Battery cell state-of-charge estimation using linear parameter varying system techniques

This paper describes a model based method for real time battery cell state-of-charge (SoC) estimation using linear parameter varying (LPV) system techniques. For this class of methods, the applicable structure is one in which the input to output dynamics of the battery can be described by a discrete parameter varying state variable model that includes the SoC as a state. Within this context, the problem of state-of-charge estimation is viewed as a state estimation problem, so that a state estimator is designed using the model. An LPV system technique, combined with input to state stability criteria, is used to analyze the stability and performance of the estimator. Compared with algorithms available in the current literature, such as those employing an extended Kalman filter and sliding mode observers, this method offers good performance with a guarantee of stability, and possesses user friendly tuning with low computational complexity for easy on-board implementation. Experimental results are given which validate the proposed methodology.

► We exam the lithium ion battery SoC estimation problem using linear parameter varying system techniques.
► This SoC estimator is designed as an output feedback state estimator based on an LPV model.
► The estimator is designed to be stable (shown analytically) with convergence characteristics that are easily tuned by the user.
► Estimator efficacy is demonstrated on two different lithium ion batteries with vastly different chemistry.