Supercapacitor thermal- and electrical-behaviour modelling using ANN

摘要: The paper presents the development of a modelling tool for evaluation thermal and electrical behaviour supercapacitors, using an artificial neural network (ANN). principle consists black-box multiple-input single-output (MISO) model. system inputs are temperature, current supercapacitor values, output is voltage. relationship between established by learning validation ANN model from experimental charge discharge cycles supercapacitors at different currents temperatures. Once training parameters known, simulator can predict operational supercapacitors. update performed Levenberg-Marquardt method to minimise error predicted output. This methodology networks may provide useful information on transient taking into account influences. Experimental results will also validate simulation results.