Abstract: A method for providing an estimated present state of charge of a battery pack includes measuring battery parameters and then calculating a first state of charge utilizing an initial state of charge and the battery parameters. An error covariance prediction is calculated for the calculated first state of charge. A correction of the first state of charge is determined by mapping the open circuit voltage of the battery to the initial state of charge to obtain an open circuit voltage-initial state of charge confidence slope, calculating a Kalman gain of the error covariance prediction, and mapping the open circuit voltage (incorporating EMF) to the first state of charge. The estimated present state of charge is calculated by correcting the first state of charge utilizing selected battery parameters, the Kalman gain, and the mapping of the resulting voltage of the battery to the first state of charge.

Abstract: Systems and methods for detecting and/or identifying an attack on a battery management system (BMS) or a battery system. The voltage and/or state of charge (SOC) of the BMS or battery system can be monitored, and one or more datasets can be obtained. A principal component analysis (PCA) based unsupervised k-means approach can be applied on the one or more datasets to monitor for irregularities that indicate an attack.

Abstract: Testbeds for battery management systems (BMSs) and/or batteries, as well as methods of using the same, are provided. A testbed can be a control-hardware-in-the-loop (CHIL) testbed and can include a simulation bench including a battery cell simulator, a temperature simulator, and/or a real-time simulator. The simulator bench can further include a programmable power supply, a relay, a resistor, and/or a communication protocol.

Abstract: Systems and methods for forecasting of State of Charge (SOC) of lithium ion batteries are provided. A multi-step forecasting process with experimentally obtained decreasing C-Rate datasets together with machine learning can be used. The multi-step approach can combine a univariate technique with machine learning techniques. An Auto Regressive Integrated Moving Average (ARIMA) and/or Holt Winters Exponential Smoothing (HWES) can be combined with each other and/or with machine learning techniques such as Multilayer Perceptron (MLP) and Nonlinear autoregressive neural network with external input (NARX-net).

Abstract: Systems and methods for forecasting of State of Charge (SOC) of lithium ion batteries are provided. A multi-step forecasting process with experimentally obtained decreasing C-Rate datasets together with machine learning can be used. The multi-step approach can combine a univariate technique with machine learning techniques. An Auto Regressive Integrated Moving Average (ARIMA) and/or Holt Winters Exponential Smoothing (HWES) can be combined with each other and/or with machine learning techniques such as Multilayer Perceptron (MLP) and Nonlinear autoregressive neural network with external input (NARX-net).