Selective harmonic elimination in multi-level inverters by using neural networks

In this study, it is aimed to eliminate the harmonics selected by selective harmonic elimination (SHE) method in a 7-level cascade multilevel inverter using artificial neural networks (ANNs). A control algorithm has been developed which 3rd and 5th or 7th can be eliminated according selection while adjusting output voltage amplitude of inverter. The required switching angles for SHE are calculated real time ANN. These were first obtained offline training ANN Newton-Raphson method. was trained MATLAB® environment data. resulting practical implementation STM32F429 ARM microcontroller® provided. Experimental results system with RL load tested.