University of Nebraska - Lincoln
DigitalCommons@University of Nebraska - Lincoln
Faculty Publications from the Department of
Electrical Engineering
Electrical Engineering, Department of
1-1-2012
A Multicell Battery System Design for Electric and
Plug-in Hybrid Electric Vehicles
Taesic Kim
University of Nebraska-Lincoln, taesickim@huskers.unl.edu
Wei Qiao
University of Nebraska–Lincoln, wqiao@engr.unl.edu
Liyan Qu
Member, IEEE, liyanqu@ieee.org
Follow this and additional works at: http://digitalcommons.unl.edu/electricalengineeringfacpub
Part of the Electrical and Computer Engineering Commons
Kim, Taesic; Qiao, Wei; and Qu, Liyan, "A Multicell Battery System Design for Electric and Plug-in Hybrid Electric Vehicles" (2012).
Faculty Publications from the Department of Electrical Engineering. Paper 176.
http://digitalcommons.unl.edu/electricalengineeringfacpub/176
This Article is brought to you for free and open access by the Electrical Engineering, Department of at DigitalCommons@University of Nebraska Lincoln. It has been accepted for inclusion in Faculty Publications from the Department of Electrical Engineering by an authorized administrator of
DigitalCommons@University of Nebraska - Lincoln.
2012 IEEE International Electric Vehicle Conference (IEVC)
Digital Object Identifier: 10.1109/IEVC.2012.6183240
A Multicell Battery System Design for Electric and
Plug-in Hybrid Electric Vehicles
Taesic Kim
Student Member, IEEE
Department of Electrical Engineering
University of Nebraska-Lincoln
Lincoln, NE 68588-0511 USA
taesickim@huskers.unl.edu
Wei Qiao
Member, IEEE
Department of Electrical Engineering
University of Nebraska-Lincoln
Lincoln, NE 68588-0511 USA
wqiao@engr.unl.edu
Abstract--The performance of electric vehicles (EVs) and
plug-in hybrid electric vehicle (PHEVs) strongly relies on their
battery storage system, which consists of multiple battery cells
connected in series and parallel. However, cell state variations
are commonly present, which reduces the energy conversion
efficiency of the battery system. Furthermore, in a large battery
system the risk of catastrophic faults of cells increases because a
large numbers of cells are used. To solve these problems, this
paper proposes a novel power electronics-enabled, self-X,
multicell battery system design. The proposed battery system
can self-heal from failures or abnormal operations of single or
multiple cells and self-balance from cell state variations. These
features are achieved by a cell switching circuit and a highperformance battery management system (BMS). The proposed
design is validated by simulation studies in MATLAB Simulink
for a battery system containing five modules connected in series,
where each module consists of 6×3 cylindrical lithium-ion cells.
The proposed design is scalable to large battery systems for
EV/PHEV applications.
I. INTRODUCTION
Electric vehicles (EVs) and plug-in hybrid vehicles
(PHEVs) use a battery system consisting of a large number
of cells to meet the required voltage, current, energy, and
power density. However, several design deficiencies are
present in current multicell battery systems, which are 1)
adopting a fixed configuration for cell connections, resulting
in low reliability and low fault-tolerance capability from
abnormal operating conditions, such as high temperature,
overcharge, over-discharge, and over-current [1]; 2) lacking
an effective method to utilize cell state variations, resulting in
nonoptimal energy conversion efficiency, and 3) lacking a
capability for flexible dynamic power management, resulting
in nonoptimal system performance.
In the existing battery design, safety circuits are
commonly used to protect cells from high temperature,
This work was supported in part by the U.S. National Science Foundation
(NSF) under CAREER Award ECCS-0954938 and the Federal Highway
Administration (FHWA) under Agreement No. DTFH61-10-H-00003. Any
opinions, findings, and conclusions or recommendations expressed in this
publication are those of the authors and do not necessarily reflect the view of
the NSF or FHWA.
Liyan Qu
Member, IEEE
Department of Electrical Engineering
University of Nebraska-Lincoln
Lincoln, NE 68588-0511 USA
liyanqu@ieee.org
overcharge, over-discharge, and over-current by monitoring
the temperature, voltage, and current of each cell [2].
However, lacking an effectively reconfigurable topology, the
safety circuits will cut off the whole battery system when any
single cell is operated in an abnormal condition. Moreover,
cell state variations are commonly present in multicell
batteries. In this case, the fixed-configuration design can only
utilize a part of the total battery capacity. To overcome this
deficiency, cell balancing circuits are used. However, most
existing balancing circuits for EVs/PHEVs use dissipative
resistors, resulting in energy loss and generating considerable
heat [3]. To reduce energy loss, active balancing circuits were
proposed by using transformers [4], capacitors [5], and dc/dc
converters [6]. The latest products of cell balancing integrated
circuits (ICs) [7] use electronic converters to transfer charge
from cell to cell during operation. However, these solutions
substantially increase the cost and volume of the battery
system and might not be able to handle faulty cells. Recently,
several reconfigurable multicell battery topologies have been
proposed to handle cell failures or to optimize the
performance of battery systems for portable electronic
devices [8]-[10]. However, these topologies are too complex
and unrealistic for the battery systems in EVs/PHEVs.
This paper extends the work in [11] by proposing a power
electronics-enabled, self-X, multicell battery design for
EVs/PHEVs. The resulting battery system can dynamically
configure itself during operation to achieve self-healing from
failures of single or multiple cells and self-balancing from
cell state variations. Compared to existing active balancing
circuits [4]-[10], the number of balancing components, such as
inductors, capacitors, and switches, in the proposed design is
significantly reduced. This reduces the cost, complexity, and
control effort of the total battery system.
II. PROPOSED SELF-X MULTICELL BATTERY DESIGN
Fig. 1 illustrates the proposed self-X multicell battery
system connected to an electric motor load in an EV. An EV
or PHEV usually requires the battery system to supply power
to the load at a constant voltage level and absorb power from
a constant voltage source, e.g., a charger or regenerative
brake. Therefore, a bidirectional dc/dc converter is
Self-X Battery
Battery Module 1
Battery Module 2
VB
●●●
Motor
Bridge
Inverter
Bidirectional
VDC
dc/dc
DC Converter
Converter
Battery Module n
Master Battery
Management System
(MBMS)
Electric Motor
Fig. 1. A self-X multicell battery supplying an electric motor load in an EV.
commonly used as the interface between the battery system
and the load/source to control charge and discharge of the
battery. In this case, the battery system can be operated with
variable voltages. The motor bridge inverter generally
requires a DC voltage at several hundred volts, e.g., 450 V.
In order to achieve optimal operation of the dc/dc converter,
a terminal voltage of 150 to 300 V is needed for the battery
[12]. Therefore, the proposed battery design consists of
several battery modules connected in series to achieve the
required terminal voltage. For example, the Chevrolet Volt’s
lithium-ion battery system consists of three modules [13]. A
Master Battery Management System (MBMS) is used to
coordinate the operations of the multiple battery modules and
the dc/dc converter. Each battery module consists of three
parts: (1) a cell pack, (2) a cell switching circuit, and (3) a
Module Management System (MMS), as shown in Fig. 2.
A. The Cell Pack
The nominal voltages and currents of most single battery
cells are limited to several volts and tens of amperes, e.g., 3.7
V and 45 A for the lithium-ion cells, which are much lower
than the voltages and currents required in EVs/PHEVs.
Therefore, the cell pack consists of m×n cells, where n cells
are connected in parallel to form a cell bank to provide a
higher current; and m banks are connected in series to step
up the voltage at the terminals of the battery module.
B. The Cell Switching Circuit
Fig. 2. Proposed self-X multicell battery module.
Fig. 3. Proposed cell switching circuit topology in the battery module.
Fig. 4. Switch implementation for Cell Cij using an n-channel power
MOSFET and for Bank i using a p-channel power MOSFET.
Fig. 3 shows the proposed cell switching circuit
topology for an m×n-cell battery module, Only m×(n+1)
controllable switches are needed to form the cell switching
circuit. Each cell only uses one switch, e.g., the switch Sij for
Cell Cij (i = 1, ···, m and j = 1, ···, n), which turn on/off
alternatively to connect/cut off the cell from the battery
module. The proposed cell switching circuit ensures that each
cell in the battery pack can be controlled independently in
three modes, i.e., off, on/charge, and on/discharge. Low-cost,
high-efficiency power MOSFETs which have a negligible
conduction loss are used for implementation of the switches
in the cell switching circuit.
Fig. 4 illustrates a switch implementation for Cell Cij (i =
1, ···, m and j = 1, ···, n) using an n-channel power MOSFET
(Sij) and for Bank i using a p-channel power MOSFET (Si)
with a gate drive circuit, which uses two opto-couplers and a
JFET. The negative terminal of the battery cell is used as the
virtual ground for the gate drive circuit. The gate signal
generated by the signal generator is applied to the gate
terminals of the power MOSFETs through the corresponding
opto-couplers to drive the power MOSFETs. Since the
ground of the gate drive circuit is separated from that of the
cell switching circuit, the switching implementation in Fig. 4
can be used for multicell batteries at any voltage levels,
Check availability of cells based on cell
condition and SOC
Determine number of banks (qMBMS) by MBMS
Determine the number of banks (q) to be used
Calculate and sort SOCb
Fig. 5. Schematic diagram of the module management system (MMS).
Determine qMMS
which facilitates modularization. In Fig. 4 when Qij turns on,
it drives Sij off. Turning on Sij is accomplished by turning off
Qij. When Qi1 turns on, it provides a gate signal to turn on Si.
Turning off Si is accomplished by turning off Qi1 while
turning on Qi2.
C. The Module Management System (MMS)
The MMS, as shown in Fig. 5, performs the functions of
sensing and condition monitoring, model-based state of
charge (SOC) tracking, control and protection, gate signal
generation, and interfacing with the MBMS. The sensing and
monitoring circuit monitors the voltage, current, and
temperature for each cell. The SOC of each cell is tracked by
a hybrid battery model-based method presented in Section III.
Fig. 6 illustrates the functional block flow chart of one
control cycle of the control and protection module. If an
abnormal condition or fault occurs in a cell, it will be cut off
immediately to prevent catastrophic failures of cells. The
remaining cells are still used to supply/store power and,
therefore, the whole battery system self-heals from the
abnormal conditions or failures of cells. The proposed
control scheme always tends to balance the SOCs of the
battery cells and modules. The MBMS determines the
number of banks, qMBMS, of each module to be used
simultaneously to balance the SOCs of the modules of the
battery system. This is called global balancing. The SOC of a
module (SOCm) is defined as
1 m
(1)
SOC m =  SOCb , j
m j =1
where SOCb is the SOC of a battery bank in the module,
which is calculated by
1 n
(2)
SOCb =  SOCi
n i =1
where SOCi is the SOC of the ith cell in the bank. If a cell is
disconnected, then its SOC is zero.
Moreover, the control and protection block in each battery
module also determines the number of series banks, qMMS, to
be used simultaneously to balance the SOCs of the banks in
the module. This is called local balancing.
q MMS = m − f ( ΔSOCb )
(3)
q = min(qMMS, qMBMS)
Select q banks with the highest/lowest SOCb
from the module in discharge/charge mode
Generate control/protection signals
Duration is over or load/
source condition is changed
N
Y
Fig. 6. The functional block flow chart of one control cycle of the
control and protection module.
where ΔSOCb = SOCb_max ‒ SOCb_min; SOCb_max and SOCb_min
are the maximum and minimum SOCs of the banks in the
module, respectively. The value of f(·) with respect to ΔSOCb
used in this work is shown in Fig. 7. Finally, q (0 < q ≤ m)
can be determined to be the smaller one of qMBMS and qMMS.
Once q is determined, the SOCs of the usable banks are
sorted in a descending order and the q banks with the highest
SOCs will be used in the discharge mode or the q banks with
the lowest SOCs will be used in the charge mode until the
SOCs of all the banks in the module become balanced. This
whole process is called a control cycle, as illustrated in Fig. 6.
The control cycle restarts with a certain predefined time
interval Ts or when the load/source condition is changed.
According to the output of the control and protection module,
the gate signal generation module generates appropriate
control signals to control the power MOSFETs through their
gate drive circuits.
The duration Ts of the control cycle will affect the
performance of the proposed battery system. Generally, the
operating time of the battery increases with the decrease of Ts.
However, using a very small Ts will result in frequent
switching of the power MOSFETs and, therefore, increase
battery. According to these two criteria, the total number of
series banks (qt) to be used in the battery system can be
determined by
(5)
qt (Vavg , w, P) = qmax − R(Vavg , w, P)

  Vavg − Vcutoff
R = floor qmax min w 
Vavg
 

Fig. 7. The value of f(·) with respect to ΔSOCb.
w (%)
100
80
0 1
5
ΔSOCm (%)
Fig. 8. The value of w with respect to ΔSOCm.
the switching loss of the cell switching circuit. In practice,
the value of Ts should be selected such that the SOCs of all m
banks in the module will be balanced before any single bank
is fully charged in a charge operation or fully discharged in a
discharge operation. In this work the value of Ts is
determined by:
Ts =
3,600
×δ
I
(4)
where Ts is in seconds; I is the normalized battery current in
C; and δ is a percentage. If the SOCs of all banks are higher
than a low threshold (e.g., 10%) in a discharge operation or
lower than a high threshold (e.g., 90%) in a charge operation,
a large δ (e.g., δ = 5%) will be used. On the other hand, if the
SOC of a bank becomes lower than the low threshold in a
discharge operation or higher than the high threshold in a
charge operation, a small δ (e.g., δ = 1% or 0.5%) will be
used.
D. The Master Battery Management System (MBMS)
The MBMS coordinates the battery modules to globally
balance the SOCs of the modules of the battery system and
controls the bidirectional dc/dc converter to safely
charge/discharge the battery system with the optimal
efficiency. Given battery states and the power demand (P)
from the load or the power supplied by the source, the
MBMS determines the optimal values of the voltage (V) and
current (I) of the battery, where P = VI, using the tradeoff of
two conditions: 1) the terminal voltage of the battery should
be set at a value such that the dc/dc converter is operated
with a voltage gain close to the optimal value; 2) the current
should be as small as possible to utilize the rate capacity
effect to improve the energy conversion efficiency of the
 
, 1 − P
  Pm
 
 
 
 
(6)
where Vavg is the average voltage of all the banks in the
battery system; Vcutoff is the discharge cut-off voltages of a
cell; qmax is the total number of banks in the battery system;
Pm is the maximum power of the battery; R is the redundancy
of the battery; min(a, b) denotes the minimum value of a and
b; floor(·) stands for rounding a number down to the nearest
integer; and w is determined by
w = g (ΔSOC m )
(7)
where ΔSOCm = SOCm_max ‒ SOCm_min; SOCm_max and
SOCm_min are the maximum and minimum SOCs of the
modules in the battery system, respectively. The value of w
with respect to ΔSOCm used in this work is shown in Fig. 8.
Once qt is calculated, the MBMS determines the number of
banks, qMBMS, for each module to be used simultaneously
based on the SOC of each module and the operating mode of
the battery. The sum of all qMBMSs is qt.
Due to protection and balancing control, the terminal
voltage of the battery system is variable, which is controlled
by regulating the duty cycle of the dc/dc converter in the
charge mode. The charge reference voltage (Vref) is
determined by the number of banks connected in the battery,
the charge cutoff voltage of a cell (Vover), and the voltage
drop caused by conduction losses of the switches.
In the charge mode, the multicell battery is firstly charged
with a constant current until the terminal voltage reaches Vref.
Thereafter, the voltage of each cell is kept constant; the
charge current is reduced as the SOC of each cell approaches
the maximum value (e.g., 95% for EVs and 90% for PHEVs).
This is called constant-current constant-voltage (CCCV)
control, under which the battery can be charged fully and
safely.
III. MODELING OF BATTERY CELLS AND SOC TRACKING
An accurate battery cell model is needed to validate the
proposed design by simulation studies. Moreover, monitoring,
control, protection, and optimization of battery systems also
need an accurate battery cell model for SOC tracking, etc. In
this paper, a hybrid battery model proposed in [14] is used,
as shown in Fig. 9. An enhanced Coulomb counting
algorithm is designed to estimate the SOC of a battery cell
based on the Kinetic Battery Model (KiBaM) [15]. The
enhanced Coulomb counting algorithm can capture the
nonlinear capacity effects, such as the recovery effect and
rate capacity effect, of the battery cell with low
computational cost, thereby is feasible for real-time
applications [14].
A voltage-controlled voltage source, Voc(SOC), is used to
bridge the SOC to the cell open-circuit voltage, Voc. The RC
circuit simulates the I-V characteristics and transient
response of the battery cell, where the series resistance, Rseries,
is used to characterize the charge/discharge energy losses of
the cell; other resistances and capacitances are used to
characterize the short-term (transient_S) and long-term
(transient_L) transient responses of the cell; and Vcell
represents the terminal voltage of the cell.
A discrete-time version of the SOC tracking is expressed
as follows:
1
C max
IV. SIMULATION RESULTS
C available (k )
= SOC ( k − 1) −
C max
(8)
[icell ( k ) ⋅ T + Cunavaiable ( k ) − Cunavaiable ( k − 1)]
C unavailabl e ( k ) = C unavailabl e ( k − 1) ⋅ e − k ′′T
+ (1 − c )
(1 − e − k ′T )
× icell ( k )
c ⋅ k′
(9)
1
Voc(SOC)
where T is the sampling period; icell (k) is the instantaneous
current of the battery cell at the time index k; k’ and c are
parameters of the KiBaM; Cmax, Cavailable, and Cunavailable are
the maximum, available, and unavailable capacities of the
cell, respectively. The initial SOC, i.e., SOC(0), is the
estimated SOC at the end of the last operating period (i.e., k
= 0).
A. Module-Level Self-Healing and Self-Balancing
A 6×3-cell battery module is built by using the proposed
design and is simulated in MATLAB Simulink. Each battery
cell is a 3.7-V, 2.6-Ah cylindrical lithium-ion cell (see
Appendix), which is represented by the hybrid battery model
in Section III. In this simulation, the battery module is
SOC (State Of Charge)
SOC ( k ) =
The hybrid battery model is implemented in MATLAB
Simulink. Fig. 10 compares the terminal voltage responses
obtained from simulations using the electrical circuit model
proposed in [16] and the hybrid model with experimental
results for a single 3.7-V, 2.6-Ah lithium-ion cell (see
Appendix) under a constant discharge current of 0.8C (2.06
A). The terminal voltage response obtained from the hybrid
model matches the experimental result better than that
obtained from the electrical circuit model, particularly when
the battery cell is close to fully discharged.
0.2
2000
4000
6000
Time (seconds)
(a)
8000
11.5
Vterm inal (Volts)
Vcell (Volts)
0.4
12
Experimental result
Proposed hybrid model
Electrical circuit model
3.8
3.6
3.4
11
10.5
10
9.5
3.2
3
0
0.6
12.5
4.2
4
0.8
0
0
Fig. 9. The hybrid battery model.
Bank 1
Bank 2
Bank 3
Bank 4
Bank 5
Bank 6
1000
2000
3000
Time (seconds)
4000
5000
Fig. 10. Comparison of simulation results of the electrical circuit model
and the hybrid model with experimental results for a single lithium-ion cell
with a constant discharge current of 0.8C (2.06 A).
9
0
2000
4000
6000
8000
Time (Seconds)
(b)
Fig. 11. The operation of a 6×3-cell battery module in discharge mode
during which one cell in Bank 4 fails at 500 s: (a) the SOCs of the six banks
and (b) the terminal voltage response of the battery module.
V. CONCLUSION
This paper has presented a novel power electronicsenabled, self-X, multicell battery design for EVs/PHEVs.
SOC (State Of Charge)
0.8
0.6
Bank 1
Bank 2
Bank 3
Bank 4
Bank 5
Bank 6
0.4
0.2
0
0
1000
2000
3000
4000
Time (seconds)
(a)
5000
1
SOC (State Of Charge)
B. System-Level Simulation with a DC/DC Converter
Five battery modules are connected in series to form a
self-X battery system with a bidirectional dc/dc converter,
where each module has 6×3 cells. The system is built and
simulated in MATLAB Simulink. The whole battery is
charged at 4.68 A, assuming that the initial SOCs of the five
modules are 20% (Module 1), 25% (Module 2), 32.5%
(Module 3), 35% (Module 4), and 40% (Module 5); the
initial SOCs of the six banks in Module 3 are 20% (Bank 1),
25% (Bank 2), 30% (Bank 3), 35% (Bank 4), 40% (Bank 5),
and 45% (Bank 6). All of the five modules are used
simultaneously; while the MBMS and the MMS of each
module determine the number of banks used in each module
in this charge operation. Figs. 12(a) and (b) show the SOCs
of the six banks in Module 3 and the SOCs of the five
modules, respectively, by using the CCCV control. The
SOCs of the six banks in Module 3 become balanced at
around 2,800 s. The SOCs of the five modules become
balanced at around 3,100 s, as shown in Fig. 12(b). The cells
are fully charged (e.g., 95% for EVs) by using the CCCV
control. Fig. 12(c) shows the terminal voltage and current of
the battery system. Under the CCCV charge control, the
battery is firstly charged with a constant current; and the
terminal voltage of the battery increases until it reaches the
cutoff value of 126 V, However, by that time the cells have
not been fully charged yet. Therefore, from that moment
onwards, the battery is charged with an adaptively reduced
current at a constant voltage until the SOC of each cell
reaches 95%. These results clearly demonstrate that the
proposed multicell battery design is capable of safe and
effective charging, discharging, and balancing operations.
1
0.8
0.6
0.4
Module
Module
Module
Module
Module
0.2
0
0
1000
2000
3000
4000
Time (seconds)
1
2
3
4
5
5000
(b)
120
Voltage (Volts) & Current (A)
operated in the discharge mode. The MMS determines that
three banks with the highest SOCs are used simultaneously
with a discharge current of 4.68 A. Assume that all of the 18
cells are healthy initially, and the initial SOCs of the six
banks are 100% (Bank 1), 95% (Bank 2), 90% (Bank 3),
85% (Bank 4), 80% (Bank 5), and 75% (Bank 6). The
interval of one control cycle is chosen to be Ts = 75 s. The
battery is continuously discharged until no bank has usable
capacity. However, one cell in Bank 4 fails at 500 s and is cut
off from the bank by the cell switching circuit. This reduces
the SOC of Bank 4 by 33.3%. Figs. 11(a) and (b) show the
SOCs of the six banks and the terminal voltage response of
the battery module, respectively, during the whole discharge
operation. The results clearly show that the MMS
successfully controls the cell switching circuit to balance the
cell banks during the battery discharge although a cell failed.
The SOCs of the six banks are balanced at about 3,100 s.
This capability cannot be achieved by using the existing
balancing circuits. These results clearly demonstrate both the
self-balancing and self-healing characteristics of the battery
module.
100
80
VB
IB
60
40
20
0
1000
2000
3000
4000
5000
Time (seconds)
(c)
Fig. 12. The operation of a five-module battery system in the charge mode:
(a) the SOCs of the six banks in Module 3; (b) the SOCs of the five
modules; and (c) the terminal voltage and charge current of the battery.
The proposed battery system can automatically configure
itself according to the condition of each cell, self-heal from
failures and abnormal operating conditions of single or
multiple cells, self-balance from cell state variations, and
self-optimize to achieve the optimal energy conversion
efficiency. These features are achieved by a cell switching
circuit and a high-performance battery management system.
The proposed design has been validated by simulation studies
in MATLAB Simulink for a battery module containing 6×3
cylindrical lithium-ion cells and a battery system consisting
of five 6×3-cell battery modules. The proposed design is
scalable to larger battery systems for EV/PHEV applications.
APPENDIX
Battery cell: Tenergy 18650; nominal voltage: 3.7 V;
nominal capacity: 2.6 Ah; discharge cutoff voltage (Vcutoff): 3
V; charge cutoff voltage (Vover): 4.2 V; maximum discharge
current: 1C (2.6 A).
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
D. Linden and T. B. Reddy, Handbook of Batteries, 3rd ed., McGrawHill, 2001.
Electronic Publication: MAX 11080 Datasheet, “12-channel, high
voltage Battery-Pack Fault Monitor,” MAXIM.
J. Cao, N. Schofiled, and. A. Emadi, “Battery balancing methods: A
comprehensive review”, in Proc. IEEE Vehicle Power and Propulsion
Conference, Sept. 3-5, 2008, Harbin, China.
J. Shin, G. Seo, C. Chun, and B. Cho, “Selective flyback balancing
circuit with improved balancing speed for series connected lithium-ion
batteries,” in Proc. 2010 International Power Electronics Conference,
21-24 June 2010, pp. 1180.
A. C. Baughman, and M. Ferdowsi, “Double-tiered switched-capacitor
battery charge equalization technique,” IEEE Trans. Industrial
Electronics, vol. 55, pp. 2277-2285, June. 2008.
S. T. Huang, D. C. Hopkins, and C. R. Mosling, “Extension of battery
life via charge equalization control,” IEEE Trans. Industrial
Electronics, vol. 40, no. 1, pp. 96-104, Feb. 1993.
Electronic Publication: BQ 102 Datasheet, “Power LAN Dual-Cell LiIon Battery Monitor with Power Pump Cell Balancing”, Texas
Instrument.
S. Ci, J. Zhang, H. Sharif, and M. Alahmad “A novel design of
adaptive reconfigurable multicell battery for power-aware embedded
networked sensing system,” in Proc. 2007 IEEE Global
Telecommunication Conference, Nov. 26-30, 2007, pp. 1043-1047.
S. K. Mandal, P. S. Bhojwani, S. P. Mohanty, and R. N. Mahapata,
“IntellBatt: Towards smarter battery design,” in Proc. 45th Annual
ACM/IEEE Design Automation Conference, June 8-13, 2008, pp. 872877.
H. Visairo and P. Kumar, “A reconfigurable battery pack for improving
power conversion efficiency in portable devices,” in Proc. 7th
International Caribbean Conference on Devices, Circuits and Systems,
Apr. 28-30, 2008, Mexico.
T. Kim, W. Qiao, and L. Qu, “Self-reconfigurable multicell batteries,”
in Proc. IEEE Energy Conversion Congress and Exposition 2011,
Phoenix, AZ, USA, Sept. 17-22, 2011, pp. 3549-3555.
C. Bonfiglio and W. Roessler, “Cost optimized battery management
system with active cell balancing for lithium ion battery stacks,” in
Proc. Vehicle Power and Propulsion Conference, Dearborn, MI, USA,
Sept. 7-10, 2009, pp. 304-309.
R. Matte, L. Turner, and H. Mettlach, “Voltec battery system for
electric vehicle with extend range,” in SAE int. j. engines 4(1) pp.
1944-1962, Apr. 2011.
[14] T. Kim and W. Qiao, “A hybrid battery model capable of capturing
dynamic circuit characteristics and nonlinear capacity effects,” IEEE
Trans. Energy Conversion, vol. 26, no. 4, pp. 1172-1180, Dec. 2011.
[15] J. Manwell and J. Mcgowan, “Extension of the kinetic battery model
for wind/hybrid power system,” in Proc. 5th European Wind Energy
Association Conference, 1994, pp. 1182-1187.
[16] M. Chen and G. A. Rincon-Mora, “Accurate electrical battery model
capable of predicting runtime and I-V performance,” IEEE Trans.
Energy Conversion, vol. 21, no. 2, pp. 504-511, June 2006.