A novel artificial neural network based selection harmonic reduction technique for single source fed high gain switched capacitor coupled multilevel inverter for renewable energy applications.

Multilevel inverters (MLIs) are commonly used in renewable energy systems for their high-quality output, low total harmonic distortion (THD), and reduced component count. This study presents a high-gain, single-source MLI designed for renewable applications like solar or wind power. It features a novel topology with twice the voltage-boosting factor, utilizing a single DC source. The inverter achieves thirteen voltage levels using just 10 power switches and three switched capacitors. The voltage gain is achieved without the need for bulky DC-DC converters or transformers. This is accomplished by configuring the switched capacitors in series and parallel arrangements to attain the desired voltage boost. Additionally, the self-balancing capacitors eliminate the need for extra sensors. Both symmetric and asymmetric variants of the extensible configuration are investigated. The suggested design lowers the total standing voltage (TSV) while achieving high gain. A selective harmonic removal technique using artificial neural networks (ANN) reduces THD by up to 6.07 %. An extensive review of recent literature reveals significant advancements and applications of ANNs in this field. The proposed system's benefits, such as gain factor, total standing voltage (TSV), and minimized device count, are assessed. Comparative analysis reveals that the proposed topology employs fewer components and features a more simplified design. Additionally, the inverter achieves an efficiency of 96.9 %. The design is validated through an experimental prototype after being confirmed with MATLAB/SIMULINK.© YEAR The Authors. Published by Elsevier Ltd. Peer-review under responsibility of the scientific committee of the Name of the Conference, Conference Organizer Name, Year or Edition of Conference.