A multi-data-driven procedure towards a comprehensive understanding of the activated carbon electrodes performance (using for supercapacitor) employing ANN technique

Abstract

Biomass resources are intensively used as economical and green-reserve precursor preparation of sustainable carbon materials used in supercapacitors. The synthetic processes of biomass-based precursors (BPs) are the most determinant proceedings for obtaining activated carbons (ACs) used in the electrode of energy storage devices. The AC-based electrode preparation and operational condition parameters can affect the capacitance performance of electrode. In the present work, the potential of Artificial Neural Network (ANN) modeling is assessed in interpreting how activation procedure, structural features, electrode synthesizing procedure, and operational condition can affect the capacitive performance of the carbon-based electrode. Radial Basis Function (RBF) model is established for the estimation of specific capacitance of biomass-based activated carbon (BAC) utilized in the electrode. Moreover, the algorithms used in RBF model performed accurate predictions of the model with the lowest error. Besides, employing the combination of quantitative and qualitative variables could perform a synergistic result. The multi-data could achieve a precise cognizance of materials participating in electrode preparation to obtain higher specific capacitance. The sensitivity analysis showed prominent effects of structural and operational characteristics (e.g. micropore to macropore carbon structure), molarity of electrolyte, binder ratio, and activation agent ratio, on Electric Double-layer capacitor performance.