Detection and diagnosis of induction motor bearing faults using multiwavelet transform and naive Bayes classifier

Abstract

Summary A novel framework is proposed for the classification of motor bearing faults by using multiwavelet transform and naive Bayes classifier. This work has explored the application of multiwavelet transform to the vibration signatures for extracting the effective fault features. Geronimo‐Hardin‐Massopust multiwavelet filter bank has been employed in this work for multiresolution analysis up to fourth decomposition level. This work has relied upon multiwavelets as the multiwavelets are enriched with the properties which wavelets do not offer simultaneously like compact support and orthogonality without compromising with symmetricity and higher order of vanishing moments. After that, fault features have been extracted from the decomposed vibration signals to utilize the fault information present at different resolution levels. The fault classification has been accomplished with high success rate by using the extracted fault features with the naive Bayes classifier due to its inherent features like robustness to noise, simplicity, and quick convergence even with less training data. The efficacy of the proposed classifier has been verified with both the offline signals and the real‐time signals having different fault severity levels. The outperformance of the proposed classifier has been shown by comparing the classification results with other established supervised classification techniques like probabilistic neural network, Gaussian‐support vector machine, polynomial‐support vector machine, and linear‐support vector machine.

The impact of engineering students' performance in the first three years on their graduation result using educational data mining

“…The discovery of hidden information is achieved by running data mining algorithms that combine statistics with computer science to mine valuable information from a seemingly meaningless data jumble. Data mining can be applied in various fields such as engineering (Adekitan et al., 2019; Saini and Aggarwal, 2018), business management (Zuo et al., 2016), marketing and product design (Jin et al., 2019), computer science (Mahendra et al., 2019), education (Ibrahim et al., 2019; Porouhan, 2018), genetics (Noreña et al., 2018), biological studies (Gu et al., 2018), facility maintenance management (Miguel-Cruz et al., 2019), health and drug development studies (Keserci et al., 2017), chemistry and toxicity analysis (Saini and Srivastava, 2019), meteorology (Kovalchuk et al., 2019), transportation safety (Divya et al., 2019) and traffic management (Amiruzzaman, 2019), fraud detection (Vardhani et al., 2019), and so forth. In the educational sector, volumes of data are daily generated from various teaching and learning activities within an institution.…”

Determining the operational status of a three phase induction motor using a predictive data mining model

“…Also, an online fault detection and performance evaluation simulation was developed [30] using the phase currents, the voltage and the motor speed for assessment. Likewise, the feasibility of using naive bayes data mining algorithm for identification and classification of motor bearing faults was demonstrated [31], while in the study [32] fuzzy logic was applied for identifying short and open circuit TPIM faults.…”

Classification of Fault Severity in Induction Machine Systems Based on Temporal Convolutions and Recurrent Networks

“…However, in the initial stages of electrical and mechanical faults and different load levels, the extracted features used for signal processing techniques such as time domain, frequency domain, and time scale cannot show the severity of fault properly [8]. Fault detection and severity identification based on machine learning methods have recently been introduced [9].…”

“…ese data-driven approaches can be regarded as time-series classification (TSC) tasks. e major problem in the fault detection process is related to the interclass variability caused by the different unknown load levels and severity of faults, which reduces the accuracy of the process [9].…”

“…ese algorithms strongly rely on the extracted features [13]. e feature-based learning methods utilize signal processing algorithms to extract features from the input signals, and consequently, the severity in different load levels is diagnosed by the classification of the extracted features [9]. Different signal processing methods based on frequency-domain, time-domain, and time-scale methods are used for the signal processing step.…”