New intelligent system for proactive bearing fault detection in electrical induction machines

In the current state of electromechanical induction machine monitoring, the systems developed are more or less generalist systems. These systems, instead of focusing on one particular type of fault, in their implementations, are listening to any type of fault that may occur. We are not saying that these systems are not efficient. However, when several faults occur at the same time, the signal to be processed is immediately overloaded, as would be the case for the analysis of the stator current of an induction machine. Thus, its primary task, which was, for example, to identify a particular fault, is blinded. From then on, we realize that we have to process the signal case by case. And even if it were designed to detect more than one fault, such a system would require more computation and decision-making time, thus distorting its nature as a real-time system and, this should not be neglected because of the damage that could occur just because of a small processing delay. So quickly we see the necessity and importance of focusing on a single type of fault. The importance is not only the time saved but it reduces the effort of other systems installed in the same machine unit. Our present study proposes an approach that, by its complexity and efficiency, would allow to monitor locally or remotely and in real time the mechanical fatigue of bearings used in rotating machines and this, by developing a full-fledged system able to detect in real time and with a good accuracy the bearing defect within the electromechanical induction machines widely employed in various fields of our life. The fusion of a powerful signal processing algorithm, the concept of fuzzy logic, a vibration sensor and the exploitation of the current induced in a coil show through the results obtained using the MATLAB software, the reliability and the efficiency of the proposed system in the early detection of the bearing defect.