Rolling bearing fault diagnostic system using fuzzy logic

Reducing maintenance costs is extremely important to raise industrial competitiveness. In this context, methodologies based on predictive maintenance are a need for optimization of plant systems diagnosis in order to increase accuracy and to reduce human errors. Diagnosis process automation directly results in improved reliability for taking decisions. The present work describes an automatic diagnosis system for detection and classification defects in rolling bearings using fuzzy logic. The designed system was developed to be able to classify three types of pre-established defects in rolling bearings operating under several shaft speeds and load conditions. The measured vibratory signals were analyzed by spectral and statistical techniques. The results demonstrate that the system is an excellent tool for identifying and classifying bearing faults.