25-08-2012, 05:04 PM
Design, Testing, and Validation of a Simplified Control Scheme for a Novel Plug-In Hybrid Electric Vehicle Battery Cell Equalizer
1Design, Testing, and Validation.pdf (Size: 660.98 KB / Downloads: 42)
Abstract
In order to meet cost targets for hybrid electric
(HEV), plug-in hybrid electric (PHEV), and all-electric vehicles
(EV), an improvement in the battery life cycle and safety is
essential. Recently, lithium batteries, in the form of lithium-ion,
lithium-polymer, or lithium iron phosphate have been explored.
Despite research initiatives, lithium-based batteries have not yet
been able to meet steep energy demands, long lifetime, and low
cost of vehicular propulsion applications. One practical approach
to improve performance is to use power electronics intensive cell
voltage equalizers, in conjunction with on-board energy storage
devices. The purpose of this paper is to introduce a simplified control
scheme, based on open-circuit voltage estimation, for a novel
cell equalizer configuration, with the potential to fulfil expectations
of the following: 1) low cost; 2) large currents; and 3) high efficiency.
Issues, such as the limitations on maximum and minimum
cell voltage, noise, and quantization errors, are explored. Finally,
a comprehensive comparison between the theoretical test results
and practical equalization test results is presented.
INTRODUCTION
IT IS KNOWN today that the batteries are the stumbling
blocks in driving electric vehicles. In fact, the issues related
to lithium rechargeable cells can be summed up by one topic:
a battery of a plug-in hybrid electric vehicle (PHEV) consists
of a long string of cells (typically 100 cells, about 360 V),
where each cell is not exactly equal to the others, in terms
of the capacity and internal resistance, because of the normal
dispersion during manufacturing. However, the most viable
solution for today’s problem might not originate merely from
the changes in the battery chemistry [1]. In fact, a smarter
solution lies in a power electronic battery cell voltage equalizer,
which can improve not only the cycle life of batteries, but also
their calendar life, power, and safety [2], [12], [15], [16].
NOVEL CELL VOLTAGE EQUALIZER TOPOLOGY
The proposed novel cell voltage equalizer topology is depicted
in Fig. 1. This cell equalizer has the ability to drive
high current with high efficiency. It may also cost less than
the typical implementations of cell equalizers [2], because it
uses only one metal–oxide–semiconductor field-effect transistor
(MOSFET) (and driver) per cell. On the other hand, the
control core is more complex than a typical equalizer, but it
may be implemented in the same microcontroller that performs
safety monitoring.
The control block calculates the timing of the MOSFETs
based on the voltage of the cells and the desired current. The
current is calculated to reduce the equalization time, taking into
account the following: 1) the voltage limit of each cell; 2) the
total equalizer current limit; and 3) the equalizer timing limits.
For some applications, the optimal efficiency might also be
considered. Some of these requirements are mutually exclusive;
for example, the best efficiency in the equalizer occurs when the
voltage of each cell is equal, but the maximum power transfer
is usually limited by the cell voltage or the equalizer current
capabilities. Based on these considerations, the controller has
to decide the amount of energy flow to/from the cells, and
calculate the MOSFET timings, in order to obtain the desired
current based on the cell voltages.
LITHIUM-ION BATTERY MODELING
There exist several equivalent models for batteries, with
varying complexities [6]–[8]. For the equalizer controller explored
in this paper, the battery cell equivalent electrical model
presented in [6] has been chosen, because of its simplicity and
high precision in the range of SOC typically used in EV (from
30% to 80% SOC) [1], [2]. The equivalent model of a battery
cell is shown in Fig. 3.
In the equivalent circuit of Fig. 3, the Rint represents the
total steady-state cell internal resistance, including connections;
Ra, Rb, Ca, and Cb model the dynamic response, in the case
tested in this paper with a time response of about 3 and 50 s;
the open-circuit voltage (VOC) has a nonlinear relationship
with SOC, temperature and age, and represents the chemically
stored energy. The internal resistance is also dependent on
SOC, temperature and age.
PROPOSED CELL EQUALIZER CONTROL STRATEGY
The cell equalizer controller is composed of several blocks:
1) cell VOC estimation, overviewed in the previous section;
2) VOC difference calculation, which is used as the error signal;
3) a proportional integrative block (PI), that will affect the
error signal; 4) desired cell current calculation, based on the
PI output; 5) cell current and voltage limiter, estimated using
the desired cell current; and 6) the timing calculations for the
MOSFETs, calculated using (1). Fig. 8(a) shows the equalizer
controller block diagram, while Fig. 8(b) shows the equalizer
controller details.
CONCLUSION
In order to control a complex multiinput, multioutput system,
like the novel battery cell equalizer for an EV/PHEV
application, with several contradictory requirements, such as
the following: 1) low cost; 2) high equalizer current; and
3) high efficiency, in a harsh environment, tradeoffs have to
be considered. Due to the cost requirements, limits have to
be set on the processing power of the microcontroller. This
paper presented a simplified control technique to estimate the
VOC of each cell, using little CPU time. A controller using
this estimator has been also presented, and its performance and
precision range have been verified. Finally, the simulation of the
modeled equalizer has been validated against the experimental
test measurements, taken from the equalizer prototype. The
remaining issues, like noise and differences in the battery
model, have been analyzed and weighted against processing
power and cost.