Aging prediction and state of charge estimation of a LiFePO4 battery using input time-delayed neural networks

•Achieve both SOC and SOH estimation under aging and temperature uncertainties.•This technique requires no prior knowledge of the battery's model or parameters.•Effectiveness is validated experimentally.•Comparison against conventional MLP. This paper presents an intelligent state of charge (SOC) and state of health (SOH) estimation method for lithium-ion batteries using an input time-delayed neural network. Unlike other estimation strategies, this technique requires no prior knowledge of the battery's model or parameters. Instead, it uses ambient temperature variations and previous battery's voltage and current data to accurately predict its SOC and SOH. The presented method compensates for the nonlinear patterns in battery characteristics such as hysteresis, variance due to electrochemical properties, and battery degradation due to aging. This technique is evaluated using a LiFePO4 battery and experimental results highlight its high accuracy, simplicity, and robustness.