Time Frequency Domain Features Research Articles

A Novel RUL-Centric Data Augmentation Method for Predicting the Remaining Useful Life of Bearings

Maintaining the reliability of rotating machinery in industrial environments entails significant challenges. The objective of this paper is to develop a methodology that can accurately predict the condition of rotating machinery in order to facilitate the implementation of effective preventive maintenance strategies. This article proposed a novel RUL-centric data augmentation method, designated as DF-MDAGRU, for the purpose of predicting the remaining useful life (RUL) of bearings. This model is based on an encoder–decoder framework that integrates time–frequency domain feature enhancement with multidimensional dynamic attention gated recurrent units for feature extraction. This method enhances time–frequency domain features through the Discrete Wavelet Downsampling module (DWD) and Convolutional Fourier Residual Block (CFRB). This method employs a Multiscale Channel Attention Module (MS-CAM) and a Multiscale Convolutional Spatial Attention Mechanism (MSSAM) to extract channel and spatial feature information. Finally, the output predictions are processed through linear regression to achieve the final RUL estimation. Experimental results demonstrate that the proposed method outperforms other state-of-the-art approaches on the FEMETO-ST and XJTU datasets.

Classification of faults in friction stir processed composites using a machine learning and ensemble learning approach.

Abstract Aluminium alloy based surface composites with hard reinforcement particles have wide scope in aerospace and automobile manufacturing industries. In this paper, the aluminium composites, manufactured by friction stir processing (FSP) with varying parameters are investigated for the faults occurred during fabrication process. It explores a machine-learning approach to detect defects of surface hybrid composites with an Al6061 alloy matrix, reinforced with copper and graphene nano-powders, using friction stir processing and a tungsten carbide tool on a milling machine. Multi-sensor time series data (vibration, force, and current) collected during fabrication, is preprocessed and labelled with normal and defective categories (e.g., pin break, brazing break, rough surface, no composite) using visual inspection. The important time domain and frequency domain features are extracted using different libraries in python. Thenafter, various types of feature selection techniques, viz filter, wrapper and embedded methods are implemented to select most relevant features. The selected subset of features from all selection methods used, are applied to different machine learning and ensemble learning classifiers and their performances are evaluated. The optimal combinations of the type of feature selection method and classifier used, are obtained for efficient classification of surface defects in composited formed by FSP. The real time monitoring and defect detection system can be developed in future for the composites developed by FSP using the developed models.

Tool wear prediction based on K-means and Adaboost auto-encoder

Abstract A new tool wear prediction model is proposed to address the tool wear issue, aimed at monitoring tool wear based on specific task requirements and guiding tool replacement during actual cutting operations. In the data preprocessing phase, tool wear states are classified using unsupervised K-means clustering. The time, frequency, and time-frequency domain features are then labeled and fused using an autoencoder neural network applied to the original set of signal features from the tool. For tool wear prediction, an enhanced autoencoder neural network leveraging AdaBoost is employed to establish the prediction model. The reconstruction error serves as the chosen loss function to assess the autoencoder's performance, taking into account data correlation and the inherent lossy nature of the autoencoder. Experimental results from real machining data obtained from a CNC milling machine demonstrate that the proposed model achieves higher prediction accuracy while reducing data dimensions.

Research on Abnormal State Detection of CZ Silicon Single Crystal Based on Multimodal Fusion

The Czochralski method is the primary technique for single-crystal silicon production. However, anomalous states such as crystal loss, twisting, swinging, and squareness frequently occur during crystal growth, adversely affecting product quality and production efficiency. To address this challenge, we propose an enhanced multimodal fusion classification model for detecting and categorizing these four anomalous states. Our model initially transforms one-dimensional signals (diameter, temperature, and pulling speed) into time–frequency domain images via continuous wavelet transform. These images are then processed using a Dense-ECA-SwinTransformer network for feature extraction. Concurrently, meniscus images and inter-frame difference images are obtained from the growth system’s meniscus video feed. These visual inputs are fused at the channel level and subsequently processed through a ConvNeXt network for feature extraction. Finally, the time–frequency domain features are combined with the meniscus image features and fed into fully connected layers for multi-class classification. The experimental results show that the method can effectively detect various abnormal states, help the staff to make a more accurate judgment, and formulate a personalized treatment plan for the abnormal state, which can improve the production efficiency, save production resources, and protect the extraction equipment.

Multilevel identity fine authentication method based on time-frequency domain feature extraction in industrial internet of things system

The cyber physical system is widely applied to industrial Internet of Things system and its security determines the safety of the industrial IoT system. This paper proposes a multilevel fine fingerprint authentication method for the security protection of key operating equipment. Firstly, Because of the acquired signals, the mathematical representation models are established separately, including the self-encoding model on the basis of the multiorder moment features and the multi-scale representation model based on the wavelet. Secondly, the multigranularity features of the time domain and frequency domain are extracted respectively based on the real-time data's self-encoding representation and multiscale representation. Then, the extracted features of different granularities can be built into the image representation based on the time-frequency fusion portrait. Based on the images of different granularities and the legitimate image library, a multiscale matching authentication method is established. Finally, simulation experiments show that the proposed method is effective.

Metal surface crack depth laser acoustic emission detection method based on multivariate feature adaptive extraction and cross-modal interaction fusion

Surface crack depth detection of metal structures is of great significance to ensure the safe of equipment. However, the commonly used detection methods currently suffer from sensitivity and imaging dimension limitations, which make it difficult to ensure accuracy and feature extraction. To this end, a laser acoustic emission crack depth detection method that combines modal decomposition and deep learning is proposed. The new method is more sensitive to elastic wave changes that detect the depth of microscopic cracks and shows higher accuracy than traditional detection methods. At the same time, frequency-adaptive feature mode decomposition of multivariate signals and color multimodal feature fusion are used to solve the difficulty of time–frequency domain feature mining in current research. The crack depth detection network CDDNet is constructed using an attention-based fusion backbone, a novel cross-modal interaction fusion, and a hierarchical information fusion strategy. Its transfer learning achieves a crack detection accuracy level of 0.05 mm and outperforms other models in terms of accuracy, reaching 93.6 %. This paper shows that the proposed transfer learning laser acoustic emission technology and supporting data processing methods can effectively establish a connection between the experimental and simulation datasets, solving the current challenges of crack depth detection in terms of micro-precision and data processing.

Enhancing gait recognition in lower limb exoskeletons: adaptive feature selection and random forest with Bayesian optimization

Abstract To improve human-machine cooperation and enhance the accuracy of gait recognition in wearable lower limb exoskeletons, an enhancement method of gait recognition based on adaptive feature selection and an optimized machine learning algorithm was proposed. In this study, surface electromyography (sEMG) signals of rectus femoris, medialis femoris, lateralis femoris, semitendinosus and biceps femoris were recorded during level-ground walking. Then, time domain, frequency domain, time-frequency domain and nonlinear features were extracted. The integrated value of electromyography, variance, root mean square and wavelength were selected as the time domain features, and the mean power frequency and median frequency were selected as the frequency domain features. Wavelet packet energy was selected as the time-frequency domain feature. Nonlinear features including approximate entropy, sample entropy and fuzzy entropy of sEMG were extracted. To reduce feature dimension and improve the calculation efficiency, adaptive feature selection was performed by particle swarm optimization combined with sigmoid function. Then, the feature matrix was determined as the input for a random forest classifier to recognize different gait phases. An adaptive optimization mechanism based on Bayesian optimization was applied to random forest. Compared with random forest, the overall performance of the optimized model was improved. Its accuracy was increased by 3.57%. The feature selection and adaptive optimization mechanisms in gait recognition not only enhance the accuracy of random forest algorithms applied to sEMG for gait prediction, but also facilitate the flexibility and consistency required for lower limb exoskeleton gait control.

Dual-Feature Fusion Learning: An Acoustic Signal Recognition Method for Marine Mammals

Marine mammal acoustic signal recognition is a key technology for species conservation and ecological environment monitoring. Aiming at the complex and changing marine environment, and because the traditional recognition method based on a single feature input has the problems of poor environmental adaptability and low recognition accuracy, this paper proposes a dual-feature fusion learning method. First, dual-domain feature extraction is performed on marine mammal acoustic signals to overcome the limitations of single feature input methods by interacting feature information between the time-frequency domain and the Delay-Doppler domain. Second, this paper constructs a dual-feature fusion learning target recognition model, which improves the generalization ability and robustness of mammal acoustic signal recognition in complex marine environments. Finally, the feasibility and effectiveness of the dual-feature fusion learning target recognition model are verified in this study by using the acoustic datasets of three marine mammals, namely, the Fraser’s Dolphin, the Spinner Dolphin, and the Long-Finned Pilot Whale. The dual-feature fusion learning target recognition model improved the accuracy of the training set by 3% to 6% and 20% to 23%, and the accuracy of the test set by 1% to 3% and 25% to 38%, respectively, compared to the model that used the time-frequency domain features and the Delay-Doppler domain features alone for recognition.

Asymmetric-Based Residual Shrinkage Encoder Bearing Health Index Construction and Remaining Life Prediction.

Predicting the remaining useful life (RUL) of bearings is crucial for maintaining the reliability and availability of mechanical systems. Constructing health indicators (HIs) is a fundamental step in the methodology for predicting the RUL of rolling bearings. Traditional HI construction often involves determining the degradation stage of the bearing by extracting time-frequency domain features from raw data using a priori knowledge and setting artificial thresholds; this approach does not fully utilize the vibration information in the bearing data. In order to address the above problems, this paper proposes an Asymmetric Residual Shrinkage Convolutional Autoencoder (ARSCAE) model. The asymmetric structure of the ARSCAE model is characterized by the soft thresholding of signal features in the encoder part to achieve noise reduction. The decoder part consists of convolutional and pooling layers for data reconstruction. This model can directly construct HIs from the original vibration signals collected, and comparisons with other models show that it constructs better HIs from the original vibration signals. Finally, experiments on the FEMTO dataset show that the results indicate that the HIS constructed by the ARSCAE model has better lifetime prediction capability compared to other methods.

An End-to-End Underwater Acoustic Target Recognition Model Based on One-Dimensional Convolution and Transformer

Underwater acoustic target recognition (UATR) is crucial for defense and ocean environment monitoring. Although traditional methods and deep learning approaches based on time–frequency domain features have achieved high recognition rates in certain tasks, they rely on manually designed feature extraction processes, leading to information loss and limited adaptability to environmental changes. To overcome these limitations, we proposed a novel end-to-end underwater acoustic target recognition model, 1DCTN. This model directly used raw time-domain signals as input, leveraging one-dimensional convolutional neural networks (1D CNNs) to extract local features and combining them with Transformers to capture global dependencies. Our model simplified the recognition process by eliminating the need for complex feature engineering and effectively addressed the limitations of LSTM in handling long-term dependencies. Experimental results on the publicly available ShipsEar dataset demonstrated that 1DCTN achieves a remarkable accuracy of 96.84%, setting a new benchmark for end-to-end models on this dataset. Additionally, 1DCTN stood out among lightweight models, achieving the highest recognition rate, making it a promising direction for future research in underwater acoustic recognition.

Rolling bearing fault diagnosis method based on wavelet packet logarithmic energy map

Aiming at the fault diagnosis problem of rolling bearings, a diagnosis method based on wavelet packet logarithmic energy map is proposed. Firstly, a new wavelet packet node logarithmic energy formula is improved and proposed to overcome the shortcomings of cumbersome parameter determination and strong subjectivity in the traditional wavelet packet energy formula, improve the recognition of high-frequency faults and the distinction of low-frequency fault categories, so as to fully extract the initial time-frequency domain features; secondly, the Gram angle and field idea is used to realize the conversion from one-dimensional features to two-dimensional image features, so as to construct the logarithmic energy map feature based on wavelet packet, which further considers the spatial information between adjacent features, thereby optimizing the initial time-frequency domain features and improving the significance of the obtained features. On this basis, the residual network is used to improve the accuracy of fault diagnosis classification results. Finally, through the simulation verification of the standard rolling bearing data set of Case Western Reserve University, it can be seen that the fault diagnosis model constructed by the proposed method has high diagnostic accuracy and strong generalization ability.

A Time-Frequency Domain Analysis Method for Variable Frequency Hopping Signal.

A variable frequency hopping (VFH) signal is a kind of frequency hopping (FH) signal that varies both in frequency and dwell time. However, in radio surveillance, the existing methods for unidentified signals using VFH cannot be effectively handled. In this paper, we proposed an improved joint analysis method based on time-frequency domain features, which adopts multi-level processing to solve the time-frequency domain feature analysis problem of the VFH signal. First, the received signal is pre-processed by Short-Time Fourier Transform (STFT) and binarization, and a highly discriminative time-frequency image is obtained; then, the fixed frequency signal is removed based on the feature of connected domains, and the conventional frequency hopping (CFH) signal is removed by density-based spatial clustering of applications with noise (DBSCAN); finally, the overlapping region is cropped by the joint energy peak time-domain continuity properties. After the above multi-level joint processing method, the problem of VFH signal processing is effectively solved. The simulation result shows that the Mean Square Error (MSE) between the output results and the time-frequency image of the original VFH signal tends to be close to 0 when the Signal-to-Noise ratio (SNR) is 5 dB.

Semi-supervised sound event detection with dynamic convolution and confidence-aware mean teacher

Recently, sound event detection (SED) has made significant advancements through the application of deep learning, but there are still many difficulties and challenges to be addressed. One of the major challenges is the diversity of sound events, leading to substantial variations in time-frequency domain features. Additionally, most existing SED models can not effectively handle sound events of different scales, particularly those of short duration. Another challenge is the lack of well labeled dataset. The commonly used solution is mean teacher method, but inaccurate pseudo-labels could lead to confirmation bias and performance imbalance. In this paper, we introduce the multi-dimensional frequency dynamic convolution, which endows convolutional kernels with frequency-adaptive dynamic properties to enhance the feature representation capability. Moreover, we propose dual self attention pooling function to achieve more precise temporal localization. Finally, to solve the incorrect pseudo-labels problems, we propose the confidence-aware mean teacher to increase pseudo-labels accuracy and train the student model with high confidence labels. Experimental results on DCASE2017, DCASE2018 and DCASE2023 Task4 dataset validate the superior performance of proposed methods.

A lightweight framework for unsupervised anomalous sound detection based on selective learning of time-frequency domain features

For industrial anomalous sound detection (ASD), self-supervised methods have achieved significant detection performance in many cases. Nevertheless, these methods typically rely on the availability of external auxiliary information, and they may not work when such information are not feasible. Unsupervised methods do not leverage auxiliary information, whereas they usually obtained lower detection performance compared to self-supervised ones. Though some unsupervised methods have shown potential performance improvements, they are at the cost of complex implementation or large model sizes. As to the issues, this paper presents an unsupervised ASD method based on spectrogram frames selection (SFS) and AutoEncoder for Frequency-feature Selection (AEFS), called SFS-AEFS. First, SFS is developed based upon the temporal characteristics of machine sounds, which aims to select spectrogram frames (SFs) that contains the primary sound information while discarding the portions that are affected by noises or interferences or do not contain the target sound. Next, AEFS is developed by introducing a Scaling Gate (SG) after AE. For the selected SF features, AEFS aims to selectively enhance the mode learning of partial frequency dimensions and weaken the rest ones. Comparative experiments with the current ASD methods were made on the DCASE 2020 Challenge Task2 dataset. The related results demonstrate that our method achieved the best performance among all relevant unsupervised methods and is comparable to the current SOTA self-supervised methods. Moreover, our method is lightweight with model parameters being only 0.08MB.

TF-REF-RNN: Time-Frequency and Reference Signal Feature Fusion Recurrent Neural Network for Underwater Backscatter Signal Separation

Underwater wireless sensor networks play an important role in exploring the oceans as part of an integrated space–air–ground–ocean network. Because underwater energy is limited, the equipment’s efficiency is significantly impacted by the battery duration. Underwater backscatter technology does not require batteries and has significant potential in positioning, navigation, communication, and sensing due to its passive characteristics. However, underwater backscatter signals are susceptible to being swamped by the excitation signal. Additionally, the signals from different reflection signals share the same frequency and overlap, and contain fewer useful features, leading to significant challenges in detection. In order to solve the above problems, this paper proposes a recurrent neural network that introduces time-frequency and reference signal features for underwater backscatter signal separation (TF-REF-RNN). In the feature extraction part, we design an encoder that introduces time-frequency domain features to learn more about the frequency details. Additionally, to improve performance, we designed a separator that incorporates the reference signal’s pure global information features. The proposed TF-REF-RNN network model achieves metrics of 28.55 dB SI-SNRi and 19.51 dB SDRi in the dataset that includes shipsEar noise data and underwater simulated backscatter signals, outperforming similar classical methods.

Monitoring the gas metal arc additive manufacturing process using unsupervised machine learning

The study aimed to assess the performance of several unsupervised machine learning (ML) techniques in online anomaly (The term “anomaly” is used here to indicate a departure from expected process behavior which may indicate a quality issue which requires further investigation. The term “defect detection” has often been used previously but the specific imperfection is often indirectly inferred.) detection during surface tension transfer (STT)-based wire arc additive manufacturing. Recent advancements in quality monitoring for wire arc manufacturing were reviewed, followed by a comparison of unsupervised ML techniques using welding current and welding voltage data collected during a defect-free deposition process. Both time domain and frequency domain feature extraction techniques were applied and compared. Three analysis methodologies were adopted: ML algorithms such as isolation forest, local outlier factor, and one-class support vector machine. The results highlight that incorporating frequency analysis, such as fast Fourier transform (FFT) and discrete wavelet transform (DWT), for feature extraction based on general frequency response and defined bandwidth frequency response, significantly improves performance, reflected in a 14% increase in F2 score, compared with time-domain features extraction. Additionally, a deep learning approach employing a convolutional autoencoder (CAE) demonstrated superior performance by processing time-frequency domain data stored as spectrograms obtained through short-time Fourier transform (STFT) analysis. The CAE method outperformed frequency domain analysis and traditional ML approaches, achieving an additional 5% improvement in F2-score. Notably, the F2-score (The F2 score is the weighted harmonic mean of the precision and recall (given a threshold value). Unlike the F1 score, which gives equal weight to precision and recall, the F2 score gives more weight to recall than to precision.) increased significantly from 0.78 in time domain analysis to 0.895 in time-frequency analysis. The study emphasizes the potential of utilizing low-cost sensors to develop anomaly detection modules with enhanced accuracy. These findings underscore the importance of incorporating advanced data processing techniques in wire arc additive manufacturing for improved quality control and process optimization.

Remaining useful life prediction of rolling bearings based on time convolutional network and transformer in parallel

Abstract Remaining useful life prediction (RUL) is crucial for maintaining the reliability and safety of industrial equipment. Recently, the Transformer has been widely used due to its ability to effectively extract global feature information in rolling bearing RUL prediction. However, the Transformer is weak in acquiring local feature information and cannot extract temporal features from the degradation process. Conversely, a temporal convolutional network (TCN) can effectively extract local features but is weak in global feature extraction. Therefore, to address the above problems, this paper proposes a prediction method based on the parallel combination of TCN and Transformer. The method first extracts the time domain, frequency domain, and time-frequency domain features from the original vibration signals. After screening through the evaluation indices, the features are fused using K-means clustering and principal component analysis (PCA) to construct health indicators (HI) that characterize the degradation of rolling bearings. Then, a TCN-Transformer parallel network model is constructed to extract both local and global features, and a feed-forward neural network (FNN) is used for the prediction of RUL. Finally, experimental validation is carried out on the PHM2012 bearing dataset, and the results show that the method achieves higher RUL prediction accuracy compared to other methods.

Nonlinear damage dynamics modeling and characteristics simulation of bevel gear set

Abstract As a key core device for power transmission in the main reducer of a helicopter, the bevel gear set affects the flight performance of the helicopter directly. A nonlinear damage dynamic model of bevel gear set was constructed using the centralized parameter method. The inherent characteristics and nonlinear behavior characteristics of bevel gear sets were studied, and the time-frequency domain characteristics under different health states were simulated. The fault response transmission law and fault mechanism of the system were deeply revealed, which provides guidance and support for the optimization design of helicopter transmission system and the diagnosis algorithms development.

The supernumerary robotic limbs of brain-computer interface based on asynchronous steady-state visual evoked potential

Brain-computer interface (BCI) based on steady-state visual evoked potential (SSVEP) have attracted much attention in the field of intelligent robotics. Traditional SSVEP-based BCI systems mostly use synchronized triggers without identifying whether the user is in the control or non-control state, resulting in a system that lacks autonomous control capability. Therefore, this paper proposed a SSVEP asynchronous state recognition method, which constructs an asynchronous state recognition model by fusing multiple time-frequency domain features of electroencephalographic (EEG) signals and combining with a linear discriminant analysis (LDA) to improve the accuracy of SSVEP asynchronous state recognition. Furthermore, addressing the control needs of disabled individuals in multitasking scenarios, a brain-machine fusion system based on SSVEP-BCI asynchronous cooperative control was developed. This system enabled the collaborative control of wearable manipulator and robotic arm, where the robotic arm acts as a "third hand", offering significant advantages in complex environments. The experimental results showed that using the SSVEP asynchronous control algorithm and brain-computer fusion system proposed in this paper could assist users to complete multitasking cooperative operations. The average accuracy of user intent recognition in online control experiments was 93.0%, which provides a theoretical and practical basis for the practical application of the asynchronous SSVEP-BCI system.

Novel GIL mechanical fault diagnosis method based on multi-sensor data feature fusion and TDEAVOA-ELM

The gas insulated transmission line (GIL), a crucial component of the power system, faces challenges in mechanical fault diagnosis including difficulties in quantifying sensor signal selection, balancing recognition accuracy, and model training speed. This paper proposes a GIL mechanical fault diagnosis method based on the feature fusion of multi-channel vibration sensor signals and an improved optimization extreme learning machine (ELM). Initially, three optimal signals are selected based on the correlation coefficients of multiple signals, and their time–frequency domain features are extracted. These features are then fused using neighborhood preserving embedding (NPE) and input into the improved optimized ELM model. The enhanced optimization algorithm reduces the likelihood of falling into local optima, thereby improving the model’s recognition performance. Experimental results demonstrate that this method effectively enhances the recognition accuracy of mechanical faults in GIL equipment and enables rapid diagnosis.