*Corresponding Author: deepprakashsinghvbspu@gmail.com 398 DOI: https://doi.org/10.52756/ijerr.2023.v32.035 Int. J. Exp. Res. Rev., Vol. 32: 398-407 (2023) Precision fault prediction in motor bearings with feature selection and deep learning Deep Prakash Singh * and Sandip Kumar Singh Department of Mechanical Engineering, VBS Purvanchal University, Jaunpur, UP, India E-mail/Orcid Id: DPS, deepprakashsinghvbspu@gmail.com, https://orcid.org/0009-0008-8931-0393; SKS, sandipkumarsingh25@gmail.com Introduction Natural neural networks are a technique in deep learning. Due to the non-invasive, affordable and helpful nature of motor performance, impulses are frequently employed as input for motor sensors (Yuan and He, 2014). Regarding the motor sensor, some observations calculate similarities for real movements compared to motor performance. The potential values of the device are changed after some activity. In various terms, motor sensors focus on different physical activities as physical motion (Bonassi et al., 2017; Munzert et al., 2009; Coyle et al., 2015; Anderson and Lenz, 2011; Phothisonothai and Nakagawa, 2008). In general, the steps in this procedure are as follows: Prior to removing undesired frequency ranges pre- processing is employed. To finish feature extraction, several mapping models are constructed for distinct feature categories. The various feature models are then categorized and decoded separately. The conventional approach necessitates many steps. The processing results will be affected if there is a requirement to rectify an intermediate step. The challenge of pattern recognition arises from the inherent complexity of analysing the minute and intertwined nature of the gathered motor sensor data (Schlögl et al., 2010; Song et al., 2013; Vidaurre et al., 2007; Woehrle et al., 2015; Duan et al., 2019). Deep neural networks (DNNs) have recently shown that they can successfully classify linguistic data, pictures, sounds, and natural texts. There are several benefits of using neural networks to decode motor sensor data. Nevertheless, due to limitations on the number of available participants, the experiment's length, and the technique's complexity, it is challenging to gather enough data for practical applications. The quantity of samples has a substantial impact on how effectively DNNs work. When training a model, small-scale datasets typically result in poor generalizability, negatively impacting classification accuracy (Song et al., 2017; Alom et al., 2018; Zhong et al., 2015; Cooney et al., 2019). We evaluate CNN, RBFN, and FFNN in this post using a variety of characteristics. The results show that Article History: Received: 18 th April., 2023 Accepted: 22 nd Aug., 2023 Published: 30 th Aug., 2023 Abstract: In the disciplines of industrial machinery, mechanical engineering is beneficial to recognize motor performance for motors with HP power, torque transducer, dynamometer, and control electronics. The motivation is to address the need for more accurate and efficient fault prediction in machinery to prevent breakdowns, reduce maintenance costs, and improve overall reliability. In this work, deep learning classifiers used to classify ball defect inner race fault, outer race fault and normal motor performance in testing. With the aid of three distinct classifiers CNN, FFNN, and RBN; these suggested relative characteristics are assessed. In comparison to other current algorithms, the suggested methodology for classifying motor performance achieved maximum accuracy in each CNN test at 95.4% and 97.7%. The correlation and chi-square algorithms are used to find out the added characteristics and rank of features. The correlation technique provides relations between attributes, and the chi-square offers the optimal balance between precision and feature space. We discovered that the performance is enhanced overall by relative power characteristics. The suggested models might offer rapid responses with less complexity. Keywords: Convolutional Neural Network, Feed Forward Neural Network and Radial Based Network, Correlation and Chi- Square.