Abstract
Accurately predicting the future capacity and remaining useful life of batteries is necessary to ensure reliable system operation and to minimise maintenance costs. The complex nature of battery degradation has meant that mechanistic modelling of capacity fade has thus far remained intractable; however, with the advent of cloud-connected devices, data from cells in various applications is becoming increasingly available, and the feasibility of data-driven methods for battery prognostics is increasing. Here we propose Gaussian process (GP) regression for forecasting battery state of health, and highlight various advantages of GPs over other data-driven and mechanistic approaches. GPs are a type of Bayesian non-parametric method, and hence can model complex systems whilst handling uncertainty in a principled manner. Prior information can be exploited by GPs in a variety of ways: explicit mean functions can be used if the functional form of the underlying degradation model is available, and multiple-output GPs can effectively exploit correlations between data from different cells. We demonstrate the predictive capability of GPs for short-term and long-term (remaining useful life) forecasting on a selection of capacity vs. cycle datasets from lithium-ion cells.
Highlights
Lithium-ion batteries (LIBs) are increasingly playing a pivotal role in applications ranging from transport to grid energy storage
We present two extensions to this basic approach which enable improved performance: (i) we use explicit mean functions based on known parametric battery degradation models to exploit prior knowledge of battery degradation behaviour and (ii) we use multi-output Gaussian process (GP) to effectively exploit available capacity data from multiple identical cells
This paper has demonstrated the applicability of GPs to battery capacity forecasting, and highlighted some of their key advantages for this application
Summary
Lithium-ion batteries (LIBs) are increasingly playing a pivotal role in applications ranging from transport to grid energy storage. We present two extensions to this basic approach which enable improved performance: (i) we use explicit mean functions based on known parametric battery degradation models to exploit prior knowledge of battery degradation behaviour and (ii) we use multi-output GPs to effectively exploit available capacity data from multiple identical cells. It is worth underscoring the fact that all the methods presented here are rigorously evaluated using different proportions of training data (i.e. using capacity data up to the current cycle for training, with various different values of the current cycle) This is in contrast to most previous studies on battery prognostics, which merely evaluate the accuracy of the predictions made at a single arbitrarily selected cycle (e.g. the first half of training data)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
