Transformada Wavelet para detecção de barras quebradas em Motores de Indução Trifásicos

Abstract

Electric machines account for 95% of the primary source of movement of industrialized nations. Out of these, 90% are three-phase induction motors (TIM), which are present in almost all industrial sectors. Due to its relevance in the productive sector it is important t that they are properly monitored in order to avoid interruptions of production and operational disasters. The aim of this work is to study and develop a system to diagnose the failure of broken bars in a three - phase induction motor, and to analyze the mechanical (vibration) and electric variables (current, air gap torque and energy efficiency). Specifically, we wanted the evaluation of the expressiveness of the signals in the presence of broken bar. We used the Fourier and wavelet transforms as analysis tools. The system should be able to track the severity of the failure at 3PIM as long as the bars are punctured (broken) prior to the total breakdown of the equipment. The installation of the system in the engine must be totally non-invasive and it can be performed with the engine running. Experimental results are presented to observe the proposed strategies, evaluating the efficiency of the detection of faults in a wide load range of the machine with direct drive of the electric network.