Browsing by Author "Eren, Levent"

Now showing 1 - 2 of 2

Citation - Scopus: 1
Improved Detection of Broken Rotor Bars by 1-D Self-Onns
(IEEE, 2022) Eren, Levent; Devecioglu, Ozer Can; Ince, Turker; Askar, Murat
Recently, machine learning techniques have been increasingly applied to the detection of both mechanical and electrical faults in induction motors. Broken rotor bars are one of the most common fault types that seriously affect the efficiency and lifetime of induction motors. In this study, compact 1-D self-organized operational neural networks (Self-ONNs) are applied to improve the detection and classification of broken rotor bars in induction motors. 1-D convolutional neural networks (CNNs) are a special case of Self-ONNs and they are usually preferred to traditional fault diagnosis systems with separately designed feature extraction and classification blocks as they provide cost-effective and practical hardware implementation. The proposed system improves the detection and classification performance of 1-D CNNs while still providing similar advantages and preserving real-time computational ability.
Citation - Scopus: 7
Improved Domain Adaptation Approach for Bearing Fault Diagnosis
(IEEE, 2022) Ince, Turker; Kilickaya, Sertac; Eren, Levent; Devecioglu, Ozer Can; Kiranyaz, Serkan; Gabbouj, Moncef
Application of domain adaptation techniques to predictive maintenance of modern electric rotating machinery (RM) has significant potential with the goal of transferring or adaptation of a fault diagnosis model developed for one machine to be generalized on new machines and/or new working conditions. The generalized nonlinear extension of conventional convolutional neural networks (CNNs), the self-organized operational neural networks (Self-ONNs) are known to enhance the learning capability of CNN by introducing non-linear neuron models and further heterogeneity in the network configuration. In this study, first the state-of-the-art 1D CNNs and Self-ONNs are tested for cross-domain performance. Then, we propose to utilize Self-ONNs as feature extractor in the well-known domain-adversarial neural networks (DANN) to enhance its domain adaptation performance. Experimental results over the benchmark Case Western Reserve University (CWRU) real vibration data set for bearing fault diagnosis across different load domains demonstrate the effectiveness and feasibility of the proposed domain adaptation approach with similar computational complexity.