Volume 2, Issue 5, May 2017 International Journal of Innovative Science and Research Technology

ISSN No: - 2456 - 2165

Energy based Battery Management System for

Microgrids using Fuzzy Logic Controller
Sushma Maney.D Eranna
Dept. of Electrical and Electronics Engineering Dept. of Electrical and Electronics Engineering
Dr. Ambedkar Institute of Technology Dr. Ambedkar Institute of Technology
Bangalore, India Bangalore, India
sushmamaney17@gmail.com eranna64@gmail.com
Abstract This paper presents an energy management by the load, the inclusion of Energy Storage System(ESS)
system to effectively mitigate grid power profile (e.g. batteries, flywheels, ultra-capacitors) and Energy
fluctuations. The system assumes that neither renewable Management Systems(EMS) are highly recommended in order
generation nor the load demand is controllable. This to improve the system stability and its performance. In
energy management system consists of a low complexity general, MGs are capable to work in both grid-connected and
Fuzzy Logic Controller embedded with 25 rules in it. The stand-alone mode. They are defined a slow voltage systems
approach is, monitoring rate of change in energy and State comprised of loads, Distributed Generation(DG) units and
of Charge (SOC) of battery thus controlling the power storage devices ,that are connected to the mains at a single
delivered/absorbed by the grid. Point of Common Coupling(PCC).

Keywords Distributed power generation, energy In short, for the case under study, the EMS should be designed
management, fuzzy control, micro grid, Renewable energy with the objective of smoothing the power exchanged with the
sources and smart grid. grid, concurrently satisfying at any time the load demand (i.e.
there is no demand side management) and the ESS constraints.
I. INTRODUCTION This heuristic knowledge suggest the use of Fuzzy Logic
Control to the design of the EMS for the case under study,
This paper presents the modeling, analysis and design of fuzzy since this approach easily integrates the experience of the user
logic controller in a Battery management system for a Wind/ rather than using a mathematical model of the system. Taking
Solar hybrid system. With the variation of wind speed, solar the same input variables as in[1], the authors presented in [2]
isolation and the load demand, the fuzzy logic controller the design of a FLC with only 25-rules which slightly
works effectively by turning on and off the batteries. The improved the battery SOC and the grid power profile obtained
entire designed system is modeled and simulated using in[1].This work presented a detailed description of the rule-
MATLAB/Simulink Environment. The control process of the base and the Membership Functions (MF) design, which
battery charging and discharging is non-linear, time varying parameters (i.e. number and mapping) were adjusted to
with time delays. It is a multiple variable control problem with optimize a set of quality criteria of the MG behavior through
unexpected external disturbances. Many parameters such as an off line learning process simulation.
the charging rate, the permitted maximum charging current,
the internal resistor, the port voltage, the temperature and Furthermore, using the same design methodology, an
moisture, etc. keep changing during the charging and improved EMS design based on FLC was presented in [3].
discharging process cant be directly obtained, so it is difficult This new design was considering the MG Net Power Trend
to achieve the optimal operation performance by using (NPT) as an additional input of the FLC, resulting in a 50-
traditional control methods. Hence a fuzzy control unit for rules FLC. Even though the results evidence a low-frequency
battery charging and discharging used in a renewable energy grid power profile with minimum fluctuations, the controller
generation system is developed. complexity was increased.

The environmental and economic benefits related to the Additionally, a common drawback of all these previous
reduction of both carbon dioxide emission and transmission designs [1], [2], [3] is that they do not operate properly when
losses have made distributed renewable generation systems the RES generation exhibits strong differences from one day
became a competitive solution for future smart grids. In this to the next one. In these cases, the battery SOC can eventually
context micro grids are considered as the key building blocks reach the undesired thresholds, thus compromising the battery
of smart grids and have aroused great attention in the last lifetime. With the aim of improving the aforementioned
decade for their potential and the impact they may have in the designs as well as simplifying the FLC complexity (i.e. to
coming future. reduce the controller inputs number and its rule-base), this
work presents a new FLC-based EMS of only two-inputs, one-
The Microgrid (MG) concept has been discussed by several output and 25-rules. As it will be seen, the key factor of the
authors. Additionally ,in a MG scenario due to the stochastic new design is to consider the MG Energy Rate - of - Change
nature of both the renewable sources and the power consumed (ERoC) as an input in order to anticipate the system behavior.

IJISRT17MY119 www.ijisrt.com 382

Volume 2, Issue 5, May 2017 International Journal of Innovative Science and Research Technology
ISSN No: - 2456 - 2165
The design methodology will follow the procedure and the the battery out of the secure limits, the EMS strategy should
optimization process developed in [2] (i.e. off-line controller cut off the power delivered / absorbed by the battery.
parameter setting process).
In these cases PBAT = 0, or equivalently according to (3) P GRID
The paper is organized as follows. Section II describes the = PLG meaning that all the power fluctuations are handled by
architecture and variables of the Energy Management System. the grid. A battery SOC estimator, shown in Fig. 2 is used to
Section III describes the Microgrid. Section IV is about the estimate the current battery SOC, which is expressed as:
proposed FLC Design and Section V presents the design of the
proposed fuzzy Implementation. The experimental validation SOC (n) = SOC (n-1) - SOC (n), (6)
of the proposed EMS design is presented in Section VI.
Finally, Section VII presents the main conclusions of this Where SOC(n) represents the battery energy variation during
paper. the sampling period Ts and can be estimated using the general
definition of the energy evolution ei of a power variable Pi
II. ENERGY MANAGEMENT SYSTEM during a period time T. Therefore, for sampled variables and
assuming equal integration and sampling periods (i.e . T =
Ts). The case under study assumes that the wind and PV
The system consists of Wind Turbine (WT), Photovoltaic modules are in charge of extracting the maximum renewable
conversion modules and a battery charger sharing the same power and that the AC load power consumption is not
DC bus. The system also includes a bidirectional inverter- controllable. In other words, PLOAD and PGEN (hence PLG)
rectifier module controlling the power exchanged with the cannot be controlled. In contrast, the power exchanged with
load. the grid PGRID will be controlled by means of the bidirectional
inverter - rectifier, where as the battery charger will handle, if
able to, the resulting battery power PBAT according to (3).

Finally the main aim of the EMS design is to control the

power inverter-rectifier in order to smooth the power profile
exchanged between the grid(i.e. minimizing the grid power
fluctuations and power peaks) while concurrently keeping the
battery SOC within secure limits.

III.MICROGRID DESCRIPTION
Fig. 1. Energy Management System for Residential Grid

The study developed in this paper is carried out for a

Referring to Fig.1, the power fluxes are considered positive microgrid with renewable energy sources and domestic load.
according to the direction of the corresponding arrows. The The MG includes a domestic AC load with a rated power of
net power ,PLG,and the grid power ,PGRID,can be expressed 4kW, a Photovoltaic (PV) array of 1 kW, a small Wind
as follows: Turbine (WT) of 2.5kW, and an ESS formed by a lead-acid
battery bank with a rated capacity of 40kWh. The grid
PLG=PLOAD PGEN, (1)
PGEN = PPV+PWT, (2) power quality criteria is defined so that the lower the criteria
PGRID = PLG - PBAT, (3) values are, the EMS performance is better.

Where PLOAD is the load power, PGEN is the renewable A. Positive and Negative Grid Power Peaks
source generated power, PPV is the photovoltaic power, PWT
is the wind turbine power and PBAT is the battery power. The positive and negative grid power peaks, PG,MAX and
PBAT depends directly to the battery SOC, which should be PG,MIN, are defined as the maximum value of power
kept at any time between a minimum and maximum limits, delivered by the grid.
SOCMIN and SOCMAX, respectively, to preserve the battery
lifetime namely: B. Maximum and Average Power Derivative

SOCMIN SOC (n) SOCMAX , (4) The Maximum Power Derivative (MPD) is defined as the
where: maximum absolute value of the slopes during one full sample
time.
SOCMIN = (1-DOD).SOCMAX, (5) The Average Power Derivative (APD) is defined as the
absolute value of the annual average value of the slopes of two
DOD being the battery Depth of Discharge. This study consecutive samples.
considers a maximum DOD of 50%, since the lifetime of this
type of battery is significantly reduced when operates at high
DOD levels[36].In order to avoid discharging/ overcharging

IJISRT17MY119 www.ijisrt.com 383

Volume 2, Issue 5, May 2017 International Journal of Innovative Science and Research Technology
ISSN No: - 2456 - 2165
IV. PROPOSED FLC DESIGN demand and to keep the battery SOC within secure limits.
Consequently, the controller output allows the interaction
An improved FLC- based EMS design is presented in this between the MG and the mains.
section with the aim of minimizing the power peaks and
fluctuations in the grid power profile while keeping the battery
SOC evolution within secure limits as well as to reduce the
FLC complexity .
s
The new fuzzy EMS design suggests computing the grid
power as the sum of the average value of the MG net power,
PAVG(n), and an additional component, PFLC(n) which is in
charge of modifying the grid power profile to keep the battery
SOC within the secure limits at any time.

Thus, the grid power profile is defined as follows:

PGRID(n) = PAVG(n) + PFLC(n), (7)

The proposed design computes by means of a FLC the

additional component PFLC (n) from the following two
inputs: Fig.3. Fuzzy EMS designed on MG energy rate- of change
and SOC block diagram
i. The SOC of the battery SOC(n),
The control block diagram of this strategy is shown in Fig.3,
ii. The MG energy rate -of- change, PAVG(n) where PAVG (n) is obtained from PLG (n) by means of a
LPF, PAVG (n) is obtained by a digital filter which
The FLC uses the SOC of the battery, SOC (n), as an input to implements and limiting the high-frequency gain and noise
check its value at any time in order to fit the constraints associated with the derivative term [41].Fig.3 also includes the
imposed by the maximum DOD of the battery and to preserve battery SOC Estimator and the fuzzy controller.
its life. Furthermore, PAVG(n) gives to the FLC the
information of the magnitude of the MG energy change of two V. FUZZY IMPLEMENTATION
consecutive samples as shown in Fig.2.
The Fuzzy implemented here is a Mamdani -based inference
and defuzzification of Center of Gravity with two - inputs,
PAVG(n) and SOC(n), and one-output PFLC(n), which
represents the second component of the grid power. Regarding
the FLC design, the adjustment of all parameters involved in
the fuzzy controller (e.g. number of MFs per Input / output,
type, mapping, rule - base).

The procedure followed is described and summarized in the

next steps:

Fig.2.Slopes produced by two consecutive samples and Step1:Set the initial FLC design.
average net power profile. Set the MF of inputs and outputs variables: number ,type and
mapping. Set the initial rule-base.
In this regard, a positive slope in Fig. 2 (e.g. m1, m4, m5, m6,
m8) means a reduction of the renewable power generation or Step2: Adjust the inputs and outputs MFs. Using the real
an increase of the load consumption in the MG. On the recorded data, adjust the inputs / outputs parameters of the
contrary, a negative slope (e.g. m2, m3, m7) corresponds to a MFs to minimize the quality criteria of Section III.
MG renewable power generation increase or a load
consumption decrease. Step3: Optimize the initial rule -base. Using the real recorded
data, adjust the initial rule base to minimize the quality
It is worth noting that PAVG (n) can be understood as the criteria.
local prediction of the battery SOC future behavior if the grid
power is not modified. For this reason, from the information By analysis of previous papers and previous experimental
of SOC(n) and PAVG(n) the FLC is in charge to modify results, the optimization process leads to optimized FLC rule-
PFLC(n) to increase, decrease or maintain the power delivered base presented in Table I.
/absorbed by the mains to concurrently satisfy the load power

IJISRT17MY119 www.ijisrt.com 384

Volume 2, Issue 5, May 2017 International Journal of Innovative Science and Research Technology
ISSN No: - 2456 - 2165
Table I :- Optimized Flc Rule-Base

As a result of this rule - base five triangular MFs shown in

Fig.4 and are defined for each input variable, and correspond
to five fuzzy subsets noted as NB, NS, ZE, PS and PB where Fig. 5 a. PGEN b. PLOAD c. PBAT
B represents Big, S Small, N Negative, P Positive and
ZE Zero. VI. EXPERIMENTAL VALIDATION

The proposed Fuzzy Logic design for Energy Management

System is programmed in Matlab/Simulink platform. In order
to experimentally validate the proposed EMS, the system has
been tested in laboratory conditions. The experimental results
are as shown in Fig. 5. The difference between generated
power and load power is stored in the battery. If the difference
is positive the battery charges and if negative battery
discharges. This is experimentally validated under laboratory
condition.

VII. CONCLUSION

In this proposed model a design of fuzzy logic control to

achieve optimization of a Battery management system for a
Wind/ Solar hybrid system is presented. According to the
variation of the load, the fuzzy logic controller works
(a) Input variable input 1
effectively by charging and discharging the battery as per the
grid status. Simulation results were obtained by developing a
detailed dynamic hybrid system model. From the simulation
results, the system achieves power equilibrium, and the battery
SOC maintains the desired value for extension of battery life.

The control process of the battery charging and discharging is

non-linear, time-varying with time delays. It is a multiple
variable control problem with unexpected external
disturbances. A fuzzy control unit for battery charging and
discharging used in a renewable energy generation system is
developed. Simulation results based on fuzzy strategies show
that the control unit has satisfied performance in a laboratory
environment.

Current work is focused on the extension of the proposed

approach by implementing predictive method of control for
(b) Input variable input 2 better system efficiency.

Fig.4. Input Variable membership functions.

IJISRT17MY119 www.ijisrt.com 385

Volume 2, Issue 5, May 2017 International Journal of Innovative Science and Research Technology
ISSN No: - 2456 - 2165
VIII. REFERENCES [15] W.A. Bissch off and R.Gouws, Energy management
system for a residential grid-tied micro-grid, in 2015
s[1] D. A. Aviles, F. Guinjoan, J. Barricarte, L. Marroyo, P. International Conference on the Domestic Use of Energy
Sanchis, and H. Valderrama, Battery management fuzzy (DUE), CapeTown, SouthAfrica,Mar.2015,pp.8591.
control for a grid- tied microgrid with renewable generation, [16] H. Zhou, T. Bhattacharya, D. Tran, T.S.T. Siew, and
in IECON 2012 - 38th Annual Conference on IEEE Industrial A.M. Khambadkone, Composite Energy Storage System
Electronics Society, Montreal, QC, Canada, Oct. 2012, pp. Involving Battery and Ultra capacitor With Dynamic Energy
56075612. Management in Microgrid Applications, IEEE Trans .Power
[2] J. J. Barricarte, I. S. Martn, P. Sanchis, and L. Marroyo, Electron., vol.26, no.3, pp.923930,Mar.2011.
Energy Management Strategies For Grid Integration Of
Microgrids Based On Renewable Energy Sources, in 10th
Int. Conf. Sustain. Energy Technol., Istambul, Turkey, Sep.
2011, pp. 47.
[3] D. Arcos-Aviles, N. Espinosa, F. Guinjoan, L. Marroyo,
and P. Sanchis, Improved Fuzzy Controller Design for
Battery Energy Management in a Grid Connected Microgrid,
in IECON 2014 - 40th Annual Conference on IEEE Industrial
Electronics Society, Dallas, TX, USA, Oct. 2014, pp. 2128
2133.N.
[4] P. Siano, C. Cecati, H. Yu, and J. Kolbusz, Real time
operation of smart grids via FCN networks and optimal power
flow, IEEE Trans. Ind. Inf., vol. 8, no. 4, pp. 944952, Nov.
2012.
[5] R. H. Lasseter, MicroGrids, in Proc. Power Eng. Soc.
Winter Meeting, Jan. 2002, vol. 1, pp. 305308.
[6] F. Katiraei, R. Iravani, N. Hatziargyriou, and A. Dimeas,
Microgrids management, IEEE Power and Energy Mag.,
vol. 6, no. 3, pp. 5465, May. 2008.
[7] Pan and S. Das, Kriging Based Surrogate Modeling for
Fractional Order Control of Microgrids, IEEE Trans. Smart
Grid, vol. 6, no. 1, pp. 3644, Jan. 2015.
[8] Q. Jiang, M. Xue, and G. Geng, Energy management of
microgrid in grid-connected and stand-alone modes, IEEE
Trans. Power Syst., vol. 28, no. 3, pp. 33803389, Aug. 2013.
[9] D. E. Olivares, C. A. Caizares, and M. Kazerani, A
centralized energy management system for isolated
microgrids, IEEE Trans. Smart Grid, vol. 5, no. 4, pp. 1864
1875, Jul. 2014.
[10] N. Hatziargyriou, Microgrids: Architectures and
Control, Chichester, UK: Wiley, 2014, pp. 124.
[11] R. Palma-Behnke, C. Benavides, F. Lanas, B. Severino,
L. Reyes, J. Llanos, and D. Saez, A microgrid energy
management system based on the rolling horizon strategy,
IEEE Trans. Smart Grid, vol. 4, no. 2, pp. 9961006, Jun.
2013.
[12] P. Malysz, S. Sirouspour, and A. Emadi, An optimal
energy storage control strategy for grid-connected
microgrids, IEEE Trans. Smart Grid, vol. 5, no. 4, pp. 1785
1796, Jul. 2014.
[13] M. Yazdanian and A. Mehrizi-Sani, Distributed Control
Techniques in Microgrids, IEEE Trans. Smart Grid, vol. 5,
no. 6, pp. 29012909, Nov. 2014.
[14] G.Bruni,S. Cordiner, V. Mulone, V. Rocco, and F.
Spagnolo, A study on the energy management in domestic
micro-grids based on Model Predictive Control strategies,
Energy Conversion and Management, vol.102,pp.50
58,Sep.2015.

IJISRT17MY119 www.ijisrt.com 386