BiLSTM을 이용한 SoC와 SoP 공동 추정

Abstract

This paper proposes a Lithium-ion battery state-of-charge and state-of-power co-estimation algorithm. In states co-estimation algorithm, battery state-of-charge is considered in state-of-power estimation and vice versa. Unlike conventional methods, the proposed method takes into account the effect of the current battery state-of-power in one-step ahead state-of-charge estimation. Since battery states are not directly measurable, a bidirectional long-short term memory model is used to co-estimate the states using the measurable battery parameters (such as voltage and current). The model is trained and tested using Urban Dynamometer Driving Schedule. The results show that the co-estimating battery state-of-charge and state-of-power has higher accuracy (approximately 28.35%) than independent estimation.