RESUMEN
This paper proposes an intelligent framework for the fault diagnosis of centrifugal pumps (CPs) based on wavelet coherence analysis (WCA) and deep learning (DL). The fault-related impulses in the CP vibration signal are often attenuated due to the background interference noises, thus affecting the sensitivity of the traditional statistical features towards faults. Furthermore, extracting health-sensitive information from the vibration signal needs human expertise and background knowledge. To extract CP health-sensitive features autonomously from the vibration signals, the proposed approach initially selects a healthy baseline signal. The wavelet coherence analysis is then computed between the healthy baseline signal and the signal obtained from a CP under different operating conditions, yielding coherograms. WCA is a signal processing technique that is used to measure the degree of linear correlation between two signals as a function of frequency. The coherograms carry information about the CP vulnerability towards the faults as the color intensity in the coherograms changes according to the change in CP health conditions. To utilize the changes in the coherograms due to the health conditions of the CP, they are provided to a Convolution Neural Network (CNN) and a Convolution Autoencoder (CAE) for the extraction of discriminant CP health-sensitive information autonomously. The CAE extracts global variations from the coherograms, and the CNN extracts local variations related to CP health. This information is combined into a single latent space vector. To identify the health conditions of the CP, the latent space vector is classified using an Artificial Neural Network (ANN). The proposed method identifies faults in the CP with higher accuracy as compared to already existing methods when it is tested on the vibration signals acquired from real-world industrial CPs.
RESUMEN
Diagnostics of mechanical problems in manufacturing systems are essential to maintaining safety and minimizing expenditures. In this study, an intelligent fault classification model that combines a signal-to-image encoding technique and a convolution neural network (CNN) with the motor-current signal is proposed to classify bearing faults. In the beginning, we split the dataset into four parts, considering the operating conditions. Then, the original signal is segmented into multiple samples, and we apply the Gramian angular field (GAF) algorithm on each sample to generate two-dimensional (2-D) images, which also converts the time-series signals into polar coordinates. The image conversion technique eliminates the requirement of manual feature extraction and creates a distinct pattern for individual fault signatures. Finally, the resultant image dataset is used to design and train a 2-layer deep CNN model that can extract high-level features from multiple images to classify fault conditions. For all the experiments that were conducted on different operating conditions, the proposed method shows a high classification accuracy of more than 99% and proves that the GAF can efficiently preserve the fault characteristics from the current signal. Three built-in CNN structures were also applied to classify the images, but the simple structure of a 2-layer CNN proved to be sufficient in terms of classification results and computational time. Finally, we compare the experimental results from the proposed diagnostic framework with some state-of-the-art diagnostic techniques and previously published works to validate its superiority under inconsistent working conditions. The results verify that the proposed method based on motor-current signal analysis is a good approach for bearing fault classification in terms of classification accuracy and other evaluation parameters.
Asunto(s)
Algoritmos , Redes Neurales de la ComputaciónRESUMEN
In the machine learning and data science pipelines, feature extraction is considered the most crucial component according to researchers, where generating a discriminative feature matrix is the utmost challenging task to achieve high classification accuracy. Generally, the classical feature extraction techniques are sensitive to the noisy component of the signal and need more time for training. To deal with these issues, a comparatively new feature extraction technique, referred to as a wavelet scattering transform (WST) is utilized, and incorporated with ML classifiers to design a framework for bearing fault classification in this paper. The WST is a knowledge-based technique, and the structure is similar to the convolution neural network. This technique provides low-variance features of real-valued signals, which are usually necessary for classification tasks. These signals are resistant to signal deformation and preserve information at high frequencies. The current signal data from a publicly available dataset for three different bearing conditions are considered. By combining the scattering path coefficients, the decomposition coefficients from the 0th and 1st layers are considered as features. The experimental results demonstrate that WST-based features, when used with ensemble ML algorithms, could achieve more than 99% classification accuracy. The performance of ANN models with these features is similar. This work exhibits that utilizing WST coefficients for the motor current signal as features can improve the bearing fault classification accuracy when compared to other feature extraction approaches such as empirical wavelet transform (EWT), information fusion (IF), and wavelet packet decomposition (WPD). Thus, our proposed approach can be considered as an effective classification method for the fault diagnosis of rotating machinery.
Asunto(s)
Redes Neurales de la Computación , Análisis de Ondículas , Aprendizaje Automático , AlgoritmosRESUMEN
Bearings are nonlinear systems that can be used in several industrial applications. In this study, the combination of a strict-feedback backstepping digital twin and machine learning algorithm was developed for bearing crack type/size diagnosis. Acoustic emission sensors were used to collect normal and abnormal data for various crack sizes and motor speeds. The proposed method has three main steps. In the first step, the strict-feedback backstepping digital twin is designed for acoustic emission signal modeling and estimation. After that, the acoustic emission residual signal is generated. Finally, a support vector machine is recommended for crack type/size classification. The proposed digital twin is presented in two steps, (a) AE signal modeling and (b) AE signal estimation. The AE signal in normal conditions is modeled using an autoregressive technique, the Laguerre algorithm, a support vector regression technique and a Gaussian process regression procedure. To design the proposed digital twin, a strict-feedback backstepping observer, an integral term, a support vector regression and a fuzzy logic algorithm are suggested for AE signal estimation. The Ulsan Industrial Artificial Intelligence (UIAI) Lab's bearing dataset was used to test the efficiency of the combined strict-feedback backstepping digital twin and machine learning technique for bearing crack type/size diagnosis. The average accuracies of the crack type diagnosis and crack size diagnosis of acoustic emission signals for the bearings used in the proposed algorithm were 97.13% and 96.9%, respectively.
Asunto(s)
Inteligencia Artificial , Lógica Difusa , Algoritmos , Retroalimentación , Aprendizaje AutomáticoRESUMEN
Estimation of the remaining useful life (RUL) of bearings is important to avoid abrupt shutdowns in rotary machines. An important task in RUL estimation is the construction of a suitable health indicator (HI) to infer the bearing condition. Conventional health indicators rely on features of the vibration acceleration signal and are predominantly calculated without considering its non-stationary nature. This often results in an HI with a trend that is difficult to model, as well as random fluctuations and poor correlation with bearing degradation. Therefore, this paper presents a method for constructing a bearing's HI by considering the non-stationarity of the vibration acceleration signals. The proposed method employs the discrete wavelet packet transform (DWPT) to decompose the raw signal into different sub-bands. The HI is extracted from each sub-band signal, smoothened using locally weighted regression, and evaluated using a gradient-based method. The HIs showing the best trends among all the sub-bands are iteratively accumulated to construct an HI with the best trend over the entire life of the bearing. The proposed method is tested on two benchmark bearing datasets. The results show that the proposed method yields an HI that correlates well with bearing degradation and is relatively easy to model.
RESUMEN
In recent years, stress analysis by using electro-encephalography (EEG) signals incorporating machine learning techniques has emerged as an important area of research. EEG signals are one of the most important means of indirectly measuring the state of the brain. The existing stress algorithms lack efficient feature selection techniques to improve the performance of a subsequent classifier. In this paper, genetic algorithm (GA)-based feature selection and k-nearest neighbor (k-NN) classifier are used to identify stress in human beings by analyzing electro-encephalography (EEG) signals. GA is incorporated in the stress analysis pipeline to effectively select subset of features that are suitable to enhance the performance of the k-NN classifier. The performance of the proposed method is evaluated using the Database for Emotion Analysis using Physiological Signals (DEAP), which is a public EEG dataset. A feature set is extracted in 32 EEG channels, which consists of statistical features, Hjorth parameters, band power, and frontal alpha asymmetry. The selected features through GA are used as input to the k-NN classifier to distinguish whether each EEG datapoint represents a stress state. To further consolidate, the effectiveness of the proposed method is compared with that of a state-of-the-art principle component analysis (PCA) method. Experimental results show that the proposed GA-based method outperforms PCA, with GA demonstrating 71.76% classification accuracy compared with 65.3% for PCA. Thus, it can be concluded that the proposed method can be effectively used for stress analysis with high classification accuracy.