Modeling, Identification, & Deep Learning of Batteries Scott Moura Assistant Professor | eCAL Director University of California, Berkeley Energy Resources Engineering Seminar Stanford University Download: https://ecal.berkeley.edu/pubs/slides/Moura-StanfordERE-Batts-Slides.pdf Scott Moura | UC Berkeley Battery Modeling, ID, Learning September 18, 2017 | Slide 1
Dong Laurel Eric Saehong Zach Dr. Hector ZHANG DUNN MUNSING PARK GIMA PEREZ Dr. Chao Bertrand Hongcai Dylan Sangjae ZHOU SUN TRAVACCA ZHANG KATO BAE Zhe Scott Moura | UC Berkeley Battery Modeling, ID, Learning September 18, 2017 | Slide 2
A Golden Era Keyword Search: Battery Systems and Control No. of Publications 3000 2000 1000 0 1985 1990 1995 2000 2005 2010 2015 Year Scott Moura | UC Berkeley Battery Modeling, ID, Learning September 18, 2017 | Slide 3
The Battery Problem Needs: Cheap, high energy/power, fast charge, long life Reality: Expensive, conservatively design/operated, die too quickly Scott Moura | UC Berkeley Battery Modeling, ID, Learning September 18, 2017 | Slide 4
The Battery Problem Needs: Cheap, high energy/power, fast charge, long life Reality: Expensive, conservatively design/operated, die too quickly Some Motivating Facts 1000 USD / kWh (2010) ∗ 485 USD / kWh (2012) ∗ 350 USD / kWh (2015) ∗∗ EV Batts 125 USD / kWh for parity to IC engine Only 50-80% of available capacity is used Range anxiety inhibits adoption Lifetime risks caused by fast charging Scott Moura | UC Berkeley Battery Modeling, ID, Learning September 18, 2017 | Slide 4
The Battery Problem Needs: Cheap, high energy/power, fast charge, long life Reality: Expensive, conservatively design/operated, die too quickly Scott Moura | UC Berkeley Battery Modeling, ID, Learning September 18, 2017 | Slide 4
The Battery Problem Needs: Cheap, high energy/power, fast charge, long life Reality: Expensive, conservatively design/operated, die too quickly Scott Moura | UC Berkeley Battery Modeling, ID, Learning September 18, 2017 | Slide 4
The Battery Problem Needs: Cheap, high energy/power, fast charge, long life Reality: Expensive, conservatively design/operated, die too quickly Some Motivating Facts 1000 USD / kWh (2010) ∗ 485 USD / kWh (2012) ∗ 350 USD / kWh (2015) ∗∗ EV Batts 125 USD / kWh for parity to IC engine Only 50-80% of available capacity is used Range anxiety inhibits adoption Lifetime risks caused by fast charging Two Solutions Design better batteries Make current batteries better (materials science & chemistry) (estimation and control) ∗ Source: MIT Technology Review, “The Electric Car is Here to Stay.” (2013) ∗∗ Source: Tesla Powerwall. (2015) Scott Moura | UC Berkeley Battery Modeling, ID, Learning September 18, 2017 | Slide 4
On-Going Research Goals Increase usable energy capacity by 20% Decrease charge times by factor of 5X Increase battery life time by 50% Decrease fault detection time by factor of 10X Scott Moura | UC Berkeley Battery Modeling, ID, Learning September 18, 2017 | Slide 5
Outline BACKGROUND 1 ELECTROCHEMICAL MODEL 2 MODEL IDENTIFICATION 3 DEEP LEARNING 4 5 SUMMARY AND OPPORTUNITIES Scott Moura | UC Berkeley Battery Modeling, ID, Learning September 18, 2017 | Slide 6
History Luigi Galvani, 1737-1798, Experiments on frog legs Physicist, Bologna, Italy “Animal electricity” Dubbed “galvanism” First foray into electrophysiology Alessandro Volta, 1745-1827 Voltaic Pile Monument to Volta in Como Physicist, Como, Italy Scott Moura | UC Berkeley Battery Modeling, ID, Learning September 18, 2017 | Slide 7
Battery Models Equivalent Circuit Model (a) OCV-R Scott Moura | UC Berkeley Battery Modeling, ID, Learning September 18, 2017 | Slide 8
Battery Models Equivalent Circuit Model (a) OCV-R (b) OCV-R-RC (c) Impedance Scott Moura | UC Berkeley Battery Modeling, ID, Learning September 18, 2017 | Slide 8
Battery Models Electrochemical Model Equivalent Circuit Model (a) OCV-R - - sep sep + + 0 L 0 L L 0 x (b) OCV-R-RC Cathode Anode Separator c s - ( r ) c s + ( r ) e - Li + e - c ss - c ss + (c) Impedance r r c e ( x ) Li x C 6 Li 1-x MO 2 Electrolyte Scott Moura | UC Berkeley Battery Modeling, ID, Learning September 18, 2017 | Slide 8
Safely Operate Batteries at their Physical Limits Electrochemical model-based limits of operation ECM-based limits of operation ECM-based limits of operation Terminal Voltage Overpotential Surface concentration Cell Current Scott Moura | UC Berkeley Battery Modeling, ID, Learning September 18, 2017 | Slide 9
ElectroChemical Controller (ECC) Measurements I r ( t ) V ( t ), T ( t ) I ( t ) EChem-based Battery Cell Controller + Innovations _ EChem-based Estimated ^ ^ x ( t ), θ ( t ) States & Params State/Param ^ ^ V ( t ), T ( t ) Estimator ElectroChemical Controller (ECC) Scott Moura | UC Berkeley Battery Modeling, ID, Learning September 18, 2017 | Slide 10
Outline BACKGROUND 1 ELECTROCHEMICAL MODEL 2 MODEL IDENTIFICATION 3 DEEP LEARNING 4 5 SUMMARY AND OPPORTUNITIES Scott Moura | UC Berkeley Battery Modeling, ID, Learning September 18, 2017 | Slide 11
Battery Electrochemistry Model The Doyle-Fuller-Newman (DFN) Model - - sep 0 L 0 sep + 0 + L L x Anode Separator Cathode c s - ( r ) c s + ( r ) e - Li + e - c ss - c ss + r r Li x C 6 c e ( x ) Li 1-x MO 2 Electrolyte Key References: K. Thomas, J. Newman, and R. Darling, Advances in Lithium-Ion Batteries. New York, NY USA: Kluwer Academic/Plenum Publishers, 2002, ch. 12: Mathematical modeling of lithium batteries, pp. 345-392. N. A. Chaturvedi, R. Klein, J. Christensen, J. Ahmed, and A. Kojic, “Algorithms for advanced battery-management systems,” IEEE Control Systems Magazine, vol. 30, no. 3, pp. 49-68, 2010. J. Newman. (2008) Fortran programs for the simulation of electrochemical systems. [Online]. Available: http://www.cchem.berkeley.edu/jsngrp/fortran.html Scott Moura | UC Berkeley Battery Modeling, ID, Learning September 18, 2017 | Slide 12
Electrochemical Model Equations well, some of them | Open Source Matlab CODE: github.com/scott-moura/fastDFN Equation Description ∂ c ± � s ) · r 2 ∂ c ± � Solid phase Li D ± s ( c ± 1 ∂ ∂ t ( x , r , t ) = s ∂ r ( x , r , t ) s r 2 ∂ r concentration � � 1 − t 0 Electrolyte Li ε e ∂ c e ∂ e ( c e ) ∂ c e i ± D eff ∂ t ( x , t ) = ∂ x ( x , t ) + c e ( x , t ) ∂ x F concentration σ eff , ± ∂φ ± Solid ∂ x ( x , t ) = i ± s e ( x , t ) − I ( t ) potential 2 RT ( 1 − t 0 c ) κ eff ( c e ) d ln f c / a Electrolyte � � κ eff ( c e ) ∂φ e ∂ ln c e ∂ x ( x , t ) = − i e ± ( x , t ) + 1 + ( x , t ) F d ln c e ∂ x potential ∂ i e ± Electrolyte ∂ x ( x , t ) = a ± s Fj n ± ( x , t ) ionic current α aF RT η ± ( x , t ) − e − α cF RT η ± ( x , t ) � Molar flux � j ± n ( x , t ) = 1 F i ± 0 ( x , t ) e btw phases � 0 + dT � T 0 ( t ) − T ( t ) � dt ( t ) = h + I ( t ) V ( t ) − 0 − a s Fj n ( x , t )∆ T ( x , t ) dx ρ c P Temperature Scott Moura | UC Berkeley Battery Modeling, ID, Learning September 18, 2017 | Slide 13
Electrochemical Model Equations well, some of them | Open Source Matlab CODE: github.com/scott-moura/fastDFN Equation Description ∂ c ± � s ) · r 2 ∂ c ± � Solid phase Li D ± s ( c ± 1 ∂ ∂ t ( x , r , t ) = s ∂ r ( x , r , t ) s (PDE in r , t ) r 2 ∂ r concentration � � 1 − t 0 Electrolyte Li ε e ∂ c e ∂ e ( c e ) ∂ c e i ± D eff ∂ t ( x , t ) = ∂ x ( x , t ) + c e ( x , t ) (PDE in r , t ) ∂ x F concentration σ eff , ± ∂φ ± Solid ∂ x ( x , t ) = i ± s e ( x , t ) − I ( t ) potential 2 RT ( 1 − t 0 c ) κ eff ( c e ) d ln f c / a Electrolyte � � κ eff ( c e ) ∂φ e ∂ ln c e ∂ x ( x , t ) = − i e ± ( x , t ) + 1 + ( x , t ) F d ln c e ∂ x potential ∂ i e ± Electrolyte ∂ x ( x , t ) = a ± s Fj n ± ( x , t ) ionic current α aF RT η ± ( x , t ) − e − α cF RT η ± ( x , t ) � Molar flux � j ± n ( x , t ) = 1 F i ± 0 ( x , t ) e btw phases � 0 + dT � T 0 ( t ) − T ( t ) � dt ( t ) = h + I ( t ) V ( t ) − 0 − a s Fj n ( x , t )∆ T ( x , t ) dx ρ c P Temperature Scott Moura | UC Berkeley Battery Modeling, ID, Learning September 18, 2017 | Slide 13
Recommend
More recommend
