Hybrid Deep Learning Approaches for Predictive Modeling of Aluminium Gallium Nitride with Rime Optimization Algorithm

High electron mobility transistor (HEMT) derived from gallium nitride (GaN) is a promising candidate for the upcoming generation of high-power and higher-frequency electronic applications. GaN power semiconductor devices have extraordinary belongings in comparison with their Silicon Carbide (SiC) and Silicon (Si) complements and possess the latent to substitute the ubiquitous Si power device. However, the main technical limitations are preventing their active usage through the power electronics communities. These are performing as an interference to the commercial usage of GaN- based electronic methods. GaN HEMTs are almost identical to conventional field-effect transistors (FET). Similar to FET it contains additionally 3 terminals are drain, gate, and source which demonstrate the subsection of voltage-controlled Power switches. The variance is, in HEMT we are applying the heterojunction in between aluminum gallium nitride (AlGaN) and GaN rather than the accepted area in FET. Therefore, this study develops a novel Hybrid Deep Learning Approach for the Predictive Modeling of Aluminium Gallium Nitride with Rime Optimization (HDLA-AlGaNRO) model. The proposed HDLA-AlGaNRO technique is designed to accurately predict the switching behaviors of the devices without the need for delving into their underlying geometry and physics. It encompasses two stages such as prediction and hyperparameter tuning. At the primary stage, the HDLA-AlGaNRO model applies a hybrid of residual convolutional neural network (CNN) with the bidirectional gated recurrent unit (RCNN-BiGRU) algorithm can be exploited for an accurate prediction process. In the final stage, the rime optimization algorithm (ROA) is used for hyperparameter tuning to enhance the predictive accuracy and optimize key parameters. The stimulated outcome investigation of the HDLA-AlGaNRO model occurs and the outcomes are examined in terms of changing features. The experimental study indicates the superiority of the HDLA-AlGaNRO model across the existing advanced approaches.