Hybrid optimization algorithm for modeling and management of micro grid connected system

Abstract

In this paper, a hybrid optimization algorithm is proposed for modeling and managing the micro grid (MG) system. The management of distributed energy sources with MG is a multi-objective problem which consists of wind turbine (WT), photovoltaic (PV) array, fuel cell (FC), micro turbine (MT) and diesel generator (DG). Because, perfect economic model of energy source of the MG units are needed to describe the operating cost of the output power generated, the objective of the hybrid model is to minimize the fuel cost of the MG sources such as FC, MT and DG. The problem formulation takes into consideration the optimal configuration of the MG at a minimum fuel cost, operation and maintenance costs as well as emissions reduction. Here, the hybrid algorithm is obtained as artificial bee colony (ABC) algorithm, which is used in two stages. The first stage of the ABC gets the optimal MG configuration at a minimum fuel cost for the required load demand. From the minimized fuel cost functions, the operation and maintenance cost as well as the emission is reduced using the second stage of the ABC. The proposed method is implemented in the Matlab/Simulink platform and its effectiveness is analyzed by comparing with existing techniques. The comparison demonstrates the superiority of the proposed approach and confirms its potential to solve the problem.

Keywords

micro grid (MG) / multi-objective function / artificial bee colony (ABC) / fuel cost / operation and maintenance cost

Introduction

For subsequent generation, the distributed power generation system is probably an important electric power supply system [1]. The requirement for more flexible electrical systems, changing regulatory and economic scenarios, energy savings and environmental brunt are offering momentum to the enhancement of micro grids (MGs) [2]. A micro grid (MG) is an element of a power system which includes one or more diesel generator (DG) units efficient of performing either in parallel with or self-governing from a large service grid, while presenting persistent power to multiple loads and end-users [3-5]. The consumption of small-modular residential or commercial units for onsite service is one of the important applications of the MG units [6].

A MG contains a low-voltage distribution network with allocated energy resources that can work either interconnected or remote from the foremost allocation grid as a controlled entity [7-9]. Distributed energy resource (DER) is a method to enhance the power quality of the distribution system [10]. For working out several matters facing electric utilities, DER is applied [11]. The basic idea of DER is to suggest for a more robust transmission system, constraints decreased, and energy efficiency increased, power quality enhanced and developed local dependability [12]. More than a few ideas of the MG systems have been suggested and learned since they have an opportunity to present high quality and/or economical electric power [13].

MGs may differ in sizes [14]. The administration of the MG units requires an accurate economic model to describe the working cost [15]. For the MG operation, special protection, control and energy management systems must be intended so as to make specific dependable, safe and inexpensive function in either grid-connected or stand-alone mode [16]. The setback of energy management in micro grids consists of the determination of the optimal (or near optimal) unit commitment (UC) and the transmission of the obtainable generators so that certain chosen purposes are accomplished [17]. Therefore, to reduce the operating costs to the minimum, optimization devices are needed.

The study devices must model the system with its three phases, the neutral conductors, the ground conductors and the connections to ground [18]. Such devices should include stable state and active models for a variety of forms of micro-sources and their interfaces. Models can be functional to predict presentation issues and simulate irregular condition [19]. A MG can be activated either in grid-linked mode or in stand-alone mode. A range of specifications such as the terminal voltage, current, grid voltage, current and fault voltages have been examined at various conditions following modeling [20].

This paper describes a hybrid approach to modeling and managing of the MG connected system. The hybrid multi-objective is applied to the environmental and the economic problem of the MG. The problem formulation takes into consideration the optimal configuration of the MGs at a minimum fuel cost, operation and maintenance costs as well as emissions reduction. The MG considered in this paper consists of a wind turbine (WT), a photovoltaic array (PV), a fuel cell (FC), a micro turbine (MT), a DG and battery storage. The hybridization is obtained as the artificial bee colony (ABC) algorithm is used in two stages. The first stage of the ABC gets the optimal MG configuration at a minimum fuel cost for the required load demand. From the minimized fuel cost functions, the operation and maintenance cost and the emission are reduced using the second stage of the ABC.

Review of recent research

To optimize the function of the micro grid, a smart energy management system (SEMS) was offered in Ref. [21]. The SEMS contains a power forecasting module, an energy storage system (ESS) management module and an optimization module. As energy storage requires optimization across multiple-time steps, regarding the power of energy price structures, the economics of the SEMS are mainly complex. As a result, the ESS module was used to find out the optimal operation approaches. The multiple-time set points of the storage tool and cost-effective presentation of the ESS were furthermore assessed. The elegant management of ESS, economic load dispatch and operation optimization of distributed generation were made simpler into a single-object optimization problem in the SEMS. After that, a matrix real-coded genetic algorithm (MRC-GA) optimization module was explained to accomplish a practical technique for load management, together with three different operation policies.

In a medium-voltage islanded micro grid, an optimization procedure that facilitates the optimal dispatching of distributed generators and storage systems was offered in Ref. [22]. The network was imagined to be contributed by programmable and nonprogrammable generators. Their optimization was executed by the niching evolutionary algorithm (NEA) that was able to locate multiple optima and the deviation of the objective function in their neighborhood. NEAs permit overcoming the presentation of standard algorithms applied for optimal power-flow calculations in power systems by keeping away from falling into local optima. The optimization procedure was executed on a test micro grid and confirmed by computer simulations. The suggested numerical results illustrate that the solutions can enhance the micro grid presentations irrespective of the network operating conditions in all of the regarded cases.

A centralized control system that matches equivalent operations of dissimilar distributed generation inverters inside a micro-grid was proposed [23]. The control design for the DG inverters used a model predictive control (MPC) algorithm that permitted faster computational time for large power systems by optimizing the steady-state and the transient control problems individually. To synchronize load sharing between different DG units during both grid-connected and islanded operations, an overall energy management system was executed for the micro grid. Under diverse test scenarios, the plan concept of the proposed control system was assessed through simulation studies. Using the suggested micro grid, the bang of the increased penetration of DG units on the distribution grid was furthermore examined. The simulation results demonstrate that the operations of the DG units inside the micro grid can be synchronized successfully.

In a MG, a technique based on the cost-benefit study for optimal sizing of an ESS was suggested in Ref. [24]. With spinning reserve for the MG, the suggested technique took the UC problem into consideration. The time series and feed-forward neural network methods were applied for predicting the wind speed and solar radiations correspondingly and the forecasting faults were furthermore considered. Two mathematical models were erected for both the islanded and grid-connected forms of MGs. The most important problem was devised as a mixed linear integer problem (MLIP), which was solved in a modeling language for mathematical programming (AMPL). The efficiency of the suggested strategy was authenticated by case studies where the optimal system energy storage ratings for the islanded and grid-connected MGs were found out. For both the grid-connected and islanded MGs, the results obtained by using the suggested method demonstrate that the optimal size of BESS subsists and fluctuates.

The current control method in the a-b-c frame was put forward in Ref. [25] for a three-phase inverter. The suggested method was applied to control the dynamic and reactive power flow from the renewable energy source to a three-phase generalized micro grid system. In the existence of typical nonlinear loads, the suggested control system not only managed the grid power flow, but also diminished the grid current total harmonic distortion. The control system formed the grid current, taking into account the grid voltage unbalance, harmonics as well as unbalance in line side inductors. By the direct technique of Lyapunov, the constancy of the control system was made certain. To develop the presentation of the current controller, a SRC was moreover suggested by estimating the periodic disturbances of the system. The suggested control system offered better performance over the conventional multiple proportional-integral and proportional-resonant control techniques owing to the lack of the transformation blocks of the PARK with phase lock loop necessary in the control structure. To take care of unbalances both in grid voltages and line side inductors, a novel inverter modeling method was furthermore offered. It was verified that the suggested method result was competent.

By utilizing DERs, the micro grid works were conversed as a local energy contributor for domestic buildings to diminish energy expenses and gas emissions [26]. The fast advances in computing and communication capabilities facilitate the idea of smart buildings. The majority of energy-consuming household tasks do not require being executed at particular times rather than inside a desired time. If these types of tasks can be synchronized between multiple homes so that they do not all happen at the same time yet still please customers’ demand, the energy cost and power peak demand could be diminished. Using a mixed integer linear programming (MILP) strategy, it was suggested the optimal scheduling of smart homes’ energy consumption was learned. Based on real-time electricity pricing, electricity the task time window and predicted renewable energy output in order to minimize a 1-day forecasted energy consumption cost, DER operation and electricity-consumption household tasks were programmed. To diminish the peak demand from grid, the peak demand charge scheme was moreover implemented.

In a distribution system based on combination of photovoltaic array, an optimized design of micro grid FC and battery bank with multiple DG units under hybrid electricity market model was offered and the results were compared with those obtained from the pool electricity market [27]. Moreover, a GA-based optimization method was applied to attain optimum power and price of the MG. After that, an intention function based on the total net present worth was regarded and GA was used to attain the maximum net present worth of the MG, during interrelated operation by optimizing the manufacture of local DGs and power replaces with the major distribution grid.

The management of distributed energy with the MG is one of the multi-objective problems in energy management. Because, perfect economic model of energy source of the MG units are needed to describe the operating cost of the output power generated, the constraints of the multi-objective optimization problem are transformed into an easier sub-problem that can then be solved and used as the basis of an iterative process. GA is one of the global optimization techniques used to solve the optimization problem. For that reason, the best solution is converged to the global solution rather than to a local solution. Nevertheless, this dissimilarity happens to be uncertain while running with multi-objective optimization, which typically involves a set of solution points. Mathematically, a single global solution to a multi-objective optimization problem is not present unless the optimal solution happens to be attainable. Besides, the optimization process is dependent on genetic operators such as crossover, mutation, reproduction and etc. So, the computational complexity and time taken to converge the solution of this algorithm are increased.

Proposed methodology

Illustration of MG architecture

The distribution system is the final stage in the delivery of electricity to end users, which contains the MG energy management problem. The management of distributed energy with the MG is one of the multi-objective problems in energy management, because perfect economic model of energy source of the MG units are needed to describe the operating cost of the output power generated. Here, the proposed hybrid model is used for the optimal management of the MGs at a minimum fuel cost. The multi-objective function is to help reduce the fuel cost, operating and maintenance cost of the distribution system. The proposed method is applied to the MG architecture, as shown in Fig. 1.

The MG architecture contains a group of radial feeders and single connection point, i.e., point of common coupling (PCC). The feeder is connected to the sensitive and non-sensitive loads. The feeders also have micro sources consisting of WT, photo voltaic cell (PV), DG, FC and MT. Moreover, the feeders have MGs like WT, PV, DG, FC and MT. The DG, FC and MT need fuel for power generation but the WT and PV work as the fuel comes from nature. The static switch is used to island feeder from the utility when the event happens. The breakers are used to avoid the system reparation whenever unexpected contingences occur. The whole structure is used to solve the power demand problem with the use of MGs and the battery storage. The battery requires a separate charge controller to limit the depth of discharging, limiting the charging current supply to the battery and preventing the overcharging of the battery. So the power generation of all the MGs is used to serve the load as well as the battery. Each component used in Fig. 1 is modeled as per the configurations presented in Ref. [28].

Overview of the hybrid technique

The proposed method is the combined operation of the ABC in two phases, i.e., ABC_ Phase I and ABC_ Phase II. The optimal configuration of the MG is determined by the first stage of the ABC; it takes the cost functions of the DG, FC and MT, because the WT and PV generate power at zero operating cost. In load demand, satisfaction at minimum cost functions are used to develop the optimum configuration of the MG. From the optimal cost functions, the operating and maintenance cost of the MG and emission cost are reduced using the second stage of the ABC.

ABC_ Phase I

The required load demand is mostly utilized with the use of the WT and PV, due to the free generation cost. Whenever it does not satisfy the condition, the DG, FC and MT are taken into account for solving the problem. The selected configuration of the MG should satisfy the load demand with a minimum fuel cost. So the DG, FC and MT fuel cost functions are considered as the multi-objective function for the first stage of the ABC. The multi-objective function is described as

(1)\mathrm{\Phi }=\underset{C\in F}{min}\{f_{\mathrm{1}}(C),f_{\mathrm{2}}(C),f_{\mathrm{3}}(C)\},

where

(2)f_{\mathrm{1}}(C)={\displaystyle \sum _{i=\mathrm{1}}^N(d_i+e_i{P}_{DG,i}+f_i{P}_{DG,i}^{\mathrm{2}})}(\mathrm{F}\mathrm{u}\mathrm{e}\mathrm{l}\cos \mathrm{t}\mathrm{o}\mathrm{f}\mathrm{t}\mathrm{h}\mathrm{e}\mathrm{D}\mathrm{G}),

(3)f_{\mathrm{2}}(C)=C_{NG}{\displaystyle \sum _J{\displaystyle \frac{P_J}{{\eta }_J}}}(\mathrm{F}\mathrm{u}\mathrm{e}\mathrm{l}\cos \mathrm{t}\mathrm{o}\mathrm{f}\mathrm{t}\mathrm{h}\mathrm{e}\mathrm{F}\mathrm{C}),

(4)f_{\mathrm{3}}(C)={\displaystyle \sum _L{\displaystyle \frac{F_{NG}{P}_L}{{\eta }_L}}}(\mathrm{F}\mathrm{u}\mathrm{e}\mathrm{l}\cos \mathrm{t}\mathrm{o}\mathrm{f}\mathrm{t}\mathrm{h}\mathrm{e}\mathrm{M}\mathrm{T}),

in which, P_{DG,i} is the output power of the generator; N is the number of generators; d_i, e_i and f_i are the fuel cost coefficients with i=\mathrm{1},\mathrm{2},\cdots n; F_{NG} and C_{NG} are the natural gas price ($/kWh) of the MT and FC, F_{NG}=C_{NG}=\mathrm{2.3}\$/kWh; P_L and P_J are the output power of the MT and FC; and {\eta }_L and {\eta }_Jis the efficiency of the MT and FC, respectively, {\eta }_L={\eta }_J=\mathrm{0.47}. The steps to optimize the configuration of the MG can be described as follows:

Step 1: Initialize the population of the MG models (X_i) such as WT, PV, DG, FC, MT, cost functions and corresponding ratings.

Step 2: Generate the random number of population cost function depending upon the load demand.

Step 3: Employ bee phase, which evaluates the fitness of the population; the required multi-objective function is given as

(5)\mathrm{\Phi }=\underset{C\in F}{min}\{f_{\mathrm{1}}(C),f_{\mathrm{2}}(C),f_{\mathrm{3}}(C)\}.

Step 4: Set the iteration count to 1, i.e., iteration K = 1.

Step 5: Repeat

Step 6: The onlooker bee attains the elite configuration of the MG combinations and improves the velocity of the populations using

(6)V_{i,j}=x_{i,j}+{\mathrm{\Phi }}_{i,j}(x_{i,j}-x_{k,j})

where k=(\mathrm{1},\mathrm{2},\cdots ,n) and j=(\mathrm{1},\mathrm{2},\cdots ,n) are the randomly chosen index.

Step 7: Apply the selection process to find the better fitness of the new solutions and determine the probability.

(7)Probability={\displaystyle \frac{\mathrm{\Phi }}{{\displaystyle \sum _{i=\mathrm{1}}^n\mathrm{\Phi }}}}.

Step 8: If better solutions are not achieved, abandon the solutions and produce the random number of scout bee solution using

(8)x_i^j=x_{\min }^j+rand[\mathrm{0},\mathrm{1}](x_{\max }^j-x_{\min }^j).

Step 9: Memorize the best solution achieved so far.

Step 10: Check the iteration range and follow the conditions that if the iteration does not reach the maximum value, increase the iteration count K = K+1, and that when the iteration reaches the maximum value, the process has been terminated.

The second stage of ABC operation process is used to reduce the operating and maintenance cost of the MG.

ABC_ Phase II

The optimized configuration of the DG, FC and MT at a minimum fuel cost is attained from the first stage of the ABC output. From the optimized MGs costs, the second stage of the ABC reduces the operating and maintenance cost. It takes into account the environmental externality costs by minimizing the emissions of oxides of nitrogen (NO_x), sulfur oxides (SO₂), and carbon oxides (CO₂). The objective function is developed according to the following requirements to minimize the cost of the operating and maintenance cost of the MG, which is described as

(9)\mathrm{\Psi }=\underset{C\in F}{min}\{F_i(C)\},

where

(10)F(C)={\displaystyle \sum _{i=\mathrm{1}}^N(C_i{F}_i+{OM}_i)+{\displaystyle \sum _{i=\mathrm{1}}^N{\displaystyle \sum _{j=\mathrm{1}}^M{\alpha }_j({EF}_{ij}{P}_j)}}},

in which C_i is the fuel cost of the generating unit in Rs/L for the diesel and Rs/kW for the natural gas; F_i is the fuel consumption rate of a generating unit; {OM}_i is the operating and maintenance cost of a generating unit, and is the externality cost of emission type j; {EF}_{ij} is the emission factor of the generating unit; M is the emission type; and N is the number of generating units. The steps to optimize the operating and maintenance cost of the MG is given as follows:

Step 1: Initialize the output population (Y_i) of the first stage of the ABC, i.e., optimum configuration of the MG at a minimum fuel cost and emission cost and externality factors.

Step 2: Generate random number of operating and maintenance cost depending on the load demand and cost functions.

Step 3: Employ bee to compute the fitness function of \mathrm{\Psi }=\underset{C\in F}{min}\{F_i(C)\}.

Step 4: Set the iteration count to 1, i.e., iteration K = 1.

Step 5: Repeat

Step 6: The optimum operating and maintenance cost for appropriate load demand is determined using the onlooker bee and the cost is improved using

(11)V_{i,j}=y_{i,j}+{\mathrm{\Psi }}_{i,j}(y_{i,j}-y_{k,j})

where k=(\mathrm{1},\mathrm{2},\cdots ,n) and j=(\mathrm{1},\mathrm{2},\cdots ,n) are the randomly chosen index.

Step 7: Apply the selection process to find the better fitness of the new solutions and determine the probability:

(12)Probability={\displaystyle \frac{\mathrm{\Psi }}{{\displaystyle \sum _{i=\mathrm{1}}^n\mathrm{\Psi }}}}.

Step 8: The better solutions are not yet found out and the scout bee develops the new solution using

(13)y_i^j=y_{\min }^j+rand[\mathrm{0},\mathrm{1}](y_{\max }^j-y_{\min }^j).

Step 9: Store the best solution found out so far.

Step 10: Increase the iteration count K=K+1, until it reaches the maximum value.

Once the process is completed, the proposed hybrid technique is ready to provide the optimal configuration of the MG at a minimum fuel cost. The operation of the proposed hybrid technique is illustrated in Fig. 2. The proposed technique is tested with the Matlab platform and the effectiveness analyzed through the comparison.

Results and discussion

The proposed hybrid model is implemented in the Matlab 7.10.0 (R2010a) platform and the simulink diagram is depicted in Fig. 3. The numerical results of the proposed method are compared with those of the other existing techniques, i.e., online MG management and ABC, which is helpful for identifying the effectiveness of the proposed method. The total costs or operating and maintenance costs are analyzed in the various load demand like 4 kW to 14 kW. The implementation MG ratings and emission factors are listed in Table 1.

The values in Table 1 are used to find the optimal management of the MG. The ABC algorithm and the proposed methods are allowed for 100 numbers of cycles and 20 numbers of bees. It takes 7.897576 s for the ABC to complete the maximum cycles, but it only takes 6.943122 s for the proposed hybrid method to do so. Then, the obtained results are presented in Table 2, which demonstrate the power generation range and total cost of the MG sources at various power demands.

Table 2 describes the total cost utilized for power generation of the MG under online optimal management of MG, ABC and proposed hybrid method. It is observed from Table 2 that the online management of the MG and ABC gets a high amount of total cost for all type of load demand variations, which is slightly higher compared to the proposed method in this paper.

The optimal configuration of the MG with a minimum cost is selected from the hourly load demand basis, because the power system load demand varies in every situation. So the MG source allocation mainly depends on the load demand, cost and emission. Figure 4 illustrates the load demand for every hour in one day and the corresponding load power remained from the PV and WT, which are plotted against time. The load demand can be satisfied with the proper management of the MG sources. Figure 5 shows the optimum selection configuration of the MG sources such as DG, FC, MT and battery for every hour. The operating and maintenance cost for the MG are taken into consideration, so it is attained from different techniques such as online optimum management of the MG, ABC and the proposed hybrid method. The fuel cost attained from the online optimal management of the MG is exhibited in Fig. 6, in which the fuel cost functions for every hour is calculated. The selection of the power generation of the MG is depending on the load demand. Similarly the fuel cost for ABC is presented in Fig. 7. Then the fuel cost of the proposed hybrid method is displayed in Fig. 8. From Figs. 7 and 8, it can be observed that the proposed method is optimally managing the load demand by optimal selection of the power generator of the MG at a minimum cost. Using the minimum fuel cost function, the operating and maintenance cost or total cost of the selected MG is analyzed in Fig. 9. Also the total cost required for the optimal management of the MG based on the iteration is illustrated in Fig. 10. The proposed hybrid method is compared with the ABC algorithm. Finally the total cost comparison is proved and the proposed hybrid method is an efficient method to manage the MG at the required demand with reduced cost and emission.

Conclusions

This paper presented a hybrid ABC approach for modeling and managing of a MG connected system. The ABC algorithm was modeled into the two stages depending on the objective functions. The first stage of the ABC provides the optimum configuration of the MG at a minimum fuel cost. Using the minimum cost function, the second stage of the ABC was attained at the minimal operating and maintenance cost. Then, the numerical results were verified with different techniques under various load demands. The hourly load demand based fuel cost, operation and maintenance cost and total cost were attained for various techniques. From the results obtained, it is clearly observed that the proposed method have attained the best selection of the power generators of the MG under various power demands at a minimum cost. The comparison results prove that the proposed method is an effective technique to model and manage the MG connected system, which is more competent over the other techniques.

References

[1] Baba J, Numata S, Suzuki S, Kusagawa S, Yonezu T, Denda A, Nitta T, Masada E. Fundamental measurements of a small scale micro grid model system. In: Proceedings of International Conference on Electrical Engineering. Kunming, China, 2005, 1-6

[2] Mohamed F A, Koivo H N. System modelling and online optimal management of microgrid using mesh adaptive direct search. International Journal of Electrical Power & Energy Systems, 2010, 32(5): 398-407

[3] Katiraei F, Iravani M R. Transients of a micro-grid system with multiple distributed energy resources. In: Proceedings of International Conference on Power Systems Transients. Montreal, Canada, 2005, Paper No. IPST05-080

[4] Kriett P O, Salani M. Optimal control of a residential microgrid. Energy, 2012, 42(1): 321-330

[5] Gerry S. Optimal rural microgrid energy management using HOMER. International Journal of Innovations in Engineering & Technology, 2013, 2(1): 113-118

[6] Mohamed F A, Koivo H N. Environmental/Economic power dispatch of microgrid using multiobjective genetic algorithms. In: Proceedings of International Conference on Renewable Energy Congress. Sousse, Tunisia: CMERP, 2010, 495-500

[7] Kremers E, Viejo P, Barambones O, de Durana J M G. A complex systems modelling approach for decentralized simulation of electrical microgrids. In: Proceedings of 15th IEEE International Conference on Engineering of Complex Computer Systems. Oxford, UK, 2010, 302-311

[8] Zhang Y, Gatsis N, Giannakis G B. Robust energy management for microgrids with high-penetration renewables. IEEE Transactions on Sustainable Energy, 2013, 4(4): 944-953

[9] Mashhour E, Moghaddas-Tafreshi S M. Mathematical modeling of electrochemical storage for incorporation in methods to optimize the operational planning of an interconnected micro grid. Journal of Zhejiang University Science, 2010, 11(9): 737-750

[10] Lasseter R H, Piagi P. Extended microgrid using (DER) distributed energy resources. In: Proceedings of IEEE Power Engineering Society General Meeting. Tampa, USA, 2007, 1-5

[11] Nichols D K, Stevens J, Lasseter R H, Eto J H, Vollkommer H T. Validation of the CERTS microgrid concept the CEC/CERTS microgrid testbed. In: Proceedings of IEEE Power Engineering Society General Meeting. Montreal, QC, Canada, 2006, 1-3

[12] Lasseter R H, Piagi P. Extended microgrid using (DER) distributed energy resources. In: Proceedings of IEEE Power Engineering Society and General Meeting. Tampa, USA, 2007, 1-5

[13] Kroposki B, Pink C, Lynch J, John V, Meor Daniel S, Benedict E, Vihinen I. Development of a high-speed static switch for distributed energy and microgrid applications. In: Proceedings of IEEE Power Conversion Conference. Nagoya, Japan, 2007, 1418-1423

[14] Gerry S. Optimal rural microgrid energy management using HOMER. International Journal of Innovations in Engineering and Technology, 2013, 2(1): 56

[15] Mohamed F A, Koivo H N. System modelling and online optimal management of microgrid using multiobjective optimization. In: Proceedings of IEEE International Conference on Clean Electrical Power (ICCEP). Capri, Italy, 2007, 148 -153

[16] Alikhani E, Ahmadian M, Salemnia A. Optimal short-term planning of a stand-alone microgrid with wind/PV/fuel cell/diesel/microturbine. Canadian Journal on Electrical and Electronics Engineering, 2012, 3(3): 135-141

[17] Olivares D E, Cañizares C A, Kazerani M. A centralized optimal energy management system for microgrids. In: Proceedings of IEEE Conference on Power Engineering Society General Meeting, 2011, 1-6

[18] Kariniotakis G N, Soultanis N L, Tsouchnikas A I, Papathanasiou S A, Hatziargyriou N D. Dynamic modeling of microgrids. In: Proceedings of International Conference on Future Power Systems. Amsterdam, Netherlands, 2005, 1-8

[19] Khamis A, Mohamed A, Shareef H, Ayob A. Modeling and simulation of small scale microgrid system. Australian Journal of Basic and Applied Sciences, 2012, 6(9): 412-421

[20] Jaganathan S, Palaniswami S, Adithya R, Kumaar M N. Synchronous generator modelling and analysis for a microgrid in autonomous and grid connected mode. International Journal of Computers and Applications, 2011, 13(5): 3-7

[21] Chen C, Duan S, Cai T, Liu B, Hu G. Smart energy management system for optimal microgrid economic operation. International Journal of Renewable Power Generation, 2011, 5(3): 258-267

[22] Conti S, Nicolosi R, Rizzo S A, Zeineldin H H. Optimal dispatching of distributed generators and storage systems for MV islanded microgrids. IEEE Transactions on Power Delivery, 2012, 27(3): 1243-1251

[23] Tan K T, Peng X Y, So P L, Chu Y C, Chen M Z Q. Centralized control for parallel operation of distributed generation inverters in microgrids. IEEE Transactions on Smart Grid, 2012, 3(4): 1977-1987

[24] Chen S X, Gooi H B, Wang M Q. Sizing of energy storage for microgrids. IEEE Transactions on Smart Grid, 2012, 3(1): 142-151

[25] Dasgupta S, Mohan S N, Sahoo S K, Panda S K. Lyapunov function-based current controller to control active and reactive power flow from a renewable energy source to a generalized three-phase microgrid system. IEEE Transactions on Industrial Electronics, 2013, 60(2): 799-813

[26] Zhang D, Shah N, Papageorgioua L G. Efficient energy consumption and operation management in a smart building with microgrid. Energy Conversion and Management, 2013, 74: 209-222

[27] Mohammadi M, Hosseinian S H, Gharehpetian G B. GA-based optimal sizing of microgrid and DG units under pool and hybrid electricity markets. International Journal of Electrical Power & Energy Systems, 2012, 35(1): 83-92

[28] Mohamed F A, Koivo H N. System modeling and online optimal management of microgrid with battery storage. In: International Conference on Renewable Energies and Power Quality. Sevilla, Spain, 2007, 1-5

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag Berlin Heidelberg