### Abstract

This article presents a framework for optimizing the thermal cycle to estimate a battery cell's entropy coefficient at 20% state of charge (SOC). Our goal is to maximize Fisher identifiability: a measure of the accuracy with which a parameter can be estimated. Existing protocols in the literature for estimating entropy coefficients demand excessive laboratory time. Identifiability optimization makes it possible to achieve comparable accuracy levels in a fraction of the time. This article demonstrates this result for a set of lithium iron phosphate (LFP) cells. We conduct a 24-h experiment to obtain benchmark measurements of their entropy coefficients. We optimize a thermal cycle to maximize parameter identifiability for these cells. This optimization proceeds with respect to the coefficients of a Fourier discretization of this thermal cycle. Finally, we compare the estimated parameters using (i) the benchmark test, (ii) the optimized protocol, and (iii) a 15-h test from the literature (by Forgez et al.). The results are encouraging for two reasons. First, they confirm the simulation-based prediction that the optimized experiment can produce accurate parameter estimates in 2 h, compared to 15-24. Second, the optimized experiment also estimates a thermal time constant representing the effects of thermal capacitance and convection heat transfer.

Original language | English (US) |
---|---|

Pages (from-to) | 18-28 |

Number of pages | 11 |

Journal | Journal of Power Sources |

Volume | 308 |

DOIs | |

State | Published - Mar 15 2016 |

### Fingerprint

### All Science Journal Classification (ASJC) codes

- Renewable Energy, Sustainability and the Environment
- Energy Engineering and Power Technology
- Physical and Theoretical Chemistry
- Electrical and Electronic Engineering

### Cite this

*Journal of Power Sources*,

*308*, 18-28. https://doi.org/10.1016/j.jpowsour.2016.01.059

}

*Journal of Power Sources*, vol. 308, pp. 18-28. https://doi.org/10.1016/j.jpowsour.2016.01.059

**Optimization and experimental validation of a thermal cycle that maximizes entropy coefficient fisher identifiability for lithium iron phosphate cells.** / Mendoza, Sergio; Rothenberger, Michael; Hake, Alison; Fathy, Hosam Kadry.

Research output: Contribution to journal › Article

TY - JOUR

T1 - Optimization and experimental validation of a thermal cycle that maximizes entropy coefficient fisher identifiability for lithium iron phosphate cells

AU - Mendoza, Sergio

AU - Rothenberger, Michael

AU - Hake, Alison

AU - Fathy, Hosam Kadry

PY - 2016/3/15

Y1 - 2016/3/15

N2 - This article presents a framework for optimizing the thermal cycle to estimate a battery cell's entropy coefficient at 20% state of charge (SOC). Our goal is to maximize Fisher identifiability: a measure of the accuracy with which a parameter can be estimated. Existing protocols in the literature for estimating entropy coefficients demand excessive laboratory time. Identifiability optimization makes it possible to achieve comparable accuracy levels in a fraction of the time. This article demonstrates this result for a set of lithium iron phosphate (LFP) cells. We conduct a 24-h experiment to obtain benchmark measurements of their entropy coefficients. We optimize a thermal cycle to maximize parameter identifiability for these cells. This optimization proceeds with respect to the coefficients of a Fourier discretization of this thermal cycle. Finally, we compare the estimated parameters using (i) the benchmark test, (ii) the optimized protocol, and (iii) a 15-h test from the literature (by Forgez et al.). The results are encouraging for two reasons. First, they confirm the simulation-based prediction that the optimized experiment can produce accurate parameter estimates in 2 h, compared to 15-24. Second, the optimized experiment also estimates a thermal time constant representing the effects of thermal capacitance and convection heat transfer.

AB - This article presents a framework for optimizing the thermal cycle to estimate a battery cell's entropy coefficient at 20% state of charge (SOC). Our goal is to maximize Fisher identifiability: a measure of the accuracy with which a parameter can be estimated. Existing protocols in the literature for estimating entropy coefficients demand excessive laboratory time. Identifiability optimization makes it possible to achieve comparable accuracy levels in a fraction of the time. This article demonstrates this result for a set of lithium iron phosphate (LFP) cells. We conduct a 24-h experiment to obtain benchmark measurements of their entropy coefficients. We optimize a thermal cycle to maximize parameter identifiability for these cells. This optimization proceeds with respect to the coefficients of a Fourier discretization of this thermal cycle. Finally, we compare the estimated parameters using (i) the benchmark test, (ii) the optimized protocol, and (iii) a 15-h test from the literature (by Forgez et al.). The results are encouraging for two reasons. First, they confirm the simulation-based prediction that the optimized experiment can produce accurate parameter estimates in 2 h, compared to 15-24. Second, the optimized experiment also estimates a thermal time constant representing the effects of thermal capacitance and convection heat transfer.

UR - http://www.scopus.com/inward/record.url?scp=84955442883&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84955442883&partnerID=8YFLogxK

U2 - 10.1016/j.jpowsour.2016.01.059

DO - 10.1016/j.jpowsour.2016.01.059

M3 - Article

AN - SCOPUS:84955442883

VL - 308

SP - 18

EP - 28

JO - Journal of Power Sources

JF - Journal of Power Sources

SN - 0378-7753

ER -