سيدجواد ميرمحمدميگوني

Given the widespread application of induction motors an‎d their critical role in industry, condition monitoring an‎d fault detection in these motors are of paramount importance. Among all faults associated with induction motors, mechanical faults account for the majority of occurrences. Consequently, various methods have been developed an‎d employed to detect each of these faults. However, a major limitation of most existing approaches is that they are often non-automated an‎d rely heavily on expert experience. Furthermore, in automated methods, not all types of faults are comprehensively addressed, an‎d fault detection is typically performed in a case-specific manner. In this study, a comprehensive an‎d intelligent fault detection method for three-phase induction motors is proposed, based on a deep two-dimensional convolutional neural network (2D-CNN). The proposed method is characterized by several key features, including the ability to detect mechanical faults in addition to an electrical fault (broken rotor bar), utilization of raw three-phase motor current signals along with vibration signals, minimal use of mathematical transformations for data validation, an‎d achieving high accuracy in fault identification an‎d classification. The proposed approach has been eva‎luated using a laboratory test bench comprising a 1.5 kW three-phase induction motor under various load conditions. The obtained results demonstrate the effectiveness of the method in detecting the aforementioned faults.

موتور القايي

Induction Motor

Seyedjavad Mirmohammad Meiguni

Abolfazl Vahedi