Dimensional Reduction Applied to an Intelligent Model for Boost Converter Switching Operation

Abstract

The dimensional reduction algorithms are applied to a hybrid intelligent model that distinguishes the switching operating mode of a boost converter. Thus, the boost converter has been analyzed and both operating mode are explained, distinguishing between Hard-switching and Soft-switching modes. Then, the dataset is created out of the data obtained from simulation of the real circuit and the hybrid intelligent classification model is implemented. Finally, the dimensional reduction of the input variables is carried out and the results are compared. As result, the proposed model with the applied dimensional reduced dataset is able to distinguish between the HS and SS operating modes with high accuracy.