A computational intelligence method to estimate capacitance loss of electrolytic capacitors based on equivalent series resistance

摘要

Capacitance loss of electrolytic eapacitors is one of aging time functions that has nonlinear behavior with equivalent series resistance. Therefore, analyzing this significant parameter with computational intelligence methods will rise the reliability and safety of the models. In this paper, a computational intelligence method to identify and estimate capacitance losses of six electrolytic capacitor based on equivalent series resistance is presented. The proposed approach contains an adaptive neuro fuzzy inference system based on subtractive clustering algorithm with six input signals. An experimental dataset of six electrolytic capacitors is used in this paper, which was carried out by a research center of NASA. The results demonstrates that the proposed approach is high accuracy for system identification of the capacitance loss and similar dynamic systems.