Artículo de publicación ISIWe present the implementation of a particle-filteringbased prognostic framework that utilizes statistical characterization of use profiles to (i) estimate the state-of-charge (SOC), and (ii) predict the discharge time of energy storage devices (lithium-ion batteries). The proposed approach uses a novel empirical statespace model, inspired by battery phenomenology, and particle-filtering algorithms to estimate SOC and other unknown model parameters in real-time. The adaptation mechanism used during the filtering stage improves the convergence of the state estimate, and provides adequate initial conditions for the prognosis stage. SOC prognosis is implemented using a particle-filtering-based framework that considers a statistical characterization of uncertainty for future discharge profiles based on maximum likelihood estimates of transition probabilities for a two-state Markov chain. All algorithms have been trained and validated using experimental data acquired from one Li-Ion 26650 and two Li-Ion 18650 cells, and considering different operating conditions.Project FONDECYT 114077

Artículo de publicación IS

Recent developments in lithium-ion technology have enabled a revolution in the automotive industry. Fully electric vehicles (EVs) operate under distinctly variable conditions, requiring high-voltage battery packs to meet their torque/power demands. Our goal is to provide a simulation engine which, for a given battery pack size, determines when recharging or battery pack replacement are needed. To that end, we study both the State-of-Charge (SOC) and the State-of-Health (SOH) indicators, using discrete state space models for both. Predictions are based on a probabilistic characterization of EV usage profiles, which in turn are a function of generic user-input, such as mission maps, vehicle mechanical characteristics,driving schedules, and battery pack configuration. State space models benefit from the incorporation of metamodels for the ohmic internal resistance and the Coulomb efficiency of the pack. Both meta-models i) effectively introduce additional phenomenology –such as dependency onthe magnitude of discharged current and depth of discharge (DoD)–, and ii) provide a link between SOC/SOH and how each discharge cycle affects the health status of the battery pack as a whole. The approach for the simulation engine presented here is stochastic in nature, meaning that prognostics for the SOC and SOH are generated in a particle filter-based scheme. Thus risk and confidence intervals can be obtained for the end-of-discharge and end-of-life respectively

We present the implementation of a particle-filtering-based framework that estimates the State-of-Health (SOH) and predicts the End-of-Life (EOL) of Lithium-Ion batteries, efficiently incorporating variations of ambient temperature in the analysis. The proposed approach uses an empirical statespace model, in which inputs are explicitly defined as the average temperature of operation and the output of an external module that detects self-recharge phenomena, on the other hand the output is a function that relates the current SOH and temperature with the Usable Capacity in that cycle. In addition, this approach allows to deal with data losses and outliers. In order to correct erroneous initial conditions in state estimates, an Outer Feedback Correction Loop is implemented. Finally, this framework is validated using degradation data from four sources: experimental degradation data from two Li-Ion 18650 cells, accelerated degradation data openly provided by NASA Ames Research Center, and artificially generated degradation data at different ambient temperatures.