Ouassima J.

Electrical Systems 15-1 (2019): 53-67

Boqtob1,*, JES
Hassan El Regular paper Journal of
Moussaoui1, Electrical
Hassane El Microgrid energy management system: Systems
Markhi1
and Tijani a state-of-the-art review
Lamhamdi1
Over the past decades, the exploitation of renewable sources has significantly increased as a
carbon-neutral mean of electricity supply. A Microgrid (MG) represents a suitable concept to
integrate renewable resources, in which local generation source and Energy Storage System
(ESS) are coordinated to cover the customer demand in any conditions. In addition, Energy
Management System (EMS) is investigated to allocate optimally the power output of the
Distributed Generator (DG) units, economically satisfy the Load Demand (LD), properly
regulate the frequency and voltage of the MG systems, and automatically assure a smooth
transition between Grid Connected (GC) and islanded operation modes. This paper reviews
MG-EMS, which has evolved in recent frameworks, discussing MG generation and storage units,
the use of Electric Vehicles (EVs) as backup device, the installation of Combined Heat and
Power (CHP) systems to supply thermal requirements, MG-EMS objectives function and system
constraints and recent optimization algorithms.

Keywords: Energy Management System, Microgrid, Electric Vehicles, Combined Heat and Power,
Optimization Algorithms.
Article history: Received 19 September 2018, Accepted 13 February 2019

1. Introduction

For a long time, fossil fuel represented the main source of electricity, which generated a
high amount of gaseous pollutant emission on our planet earth. Clean energy generated by
renewables has been the best suitable choice to mitigate the fossil fuel dependent and so to
lower the Greenhouse Gas (GHG) from our atmosphere. Renewable sources produce
electricity from natural resources like wind, biomass, solar, geothermal, ocean and hydro.
According to renewables global status report of 2017, the sector of power remains the
main concern of renewable technologies policy support to facilitate its integration into
energy systems. Figure 1 depicts the global electricity production until the end of 2016,
where 24.5% represent the Renewable Energy (RE) production, which is estimated to be
2017GW from 6660GW of total global electric capacity, increased by 9% compared to
2015. It is clearly shown that the most produced RE is from the hydropower (54.34%),
followed by wind power (24.15%), Photovoltaic (PV) power (15.02%), bio-power (5.55%)
and others. However, in 2016, 47% of RE newly implemented is solar PV, 34% of wind
power and 15.5% of hydropower [1].
In addition, renewable consumption is estimated to raise by 28% between 2015 and
2040, Figure 2 depicts the growth of different energy sources (renewable, coal, natural gas,
nuclear, and other liquids) from 1990 to 2040 [2].
Extensive research has concentrated on the integration of a MG as a suitable renewable
integration platform that can be installed by multiple stakeholders worldwide. A MG might
*
Corresponding author: O. Boqtob, Signals, Systems and Components Laboratory - Faculty of Sciences and
Technologies-USMBA- B.P. 2202 – Route d’Imouzzer, Fez, Morocco, E-mail: [email protected]
1
Signals, Systems and Components Laboratory - Faculty of Sciences and Technologies-USMBA- B.P. 2202 –
Route d’Imouzzer, Fez, Morocco

Copyright © JES 2019 on-line : journal/esrgroups.org/jes

 O. Boqtob et al: Microgrid energy management system:: a state-of-the-art review fi

be simply defined as : ‘’a distribution network that incorporates a variety of possible

distributed energy resource that can be optimized and aggregated into single system that
can balance loads and generation with or without ESS and is capable of islanding whether
connected or not connected to a traditional utility power grid.”[3].
Therefore, a MG is an electrical network having local generation sources, located in the
downstream of the distribution grid through a Point of Common Coupling (PCC). A MG
can consist of renewable distributed generation unit, conventional generators, electricity
storage system, hydrogen storage system, thermal ES, Thermal Loads (TLs) and Electric
Loads (ELs). Renewable Energy Sources (RES) can include Wind Turbine (WT), PV
power generation, biomass and hydrokinetic (HKT), unlike the conventional generators that
can comprise of Micro Turbine (MT) and Diesel Generator Set (DGS).
A MG can operate in two modes, GC mode means that a MG is linked to the main grid,
and can participates in the energy market by exchanging energy with the utility as buyer or
seller. However, in islanded mode, a MG operates as an autonomous entity, which is
disconnected from the utility network, due to several causes such as brownout, geography
position or economic issues.

Fig. 1. Global estimated power capacity end 2016

Fig. 2. Growth of different energy sources from 1990 to 2040 [2].

Kirkham et al. [4] have defined an EMS as « a collection of control strategies and
operational practices, together with the software and hardware to accomplish the objectives

54
 J. Electrical Systems 15-1 (2019): 53-67

of energy management». An EMS can optimally dispatch the power output of the DG units
to supply economically the load, properly regulate the frequency and voltage of the MG
systems, and automatically provide a smooth transition between GC operation mode and
isolated mode, accordingly to real-time operating conditions and MG components. A MG-
EMS has to provide high power quality, sustainable, reliable, and environmentally friendly
power source by meeting cost-effective benefits. EMS is one of the major integration
subjects of RES in a MG.
Due to intermittence of RES, researchers started focusing on EMS strategies. Extensive
research has developed an EMS for different MG scales ranging from small-scale (single
home) to large-scale MG (cities) as depicted in Fig.3.

Fig. 3. Different scales of MG.

Research has focused on modelling and studying the feasibility of MG components,

optimum scheduling of DGs, forecasting environmental data (wind speed, solar irradiance
and water speed), forecasting consumption profile, EMS based optimization algorithms, for
the purposes of getting economic benefits with less impact on the environment. Fig.4
outlines the main MG structure and components related to EMS.
Figure 5 shows the increasing interest in the field. It depicts that the research
predominantly focuses on the solution of different concerns related to the MG by using
EMS.
This paper reviews different aspects MG-EMS that are evolved in recent topics, ranging
from MG studied units to the recent optimization algorithms based objective functions and
system constraints for EMS.
The paper is managed as follows: section 2 discusses a review of the MG configuration
and structure used for EMS with different distributed energy source, ESS and different
types of consumer loads, the incorporation of EVs as backup source, and CHP to co-
generate the heat with electricity and supply thermal requirements. Section 3 focuses on
MG-EMS objectives functions and system constraints in isolated and connected MG
operation mode. Section 4 discuses a review of recent optimization algorithms which are
applied in EMS. Finally, section 5 concludes the paper.

55
 O. Boqtob et al: Microgrid energy management system:: a state-of-the-art review fi

Fig. 4. Energy management system of MG.

Fig. 5. Evolution of MG- EMS published paper during the last ten years.

56
 J. Electrical Systems 15-1 (2019): 53-67

2. Microgrid generation and storage units

2.1. Alternative power generation

Nowadays, a significant number of researches are addressing the issue of a MG-

EMS. Several configurations of MG structure have been studied in the literature. The
combination of different DGs is based on economic benefits, geographic characteristic,
environment condition and impact. Syed & Raahemifar [5] have discussed a predictive
EMS, control issue and communication of MG connected to independent electricity-
system-operator-grid, the studied MG includes 1.5 MW of wind energy assisted by battery
ESS. Kusakana [6] used a hybrid HKT-battery in GC mode. The surplus HKT power will
charge the battery or be sold to the grid accordingly to the battery’s State Of Charge (SOC)
and to the pricing bid. The grid can serve charging the battery in the off-peak period, while
be discharged in the peak period to take profit of the low price period. However, the grid
supply directly the LD when cannot be completely met by the HKT and the battery.
The commonly ES device used in the previous papers is the battery. However, the
battery system has some disadvantages, as unreliable lifespan, limitations in charge and
discharge current issues. Hence, techno-economic indexes of the MG based hybrid RES
may be significantly enhanced by implementing the battery with the DGS [7]. Giday [8]
studied economic and technical assessment of hybrid power sources with PV, battery and
DGS to supply the LD of rural school in the south region of Ethiopia. The paper has
compared the hybrid combination to only DGS system in terms of total Net Present Cost
(NPC) and energy cost. The author used Homer software for analysis purposes. The result
has demonstrated that hybrid PV/battery/DGS has economic and technical advantages
unlike DGS system.
Recently, with the increasing interest of hydrogen utilization technology, due to its
advantages as longer lifetime of hydrogen tanks and cheaper cost as compared to battery,
hydrogen DC-MGs has been addressed in several frameworks, which use batteries as a
short-term energy storage while Fuel Cell (FC) runs for long-term one. Thirumalaisamy &
Jegannathan [9] proposed EMS for hybrid MG, taking as RES a PV and WT, while
FC/Ultra-Capacitor/battery are considered as ESS. Table 1 summarizes MG generation
units and storage devices, and the load type related to EMS topics. In this table, different
combination of RES with conventional units and storage device have been used, taking in
consideration load type that can be thermal and/or electrical requirement, Critical Load
(CL) and Non-Critical Load (NCL).

2.2. Integration of electric vehicle in microgrid as backup device

Recently, the incorporation of EV within a MG has become popular since their

impacts on reducing gaseous emission, inexpensive charging and decreasing fossil fuels
usages. The EV in MG can operate under two modes, Grid-to-Vehicle (G2V) or/and
Vehicle-to-Grid (V2G) modes. Panwar et al. [25] presented a strategic EMS in a MG
installing two types of storage elements: Unitised Regenerative FC and EV. Scheduling
strategy of the micro grid with Unitised Regenerative FC and EV in V2G mode offered less
line loss and cost of operation. The proposed EV is equipped with Lithium iron phosphate
battery. Two cases for scheduling the EV are considered, charging and V2G cases. The

57
 O. Boqtob et al: Microgrid energy management system:: a state-of-the-art review fi

results outline that V2G operation achieved the least objective function rate compared to
MG operation strategies without V2G and MG operation strategy with conventional
vehicle. Fathabadi [26] proposed a large-scale EV charging station, in which solar and wind
are combined to generate its requirements power. In this framework, both PV array and
wind work with a unified maximum power point tracking technique. The author concluded
that the invested money of the EV charging station construction has returned after only one
year. Aryanezhad [27] integrated the plug-In hybrid EV-V2G charger in the distributed grid
through an AC/DC converter and a DC/DC converter, which are connected by a DC link
capacitor. The proposed V2G charger configuration can regulate the grid voltage by
implementing the DC link capacitor to provide reactive power compensation, to avoid the
high risk of voltage drop in the grid. The plug-In hybrid EV battery energy is applied for
load levelling and peak load shaving. Sbordone et al. [28] described a real implementation
of EVs charging station associated with an ESS, including Li-polymer battery. An
experimental test is carried out at Energy and Sustainable Economic Development labs of
Italian National Agency for New Technologies. The results demonstrate the great
performance of the EVs in the peak shaving function, compared to the grid.
In addition, RES and EVs make good alternatives. Balasubramaniam et al. [29]
presented a resiliency-based EMS for MG during islanded operation. The studied MG
comprise of PV farms and WT farms both equipped by battery ESS and DGS. Based on
literature recommendation, the authors used 25 plug-In hybrid EV as adjustable demand
and generation to model the concept of demand response in the optimization process, so
plug-In hybrid EV in G2V mode are used as adjustable demand, while in V2G mode are
applied as adjustable generation. Wu et al. [30] discussed a stochastic EM of a smart home
with PV panels and Plug-in EV (PEV) energy storage, to reduce energy demand charging
under a time-of-use tariff, while supplying PEV charging requirements and home LD. The
PEV battery is designed with a Li-ion battery and is controlled by DC/AC converter,
allowing both unidirectional and bidirectional power flow. The PEV can plugs-out and
plugs-in once a day. Markov Chain model is used to model the PEV mobility. The author
has used predictive models to model PV power and home LD. The author demonstrated that
the proposed system could offer significant cost savings for customers. The power flow
between the PEV battery, home loads, PV array and utility grid is managed by the proposed
EMS.

2.3. Integration of combined heat and power system

The heat can also be cogenerated to electricity by installing CHP systems, which can
provide heat for different users: residential buildings space, commercial customers and
industry processes. In addition, efficiency gains can be improved through CHP, which uses
waste heat for thermal applications, and reduces distribution and transmission losses. The
Combined-Cooling- Heating-Power (CCHP), also known as tri-generation MGs, can suffice
heat, cooling and power demand of users. Hussain et al. [31] discussed the optimization
based modelling of tri-generation microgrid CCHP. The paper has taken in consideration
the uncertainty of heat, cooling and electrical demand. Moradi et al. [32] discussed an EMS
for CHP of GC-MG to optimally planning the grid operation and CHP dispatch. The MG is
considered to supply electric and thermal requirements. The proposed method has been

58
 J. Electrical Systems 15-1 (2019): 53-67

tested using a typical hospital located in the Northern Iran as case study, due to their
relatively equal and large demand for electricity and heat. Hussain et al. [33] proposed an
optimal EM for three GC-MGs. The MG in this paper is modelled as thermally networked
MGs containing CHP generator, Controllable DG (CDG), wind mill, battery ESS, heat only
boiler, thermal ESS, EL, and TL. The author assumed that each MG used own local CHP,
CDG, and battery ESS units to meet the electric LD, while the electricity deficit can be
bought from the grid and the excess electricity production can be sold to the grid. For the
thermal demand, the three MGs can exchange between them the thermal energy in the
network to maintain greater thermal stability and reduce thermal energy wastage. In [16],
the combination of PV and CHP system in a MG with storage devices is applied to meet the
requirement TLs and ELs of a single-family home in Germany. The PV-CHP hybrid system
is developed to reach a higher level of reliability, flexibility, efficiency and lower energy
costs. Waqar et al. [34] proposed a CHP plant installed in MG as better option to meet both
ELs and TLs simultaneously, to cover natural gas depletion and prevent the use of
electricity grid to feed the TLs. The studied MG included A CHP plant, a PV array, DGS
and batteries. Li et al. [35] studied a MG based natural gas CHP generators, boilers, PV
panels, WTs and battery storage, to supply the electricity, heat and fuel demand of the
ammonia plant in central Inner Mongolia in China. The CHP and PV panels are used to
meet the electricity requirements. In addition, The CHP generators and the boiler are used
to meet the heat demand. While the hydrogen can be generated by natural gas reforms and
water electrolysis using the solid polymer electrolyte. The author compared the existing
energy supply system to the natural gas based MGs, and concluded that the last structure
can reduce the life cycle total energy by 6.13%, the GHG emissions by 30.44% and the
fossil fuel energy by 6.50%. Wang et al. [36] proposed an integrated scheduling approach
for MG with industrial customers. The studied MG includes CCHP, Ground Source Heat
Pump (GSHP), the PV generation and BB. The TL is supplied by the CCHP, and the
remaining required heating/cooling energy can be guaranteed by the GSHP. While
electricity loads are fulfilled by the PV and the CCHP. For the excess electricity, it is used
to charge the Battery or sell to the main grid, and vice versa.

Table 1. MG generation units and storage devices used to supply specific LD related to
EMS.
Renewab Storage MG Demand
Paper Notes
le source device type nature
[6] Hydrokin Battery GC Household - The Hydrokinetic turbine has a rated power of
etic load 4 kW at 1.4 m/s water velocity.
turbine - Hydrokinetic turbine produces electricity from
flowing water at low velocity without needing a
dam or barrage.
small-
PV/WT/ Home loads are classified as interruptible loads
[10] ES scale Home
MT and no interruptible loads.
GC
PV/Hydr domestic
o load/
Power/Bi community - MG is composed of seven unelectrified remote
Battery remote
[11] omass/ load/ areas of Uttarakhand in India
Bank rural
DGS commercial - In this area, RES are available in high quantity.
/Biogas load/
Generato industrial

59
 O. Boqtob et al: Microgrid energy management system:: a state-of-the-art review fi

r load

[12] PV Battery/W GC in - Batteries are used to store or compensate the

ater desert PV energy production. While water hydraulic
pumping area pumping is activated to supply water demand for
storage agriculture and farmer in case of higher PV
production.
[13] PV/Hydr Battery Stand- - The MG is at an altitude above 4000 m, in
o alone which solar energy is abundant and more stable.
Generatio - the MG is composed of four 1,6MW small-
n Plant/ hydro generators, four 2,5MW DGS, 5 MW PV
DGS and 5 MWh battery ESS.
- The battery ESS is used to smooth the power
volatility and transfer the excess production of
the PV system.
[14] PV FC/ Stand- Residential - PV is considered as primary power source, FC
Battery alone. load as secondary power source and battery as
backup device.
[15] PV Lead acid DC - The FC operates at constant voltage mode, to
battery stand- maintain the balance of the power flow and the
bank/ alone stability of the DC bus voltage.
Proton - The author addressed the EM strategy to
Exchange control the BB for distributing power between
Membran the battery and FC.
e FC
[16] PV/CHP Li-Ion GC Household’s The MG can compensate the lack of electrical
battery TLs and ELs energy by purchasing electricity from the grid,
andtherm while the surplus generation is sold to the grid.
al energy However, the battery is not allowed to exchange
storage power directly with the grid
[17] PV/WT/C FC GC Residential - Hybrid power sources are more cost effective
HP load to systems with single energy source.
- The most cost-effective and reliable to supply
the residential use is the use of thermal recovery
without thermal storage.
[18] PV/WT/ ESS GC Park with - The implementation of RES and CCHP have
Natural ELs, thermal great impact to reduce the dependence on the
gas/CCH and cooling grid to supply TLs/ELs.
P loads.
[19] PV/WT/ stand- AC loads The DGS is assumed to be available in
DGS alone continuous time.
[20] PV/WT Battery GC Residential The combination of PV/WT/battery can reduce
demand the fossil energy consumption, environment
pollution and the payment of electricity.
[21] Vanadium GC industrial - The minimum and maximum SOC of the all
Redox park the ES units is 20% and 100%, respectively.
flow - The MG sells the excess energy to the grid,
Battery when the batteries are fully charged. However,
(VRB) the MG purchases power from the grid when the
and power produced from the RES and the batteries
lithiumion is insufficient to supply the whole demand.
(Li-Ion)
battery

60
 J. Electrical Systems 15-1 (2019): 53-67

[22] PV Battery/ GC - The power exchange between different MG

FC units is based on purchasing energy from
upstream grid to supply LD and charging the
battery storage.
- The produced power by PV can be used to
supply load, charge battery and to be sold to the
upstream grid, also the produced power by the
FC is used to supply load and charge battery
storage. However, the energy stored in the
battery can only serve the LD.
[23] PV Battery GC Residential The BB is modeled with fixed-final state
loads (CL condition, based on equating the SOC at the last
and NCL) sampling interval to the initial condition. This
means that the remaining quantity of the battery
energy at the end of the control timescale should
be equal to its energy amount at the beginning of
the control horizon.
[24] PV/hydro Battery Stand- The MG is simulated with a 1 MW PV array, a 2
power/ alone MVA hydro generation units, a 500 kW battery
DGS and a 500 kW DGS.

3. Energy management system

Extensive research has focused on the implementation of EMS for MG in isolated and
GC operation mode. For an efficient EMS, it is necessary to define the problem objective
function and take in consideration system’s constraints in accordance with MG
components and operational mode. The EMS has a great impact on environment condition,
the lifetime of the power generator’s units, system performance and customer’s
satisfaction.

3.1. Objectives function and system constraints for islanded microgrid

Upadhyay & al. [11] have studied the EMS of AC/DC MG with RES. The author has
considered as optimization objective the minimization of cost of energy, which is described
as the ratio of total analysed cost of hybrid power sources to the annual electrical energy
produced. However, the author has subjected the optimization objective to several
constraints, which are defined as reliability criteria, economic criteria, renewable fraction
and environment conditions. The reliability criteria are indexed by the amount of energy
not supplied at the time when demand exceeds generation. The economic criteria are
examined by total NPC taking in consideration investment cost and total non-recurring and
recurring cost to demonstrate the viability of an investment project. The renewable fraction
has been considered to quantify the amount of demand energy that is supplied by RES.
Emission of CO2 related to DGS output is considered as environment constraints.
Han & al. [15] have implemented an EMS to fulfil different goals. The multi-objective
optimization is formulated to keep power balance between generated and consumed energy,
to ensure the reliability of bus voltage, to extract the maximum PV array output power at
any environment states, to protect the battery by keeping SOC in a suitable range, to avoid
the fluctuation of the PEMFC output power and to reduce the hydrogen consumption.
Marzband & al. [37] have presented an EMS for PV/WT/MT/ES standalone MG to
minimize the total system cost. The total cost has defined as cost of energy injected by
dispatchable and non-dispatchable generators, the cost of storage unit energy during

61
 O. Boqtob et al: Microgrid energy management system:: a state-of-the-art review fi

discharging/charging states, taking in consideration cost of responsive LD energy and

undelivered power penalty cost. While the optimization issue has been subjected to power
balance, non-dispatchable generator power limits, dispatchable generators operating time
limits, dispatchable ramp-up and ramp-down levels, ESS constraints as battery SOC limits
and electric water heater limits.
In Ref. [29], the authors have developed an EMS for WT/PV/PHEV/DGS/battery
islanded MG. The EMS has evolved to maximize the amount of adjustable loads (PHEV),
CL & NCL that are served during the islanded operation mode. The optimization issue has
taken in consideration power balance constraints, adjustable load constraint, generator
power boundaries and battery constraint.
In Ref. [38], the authors have discussed an economic dispatch of thermal power plant.
The optimization objective has been defined to decrease the fuel cost for thermal
generators, considering the valve’s point loading effect. The optimization issue has been
subjected to power balance, power generator’s boundaries, prohibited generators operating
ranges and its ramp rate limits.

3.2. Objectives function and system constraints for grid-connected microgrid

In Ref. [6], an EMS of GC-HKT system with battery storage has been presented to
minimize cost function that is explained by three main costs. The first one is the cost of
purchasing energy from the grid to satisfy LD and to charge the battery. The second one is
the revenue from selling electricity to the grid during the peak pricing time. The third one is
the wearing cost of the system. The authors have taken as optimization constraints power
balance, HKT’s output limits and battery SOC levels.
Marzband & al. [10] have optimized GC home MG with dispatchable and non-
dispatchable generators, ESS and responsive LD. The objective function has included the
overall cost energy, which is presented as cost of fuel consumption, cost energy of battery
during charging/discharging states, cost energy produced by dispatchable and non-
dispatchable generators, cost energy generated by buying and selling power between home
MG and the main grid, and cost energy to move load power demand throughout the
scheduling horizon. The authors have integrated environment and technical factors into
optimization constraints and balancing criteria. The amount of energy produced by all
generators must be equal to the amount of energy consumed. The consumer and supply
electricity bids have also taken in consideration the dispatchable and non-dispatchable
generators technical constraints and boundaries. Energy storage technical constraints that
includes charge/discharge limits with or without the stored energy, battery energy balance,
energy stored boundaries and energy storage SOC. The author has proposed to transfer
interruptible loads at the power shortage time to another period to ensure system stability.
The last constraint was the energy grid boundaries that the amount of energy exchanged
between the grid and home MG.
Maleki & al. [17] have optimized a hybrid PV/WT/FC/CHP system in GC mode. The
authors have proposed to reduce the operation cost while maximizing the system income.
As system cost, the authors have considered the fuel cost, the cost of energy purchased
from the gird, the start-up cost of hybrid system, the operation & maintenance cost of
PV/WT/FC generators and the gas cost. For the system income, the authors have included

62
 J. Electrical Systems 15-1 (2019): 53-67

the gain of selling excess thermal and electrical energy to the grid when production of the
hybrid system exceeds the total demand.
In Ref.[33], the authors have developed an EMS for GC thermally network MGs with
CHP, controllable DG, BESS and thermal ESS, wind mill, heat only boiler(HOB), TLs and
ELs. The proposed EMS has been investigated to minimize operation cost system. The
authors have considered the cost energy of CHP/CDG/HOB units, cost of buying/selling
energy from/to the man grid, and start-up cost of CDG/HOB units. The optimization issue
has been subjected to power and thermal balance constraints. The power balance ensures
that the energy production with charging/discharging battery energy must be equal to load
consumption and energy traded with the grid. The thermal balance ensures that the thermal
energy production of local generators with charging/discharging thermal storage must be
equal to local TL demand and thermal energy shared with other MGs, considering the
thermal energy lost. Another constraint has been considered in the paper that explains the
capacity of each MG components. CHP, CDG and HOB must operate in their specified
boundaries. The electric line capacity limits the amount of energy shared with the main
grid. Also, the thermal line capacity limits the amount of thermal energy shared between
two MGs. The last constraint is thermal energy trading between thermal networks MGs,
which has been considered to reduce thermal energy waste when local thermal production
exceeds local demands.

4. Recent optimization algorithm used for energy management system

In the MG, power generated from RES, LDs and market prices are known as stochastic
variables due to their uncertain nature. Thus, it is necessary to use a stochastic model to
obtain efficient results in EM strategies. Most of recent researches have used
metaheuristic/stochastic algorithm to solve optimization scheduling in MG. A Stochastic
programming can be defined as mathematical programming having a stochastic element in
the data and the result solution is dependent on the set of random generated variables [39].
Recent research papers discussing MG-EMS that have implemented advanced techniques
based on hybridization of various optimization techniques or modified basic algorithms, to
reach the optimal scheduling solution for EMS of DGs in different MG.
Yu et al. [40] have analyzed the inertial weight factor role to balance convergence and
exploration ability of the Particle Swarm Optimization (PSO). Although, larger values of
the inertial weight factor can avoid the algorithm to derive a local optimal solution, a very
slow speed convergence can be achieved. To circumvent this problem, the author has
proposed improved PSO method to dynamically adjust and linearly decrease the inertial
weight factor during the iterations process. Thus, the improved PSO can achieve the global
optimum solution without trapping into a local minimum. The proposed method has been
used for power dispatched based EMS of a hybrid power source.
Advanced metaheuristic methods have used chaotic sequences instead of random
numbers to upgrade their performance. Adarsh et al. [38] have proposed Chaotic Bat
Algorithm (CBA) for optimizing the economic dispatch of the studied system. The
proposed algorithm incorporated chaotic sequences based sinusoidal map (or sequence) in
the basic BA to improve the algorithm performance for reaching the global optimal results.
Marzband et al. [37] have proposed a new Multi-Layer Ant Colony Optimization
(MACO) for real time implementation of EMS in standalone MG. The MACO is developed

63
 O. Boqtob et al: Microgrid energy management system:: a state-of-the-art review fi

from the basic Ant Colony Optimization (ACO) equalling the number of layers to the
number of design variables and number of nodes in each particular layer to the number of
allowable values of each variable.
To upgrade the performance of fireworks algorithm (FA), Wang et al. [41] have
proposed a new hybrid Multi-objective based FA and Gravitational Search Operator
(MFAGSO) to solve the multi-variable non-linear programming problem subjected to
multiple hybrid constraints. The suggested algorithm used gravitational search operator to
guide the sparks into the clustered region to exchange location information with
Pareto optimal solutions at each generator’s process to reach best results.
Roy et al. [42] have exposed an Improved Artificial Bee Colony algorithm (IABC)
algorithm to optimize a hybrid GC-MG operation. The author has modified the basic ABC
by generating the scout bee with Gravitational Search Algorithm (GSA) to upgrade
searching accuracy, and hence improving the global optimal solution of optimization
problem. Lin et al. [43] have used Enhanced Bee Colony Optimization (EBCO) algorithm
for EMS of MGs with RES and ESS. EBCO algorithm differs from the basic BCO because
of its inclusion of the self-adaption repulsion factor in the bee swarm, for improving the
performance of each bee swarm and increasing the search accuracy in high dimensional
problems.
Other optimization methods are found in literature as presented in Table 2, which
summarizes recent frameworks on the field of EMS of MG, including MG components, the
optimization algorithm, the simulation tests and results, and convergence speed of the
algorithm.

5. Conclusion

This paper discusses a review of recent papers in the literature discussing the
integration of energy management system in different scales of microgrid with several
components and different load type, to optimize the whole system, meeting the defined
objectives with consideration of the system constraints. The paper reviews the novelty to
integrate the electric vehicle as energy resource, the installation of combined heat and
power systems to co-generate the heat with electricity and supply thermal requirements, and
the implementation of hybrid optimization algorithms that can be a good alternative than
single algorithm.

Table 2: Recent topics using an optimization algorithm to solve MG-EMS.

Paper Algorithm MG information Simulation tests and results Conver
gence
speed
[10] Mixed- GC home MG with - The proposed EMS scheduling assumed that slow
Integer PV (6 kW), WT (8 the MT is used for 83% of the times, and the
Linear kW), MT (12 kW), WT has operated among 92% of the times,
Programm ES (2 kWh), contrary to the PV with being in service only
ing responsive/non- 46% of the time. For the ES operation modes,
(MILP) responsive LD. about 42% of the horizon times ES has been
used in discharging mode to supply the LD,
12% of the times in repos mode, and about
46% of the times ES has been operated in the
charging mode.
- The computation time of MILP is 30s.
[17] PSO GC residential - Compared to GA, GA outperforms PSO in slow

64
 J. Electrical Systems 15-1 (2019): 53-67

loads with term of Min, Max, and Mean of fitness

PV/WT/CHP/ function and standard deviation indexes.
FC/TL & EL. - However, PSO outperforms GA in term of
computation time.
[21] Regroupin WT (2500kW), PV -Compared to GA, global optimum solutions fast
g PSO (480kW), VRB, Li- at 2.1152s unlike GA with 16.3456s.
(RegPSO) Ion battery, - Daily cost balance |Profit-Expense| by
industrial demand. RegPSO is 341.9$ better than 348.07$ of GA.
[38] CBA Five different tests Different tests have been used to validate the fast
are taken with proposed CBA.
different number of - test 1: 6 generators and meeting a LD of
generators and LD. 1263 MW.- test 2: 13 generators meeting a
LD of 1800MW. It includes valve point
loading. - test3: 20 generators supplying a
demand of 2500 MW.
- test4: 40 generators supplying a demand of
10,500 MW. It includes valve point loading
effects.
- test5: 100 and 60 generators, supplying a
demand of 43,200MW, with multiple fuel
options and valve point loading effects.
Result: the CBA shows a good performance
against several methods.
[41] MFAGSO GC-MG including 3 cases are simulated: Case 1: All the units fast
3 feeders: are dispatched in their limited capacity. Case
Feeder1: 2: PV and WT act at their maximum output
Residential loads: powers of their forecasted values. Case 3: A
MT (30KW), WT larger range of power exchange between the
(15KW max), PV MG and utility is allowed.
(25KW), battery Result: MFAGSO shows a high performance
(30KW) and loads. in term of convergence speed than other
Feeder2: Industrial methods.
loads: FC (30KW)
and loads.
Feeder3:
Commercial loads.
[42] IABC PV/WT/MT/DGS/ - Execution time for 100 iteration of IABC is Fast
battery. 5.955264s, whereas ABC and ABC–ABC
need 7.897576s and 6.943122 s, respectively.
[44] Modified A neighborhood of - Compared to gradient descent algorithm, mediu
Simulated six single-family The gradient descent algorithm shows fast m
Annealing houses, each house computation time, However, it has the
(MSA) consists of: drawback of getting stuck in local minima for
PV/battery/load/ most of the test scenarios.
interface to the - MSA has fast computational effort than the
neighbor building total state space search.
and grid - The MSA found the global optimum in 6 out
connection. of the 8 tested scenarios. For the remaining 2
scenarios, it found a local optimum that was
very close to the global optimum.
[45] Multiperi PV/WT/MT/battery - Execution time of MABC is 10.14s mediu
od / responsive compared with 27.45s of PSO for case study m
Artificial loads/non- corresponding to 100 iteration.
Bee responsive loads.
Colony
(MABC)
[46] Improved Systems of 5 - Compared to many hybrid algorithms, I DE Mediu
differentia generation units shows good performance in term of fuel cost m
l and 10 generation and convergence.
evolution units.
(IDE)

65
 O. Boqtob et al: Microgrid energy management system:: a state-of-the-art review fi

References

[1] REN21. 2017. Renewables 2017 Global Status Report (Paris: REN21 Secretariat).
[2] International Energy Outlook, US energy information administration, September 14, 2017
[3] Navigant Research, Report 1Q 2016.
[4] H. Kirkham, D. Nightingale, and T. Koerner, “Energy management system design with dispersed storage
and generation,” Power Apparatus and Systems, IEEE Transactions on, vol. PAS-100, no. 7, pp. 3432–
3441,(1981).
[5] Syed, I. M., & Raahemifar, K. Predictive energy management, control and communication system for grid
tied wind energy conversion systems. Electric Power Systems Research, 142, 298-309. (2017).
[6] Kusakana, K. Energy management of a grid-connected hydrokinetic system under Time of Use tariff.
Renewable Energy, 101, 1325-1333. (2017).
[7] Kusakana K. Optimal scheduled power flow for distributed photovoltaic/wind/diesel generators with battery
storage system. IET Renew Power Generation; 9(8):916–24. (2015).
[8] Giday, Z. G. Technical and economic assessment of solar PV/diesel hybrid power system for rural school
electrification in Ethiopia. International Journal of Renewable Energy Research (IJRER), 3(3), 735-744.
(2014).
[9] Thirumalaisamy, B., & Jegannathan, K. A novel energy management scheme using ANFIS for independent
microgrid. International Journal of Renewable Energy Research (IJRER), 6(3), 735-746. (2016).
[10] Marzband, M., Alavi, H., Ghazimirsaeid, S. S., Uppal, H., & Fernando, T. Optimal energy management
system based on stochastic approach for a home Microgrid with integrated responsive load demand and
energy storage. Sustainable Cities and Society, 28, 256-264. (2017).
[11] Upadhyay, S., & Sharma, M. P. Selection of a suitable energy management strategy for a hybrid energy
system in a remote rural area of India. Energy, 94, 352-366. (2016).
[12] Eddine, B. S., & Djallel, K. E. R. D. O. U. N. Power Management of Grid Connected Photovoltaic
Installation Assisted by Batteries and Water Pumping Energy Storage in Desert Location. International
Journal of Renewable Energy Research (IJRER), 7(4), 2139-2150. (2017).
[13] Zhao, B., Xue, M., Zhang, X., Wang, C., & Zhao, J. An MAS based energy management system for a stand-
alone microgrid at high altitude. Applied Energy, 143, 251-261. (2015).
[14] Slouma, S., Skander Mustapha, S., Slama Belkhodja, I., & Orabi, M. (2015). An improved simple Fuel Cell
model for Energy Management in residential buildings. Journal of Electrical Systems, 11(2).
[15] Han, Y., Chen, W., & Li, Q. Energy Management Strategy Based on Multiple Operating States for a
Photovoltaic/Fuel Cell/Energy Storage DC Microgrid. Energies, 10(1), 136. (2017).
[16] Kneiske, T. M., Braun, M., & Hidalgo-Rodriguez, D. I. A new combined control algorithm for PV-CHP
hybrid systems. Applied Energy, 210, 964-973. (2018).
[17] Maleki, A., Hafeznia, H., Rosen, M. A., & Pourfayaz, F. Optimization of a grid-connected hybrid solar-
wind-hydrogen CHP system for residential applications by efficient metaheuristic approaches. Applied
Thermal Engineering, 123, 1263-1277. (2017).
[18] Wenjin, M., Hongwei, L., Hairong, Z., & Hao, Y (2018). Study on Multi-energy Flow Management for
Multi-energy Network System. Journal of Electrical Systems, 14(3).
[19] Thanh, T. D., Phung, Q. N., & Duc, H. N. (2017). Power flow analysis for islanded microgrid in hierarchical
structure of control system using optimal control theory. Journal of Electrical Systems, 13(4).
[20] Liu, Z., Li, W. Z., & Liu, X. Optimal Scheduling Algorithm for Residential Energy Consumption
Considering the Wind-Photovoltaic Power and the Electricity Price of Purchase/Selling Electricity.
International Journal of Smart Home, 10(11), 25-38. (2016).
[21] Eseye, A. T., Zheng, D., Li, H., & Zhang, J. Grid-Price Dependent Optimal Energy Storage Management
Strategy for Grid-Connected Industrial Microgrids. In Green Technologies Conference (GreenTech), 2017
Ninth Annual IEEE (pp. 124-131). IEEE. (2017), March.
[22] Majidi, M., Nojavan, S., & Zare, K. Optimal stochastic short-term thermal and electrical operation of fuel
cell/photovoltaic/battery/grid hybrid energy system in the presence of demand response program. Energy
Conversion and Management, 144, 132-142. (2017).
[23] Numbi, B. P., & Malinga, S. J. Optimal energy cost and economic analysis of a residential grid-interactive
solar PV system-case of eThekwini municipality in South Africa. Applied Energy, 186, 28-45. (2017).
[24] Karimi, M., Azizipanah-Abarghooee, R., Uppal, H., Hong, Q., Booth, C., & Terzija, V. Smart integrated
adaptive centralized controller for islanded microgrids under minimized load shedding. In Smart Grid and
Cities Congress and Fair (ICSG), 2017 5th International Istanbul (pp. 41-45). IEEE. (2017), April.
[25] Panwar, L. K., Reddy, K. S., Kumar, R., Panigrahi, B. K., & Vyas, S. Strategic Energy Management (SEM)
in a micro grid with modern grid interactive electric vehicle. Energy Conversion and Management, 106, 41-
52. (2015).
[26] Fathabadi, H. Novel grid-connected solar/wind powered electric vehicle charging station with vehicle-to-
grid technology. Energy, 132, 1-11. (2017).
[27] Aryanezhad, M. Optimization of grid connected bidirectional V2G charger based on multi-objective
algorithm. In Power Electronics, Drive Systems & Technologies Conference (PEDSTC), 2017 8th (pp. 519-
524). IEEE. (2017), February.

66
 J. Electrical Systems 15-1 (2019): 53-67

[28] Sbordone, D., Bertini, I., Di Pietra, B., Falvo, M. C., Genovese, A., & Martirano, L. EV fast charging
stations and energy storage technologies: A real implementation in the smart micro grid paradigm. Electric
Power Systems Research, 120, 96-108. (2015).
[29] Balasubramaniam, K., Saraf, P., Hadidi, R., & Makram, E. B. Energy management system for enhanced
resiliency of microgrids during islanded operation. Electric Power Systems Research, 137, 133-141. (2016).
[30] Wu, Z., Wang, B., & Xia, X. Large-scale building energy efficiency retrofit: Concept, model and control.
Energy, 109, 456-465. (2016).
[31] Hussain, A., Bui, V. H., Kim, H. M., Im, Y. H., & Lee, J. Y. Optimal operation of tri-generation microgrids
considering demand uncertainties. Int. J. Smart Home, 10(10), 131-144. (2016).
[32] Moradi, M. H., Hajinazari, M., Jamasb, S., & Paripour, M. An energy management system (EMS) strategy
for combined heat and power (CHP) systems based on a hybrid optimization method employing fuzzy
programming. Energy, 49, 86-101. (2013).
[33] Hussain, A., Lee, J. H., & Kim, H. M. An optimal energy management strategy for thermally networked
microgrids in grid-connected mode. Int. J. Smart Home, 10, 239-258. (2016).
[34] Waqar, A., Shahbaz Tanveer, M., Ahmad, J., Aamir, M., Yaqoob, M., & Anwar, F. Multi-Objective
Analysis of a CHP Plant Integrated Microgrid in Pakistan. Energies, 10(10), 1625. (2017).
[35] Li, M., Zhang, X., Li, G., & Jiang, C. A feasibility study of microgrids for reducing energy use and GHG
emissions in an industrial application. Applied energy, 176, 138-148. (2016).
[36] Wang, H., Wang, W., Cui, L., Sun, H., Zhao, J., Wang, Y., & Xue, Y. A hybrid multi-objective firefly
algorithm for big data optimization. Applied Soft Computing. (2017).
[37] Marzband, M., Yousefnejad, E., Sumper, A., & Domínguez-García, J. L. Real time experimental
implementation of optimum energy management system in standalone microgrid by using multi-layer ant
colony optimization. International Journal of Electrical Power & Energy Systems, 75, 265-274. (2016).
[38] Adarsh, B. R., Raghunathan, T., Jayabarathi, T., & Yang, X. S. Economic dispatch using chaotic bat
algorithm. Energy, 96, 666-675. (2016).
[39] Blum, C., & Roli, A. Metaheuristics in combinatorial optimization: Overview and conceptual comparison.
ACM Computing Surveys (CSUR), 35(3), 268-308. (2003).
[40] Yu, J., Dou, C., & Li, X. MAS-based energy management strategies for a hybrid energy generation system.
IEEE Transactions on Industrial Electronics, 63(6), 3756-3764. (2016).
[41] Wang, Z., Zhu, Q., Huang, M., & Yang, B. Optimization of economic/environmental operation management
for microgrids by using hybrid fireworks algorithm. International Transactions on Electrical Energy
Systems, 27(12). (2017).
[42] Roy, K., Mandal, K. K., & Mandal, A. C. Modeling and managing of micro grid connected system using
improved artificial bee colony algorithm. International Journal of Electrical Power & Energy Systems, 75,
50-58. (2016).
[43] Lin, W. M., Tu, C. S., & Tsai, M. T. Energy management strategy for microgrids by using enhanced bee
colony optimization. Energies, 9(1), 5. (2015).
[44] Velik, R., & Nicolay, P. Energy management in storage-augmented, grid-connected prosumer buildings and
neighborhoods using a modified simulated annealing optimization. Computers & Operations Research, 66,
248-257. (2016).
[45] Marzband, M., Azarinejadian, F., Savaghebi, M., & Guerrero, J. M. An optimal energy management system
for islanded microgrids based on multiperiod artificial bee colony combined with Markov chain. IEEE
Systems Journal. (2015).
[46] Balamurugan, R., & Subramanian, S. (2007). An improved differential evolution based dynamic economic
dispatch with nonsmooth fuel cost function. Journal of Electrical Systems, 3(3), 151-161.

67
 © 2019. This article is published under
https://creativecommons.org/licenses/by-nc/4.0/(the “License”).
Notwithstanding the ProQuest Terms and Conditions, you may use this
content in accordance with the terms of the License.