An optimized controller for novel asymmetrical 51-level inverter in hybrid solar and wind turbine-connected power system

The necessity of lower carbon emissions with uninterruptable power generation is the driving force for the invention of renewable energy sources (RES). In most cases, solar and wind turbine-connected power systems are developed to satisfy the escalating power demand. Variations in renewable energy sources caused by climatic changes can be predicted but not controlled. Thus, the adaptive neuro fuzzy inference system (ANFIS)-based maximum power tracking controller is used. In this proposed work, the voltage and current from solar and wind sources are sent to the ANFIS controller that regulates optimum switching pulses to the DC–DC converter. Moreover, the complexities are increased during power conversion through multi-level inverters (MLI). One of the major issues in the MLI is the higher switch count and DC sources. Hence, the walruses optimized deep belief neural network (WOA-DBN)-controlled 51-level inverter is proposed in this work. The proposed inverter has used only 13 switches and 5 DC voltage sources. The performance of the proposed methodology is verified under varying and constant inputs to the solar and wind sources in the MATLAB/SIMULINK tool. The experimental validation is established using HIOKI PW3337 power meter to validate the proposed method. The results showed that the proposed inverter has 0.75% total harmonic distortion (THD), lower than the existing 51-level inverter topologies. The comparison of output voltage from both simulation and experimental result is obtained equally, i.e. 400 V, and output current is 20 A.