Optimization of an Asymmetrical 5-level Cascaded H-Bridge Inverter Modulation by Evaluation of Dead Times Impact

Artificial intelligence, born after neural networks, finds application in different fields of science. The algorithms for evaluating the best solutions can be applied in different aspects of power quality. Specifically, the paper presents the possibility of using asymmetric voltages, due for example to strings of different photovoltaic panels, and emulating the behavior of a 7-level staircase with a 5 level CHBMLI. The optimization algorithm exploits the command of the insertion of the inverter legs to eliminate or mythologize the harmonics that can be injected into the network. Attention is placed on dead times, essential to avoid short circuits and makes use of the black widow technique.