Extreme Learning Machine Optimization based on Hippopotamus Optimization Algorithm for Gear Fault Diagnosis

• Published in: Journal of physics. Conference series Vol. 2933; no. 1; pp. 12019 - 12028
• Main Authors: Amirulaminnur, R., Firdaus Isham, M., Harith, M. K., Saufi, M. S. R., Saad, W. A. A., Hasan, M. D. A., Talib, M. H. A.
• Format: Journal Article
• Language: English
• Published: Bristol IOP Publishing 01.01.2025
• Subjects: Algorithms; Component reliability; Downtime; ELM; Fault Diagnosis; Gear; Machine learning; Optimization; Optimization algorithms; Sustainability
• ISSN: 1742-6588; 1742-6596; 1742-6596
• DOI: 10.1088/1742-6596/2933/1/012019

It is crucial to ensure the dependability, reliability, and sustainability of machines to optimize industrial productivity and efficiency. Any malfunction or breakdown of machine components or mechanical equipment may result in unexpected downtime and financial losses. This study presents a maintenance strategy for mechanical equipment, primarily focusing on a gear failure diagnosis approach using an extreme learning machine optimization based on the hippopotamus optimization algorithm (HO). The proposed method was evaluated using sets of gear vibration signals obtained from an online database, that included both healthy and malfunctioning data. The HO approach was employed to determine an optimal parameter for the ELM method, specifically the number of neurones, input weight, and bias range values. The findings indicate that the proposed approach enhances the classification efficacy of ELM by 12% compared to traditional ELM. The proposed strategy can be applied in any relevant industry to improve the sustainability and dependability of its plants.