A Comprehensive Review of Lithium-Ion Cell
Temperature
Estimation Techniques Applicable to Health-
Conscious Fast
Charging and Smart Battery Management Systems
Abstract: Highly nonlinear characteristics of lithium-ion batteries (LIBs) are
significantly influenced
by the external and internal temperature of the LIB cell. Moreover, a cell
temperature beyond the
manufacturer’s specified safe operating limit could lead to thermal runaway and
even fire hazards
and safety concerns to operating personnel. Therefore, accurate information of
cell internal and
surface temperature of LIB is highly crucial for effective thermal management
and proper operation
of a battery management system (BMS). Accurate temperature information is
also essential to BMS for
the accurate estimation of various important states of LIB, such as state of
charge, state of health and so
on. High-capacity LIB packs, used in electric vehicles and grid-tied stationary
energy storage system
essentially consist of thousands of individual LIB cells. Therefore, installing a
physical sensor at each
cell, especially at the cell core, is not practically feasible from the solution cost,
space and weight
point of view. A solution is to develop a suitable estimation strategy which led
scholars to propose
different temperature estimation schemes aiming to establish a balance among
accuracy, adaptability,
modelling complexity and computational cost. This article presented an
exhaustive review of these
estimation strategies covering recent developments, current issues, major
challenges, and future
research recommendations. The prime intention is to provide a detailed
guideline to researchers
and industries towards developing a highly accurate, intelligent, adaptive, easy-
to-implement and
computationally efficient online temperature estimation strategy applicable to
health-conscious fast
charging and smart onboard BMS.
Keywords: electric vehicles; machine learning; Kalman filter; thermal
modelling; online prediction;
Battery Management Systems—Challenges
and
Some Solutions
State of Charge Estimation
Coulomb counting is the easiest approach to estimate the state of charge (SOC) of a
battery [2,3].
Figure 1a gives the approximate Coulomb counting equation that is used to compute
SOC in a
recursive manner. However, Coulomb counting method suffers from the following
sources of errors:
1. Initial SOC error. Since it is a recursive integration, any errors in the initial SOC
assumption will
remain as a bias.
2. Current measurement error. Current sensors are corrupted by measurement noise;
simple, inexpensive
current sensors are likely to be more noisy and possibly biased.
3. Current integration error. Coulomb counting methods employ a simple, rectangular
approximation for
current integration. Such an approximation results in errors that increase with sampling
interval as
the load changes rapidly.
4. Uncertainty in the knowledge of battery capacity [14]. Coulomb counting method
assumes perfect
knowledge of the battery capacity, which is known to vary with temperature, usage
patterns and
time (age of the battery) [15,16].
5. Timing oscillator error. Timing oscillator provides the clock for (recursive) SOC update,
that is, the
measure of time comes from the timing oscillator. Any error/drift in the timing oscillator
will have an
effect on the measured Coulombs.
Alternatively, the open circuit voltage (OCV) can be modeled as a function of the SOC of
the battery.
This OCV-SOC model [17] can be exploited to estimate the SOC based on voltage
measurements. However,
measuring the OCV in real-time during battery operation is not feasible because the
battery needs to be
rested for several hours before the OCV can be measured. While the battery is
operational a measure of
OCV can be obtained by estimating the voltage across the battery ECM; this requires the
estimation of the
ECM parameters as well. Once the OCV is estimated, the SOC can be looked-up [17]
using the OCV-SOC
characterization parameters. Figure 1b summarizes the voltage based approach to SOC
estimation.
The following errors are encountered by the OCV-SOC based state of charge estimation
approach:
1. Errors in the parameters estimated for the electrical ECM of the battery.
2. Voltage and current measurement error.
Most of the advanced BFG’s use a fusion based approach where both the Coulomb
counting method
and the OCV-lookup method a combined in an efficient manner.
State of charge estimation. The fusion based approach is one of the most robust approaches to
accurate battery SOC estimation.
Th fusion approach to SOC estimation (more appropriately, SOC tracking) is modelled as
a recursive
Bayesian estimation problem and by employing a nonlinear filtering approach (such as
an extended
Kalman filter) for online SOC tracking [2,3]. A complete SOC tracking solution involves
the following:
Estimation of the OCV parameters that form part of the state space model through offline
OCV characterization:
The OCV-SOC characterization is stable over temperature changes and aging of the
battery.
Once estimated, these parameters form part of a state-space model with known
parameters.
(ii) Estimation of the dynamic ECM parameters: These parameters can change depending
on the battery age,
temperature, and SOC, therefore, they must be estimated in real time.
(iii) Estimation of battery capacity: Even though the the manufacturer provides the
nominal capacity
of the battery, it changes over time. Some important factors that cause capacity fading
are,
elevated temperature, cycling (usage), depth of discharge patterns, and calendar aging.
Due to
this, the battery capacity needs to be estimated in real-time for an accurate BFG.
Capacity estimation
is still being actively investigated in the literature [14].
(iv) Model parameter-conditioned SOC tracking: As soon as the model parameters are
estimated, a filtering
approach can be used to track the SOC using the state-space model discussed above. In
order to
do this, numerous filtering approaches, including extended Kalman filter, Unscented
Kalman filter
and particle filter, were experimented in the literature. However, it is observed that the
resulting
state-space model contains correlated process and measurement noise processes.
Properly addressing
the effect of these correlations will yield better SOC tracking accuracy.
