Electrochemical cell prognostics using online impedance measurements and model-based data fusion techniques

A method to accurately assess the state-of-charge (SOC) and the state-of-health (SOH) of electrochemical energy sources. such as primary and secondary batteries, provides significant benefit in operational systems. The model-based effort described here is focused on prognostics for primary and secondary batteries. However, this approach can also be applied to other electrochemical energy sources such as %el cells for performance monitoring. This method is based on accurate modeling of the transport mechanisms within the battery that requires careful development of the electrochemical models. These models have heen used to develop new features to be used to assess the condition of the battery. Data fusion of feature vectors is used to develop inferences about the state of the system. The resulting output and any usage information available about the battery is then evaluated using hybrid automated reasoning schemes consisting of neural network, fuzzy logic, and decision theoretic methods. The focus of this paper is on model identification and data fusion of the monitored and virtual sensor data. In addition to modeling the main electrochemical processes, modeling of competing failure mechanisms is incorporated. Since these mechanisms dictate the remaining useful life of the battery (i.e., state-of-life) their proper identification is a critical step for prognostics. Also, data fusion processing can be used to associate the monitored and feature data to the different critical mechanisms. Examples using test data 6om various types of batteries such as lead-acid, nickelcadmium, and lithium chemistries are also presented.