Generic Stacked Ensemble Health Classifier for Predictive maintenance in Electric Motors

作者：S. S, Jothi N T, Muruga Lal Jeyan J V, Karthikeyan Selvaraj · 发表于：Proceedings of the 7th International Conference on Information Management & Machine Intelligence · 年份：2025 · DOI：10.1145/3793449.3793457

Prognostics and Health Management (PHM) is gaining rapid momentum in Industry 4.0. The pivotal focus of PHM is early detection of the deviated operating profiles of the machinery by Condition Monitoring (CM) eventually preventing the disruptive faults and failures. The abundance of data and rampant progress in Machine Learning (ML) and Deep Learning (DL) are aspiring factors in developing data driven PHM systems. This work focuses on developing an efficient fault prognosis method by constructing a fresh Generic Stacked Ensemble Health (GSEH) Classifier by integrating the predictive prowess of individual ML learners. Random Forest model is stacked to aggregate the results. In addition to this, the work uncovers important domain specific relations within the data. The base learners used in the model can explore the data space in their unique way, thus eliminating the need for feature selection mechanism. The proposed model is validated on AI4I 2020 Predictive Maintenance Dataset, which is a synthetic data acquired from electric motors that are more Computing methodologies susceptible to various electrical and mechanical failures.