Time-frequency Scales Research Articles

The forecasting of surface displacement for tunnel slopes utilizing the WD-IPSO-GRU model

To quickly assess slope stability based on field displacement monitoring data, this paper constructs a hybrid optimization model that predicts surface displacement during tunnel excavation in base-overburden slopes. The model combines Wavelet Decomposition (WD) with a Gated Recurrent Unit (GRU), and the GRU's hyperparameters are optimized using an Improved Particle Swarm Optimization algorithm (IPSO). The specific steps are as follows: First, the Wavelet Decomposition (WD) technique is applied to decompose the raw displacement data, extracting features at different time–frequency scales. Next, the Dropout technique is incorporated into the GRU model to prevent overfitting. Additionally, nonlinear inertia weight ω improved cognitive factor c1, and social factor c2 are introduced. The PSO algorithm is improved by integrating crossover and mutation concepts from genetic algorithms. Finally, the IPSO is used to optimize the number of neural units hN, HN, LN and dropout rates D1 and D2 in the GRU network architecture. After constructing the WD-IPSO-GRU model, a comprehensive comparison is made with various swarm intelligence algorithms and state-of-the-art models. The experimental results demonstrate that the WD-IPSO-GRU model significantly improves the prediction accuracy of surface displacement in slopes during tunnel excavation. Compared to directly using raw data for prediction, the introduction of the WD preprocessing technique improved the prediction accuracy at measurement points 01 and 02 by 28% and 45.9%, respectively. Additionally, with the model optimized by IPSO, the prediction accuracy at measurement points 01 and 02 increased by 76% and 56.7%, respectively. The WD-IPSO-GRU model effectively addresses the challenges of extracting features from univariate displacement time-series data and determining the parameters of the GRU network. It improves the prediction accuracy of surface displacement in base-overburden type slopes and demonstrates excellent generalization ability and reliability. The research results validate the potential application of the model in geotechnical engineering and provide strong support for assessing slope stability during tunnel excavation.

Propagation Dynamics from Meteorological to Agricultural Drought in Northwestern China: Key Influencing Factors

Meteorological and agricultural droughts are inherently correlated, whereas the propagation mechanism between them remains unclear in Northwestern China. Investigating the linkages between these drought types and identifying the potential influencing factors is crucial for effective water resource management and drought mitigation. This study adopted the Standardized Precipitation Evapotranspiration Index (SPEI) and Standardized Soil Moisture Index (SSMI) to characterize the meteorological and agricultural droughts from 1960 to 2018. The propagation time between these droughts was detected using the Pearson correlation analysis, and the cross-wavelet transform and wavelet cross-correlation were utilized to describe their linkages across the time–frequency scales. The grey relational analysis was applied to explore the potential factors influencing the propagation time. The results revealed that the agricultural drought typically lagged behind the meteorological drought by an average of 6 months in Northwestern China, with distinct seasonal and regional characteristics. The shortest propagation time occurred in the summer (3 months), followed by the autumn (4 months), and the propagation time was longer in the winter (8 months) and spring (9 months). Additionally, the average propagation time was longer in the plateau climate zone (8 months) than in the southeastern climate zone (6 months) and the westerly climate zone (4 months). There was a multi-timescale response between the meteorological and agricultural droughts, with a relatively stable and significant positive correlation over long timescales, whereas the correlation was less clear over short timescales. The key factors influencing the propagation time were soil moisture, elevation, precipitation, and potential evapotranspiration. Furthermore, the wavelet cross-correlation between agricultural and meteorological droughts was relatively high, with a lag of 0 to 3 months; as the timescale increased, the fluctuation period of their cross-correlation also increased.

Pipeline Leak Detection: A Comprehensive Deep Learning Model Using CWT Image Analysis and an Optimized DBN-GA-LSSVM Framework.

Detecting pipeline leaks is an essential factor in maintaining the integrity of fluid transport systems. This paper introduces an advanced deep learning framework that uses continuous wavelet transform (CWT) images for precise detection of such leaks. Transforming acoustic signals from pipelines under various conditions into CWT scalograms, followed by signal processing by non-local means and adaptive histogram equalization, results in new enhanced leak-induced scalograms (ELIS) that capture detailed energy fluctuations across time-frequency scales. The fundamental approach takes advantage of a deep belief network (DBN) fine-tuned with a genetic algorithm (GA) and unified with a least squares support vector machine (LSSVM) to improve feature extraction and classification accuracy. The DBN-GA framework precisely extracts informative features, while the LSSVM classifier precisely distinguishes between leaky and non-leak conditions. By concentrating solely on the advanced capabilities of ELIS processed through an optimized DBN-GA-LSSVM model, this research achieves high detection accuracy and reliability, making a significant contribution to pipeline monitoring and maintenance. This innovative approach to capturing complex signal patterns can be applied to real-time leak detection and critical infrastructure safety in several industrial applications.

A time-varying analysis between economic uncertainty and tourism development in Singapore.

Tourism development (TO) is seen as a viable solution to address economic policy uncertainty (EPU) risks. However, previous studies have largely ignored the relationship between short, medium, and long term by decomposing TO and EPU index at different time-frequency scales, especially in Singapore. In this study, the Wavelet tools analysis and a rolling window algorithm are employed to re-visit the causal relationship between EPU, industrial production index (IPI), government revenue (GR), and tourism development (TO) in Singapore from January 2003 to February 2022. The findings revealed the heterogeneous effects of EPU on TO at different time horizons in terms of importance and magnitude over time. A rise in EPU results in a decline in TO at the low frequencies, indicating that EPU has a detrimental effect on TO over the short term. Conversely, in the long term, an increase in TO results in a decrease in EPU. Furthermore, the outcome also indicated that there is a uni-directional causality running from TO to EPU, GR and IPI. Expressly, we confirm that the negative co-movement is more pronounced in the aftermath of the COVID-19 crisis, particularly for EPU, and GR at low-medium frequencies throughout the research period. The findings provide tourism policymakers with insight to develop strategic plans for tourism development that consider the effects of economic policy uncertainty. By understanding how uncertainty impacts tourism, governments can tailor development strategies to mitigate risks and capitalize on opportunities.

Source tracing and contagion measurement of carbon emission trading price fluctuation in China from the perspective of major emergencies.

Based on monthly economic data spanning from January 2015 to December 2022, we have established an analytical framework to examine the "Russia-Ukraine conflict-financial market pressure and energy market-China carbon emission trading prices." To achieve this objective, we developed indices for financial system pressure, the energy market, and investor sentiment, applying a mediation effects model to validate their transmission mechanisms. Subsequently, the TVP-SV-VAR model was employed to scrutinize the nonlinear impact of the Russia-Ukraine conflict on the valuation of China's carbon emission trading rights. This model integrates time-varying parameters (TVP) and stochastic volatility (SV), utilizing Markov Chain Monte Carlo (MCMC) technology for parameter estimation. Finally, various wavelet analysis techniques, including continuous wavelet transform, cross-wavelet transform, and wavelet coherence spectrum, were applied to decompose time series data into distinct time-frequency scales, facilitating an analysis of the lead-lag relationships within each time series. The research outcomes provide crucial insights for safeguarding the interests of trading organizations, refining the structure of the carbon market, and mitigating systemic risks on a global scale.

Daily fluctuations in COVID-19 infection rates under Tokyo's epidemic prevention measures - new evidence from adaptive Fourier decomposition.

The COVID-19 pandemic has witnessed widespread infections and variants. Particularly, Tokyo faced the challenge of seven waves of COVID-19, during which government interventions played a pivotal role. Therefore, gaining a comprehensive understanding of government control measures is of paramount importance, which is beneficial for health authorities in the policy development process. Our study analysis the daily change data of the daily COVID-19 infection count in Tokyo from January 16, 2020 to September 30, 2022. We utilized adaptive Fourier decomposition (AFD) for analyzing the temporal trends within COVID-19 data. It extends the conventional AFD approach by constructing new components base on multiple individual components at various time-frequency scales. Furthermore, we conducted Pearson correlation assessments of the first to third-order synthesis results, along with comparative analyses against other signal analysis techniques. Ultimately, these new components are integrated with policy data spanning different time periods for a comprehensive analysis. The analysis of daily COVID-19 data in Tokyo using AFD reveals how various government policies impacted infection rates across seven distinct fluctuation periods. In the decomposition results, the reduction of business hours policy correlated with high-frequency components in the first four waves, while the low-frequency components for the sixth wave suggested a decline in its relevance. The vaccination policy initially displayed a mid-frequency correlation with the fifth wave and continued with a low-frequency correlation in the last wave. Moreover, our statistical analysis (value of p < 0.05) demonstrated that 75% of the third-order AFD components exhibited significant positive correlations with the original infections, while the correlation coefficients of most components in EMD and VMD did not attain significance. In the time-frequency domain, AFD demonstrates superior performance compared to EMD and VMD in capturing crucial data related to epidemic control measures. The variations in daily COVID-19 infection counts during these seven periods under various policies are evident in distinct third-order AFD components. These findings guide the formulation of future public health policies and social measures.

Time-frequency return connectedness between Chinese coal futures and international stock indices

This paper explores the return connectedness between Chinese coal futures (thermal coal and coking coal) and 16 representative international equity indices from Asia, America, Australia, and Europe. We employ a time-frequency connectedness methodology to capture time-frequency features of the return connectedness between Chinese coal futures and stock indices. The empirical results suggest that return connectedness between Chinese coal futures and equity indices is time-varying and heterogeneous at different frequencies. The evidence also indicates that the short-term (high-frequency) return connectedness is stronger than long-term (low-frequency) during the sample period, implying that high-frequency return connectedness dominates total return connectedness. Meanwhile, Chinese thermal coal futures (TCF) and Chinese coking coal futures (CCT) serve as net receivers of return spillover in the coal-stock connectedness system, whether long-term or short-term. Finally, we detect that the return connectedness between Chinese coal futures and emerging equity markets exceeds developed equity markets. Summarily, the return connectedness between Chinese coal futures and the international stock indices varies depending on the development degree of the equity market, coal varieties, and time-frequency scales. These new findings offer implications for investors, policymakers, and market participants.

Spatio-temporal Variations in PM2.5and Its Influencing Factors in the Yangtze River Delta Urban Agglomeration

To coordinate the contradiction between economic development and environmental pollution and achieve the sustainable development of the economy and society, the spatio-temporal variations in PM2.5 were analyzed based on PM2.5 concentration and meteorological data of the Yangtze River Delta (YRD) urban agglomeration. Wavelet transform coherence (WTC), partial wavelet coherence (PWC), and multiple wavelet coherence (MWC) were used to analyze the multi-scale coupling oscillation between PM2.5 and meteorological factors in the time-frequency domain. The results showed that:① the concentration of PM2.5 in the YRD decreased from northwest to southeast, and the spatial range with high PM2.5 concentration decreased annually. The spatial distribution characteristics of the seasonal average PM2.5 concentration were similar to those of the annual average PM2.5 concentration. PM2.5 concentration exhibited the seasonal variation characteristics of high in winter, low in summer, and transitioning between spring and autumn. ② PM2.5 concentration decreased from 2015 to 2021, and the compliance rate increased. The difference in annual average PM2.5 concentration was decreased with dynamic convergence characteristics. The convergence of PM2.5 concentration in summer was greater than that in winter. During the whole study period, the daily average PM2.5 concentration showed a "U" distribution, and the proportion of days with excellent and good PM2.5 levels were 49.72% and 41.45%, respectively. ③ The wavelet coherence between PM2.5 and meteorological factors was different in different time-frequency domains. The main factors affecting PM2.5 were different in different time-frequency scales. At all time-frequency scales, WTC and PWC showed that wind speed and temperature were the best explanatory variables of PM2.5 variation, respectively. ④ The larger the time-frequency scale, the stronger the interaction of multi-factor combinations to explain PM2.5 variations. The synergistic effect of temperature and wind speed could better explain the variation in PM2.5. These results can provide reference for air pollution control in the YRD.

A Hybrid Deep Learning Approach: Integrating Short-Time Fourier Transform and Continuous Wavelet Transform for Improved Pipeline Leak Detection.

A hybrid deep learning approach was designed that combines deep learning with enhanced short-time Fourier transform (STFT) spectrograms and continuous wavelet transform (CWT) scalograms for pipeline leak detection. Such detection plays a crucial role in ensuring the safety and integrity of fluid transportation systems. The proposed model leverages the power of STFT and CWT to enhance detection capabilities. The pipeline's acoustic emission signals during normal and leak operating conditions undergo transformation using STFT and CWT, creating scalograms representing energy variations across time-frequency scales. To improve the signal quality and eliminate noise, Sobel and wavelet denoising filters are applied to the scalograms. These filtered scalograms are then fed into convolutional neural networks, extracting informative features that harness the distinct characteristics captured by both STFT and CWT. For enhanced computational efficiency and discriminatory power, principal component analysis is employed to reduce the feature space dimensionality. Subsequently, pipeline leaks are accurately detected and classified by categorizing the reduced dimensional features using t-distributed stochastic neighbor embedding and artificial neural networks. The hybrid approach achieves high accuracy and reliability in leak detection, demonstrating its effectiveness in capturing both spectral and temporal details. This research significantly contributes to pipeline monitoring and maintenance and offers a promising solution for real-time leak detection in diverse industrial applications.

Measuring the multi-scale price transmission effects from crude oil to energy stocks: A cascaded view

The extant literature confirms that international crude oil price volatility has a transmission effect on the energy stock market in distinct time-frequency scales. However, the transmission process might spread step by step and lead to direct and indirect impacts, which are defined as cascaded transmission in this paper. To reveal the cascaded transmission effects from crude oil price volatility to the energy stock market at different time-frequency domains, this paper presents a systematic research framework combining the maximal overlap discrete wavelet transform method, BEKK-GARCH model, Diebold-Yilmaz spillover index and the network analysis method. Taking the Chinese energy stock market as an example, the empirical results show that: (1) Crude oil price volatility has different impacts on Chinese energy stocks across the six time scales, with the largest impact in D6. Moreover, crude oil price volatility does not directly affect all energy stocks, but it does indirectly. (2) The proportion of indirect impact as a percentage of total impact ranges from 22.66% to 60.96%. Thus, indirect effects should not be overlooked when analyzing the impact of crude oil price volatility on Chinese energy stock markets. (3) The impact of crude oil on energy stocks varies by industry, and “Energy Mining Equipment and Services” stocks are the most significantly affected in time scales except D3 and D6. This study could provide insights for energy stock market participants with different risk preferences to accurately make decisions under distinct time scales.

Influencing factors of PM2.5 concentration in the typical urban agglomerations in China based on wavelet perspective

PM2.5 is one of the most harmful air pollutants affecting sustainable economic and social development in China. The analysis of influencing factors affecting PM2.5 concentration is significant for the improvement of air quality. In this study, three typical urban agglomerations in China (Beijing‒Tianjin‒Hebei [BTH], the Yangtze River Delta [YRD], and the Pearl River Delta [PRD]) were studied using innovative trend analysis, a Bayesian statistical model, and partial wavelet and multiwavelet coherence to analyze PM2.5 concentration variations and multi-scale coupled oscillations between PM2.5 concentration and air pollutants/meteorological factors. The results showed that: (1) PM2.5 concentration time-series showed significant downward trends, which decreased as follows: BTH > YRD > PRD. The higher the pollution level, the greater the change trend. In BTH and the PRD, PM2.5 had obvious trends and seasonal change points; whereas, the PM2.5 time-series change point in the YRD was not obvious. (2) PM2.5 had significant intermittent resonance cycles with air pollutants and meteorological factors in different time domains. There were differences in the main controlling factors affecting PM2.5 among the three urban agglomerations. (3) The explanatory ability of air pollutant combinations for variations in PM2.5 was higher than that of meteorological factor combinations. However, the synergistic effect of air pollutants/meteorological factors could better explain the PM2.5 concentration variations on all time-frequency scales. The results of this study provide a reference for ecological improvement as well as collaborative governance of air pollution.

Response of Runoff Change to Extreme Climate Evolution in a Typical Watershed of Karst Trough Valley, SW China

Identifying the response of runoff changes to extreme climate evolution was of great scientific significance for the rational regulation of watershed water resources and the prevention of hydrological disasters. However, the time–frequency response relationships were not clear. The Yinjiang River watershed, a typical watershed with karst trough valley areas, was chosen to identify the impact of different climatic driving factors on runoff changes from 1984 to 2015. Continuous wavelet transform (CWT), cross-wavelet transform (XWT), and wavelet coherence transform (WTC) were performed to study the response relationship and time–frequency effect between runoff changes and extreme climate change at different time scales. The main results showed that: (1) Twelve extreme climate indices (ECIs) were detected to have a significant impact on runoff changes, mainly on a 6-year time scale; (2) The R10 and Rx1day in extreme precipitation index and SU34.4 and TNx in the extreme temperature index were the main driving factors of runoff changes, which had relatively large impacts on runoff changes in high and low energy vibration regions. However, the remaining eight ECIs that passed the 0.05 confidence level showed relatively large impacts on runoff changes only in low energy vibration regions; (3) The transition of the interaction between ECIs and runoff changes in high and low time–frequency scales was related to the abrupt change characteristics of the ECIs. The correlation of abrupt change was an important reason for the emergence of highly correlated regions that trigger high and low energy vibrations; (4) As a whole, the extreme precipitation events were ahead of runoff changes at the high time–frequency scale and exhibited small lag effects at the low time–frequency scale, while extreme temperature events were mainly ahead of runoff changes. This study has effectively revealed the impact of climate factors at different scales on runoff changes, and provides a theoretical understanding for regulating and managing water resources in karst basins.

Measurements and estimation of frequency-dependent multi-scale viscoelastic damping properties of materials

The state-of-art viscoelastic property measurement and estimation methods at different frequencies/length scales inherently couple structural effects or sometimes lose accuracy for relevant parameters at smaller length scales in the material. We report a unified approach to measuring and estimating the frequency-dependent viscoelastic damping properties across different frequency scales independently of the specimen size effect and boundary conditions. The Dynamic Mechanical Analysis (DMA) method with three-point bending configuration is extended to higher-fidelity modeling in terms of time–frequency scales and material micro-mechanical scales together. The methodology demonstrates ways to identify the possible sources of undesired dynamics in the measurement process and eliminate those effects. It includes exciter tip contact dynamics in three-point bending configuration, anharmonic energy transfer to different frequencies (fractal bifurcation), and unmodelled artifacts or noise. Critically reviewing the standard DMA procedure with normal mode synthesis, this paper proposes a time–frequency Fourier spectral method that resolves a broader frequency range that extends to even acoustic/ultrasonic wave regimes of viscoelastic damping. An algorithm proposed in this paper with the advanced viscoelastic model with multi-parameters (standard linear solid model) estimates the fundamental material parameters at small length scales, which is more insightful to obtain a frequency-dependent correlation of material response to general dynamic loading. Stiffness and damping properties are estimated from measured data and validated for self-consistencies via different controls regarding static preload, specimen resonance, filtered noise, and model fidelity across the frequency range, and certain consistency and convergence properties of a series solution. The unified approach with a single test will be helpful in the design and evaluation of a wide variety of materials and their applications in vibration and acoustic problems. The proposed technique further demonstrates a novel way to extend low-frequency measurement to high-frequency response modeling.

Multiscale Monitoring Using Machine Learning Methods: New Methodology and an Industrial Application to a Photovoltaic System

In this study, a multiscale monitoring method for nonlinear processes was developed. We introduced a machine learning tool for fault detection and isolation based on the kernel principal component analysis (PCA) and discrete wavelet transform. The principle of our proposal involved decomposing multivariate data into wavelet coefficients by employing the discrete wavelet transform. Then, the kernel PCA was applied on every matrix of coefficients to detect defects. Only those scales that manifest overruns of the squared prediction errors in control limits were considered in the data reconstruction phase. Thus, the kernel PCA was approached on the reconstructed matrix for detecting defects and isolation. This approach exploits the kernel PCA performance for nonlinear process monitoring in combination with multiscale analysis when processing time-frequency scales. The proposed method was validated on a photovoltaic system related to a complex industrial process. A data matrix was determined from the variables that characterize this process corresponding to motor current, angular speed, convertor output voltage, and power voltage system output. We tested the developed methodology on 1000 observations of photovoltaic variables. A comparison with monitoring methods based on neural PCA was established, proving the efficiency of the developed methodology.

A Method for Pipeline Leak Detection Based on Acoustic Imaging and Deep Learning.

This paper proposes a reliable technique for pipeline leak detection using acoustic emission signals. The acoustic emission signal of a pipeline contains leak-related information. However, the noise in the signal often obscures the leak-related information, making traditional acoustic emission features, such as count and peaks, less effective. To obtain leak-related features, first, acoustic images were obtained from the time series acoustic emission signals using continuous wavelet transform. The acoustic images (AE images) were the wavelet scalograms that represent the time–frequency scales of the acoustic emission signal in the form of an image. The acoustic images carried enough information about the leak, as the leak-related information had a high-energy representation in the scalogram compared to the noise. To extract leak-related discriminant features from the acoustic images, they were provided as input into the convolutional autoencoder and convolutional neural network. The convolutional autoencoder extracts global features, while the convolutional neural network extracts local features. The local features represent changes in the energy at a finer level, whereas the global features are the overall characteristics of the acoustic signal in the acoustic image. The global and local features were merged into a single feature vector. To identify the pipeline leak state, the feature vector was fed into a shallow artificial neural network. The proposed method was validated by utilizing a data set obtained from the industrial pipeline testbed. The proposed algorithm yielded a high classification accuracy in detecting leaks under different leak sizes and fluid pressures.

The price fluctuation in Chinese carbon emission trading market: New evidence from adaptive Fourier decomposition

This paper studies the carbon price fluctuation in China through the adaptive Fourier decomposition (AFD). Apart from the transient time-frequency distribution of the original AFD model, we also reconstruct the mono-components of this model to obtain the components in different time-frequency scales. Our empirical results based on the carbon price in Hubei Province demonstrate that there are three periods when the price fluctuates dramatically, mainly affected by the governmental policies about carbon emission and the development of clean energies, as well as the outbreak of COVID-19. Furthermore, the fluctuations of the price in the three identified periods are reflected in different scales. The comparison of the decomposition results and those of EMD and VMD shows that the AFD performs best in absorbing the price’s useful information extracted through all these methods.

Multiple-scale intermuscular coupling network analysis

In order to more accurately and effectively understand the intermuscular coupling of different temporal and spatial levels from the perspective of complex networks, a new multi-scale intermuscular coupling network analysis method was proposed in this paper. The multivariate variational modal decomposition (MVMD) and Copula mutual information (Copula MI) were combined to construct an intermuscular coupling network model based on MVMD-Copula MI, and the characteristics of intermuscular coupling of multiple muscles of upper limbs in different time-frequency scales during reaching exercise in healthy subjects were analyzed by using the network parameters such as node strength and clustering coefficient. The experimental results showed that there are obvious differences in the characteristics of intermuscular coupling in the six time-frequency scales. Specifically, the triceps brachii (TB) had relatively high coupling strength with the middle deltoid (MD) and posterior deltoid (PD), and the intermuscular function was closely connected. However, the biceps brachii (BB) was independent of other muscles. The intermuscular coupling network had scale differences. MVMD-Copula MI can quantitatively describe the relationship of multi-scale intermuscular coupling strength, which has good application prospects.

Gearbox Fault Identification Framework Based on Novel Localized Adaptive Denoising Technique, Wavelet-Based Vibration Imaging, and Deep Convolutional Neural Network

This paper proposes an accurate and stable gearbox fault diagnosis scheme that combines a localized adaptive denoising technique with a wavelet-based vibration imaging approach and a deep convolution neural network model. Vibration signatures of a gearbox contain important fault-related information. However, this useful fault-related information is often overwhelmed by random interference noises. Furthermore, the varying speed of gearboxes makes it difficult to distinguish the fault-related frequencies from the interference noises. To obtain a noise-free signal for extraction of fault-related information under variable speed conditions, first, a new localized adaptive denoising technique (LADT) is applied to the vibration signal. The new localized adaptive denoising technique results in optimized vibration sub-bands with negligible background noise. To obtain fault-related information, the wavelet-based vibration imaging approach (WVI) is applied to the denoised vibration signal. The wavelet-based vibration imaging approach decomposes the vibration signal into different time–frequency scales, these scales are reflected by a two-dimensional image called a scalogram. The scalograms obtained from the wavelet-based vibration imaging approach are provided as an input to the deep convolutional neural network architecture (DCNA) for extraction of discriminant features and classification of multi-degree tooth faults (MDTFs) in a gearbox under variable speed conditions. The proposed scheme outperforms the already existing state-of-the-art gearbox fault diagnosis methods with the highest classification accuracy of 100%.

Spatiotemporal variations in extreme precipitation on the middle and lower reaches of the Yangtze River Basin (1970–2018)

Extreme precipitation events are likely occur more frequently and lead to casualties and socio-economic losses as global warming intensifies. Therefore, exploring the characteristics of extreme precipitation events is essential. Based on meteorological station data from the middle and lower reaches of the Yangtze River Basin (MLYRB) from 1970 to 2018, we analyzed the spatiotemporal distributions of 11 extreme precipitation indices and considered the correlation between altitude and large-scale atmospheric circulation with extreme precipitation. To this end, we employed Sen's trend analysis, rescaled range analysis, the Mann-Kendall test, and continuous wavelet transform. The results indicate that on this spatial scale, extreme precipitation indices in the northwestern region of the MLYRB have mainly decreased, but they have increased along the eastern coastal region. From the perspective of the entire region, Hurst values of extreme precipitation indices are >0.5. Areas with consistent increases are mainly concentrated in the central region, while those with consistent decreases are mainly scattered in the northern and southern regions. Abrupt change in extreme precipitation indices mostly occurred in the 1980s and 1990s. The periodic oscillations of most indices tendtoward high time-frequency scales (2.3 years or less). Consecutive wet days increased with increasing altitude; other indices changed similarly with increasing altitude, first increasing then decreasing before increasing again. Additionally, we found significant resonance periods between the extreme precipitation indices and the atmospheric circulation index, however, there were obvious differences in time domains.

A Fault Diagnosis Framework for Centrifugal Pumps by Scalogram-Based Imaging and Deep Learning

Centrifugal pumps are the most vital part of any process industry. A fault in centrifugal pump can affect imperative industrial processes. To ensure reliable operation of the centrifugal pump, this paper proposes a novel automated health state diagnosis framework for centrifugal pump that combines a signal to time-frequency imaging technique and an Adaptive Deep Convolution Neural Network model (ADCNN). First, the vibration signals corresponding to different health conditions of the centrifugal pump are acquired. Vibration signals obtained from the centrifugal pump carry a great deal of information and generally, statistical features are extracted from the vibration signals to retain meaningful fault information. However, these features are either insensitive to weak incipient faults or unsuitable for tracking severe faults, thus, decreasing the fault classification accuracy. To tackle this problem, a signal to time-frequency imaging technique is applied to the pump vibration signals. For this purpose, Continuous Wavelet Transform (CWT) is applied to decompose the vibration signals over different time-frequency scales and extract the pump fault information in both the time and frequency domains. The CWT scales form two-dimensional time-frequency images commonly referred to as scalograms. The CWT scalograms are then converted into grayscale images (SGI). Over the past few decades, CNN models have been established as an effective practice to process images for classification and pattern recognition. Consequently, the extracted CWTSGIs are finally provided as inputs to the proposed ADCNN architecture to achieve feature extraction and classification for centrifugal pump faults. The performance of the proposed diagnostic framework (CWTSGI + ADCNN) is validated with a vibration dataset collected from a testbed specifically designed for centrifugal pump diagnosis. The experimental results suggest that the proposed technique based on CWTSGI and ADCNN outperformed existing methods with an average performance improvement of 4.7 – 15.6%.