Summary Estimating the dynamic status parameters of a battery, such as its state of health (SoH) and remaining useful life (RUL), is still a very difficult and complex task. In this paper we perform a structured review of the most relevant state of the art models, algorithms and commercial devices employed in the estimation of the SoH/RUL battery performance figures, in the context of embedded applications. The models and estimation techniques are thoroughly classified and, for each taxonomy class, a presentation of the working principles is made. A comprehensive set of metrics is then introduced for the evaluation of the SoH/RUL estimation techniques from the perspective of their implementation and operation efficiency in embedded systems. These algorithms are then analyzed and discussed in a comparative manner, with concrete figures and results. The capability and the performance of the different types of off‐the‐shelf fuel gauges to estimate the battery SoH/RUL parameters are also evaluated in this paper. Copyright © 2016 John Wiley & Sons, Ltd.

No abstract

Summary Online state of health (SOH) prediction of lithium‐ion batteries remains a very important problem in assessing the safety and reliability of battery‐powered systems. Deep learning techniques based on recurrent neural networks with memory, such as the long short‐term memory (LSTM) and gated recurrent unit (GRU), have very promising advantages, when compared to other SOH estimation algorithms. This work addresses the battery SOH prediction based on GRU. A complete BMS is presented along with the internal structure and configuration parameters. The neural network was highly optimized by adaptive moment estimation (Adam) algorithm. Experimental data show very good estimation results for different temperature values, not only at room value. Comparisons performed against other relevant estimation methods highlight the performance of the recursive neural network algorithms such as GRU and LSTM, with the exception of the battery regeneration points. Compared to LSTM, the GRU algorithm gives slightly higher estimation errors, but within similar prediction error range, while needing significantly fewer parameters (about 25% fewer), thus making it a very suitable candidate for embedded implementations.

No abstract

No abstract