Time-frequency Image Features Research Articles

An End-to-End Adaptive Method for Remaining Useful Life Prediction of Rolling Bearings Using Time–Frequency Image Features

The deep learning model has attracted widespread attention in the field of rolling bearing remaining useful life (RUL) prediction due to its advantages of less reliance on prior knowledge, high accuracy, and strong generalization. However, a large number of prediction models use very complicated artificial feature extraction and selection methods to build the original input features of the deep learning model and health indicator. These approaches do not fully exploit the capabilities of deep learning models as they continue to heavily rely on prior knowledge, The accuracy of their predictions largely hinges on the quality of the input features, and the generalization of manually crafted features remains uncertain. To address these challenges, in this paper, an end-to-end prediction model for the remaining useful life of rolling bearings is proposed, which is divided into three modules. First, a short-term Fourier transform module is incorporated into the model to automatically obtain the time–frequency information of the signal. Then, the convolutional next (ConvNext) module, which is a simple and efficient pure convolutional neural network, is utilized to extract features from the spectrogram. Finally, we capture the short-term dependence and long-term dependence by two parallel channels Transformer and self-attention convolutional long short-term memory (SA-ConvLSTM), and the self-attention mechanism is employed for the adaptive prediction of the bearing’s remaining useful life. Through integration with artificial intelligence, this method proposes a high-performance solution for predicting the remaining useful life of bearings. It has minimal reliance on manual labor, stronger fitting capabilities, and can be widely used for predicting the remaining useful life of bearings.

Impact location of metal structures based on time–frequency image features and deep residual network

Impact location detection plays an important role in the structural health monitoring of metal materials. However, the methods of metal material impact location detection based on physical analysis are often limited by the extraction accuracy of some parameters such as material and structure parameters and time difference calculation. Therefore, this paper develops a deep residual network method for impact location detection, time–frequency characteristic deep residual network (TF-DRN). This method takes the four-channel short-time Fourier transform time–frequency graph as input, uses the unique residual network architecture to automatically extract the advanced features, and then uses the global average pooling layer and the full connection layer to establish the mapping between the advanced features and the impact location, so as to detect the impact location. By introducing regularization and batch normalization, the problems of gradient disappearance and gradient explosion are alleviated, and the generalization and efficiency of impact location detection are further improved. The experimental results show that on an 800 mm × 800 mm × 2.5 mm aluminum plate, the average error of the validation set and the test set are 0.85 cm and 1.33 cm respectively, and the performance of the method is significantly better than that of CNN, ResNet18 and ResNet33 networks.

Research on the Application of Image Feature Extraction in Mechanical Structure Recognition and Fault Diagnosis

Abstract With the rapid development of modern industry and science and technology, in recent years the fault diagnosis method based on image processing has become a research hotspot in the field of mechanical fault diagnosis. In this paper, image characteristics are extracted from multiple aspects such as image texture, color, shape, etc. A grayscale symbiotic matrix image feature extraction method is proposed. On this basis, the algorithm for extracting gray symbiotic matrix time-frequency image features is designed. At the same time, the algorithm and parameters of mechanical structure identification are optimized to identify and diagnose mechanical faults. The results show that the grayscale symbiotic matrix time-frequency image feature extraction algorithm is able to accurately diagnose the wear-type faults, overwork-type faults, and short-circuit-type fault behavior of the mechanical equipment. All of them are able to obtain more than 80% accuracy, and all of them are able to reach 99.99% accurate detection of mechanical faults, which proves the effectiveness of the method of this research.

Deep transfer learning rolling bearing fault diagnosis method based on convolutional neural network feature fusion

Current deep-learning methods are often based on significantly large quantities of labeled fault data for supervised training. In practice, it is difficult to obtain samples of rolling bearing failures. In this paper, a transfer learning-based feature fusion convolutional neural network approach for bearing fault diagnosis is proposed. Specifically, the raw vibration signal features and the corresponding time-frequency image features of the input data are extracted by a one-dimensional convolutional neural network and a pre-trained ConvNeXt, respectively, and connected by a feature fusion strategy. Then, the fine-tuning method based on transfer learning can effectively reduce the reliance on labeled samples in the target domain. A wide convolution kernel is introduced in the time-domain signal feature extraction to increase the receptive field, which is combined with the channel attention mechanism to further optimize the feature quality. Finally, two common bearing datasets are utilized for fault diagnosis experiments. The experimental results show that the proposed model achieves an average accuracy of more than 98.63% in both cross-working conditions and cross-device diagnosis tasks. Meanwhile, anti-noise experiments and ablation experiments further validate the accuracy and robustness of the proposed method.

Research on Diesel Engine Fault Status Identification Method Based on Synchro Squeezing S-Transform and Vision Transformer.

The reliability and safety of diesel engines gradually decrease with the increase in running time, leading to frequent failures. To address the problem that it is difficult for the traditional fault status identification methods to identify diesel engine faults accurately, a diesel engine fault status identification method based on synchro squeezing S-transform (SSST) and vision transformer (ViT) is proposed. This method can effectively combine the advantages of the SSST method in processing non-linear and non-smooth signals with the powerful image classification capability of ViT. The vibration signals reflecting the diesel engine status are collected by sensors. To solve the problems of low time-frequency resolution and weak energy aggregation in traditional signal time-frequency analysis methods, the SSST method is used to convert the vibration signals into two-dimensional time-frequency maps; the ViT model is used to extract time-frequency image features for training to achieve diesel engine status assessment. Pre-set fault experiments are carried out using the diesel engine condition monitoring experimental bench, and the proposed method is compared with three traditional methods, namely, ST-ViT, SSST-2DCNN and FFT spectrum-1DCNN. The experimental results show that the overall fault status identification accuracy in the public dataset and the actual laboratory data reaches 98.31% and 95.67%, respectively, providing a new idea for diesel engine fault status identification.

Gear Fault Diagnosis Based on GPU-CNN

Gear is widely used in mechanical transmission system, but it is prone to failure, which seriously affects the performance of equipment. In order to realize the diagnosis and classification of gear faults, we used Convolutional Neural Networks (CNN) to extract time-frequency image features of its vibration signal. CNN can extract the characteristics of time-frequency signals from vibration signals and identify gear faults accurately. However, due to the large training data set, CNN training costs too much time. According to the characteristics of Graphics Processing Unit (GPU), Compute Unified Device Architecture (CUDA) can improve the speed of CNN algorithm and reduce the time consumption. Therefore, this paper proposes a method based on GPU-CNN for gear fault diagosis. The experimental results show that the method can effectively shorten the training time and significantly improve the operation efficiency.

Internal Combustion Engine Fault Diagnosis Method Based MICEEMD-PWVD

As to multi-signals, the figure of time-frequency distribution becomes difficult to understand due to cross-terms in the Wigner-Ville distribution (WVD). At present time, cross-terms suppression of time-frequency analysis is one of the top interests in signal processing, and more and more research had been put on this topic, but the present research on cross-terms suppression exist conflicting problem that these methods couldn’t suppress the cross-terms while holding high time-frequency resolution. A novel method of cross-terms suppression in the WVD based on Complementary Ensemble Empirical Mode Decomposition (CEEMD) was proposed to solve the above problem. Firstly, multi-component signals were decomposed into intrinsic mode functions (IMFs) which were single component signal by CEEMD; Secondly, the Mutual Information values of IMFS and original signal were calculated to judge and delete the CEEMD false components; Thirdly, these Wigner-Ville distributions of true component of IMFs were computed; Finally, WVDs of each single component signal were added linearly to reconstruct the WVD of original signal. Test on synthetic data shows the effectiveness of the proposed method, which can suppress the cross-terms while holding high time-frequency resolution. Apply MICEEMD - PWVD in internal combustion engine valve clearance fault diagnosis, useTD-2DPCA method to extract MICEEMD-PWVD time-frequency image feature, then classify the characteristic parameters with the nearest neighbor classifier. The results show that the IC Engine Fault Diagnosis Method Based MICEEMD-PWVD and TD-2DPCA can accurately diagnose IC engine valve fault.

Intelligent fault diagnosis of rotating machinery via wavelet transform, generative adversarial nets and convolutional neural network

Abstract The fault detection of rotating machinery systems especially its typical components such as bearings and gears is of special importance for maintaining machine systems working normally and safely. However, due to the change of working conditions, the disturbance of environment noise, the weakness of early features and various unseen compound failure modes, it is quite hard to achieve high-accuracy intelligent failure monitoring task of rotating machinery using existing intelligent fault diagnosis approaches in real industrial applications. In the paper, a novel and high-accuracy fault detection approach named WT-GAN-CNN for rotating machinery is presented based on Wavelet Transform (WT), Generative Adversarial Nets (GANs) and convolutional neural network (CNN). The proposed WT-GAN-CNN approach includes three parts. To begin with, WT is employed for extracting time-frequency image features from one-dimension raw time domain signals. Secondly, GANs are used to generate more training image samples. Finally, the built CNN model is used to accomplish the fault detection of rotating machinery by the original training time-frequency images and the generated fake training time-frequency images. Two experiment studies are implemented to assess the effectiveness of our proposed approach and the results demonstrate it is higher in testing accuracy than other intelligent failure detection approaches in the literatures even in the interference of strong environment noise or when working conditions are changed. Furthermore, its result in the stability of testing accuracy is also quite excellent.

Application of time-frequency image feature extraction method based on gray level co-occurrence matrix in coal mine pipeline leakage detection

The underground mining environment of coal mine is complex. Once the water supply or gas supply pipes leak, the location and time of leakage are unpredictable. If the leakage can not be found in time, it will seriously threaten the life safety of underground workers. Therefore, it is very necessary to automatically detect the pipeline leakage. In order to solve the problem of automatic detection of coal mine pipeline leakage, this paper proposes an image processing method based on gray level co-occurrence matrix to extract the features of time-frequency image of vibration signal. Firstly, the data in a relatively short time is divided into one frame to reduce the amount of calculation. Secondly, each frame of signal is processed by SPWVD to get the time-frequency image. Then, the time-frequency image after time-frequency analysis is transformed into gray-scale image, and the gray-scale co-occurrence matrix (GLCM) of the image is extracted. Finally, the energy, contrast and correlation of the co-occurrence matrix are extracted and input to SVM for learning and training. Through the design and build of the experimental pipeline to test the algorithm proposed in this paper, the detection result of SVM is 96.67%, which shows that this method can effectively detect the pipeline leakage.

Fault diagnosis of rotating machinery based on time-frequency image feature extraction

Aiming at the characteristics of time-frequency analysis of unsteady vibration signals, this paper proposes a method based on time-frequency image feature extraction, which combines non-downsampling contour wave transform and local binary mode LBP (Local Binary Pattern) to extract the features of time-frequency image faults. SVM is used for classification and recognition. Finally, the method is verified by simulation data. The results show that the classification accuracy of the method reaches 98.33%, and the extracted texture features are relatively stable. Also, the method is compared with the other 3 feature extraction methods. The results also show that the classification effect of the method is better than that of the traditional feature extraction method.

Feature extraction using dominant frequency bands and time-frequency image analysis for chatter detection in milling

Chatter is a cause of low surface quality and productivity in milling and crucial features need to be extracted for accurate chatter detection and suppression. This paper introduces a novel feature extraction approach for chatter detection by using image analysis of dominant frequency bands from the short-time Fourier transform (STFT) spectrograms. In order to remove the environmental noises and highlight chatter related characteristics, dominant frequency bands with high energy are identified by applying the squared energy operator to the synthesized fast Fourier transform (FFT) spectrum. The time-frequency spectrogram of the vibration signal is divided into a set of grayscale sub-images according to the dominant frequency bands. Statistical image features are extracted from those sub-images to describe the machining condition and assessed in terms of their separability capabilities. The proposed feature extraction method is verified by using dry milling tests of titanium alloy Ti6Al4V and compared with two existing feature extraction techniques. The results indicate the efficiency of the time-frequency image features from dominant frequency bands for chatter detection and their better performance than the time domain features and wavelet-based features in terms of their separability capabilities.

DNN Filter Bank Improves 1-Max Pooling CNN for Single-Channel EEG Automatic Sleep Stage Classification.

We present in this paper an efficient convolutional neural network (CNN) running on time-frequency image features for automatic sleep stage classification. Opposing to deep architectures which have been used for the task, the proposed CNN is much simpler However, the CNN's convolutional layer is able to support convolutional kernels with different sizes, and therefore, capable of learning features at multiple temporal resolutions. In addition, the 1-max pooling strategy is employed at the pooling layer to better capture the shift-invariance property of EEG signals. We further propose a method to discriminatively learn a frequency-domain filter bank with a deep neural network (DNN) to preprocess the time-frequency image features. Our experiments show that the proposed 1-max pooling CNN performs comparably with the very deep CNNs in the literature on the Sleep- EDF dataset. Preprocessing the time-frequency image features with the learned filter bank before presenting them to the CNN leads to significant improvements on the classification accuracy, setting the state- of-the-art performance on the dataset.

Performance evaluation of time-frequency image feature sets for improved classification and analysis of non-stationary signals: Application to newborn EEG seizure detection

This study demonstrates that a time-frequency (TF) image pattern recognition approach offers significant advantages over standard signal classification methods that use t-domain only or f-domain only features. Two approaches are considered and compared. The paper describes the significance of the standard TF approach for non-stationary signals; TF signal (TFS) features are defined by extending t-domain or f-domain features to a joint (t, f) domain resulting in e.g. TF flatness and TF flux. The performance of the extended TFS features is comparatively assessed using Receiver Operating Characteristic (ROC) analysis Area Under the Curve (AUC) measure. Experimental results confirm that the extended TFS features generally yield improved performance (up to 19%) when compared to the corresponding t-domain and f-domain features.The study also explores a second approach based on novel TF image (TFI) features that further improves TF-based classification of non-stationary signals. New TFI features are defined and extracted from the (t, f) domain; they include TF Hu invariant moments, TF Haralick features, and TF Local Binary Patterns (LBP). Using a state-of-the-art classifier, different metrics based on confusion matrix performance are compared for all extended TFS features and TFI features. Experimental results show the improved performance of TFI features over both TFS features and t-domain only or f-domain only features, for all TF representations and for all the considered performance metrics. The experiment is validated by comparing this new proposed methodology with a recent study, utilizing the same large and complex data set of EEG signals, and the same experimental setup. The resulting classification results confirm the superior performance of the proposed TFI features with accuracy improvement up to 5.52%.

Robust acoustic event classification using deep neural networks

Support vector machines (SVMs) have seen an increased usage in applications of acoustic event classification since its rise to popularity about two decades ago. However, in recent years, deep learning methods, such as deep neural networks (DNNs), have shown to outperform a number of classification methods in various pattern recognition problems. This work starts by comparing the classification performance of DNNs against SVMs with a number of feature representations which fall into two categories: cepstral features and time-frequency image features. Unlike related work, the classification performance of the two classifiers is also compared with feature vector combination and the training and evaluation times of the classifiers and features are also compared. The performance is evaluated on an audio surveillance database containing 10 sound classes, each class having multiple subclasses, with the addition of noise at various signal-to-noise ratios (SNRs). The experimental results shows that DNNs have a better overall classification performance than SVMs with both individual and combined features and the classification accuracy with DNNs is particularly better at low SNRs. The evaluation time of the DNN classifier was also determined to be the fastest but with a slow training time.

Fault Identification of Rotor System Based on Classifying Time-Frequency Image Feature Tensor

In the field of rotor fault pattern recognition, most of classical pattern recognition methods generally operate in feature vector spaces where different feature values are stacked into one-dimensional (1D) vector and then processed by the classifiers. In this paper, time-frequency image of rotor vibration signal is represented as a texture feature tensor for the pattern recognition of rotor fault states with the linear support higher-tensor machine (SHTM). Firstly, the adaptive optimal-kernel time-frequency spectrogram visualizes the unique characteristics of rotor fault vibration signal; thus the rotor fault identification is converted into the corresponding time-frequency image (TFI) pattern recognition. Secondly, in order to highlight and preserve the TFI local features, the TFI is divided into some TFI subzones for extracting the hierarchical texture features. Afterwards, to avoid the information loss and distortion caused by stacking multidimensional features into vector, the multidimensional features from the subzones are transformed into a feature tensor which preserves the inherent structure characteristic of TFI. Finally, the feature tensor is input into the SHTM for rotor fault pattern recognition and the corresponding recognition performance is evaluated. The experimental results showed that the method of classifying time-frequency texture feature tensor can achieve higher recognition rate and better robustness compared to the conventional vector-based classifiers, especially in the case of small sample size.

Subband Time-Frequency Image Texture Features for Robust Audio Surveillance

In this paper, we utilize time-frequency image representations of sound signals for feature extraction in an audio surveillance application. Starting with the conventional spectrogram images, we consider a new feature which is based on image texture analysis. It utilizes the gray-level co-occurrence matrix, which captures the distribution of co-occurring values at a given offset. We refer this as the spectrogram image texture feature. Texture analysis is carried out in subbands and experimented on a sound database containing ten classes with each sound class containing multiple subclasses. The proposed feature was seen to be more noise robust than two commonly used cepstral features, mel-frequency cepstral coefficients and gammatone cepstral coefficients, the spectrogram image feature (SIF), where central moments are extracted as features, and a variation of SIF with reduced feature dimension. In addition, we achieved a significant improvement in classification accuracy for the three time-frequency image features by utilizing a gammatone filter-based time-frequency image, referred as cochleagram image, for feature extraction instead of the spectrogram image. A combination of cepstral and cochleagram image features also gave improvement in the classification performance.

An investigation into machine pattern recognition based on time-frequency image feature extraction using a support vector machine

In this article, a new method of pattern recognition for machine working conditions is presented that is based on time-frequency image (TFI) feature extraction and support vector machines (SVMs). In this study, the Hilbert time-frequency spectrum (HTFS) is used to construct TFIs because of its good performance in non-stationary and non-linear signal analysis. Cyclostationarity signal analysis is a pre-processing method for improving the performance of the HTFS in the construction of TFIs. Feature extraction for TFIs is investigated in detail to construct a feature vector for pattern recognition. Gravity centre and information entropy of TFIs are used to construct the feature vector for pattern recognition. SVMs are used for different working conditions classification by the constructed feature vector because of its powerful performance even for small samples. In the end, rolling bearing pattern recognition is used as an example to testify the effectiveness of this method. According to the result analysis, it can be concluded that this method will contribute to the development of preventative maintenance.