A generalised differential sparsity measure for reconstructing compressively sampled signals

Maronidis, Anastasios; Chatzilari, Elisavet; Nikolopoulos, Spiros; Kompatsiaris, Ioannis

doi:10.1016/j.dsp.2017.11.016

Cited by 4 publications

(2 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In addition to kurtosis, sparsity indicators also play an important role in rotating machinery fault diagnosis. Common sparsity indices mainly include l 0 norm, l 2 / l 1 norm, and Gini index [32]. Among them, the Hoyer index, which is a normalized version of the l 2 / l 1 measure, has been utilized in the bearing health management [33].…”

Section: Demodulation Frequency Band Selection Based On Hoyer Indexmentioning

confidence: 99%

Frequency Hoyer attention based convolutional neural network for remaining useful life prediction of machinery

Huang¹,

Zhang²,

Shi³

et al. 2021

Meas. Sci. Technol.

View full text Add to dashboard Cite

Remaining useful life (RUL) prediction based on vibration data is a vital part of bearing prognostic and health management, which can be applied to formulate a suitable maintenance strategy. Recently, attention mechanism has been widely studied and applied in the field of prognostics, which can adaptively enhance the features for RUL prediction and weaken the features interfering with the accurate estimation of the condition. However, insufficient priori information has been provided to the deep learning model via the attention mechanism. The domain knowledge of the special structure and the characteristics of the bearing vibration signal is underutilized in the conventional attention mechanism. An innovative attention based RUL prediction model, called frequency Hoyer attention based convolutional neural network (FHA-CNN), is proposed in this study, which combines a deep learning model and signal processing method organically. The 1D convolutional layer and isometric empirical wavelet transform are developed to extract the latent representation of vibration signals from different scales. The proposed FHA is applied to calculate the weight of the feature map adaptively, in which three types of Hoyer index are adopted to comprehensively evaluate the contribution of each frequency part to the degradation of rolling bearings from the frequency domain perspective. To verify the superiority of the proposed method, two run-to-failure experimental dataset case studies are analysed. The obtained results indicate that the proposed FHA-CNN model exhibits a better performance than conventional deep learning-based RUL prediction methods. In addition, the proposed method concentrates on the special structure of the bearing vibration signal and provides a novel insight into the decision-making processes of deep neural networks.

show abstract

Section: Demodulation Frequency Band Selection Based On Hoyer Indexmentioning

confidence: 99%

Frequency Hoyer attention based convolutional neural network for remaining useful life prediction of machinery

Huang¹,

Zhang²,

Shi³

et al. 2021

Meas. Sci. Technol.

View full text Add to dashboard Cite

show abstract

“…It is divided into three main parts: integrated energy data collection for regional buildings, data relay, and integrated energy data analysis. Data collection for the integrated energy use of regional buildings utilises a terminal for real-time sampled data (such as a smart metre), temporarily stores the data in the data relay, and then performs selection and compression on the sampled data [10][11][12]. Next, it reconstructs the data and analyses the load characteristics by applying the compressed sensing theory and the constructed over-complete dictionary.…”

Section: System Architecture For Integrated Energy-load Characteristimentioning

confidence: 99%

Analysis of integrated energy‐load characteristics based on sparse clustering and compressed sensing

Wang

Hou

2019

IET Energy Systems Integration

View full text Add to dashboard Cite

An integrated energy system not only provides a platform for multi‐energy coupling utilisation but also satisfies users' diversified energy demands. However, in view of the enormous amount of integrated energy data and the difficulty of analysing the characteristics of that data, an integrated energy‐analysis method based on sparse clustering and compressed sensing is proposed in this study. This method uses the fuzzy c‐means algorithm to construct an over‐complete dictionary and then compresses, collects, and reconstructs the integrated energy data using the compressed sensing theory method. This process analyses integrated energy‐load characteristics accurately and also solves the problem of low data‐transmission efficiency. Simulation results show that the method is suitable for analysing and processing integrated load data in integrated energy systems.

show abstract

Convolutional neural network classifier for the output of the time-domain $\mathcal{F}$-statistic all-sky search for continuous gravitational waves

Morawski

Bejger

Cieciela̧g

2020

Mach. Learn.: Sci. Technol.

View full text Add to dashboard Cite

A generalised differential sparsity measure for reconstructing compressively sampled signals

Cited by 4 publications

References 31 publications

Frequency Hoyer attention based convolutional neural network for remaining useful life prediction of machinery

Frequency Hoyer attention based convolutional neural network for remaining useful life prediction of machinery

Analysis of integrated energy‐load characteristics based on sparse clustering and compressed sensing

Convolutional neural network classifier for the output of the time-domain $\mathcal{F}$-statistic all-sky search for continuous gravitational waves

Contact Info

Product

Resources

About