Deep learning classification for improved bicoherence feature based on cyclic modulation and cross-correlation

Improved cyclostationary analysis method based on TKEO and its application on the faults diagnosis of induction motors

Cyclostationary analysis has been strongly recognized as an effective demodulation tool in identifying fault features of rotating machinery based on vibration signature analysis. This study improves two current mature cyclostationary approaches, cyclic modulation spectrum (CMS) and fast spectral correlation (Fast-SC), combined with the novel frequency-domain application of Teager Kaiser energy operator (TKEO). They can enhance fault feature identification with the lower computational burden. Firstly, the raw vibration signal is transformed into the time-frequency domain through the short-time Fourier transform (STFT) to realize the conversion of the multi-carrier signal to a multiple signal-carrier signal. Secondly, the TKEO is utilized to enhance the fault feature by taking full advantage of demodulating the mono-component. Finally, the spectral coherence and enhanced envelope spectrum (EES) are calculated to effectively exhibit fault features. The superiority of the proposed methods is successfully validated by the simulation study and diagnosing the broken rotor bar (BRB) and bearing outer race faults of induction motors (IMs) under various operating conditions.

Underwater single-channel acoustic signal multitarget recognition using convolutional neural networks

The radiated noise from ships is of great significance to target recognition, and several deep learning methods have been developed for the recognition of underwater acoustic signals. Previous studies have focused on single-target recognition, with relatively few reports on multitarget recognition. This paper proposes a deep learning-based single-channel multitarget underwater acoustic signal recognition method for an unknown number of targets in the specified category. The proposed method allows the two subproblems of recognizing the unique class and duplicate categories of multiple targets to be solved. These two tasks are essentially multilabel binary classification and multilabel multiple value classification, respectively. In this paper, we describe the use of real-valued and complex-valued ResNet and DenseNet convolutional networks to recognize synthetic mixed multitarget signals, which was superimposed from individual target signals. We compare the performance of various features, including the original audio signal, complex-valued short-time Fourier transform (STFT) spectrum, magnitude STFT spectrum, logarithmic mel spectrum, and mel frequency cepstral coefficients. The experimental results show that our method can effectively recognize synthetic multitarget ship signals when the magnitude STFT spectrum, complex-valued STFT spectrum, and log-mel spectrum are used as network inputs.

Deep transfer learning for underwater direction of arrival using one vector sensor

A deep transfer learning (DTL) method is proposed for the direction of arrival (DOA) estimation using a single-vector sensor. The method involves training of a convolutional neural network (CNN) with synthetic data in source domain and then adapting the source domain to target domain with available at-sea data. The CNN is fed with the cross-spectrum of acoustical pressure and particle velocity during the training process to learn DOAs of a moving surface ship. For domain adaptation, first convolutional layers of the pre-trained CNN are copied to a target CNN, and the remaining layers of the target CNN are randomly initialized and trained on at-sea data. Numerical tests and real data results suggest that the DTL yields more reliable DOA estimates than a conventional CNN, especially with interfering sources.

Line spectrum extraction based on autoassociative neural networks

Line spectrum is an important feature for the detection and classification of underwater targets. This letter presents a method for extracting the line spectrum submerged in underwater ambient noise through autoassociative neural networks (AANN). Compared with the traditional methods, the proposed method based on AANN can directly enhance the line spectrum from the raw time-domain noise data without relying on prior information and spectral features. Moreover, the proposed method can suppress the background noise while extracting the line spectrum. Both the numerical simulation and experimental data test results demonstrate that the proposed method provides a good ability to extract the line spectrum from the strong background noise.