Analysis of Battery Management Systems

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 09 Issue: 06 | June 2022 www.irjet.net p-ISSN: 2395-0072
Analysis of Battery Management Systems
Dr. Venugopal P4
1,2,3Student, Vellore Institute of Technology, Vellore, Tamil Nadu
4Professor, Vellore Institute of Technology, Vellore, Tamil Nadu
------------------------------------------------------------------------------***---------------------------------------------------------------------------
ABSTRACT: The battery management systems also
referred to as BMS is a battery management unit that is
not only responsible to measure the state of battery
accurately, but it also ensures safe operation and also a
prolonged battery life. Because of their high energy
density, lifespan, nominal voltage, power density, and low
cost, lithium-ion (Li-ion) batteries have received a lot of
attention in the EV industry. A smart battery management
system (BMS) is an essential component in electric
vehicles it not only measures the states of the battery
accurately, but it also ensures safe operation and extends
battery life. Estimating the state of charge (SOC) of a Li-
ion battery accurately is difficult due to the battery's
highly time variant, non-linear, and complex
electrochemical system. So in this paper we will study
about the various parameters related to battery
performance like the state of charge, state of health and
also look at over charge and under charge circuits that
were designed to monitor the battery health.
KEYWORDS: State of charge, state of health, lithium ion
batteries, matlab.
I. INTRODUCTION
What is BMS
The lithium ion batteries that are used now a days are
very fragile in nature when compared to lead-acid, or
NiCd batteries which makes them useful in various
applications like cell phones or industrial equipment etc.
these lithium ion batteries that are used, needs a
management system to monitor its health by calculating
various parameters. The complexity of any battery
management system totally depends on the type of
application for which this system has to be designed. for
applications like mobile phones or e-book readers etc, a
simple BMS IC can be used, this IC will monitor the
voltage, current, state of charge and various other
parameters. these parameters will give us the overall
health of the battery. But for bigger applications like
electronic vehicles, we need a better alternative. A normal
IC won’t be sufficient. We need very high-end algorithms
and processing to measure various parameters. We will
go through these in detail as we move further with our
paper.
First in this paper we will go through various battery
management system requirements in detail to
understand how can we actually optimise the
performance of the lithium ion batteries in any system.
Battery Management System Requirements
Several factors must be considered when designing a
bms. There are various Relevant parameters include
temperature acquisition, voltage and current acquisition,
battery pack etc that are important to ensure proper EMI
robustness, functional safety in terms of redundancy,
galvanic isolation, balancing, and power consumption.
Temperature acquisition: One of the most difficult tasks
in BMS design is determining the most
accurate temperature possible. It is sometimes necessary
to obtain peripheral temperatures, such as those of
contactors, fuses, or even the electric busbar. Heavy
copper bus bars should be avoided as much as possible.
Temperature requirements must take into account three
use cases: charging, discharging, and storage.
Temperature sensor to voltage sensor channel ratios
typically range from 2:3 to 2:12. At too high current rates,
the significant effect of lithium plating can occur in the
normal temperature range.
Voltage acquisition: A battery management system for
lithium based batteries needs to have at least one voltage
acquisition channel per serially connected cell. High data
acquisition rates are advisable, if oversampling is chosen
to have a positive effect on the system behaviour. When
large pulses occur in an application, faster monitoring can
be necessary.
Current acquisition: For dynamic SOC determination,
coulomb counting is used. This method simply integrates
the current flowing into or out of a battery. The
determined SOC will be different from the actual SOC due
to noise and offsets. There are approaches using
algorithms and parametrised models to retrieve usable
SOC values.
Balancing: There are several factors that have to be taken
care of before a battery can be charged and discharged,
e.g., resulting from weight constraints, or required
charging currents that lead to balancing currents.
Rishita Singh1 Shubham Adgaonkar2 Pranav Dabhade3
© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 714

So in this paper our main objective is to study about the
the state of charge and state of health of the battery. Once
studied, the simulation results are then observed using
MATLAB (Matrix Laborotory). The flow of the paper will
be such that we will first Monitor the cell voltage and
temperature. Then the values of state-of-charge and state-
of-health are estimated. Once that is done for those values
we will limit thepower input and output for thermal and
overcharge protection.
As lithium ion cells are used so we need to balance the
state of charge of these individual cells. After this is done
we need to isolate the battery pack from the load
whenever necessary to avoid unnecessary loading effects
in the circuit. Then the entire design was implemented on
the software using MATLAB and Simulink. Once the
simulation was done, the algorithm was deployed on a
raspberry pi microcontroller and finally it was tested on a
lithium ion battery pack.
Literature Survey
[1]Many battery-operated industrial and commercial
equipment use battery management systems (BMSs) to
make battery operation more efficient and battery
estimates nondestructive state. The study is related to
existing BMS methodologies and proposes a new design
methodology for a generalized dependable BMS. This
BMS has a number of advantages over existing systems,
the most important of which is that it is fault-tolerant and
has battery protection. The proposed BMS is made up of a
series of smart battery modules (SBMs), each of which
performs battery equalization, monitoring, and
protection for a string of battery cells. In the lab, an
evaluation SBM was designed and tested, and the
practical findings back up the theoretical predictions.[2]
The hardware components of battery management
systems (BMS) for electric vehicles and stationary
applications are the subject of this research. The goal is to
provide a high-level overview of existing principles in
cutting-edge systems, allowing the reader to estimate
what has to be considered when creating a BMS for a
specific application. Following a brief review of general
needs, various different battery pack topologies and their
implications for the BMS' complexity are considered. As
examples, four battery packs from commercially available
electric vehicles are displayed. Later on, implementation
difficulties such as measuring required physical variables
(voltage, current, temperature, etc.) as well as balance
issues and solutions are examined. Finally, safety and
dependability factors are explored.[3] The battery is an
important component of a hybrid electric vehicle's (HEV)
drivetrain. The battery's unpredictable internal
resistance change throughout the operating cycle is a
significant constraint to its effective performance and
reliability. Because temperature has a significant impact
on internal resistance, the battery management system
keeps track of cell and battery pack temperature in
relation to the state of charge to avoid thermal runaway.
In order to improve their performance, Li-ion batteries,
which provide prospective solutions to HEVs' energy and
power density demands, require a good thermal
management system. The goal of this work is to create a
battery pack model that looks at how internal resistance
changes as a function of temperature.[4]A battery
management system (BMS) is a system that controls the
operation of a rechargeable battery (cell or battery pack)
by preventing the battery from exceeding its safe
operating limits and monitoring its state of charge (SoC)
and condition of health (SoH). BMS has been a vital
component of hybrid and electric vehicles (HEVs) for
many years (EVs). With run-time battery monitoring for
any critical hazard circumstances, the BMS ensures the
system's and user's safety. The design and simulation of
BMS for EVs are presented in this paper. The whole BMS
model, as well as all other BMS functional blocks, are
implemented in the MATLAB R2012a Simulink toolbox.
The Neural Network Controller (NNC), Fuzzy Logic
Controller (FLC), and Statistical Model are the BMS given
in this research study. The testing results are used to
extract the battery parameters needed to construct and
simulate the BMS, which are then incorporated into the
model. The SoC is estimated using the Neuro-Fuzzy
technique, which is utilized to mimic the electrochemical
behavior of the Lead-acid battery (chosen for case study).
The Statistical model is utilized to address the SoH of the
battery. The findings of the battery cycle tests were used
for initial model design, Neural Network training, and
later, BMS design and simulation using Simulink.
Experiments and MATLAB/Simulink simulations are used
to validate the simulation results. This model has a SoC
accuracy of about 97 percent and a SoH accuracy of about
97 percent.[5]The qualities and performance of batteries
under various operating situations are critical in their
applications, particularly in the field of electric vehicles
(EVs). It is possible to forecast and optimize battery
performance using an accurate and efficient battery
model, particularly in practical runtime applications such
as Battery Management Systems (BMS). Before building a
solid battery model, an accurate method for predicting
battery parameters is required. For battery modeling,
three well-known battery models are often used: the
Partnership for a New Generation of Vehicles (PNGV),
Thevenin, and the second order Battery Model. For the
prediction of three battery models, a new technique of
parameter estimation was developed utilizing the
Matlab/Simulink® parameter estimation tool. A 12Ah
3.7V Lithium Polymer (Li-Po) battery is put through a
series of conventional tests at various states of charge
(SoC), temperatures, and current rates, and the three
model parameters are estimated based on the results. The
spreadsheet approach for PNGV battery parameter
estimate is compared to the Matlab/Simulink method for
parameter estimation.

2. Technical specifications
In this paper there are 2 methods that we used to
measure the State Of Charge(SoC). First is the Open
Circuit Voltage Measurement method. Here we collect
various parameters like voltage and current and then
finally using MATLAB simulation results are seen.
1.1 Data collection circuits for current measurement
First we will need the current and the voltage values for
further data processing. In the circuit shown below, the
current measurement is been carried out
Fig.no. 1a
Here battery connected the to a 1 Ω power resistor and a
60 Ω test load using a setup similar to the Arduino
voltmeter. The Arduino analyses the voltage drop across
the power resistor before using Ohm's Law: I=V/R. The
Arduino then uses the Serial monitor to send the current
value to Python.
This is the complete circuit which has a relay and we will
see the functioning of this circuit in detail
Fig.no.1b
Relay Off: Current Measurement: The battery is
connected to load and discharges through R-Power.
Arduino measures the voltage drop across R-Power by
measuring the drop between A0 and GND.
Relay On: OCV Measurement: Arduino writer D2 to HIGH,
which energizes the relay and disconnects the batteries
from load. Arduino measures the voltage drop between
A1 and GND to determine the OCV of the batteries. R1
and R2 create a voltage divider so the OCV reading stays
within the Arduino’s 5V measurement bounds.
Fig.no. 1c
Over-Charge Protection circuit
Over charge protection is a very important step to ensure
that our battery performance is optimised.
The circuit given above cuts off the charging half of the
circuit by reverse biasing a Zener diode.
The batteries charge when the relay is in its non-
energized state, so to stop charging the cells, the relay
must become energized when the voltage drop across the
cells reaches 8.4V. To prevent the relay from energizing
before then, the ground terminal is connected to the
collector of a transistor whose base is connected to a
Zener diode, a Schottky diode, and a resistor.
When the batteries are below 8.4V, the voltage drop
across the components in the base is insufficient to
reverse bias the Zener, which prevents current flow and
keeps the transistor in its non-conducting state (which
acts as an open circuit to the relay).
When the batteries reach 8.4V, though, base current flows
which turns on the transistor, which allows current to
flow through the relay, which opens the switch and
disconnects the batteries. An LED then turns on to
indicate that the batteries are fully charged.

Fig.no.1d
Over-Discharge Protection circuit
A 5V regulator needs at least 5V to operate, and we need
to disconnect the batteries at 2.5V So, since the 5V
regulator is the only component the batteries are
connected to, the circuit will open if the batteries fall
below 5V as the regulator will turn off and produce no
output.
2.2a Deployment of the model on raspberry pi and testing
it on lithium ion battery pack
We trained a model on raspberry pi . the inputs to this
model are voltage, current and the charging , discharging
rate of the battery. We took this as an input since these
are some parameters that has the maximum effect on the
battery performance.
Finally the output can be observed a display/monitor
connected to Raspberry Pi 3. The following parameters
line the SOC(State Of Charge), SOH(State Of Health) , Qm
(measured maximum capacity in A-hr) and EoL are
observed as the output.
2.3 Data Analysis
SOC based on OCV measurements
• During data collection, the Arduino collected over
50k data points during a ~21 hour discharge.
• OCV measurements as [measurement number, OCV],
normalized the data points to a 0 to 100% scale, from
which a polynomial function is fitted to get the state
of charge as a function of open circuit voltage(OCV).
SOC based on Current measurements
• The 18650 cell is rated at 1.2Ah, so we can easily
calculate the total number coulombs the battery can
hold.
• We can then calculate the coulombs used from our
current reading. By dividing the coulombs used over
our total coulombs, we get the state of charge used.
And by subtracting/adding that to the initial state of
charge, we end up with the state of charge as a
function of time and current.
Total Coulombs Charge of the system = 1.2
Amphr*(60mins/1hr)*(60sec/1min)
The value of the above equation comes out to be 4320
coulombs.
Therefore, the total coloumbs used will be given by:
l*(change in t)
The SoC(State of charge) used in percentage is given by ,
(coulombs used/ total coulombs)*100
So finally SoC(t,l)= SoC0 - SoCused
SoC - (l*(change in t/4320))*100
This is how the state of charge and the state of health
of the lithium-ion battery back is determined.
Once the value of the SoC is found by using both the
methods mentioned above, we use a Kalman Filter
model to check which output is the best for our
application. As shown in the diagram below, first the
data is collected and then SoC is found as a function
of both current and voltage. Then these 2 values are
compared and the best is then given at the output.

Fig 3: Flow of proposed model
3. Results:
In the OCV method, for simulation Tamb ambient
temperature is considered. There are 2 lithium-ion cells
that are connected to a voltage load in parallel. This is
used to calculate the battery voltage. The input current is
also measured as shown in the figure. Finally, these are
the simulation results that were obtained. Figure below
shows 2RC model taken as a reference from mathworks
website and changed few parameters as per condition.
Fig.no.4
To calculate the SoC based on current measurements,
mathematical analysis is done which is mentioned in
section 2.3. So finally, after both these results are
obtained, we pass them through a Kalman Filter model to
find the best value of SoC.
After the SoC is found we need to find the State of Health.
It is expressed as a percentage and is defined as the ratio
of the maximum battery charge to the rated capacity.
Fig 5
When the battery is manufactured, the SoH of the battery
is 100% but as time goes by and battery is used, the
charges in the battery are permanently lost. This affects
the battery performance thereby affecting the battery
health.
Using the values of the SoC this graph of SoH was
obtained.
In the above graph we can clearly see that as the number
of charge- discharge cycles increase, the SoH of the
battery keeps on reducing. Once this value reaches below
a threshold value battery should no longer be used for
any application.
References:
[1]Chatzakis, J., Kalaitzakis, K., Voulgaris, N. C., &
Manias, S. N. (2003). Designing a new generalized
battery management system. IEEE transactions on
Industrial Electronics, 50(5), 990-999.

[2]Lelie, M., Braun, T., Knips, M., Nordmann, H.,
Ringbeck, F., Zappen, H., & Sauer, D. U. (2018). Battery
management system hardware concepts: an overview.
Applied Sciences, 8(4), 534.
[3]Sen, C., & Kar, N. C. (2009, September). Battery pack
modeling for the analysis of battery management
system of a hybrid electric vehicle. In 2009 IEEE
Vehicle Power and Propulsion Conference (pp. 207-
212). IEEE.
[4]Kumar, B., Khare, N., & Chaturvedi, P. K. (2015,
October). Advanced battery management system
using MATLAB/Simulink. In 2015 IEEE International
Telecommunications Energy Conference (INTELEC)
(pp. 1-6). IEEE.
[5]Daowd, M., Omar, N., Verbrugge, B., Van Den
Bossche, P., & Van Mierlo, J. (2010, November).
Battery models parameter estimation based on
MATLAB/Simulink®. In Proceedings of the 25th
Electric Vehicle Symposium (EVS-25), Shenzhen,
China
[6] Huria, Tarun & Jackey, Robyn & Gazzarri, Javier &
Saginaw, Michael & Sanghvi, Pravesh. (2013). Battery
Model Parameter Estimation Using a Layered
Technique: An Example Using a Lithium Iron
Phosphate Cell. SAE Technical Papers. 2.
10.4271/2013-01-1547.

Analysis of Battery Management Systems

More Related Content

Similar to Analysis of Battery Management Systems (20)

More from IRJET Journal (20)

Recently uploaded (20)

Analysis of Battery Management Systems