China 
Africa 
Explore chapters and articles related to this topic
Single-sensor Real Time Damage Detection Techniques: RSSA and its Variants
Published in Basuraj Bhowmik, Budhaditya Hazra, Vikram Pakrashi, Real-Time Structural Health Monitoring of Vibrating Systems, 2022
Basuraj Bhowmik, Budhaditya Hazra, Vikram Pakrashi
To operate in real time for separation of line noise from signal and simultaneous damage detection in a single framework the proposed algorithm primarily comprises of two segments: (i) Filtering and (ii) Damage detection. The Filtering utilizes Recursive Multichannel Singular Spectrum Analysis (RMSSA) to filter out the noise components manifested by signal noises, environmental and operational conditions. The mathematical structure of proposed RMSSA algorithm is analogous to traditional Singular Spectrum Analysis (SSA) but utilized the principles of first order eigenperturbation technique (FOEP) for real time series analysis, primarily for online separation of unwanted components. The RMSSA algorithm is also an advanced structure of recently established RSSA [4] algorithm with built-in multivariate analysis feature that was lacking in RS SA algorithm therefore can account for multivariate analysis without much computational cost. The RMSSA algorithm produces a set of filtered data series having a lower model order than the original signal. The reconstructed signal has significant information regarding characteristics of source and a better representative of structural damage. The filtered response of RMSSA is taken as input to the damage detection that utilizes Recursive Principal Component Analysis (RPCA) on the colored signal (filtered) for identification of damage events. The RPCA is the recursive counterpart of Principal Component Analysis (PCA) and has shown great damage identification in real time in recently published literature [5, 6].
A comparison between ARIMA, LSTM, ARIMA-LSTM and SSA for cross-border rail freight traffic forecasting: the case of Alpine-Western Balkan Rail Freight Corridor
Published in Transportation Planning and Technology, 2023
Miloš Milenković, Miloš Gligorić, Nebojša Bojović, Zoran Gligorić
Singular Spectrum Analysis (SSA) represents a nonparametric spectral estimation method based on the combination of time series analysis, dynamical systems, and signal processing (Harris and Yuan 2010; Hassani 2007; Hassani and Mahmoudvand 2013; Hassani and Zhigljavsky 2009; Stratigakos et al. 2021). The algorithm is composed of two main stages: decomposition and reconstruction. In the first stage, the time series is represented as a spectrum of independent components such as trend, periodic oscillatory, and noise. At the embedding step, the univariate time series is transformed into a trajectory matrix, which has the properties of a Henkel matrix with equal elements on the antidiagonal. Consider that there are realizations of a stochastic process . The trajectory matrix is obtained as follows: where represents the length of the selected window and are the selected lagged vectors.
An EEMD-based method for removing residual blood oxygen signal noise by combining wavelet and singular spectrum analysis
Published in Journal of Modern Optics, 2023
Zhiming Xing, Yan Cao, Xinzhi Shan, Lingyu Wang, Xiumin Gao
SSA (Singular Spectrum Analysis) is a special case of principal component analysis, which is particularly suitable for analysing one-dimensional time series, and can effectively extract useful information such as trend terms, periodic terms and semi-periodic terms in the time series, and can also achieve data denoising, interpolation and extrapolation, etc. It is an extremely widely used method for time series analysis. SSA was applied to the EEMD-LWT denoising results to remove the residual noise component. Figure 10 shows the singular entropy increments of the EEMD-LWT denoising results. The large value of the singular entropy increment represents the signal component and the small value of the singular entropy increment represents the noise component. From the figure, it can be seen that the signal mutates at the fifth component, so the first four components are retained and the reconstruction results of the first four components are taken as the denoising results. The comparison between EEMD-LWT-SSA denoising results and original signals is shown in Figure 11 (a), and Figure 11 (b) is the spectrum diagram of EEMD-LWT-SSA denoising results. It can be seen from Figure 11 that the recovered signal after EEMD-LWT-SSA denoising is consistent with the original signal to the greatest extent in terms of both signal amplitude and signal composition, and the residual interference after filtering is also well suppressed. The denoising effect is better than that of EEMD-LWT.
Accuracy improvement of various short-term load forecasting models by a novel and unified statistical data-filtering method
Published in International Journal of Green Energy, 2020
Duong Minh Bui, Phuc Duy Le, Minh Tien Cao, Trang Thi Pham, Duy Anh Pham
Finally, similar to the Wavelet transform-based filter, a Singular Spectrum Analysis (SSA)-based filter is also a signal decomposition technique, which can be considered as a trustful time-series analyzing technique to detect and extract trends, periodic components, and noise from the database (Zhang, Wang, and Zhang 2017). The SSA-based filter, a non-parametric technique, is a powerful tool in short-term load forecasting when considering the elements of classical time-series analysis, multivariate statistics, geometry, dynamical systems, and signal processing. However, this non-parametric technique might not be ubiquitously used for the time-series forecasting.