High-Performance Hybrid Cascaded Modular Multilevel Inverter with Reduced Switch and Conduction Losses Featuring Inherent DC-Link Voltage Balancing for Renewable Energy-Based Smart Grid Applications

Cascaded Module Multilevel Inverter (CMMI) minimizes power losses and maintains DC-link voltage balance, improving efficiency and reliability for smart grid applications. However, the high-performance hybrid CMMI requires precise coordination among its modules to ensure effective voltage balancing and reduced losses. This work proposes a novel method that combines the Opposition-based Banyan Tree Growth Optimization (OBTGO) with a Deep Dendritic Artificial Neural Network (DDANN) for optimal control of hybrid in 13 level cascaded modular multilevel inverters in renewable smart grids. The proposed method's primary objectives are to lower switching losses and Total Harmonic Distortion (THD), which will enhance power quality by utilizing adaptive gate pulse generation and optimized switching state prediction. The inverter's switching states are predicted by the DDANN, and the gate pulses for level generator switches are optimized by the OBTGO. The proposed technique is assessed and related to other existing techniques by MATLAB model. The proposed method demonstrates better performance than existing techniques, including Recalling-Enhanced Recurrent Neural Network (RERNN), Feed forward neural network (FNN) and mother optimization (MO). The proposed technique achieves a higher efficiency of 95%, a reduced THD of 1.2%, and a lower computation time of 0.22 seconds, indicating improved performance compared to existing methods. The proposed OBTGO-DDANN method effectively optimizes control of 13-level cascaded modular multilevel inverters, minimizing THD and switching losses to develop power quality (PQ) and efficiency in renewable smart grids