Abstract: Growing industrial need makes the choice of fast response, accurate and efficient systems. The Multilevel inverter based Induction Drives (MLID) are the best solution for the industrial drive needs which reduces the harmonics and increases the efficiency of the system. And also the need for hybrid electric vehicles increases the need of an efficient traction system with the use of multilevel inverters. As multilevel inverter has many semi-conductor switches, it is difficult to identify the fault in it. In this study, a new fault diagnosis method by using the Total Harmonic Distortion (THD) of the voltage waveform and to classify the fault, Neural Network (NN) trained with back propagation, Genetic Algorithm (GA) and also with Particle Swarm Optimization (PSO) is applied and the results are compared. Results shows that NN trained with PSO gives fast response in training algorithm when compared to BP and GA. Here, a cascaded multi-level inverter based three-phase induction motor drive is taken as the test system. MATLAB Software is used to analyse the effectiveness of the test system and results are tabulated.
M. Sivakumar and R.M.S. Parvathi, 2014. Application of Neural Network Trained with Meta-Heuristic Algorithms on Fault Diagnosis of Multi-Level Inverte. Research Journal of Applied Sciences, 9: 369-375.