by frequency decomposition. However, classical methods such as Fourier analysis are limited by spectral leakage, particularly in longitudinal data with shifting periodicity or unequal intervals. Wavelet transformation provides a workable compromise by decomposing the signal in both time and frequency through translation and scaling of a basis function followed by correlation or convolution with the original signal. This study aimed to compare the accuracy of predictive models in mental arithmetic in time and frequency domains. Analysis of the author's response time at mental arithmetic using a soroban was modeled for two periods, an initial period (TI = 68 days), and a return period (TR = 170 days) both separated by an interval of 370 days. The median (min,max) response times in seconds (s) was longer for all tasks during the TI compared to the TR period (p < 0.001), for addition [CTAdd 62 (45, 127) vs 50 (38, 75) s] and summation [CTSum 68 (47, 108) vs 57(43, 109) s]. Response times were longer for errors regardless of the study period or task. There was an increasing phase difference for the addition and summation tasks during the TI period toward the end of the series 49.65o compared to the TR period where the phase difference between the two tasks was only 2.05o, indicating that both tasks are likely demonstrating similar learning rates during the latter study period. A comparison between time and time/frequency domain forecasts for an additional 100 tasks demonstrated higher accuracy of the maximum overlap discrete wavelet transform (MODWT) model, where the mean absolute percentage error ranged between 5.48 and 8.19% and that for the time domain models [autoregressive integrated moving average (ARIMA), generalized autoregressive conditional heteroscedasticity (GARCH)] was 6.16–10.80%.

CNC machine tool (CNCMT) is the mother machine of industry, which plays an important role in the coming smart manufacturing. The intelligence of CNCMT has a big significance, which will enables its self-sensing, self-prediction and self-maintenance without user concerns. In order to realize the intelligence of CNCMT, a Digital Twin (DT) modeling method for CNCMT is researched, including a multi-domain unified modeling method, a mapping method and an autonomous strategy. This paper provides a demonstration of DT modeling method for CNCMT.

Influence of water activity on acoustic emission of flat extruded bread

Wavelet transform is an analysis method that combines time, frequency (or scale) domains. It has: (1) Multi-resolution; (2) The relative bandwidth is constant; (3) Proper selection of basic wavelet can make wavelet have the ability to represent the local characteristics of signals in both time and frequency domains, and it is known as "the microscope for analyzing signals". An analysis method of EEG signal based on autoregressive model (ARM) and wavelet transform is proposed, and it is used to eliminate noise interference in EEG signal. Wavelet transform is a multi-resolution time scale analysis method, which can divide the signal into sub-band signals of different frequency bands. According to this characteristic of wavelet transform, the EEG signals obtained by sampling are decomposed and denoised at various scales, and the results of decomposition and denoising at various scales are given. Wavelet transform can effectively remove noise interference from EEG signals. Wavelet transform is a multi resolution time scale analysis method that can divide signals into subband signals of different frequency bands. According to this characteristic of wavelet transform, the sampled EEG signal is decomposed and denoised at various scales, and the decomposition and denoising results at each scale are given. Wavelet transform can effectively remove noise interference from EEG signals.

The high penetration of power electronic converters with complex control systems is changing the power system dynamics, introducing new challenges such as converter-driven stability incidents. Traditional stability analysis methods, suitable for classical problems like voltage, frequency, and rotor angle stability in large systems, are insufficient for addressing the fast control dynamics of converters, which involve electromagnetic phenomena. These phenomena require detailed converter and network modeling, which can be performed in both the frequency and time domains, enabling the respective stability analyses to be carried out. However, frequency domain methods, based on small-signal impedances linearized at a single operating point, inherently ignore time domain phenomena like switching events and nonlinear behaviors. In contrast, time domain electromagnetic transient (EMT) simulations are effective for analyzing converter-driven stability but are computationally intensive when applied to large transmission systems with numerous use cases. Therefore, to reduce the simulation complexity in EMT tools, a complexity reduction procedure is proposed in this paper. Leveraging the advantages of the frequency domain, such as faster simulation times and information on wideband frequency characteristics of the system, this procedure utilizes the small-signal impedances and introduces a method for network reduction. The procedure also uses the frequency domain stability analysis method to screen for critical network use cases. Primarily, this procedure is a frequency domain toolchain encompassing frequency domain stability analysis and frequency domain network reduction. The result of the toolchain is a reduced network size and reduced network use cases that can be used for EMT simulations. The procedure is applied to an IEEE 39 bus system, where converter-driven stability is evaluated for two use cases. Furthermore, the network reduction method is tested on a critical use case, demonstrating reductions in network size and computation times without compromising the quality of stability analysis results.

Automatic speech recognition (ASR) systems have been widely deployed in modern smart devices to provide convenient and diverse voice-controlled services. Since ASR systems are vulnerable to audio replay attacks that can spoof and mislead ASR systems, a number of defense systems have been proposed to identify replayed audio signals based on the speakers' unique acoustic features in the frequency domain. In this paper, we uncover a new type of replay attack called modulated replay attack, which can bypass the existing frequency domain based defense systems. The basic idea is to compensate for the frequency distortion of a given electronic speaker using an inverse filter that is customized to the speaker's transform characteristics. Our experiments on real smart devices confirm the modulated replay attacks can successfully escape the existing detection mechanisms that rely on identifying suspicious features in the frequency domain. To defeat modulated replay attacks, we design and implement a countermeasure named DualGuard. We discover and formally prove that no matter how the replay audio signals could be modulated, the replay attacks will either leave ringing artifacts in the time domain or cause spectrum distortion in the frequency domain. Therefore, by jointly checking suspicious features in both frequency and time domains, DualGuard can successfully detect various replay attacks including the modulated replay attacks. We implement a prototype of DualGuard on a popular voice interactive platform, ReSpeaker Core v2. The experimental results show DualGuard can achieve 98% accuracy on detecting modulated replay attacks.

Recently, along with several technological advancements in cyber-physical systems, the revolution of Industry 4.0 has brought in an emerging concept named digital twin (DT), which shows its potential to break the barrier between the physical and cyber space in smart manufacturing. However, it is still difficult to analyze and estimate the real-time structural and environmental parameters in terms of their dynamic changes in digital twinning, especially when facing detection tasks of multiple small objects from a large-scale scene with complex contexts in modern manufacturing environments. In this article, we focus on a small object detection model for DT, aiming to realize the dynamic synchronization between a physical manufacturing system and its virtual representation. Three significant elements, including equipment, product, and operator, are considered as the basic environmental parameters to represent and estimate the dynamic characteristics and real-time changes in building a generic DT system of smart manufacturing workshop. A hybrid deep neural network model, based on the integration of MobileNetv2, YOLOv4, and Openpose, is constructed to identify the real-time status from physical manufacturing environment to virtual space. A learning algorithm is then developed to realize the efficient multitype small object detection based on the feature integration and fusion from both shallow and deep layers, in order to facilitate the modeling, monitoring, and optimizing of the whole manufacturing process in the DT system. Experiments and evaluations conducted in three different use cases demonstrate the effectiveness and usefulness of our proposed method, which can achieve a higher detection accuracy for DT in smart manufacturing.