تشخیص خطا مبتنی برسیگنال‌های صوتی فن‌های تهویه با استفاده از الگوریتم های یادگیری ماشین

This paper introduces a practical approach for fault diagnosis of Ventilation Fans using au-dio signals captured through laboratory microphone, aimed at enhancing efficiency while minimizing costs in industrial settings and personal ones. Ventilation Fans are important in industries, labs and personal use. Our research employs machine learning classification al-gorithms to process audio signals, eliminating the need for specialized devices or expensive sensors and making the methodology accessible through common microphone. We system-atically collected data from ventilation fans in three distinct operational states and extracted features from the audio signals for analysis. Classification models, including k-Nearest Neighbors (KNN), logistic regression and decision trees. This study contributes to the field of fault diagnosis by offering a practical and accessible methodology for identifying machines faults using laboratory microphone -captured audio signals. The findings underscore the efficacy of machine learning techniques in industrial and personal applications and provide valuable insights for optimizing machines performance while minimizing costs.