A New Centroid Based Fault Detection Method for 3-Phase Inverter-Fed Induction Motors

A novel, simpler, and non-invasive fault detection and data reduction method for inverter-fed induction motors is developed in this paper. Using a centroid determination method, location and type of the fault in a three-phase system, specifically in an induction motor and the PWM current-controlled inverter that feeds it, is identified. MATLAB simulations and the DSpace DSP based experiments are discussed in detail to prove the effectiveness of the proposed new method. The fault monitoring algorithms use pattern symmetry across the positive and negative alpha-beta axes after the three-phase currents are transformed from a-b-c to alpha-beta plane using Concordia transform. This allows detection method to view motor and inverter AC currents in the time independent realm by using cycle-by-cycle single point symmetry of the current spectrum. The proposed method detects a variety of probable faults in real-time and requires reduced computational efforts. Therefore, it is simpler, cost-effective, and free from hardware complexity associated with traditional fault diagnosis methods