Optimization techniques applied for optimal planning and integration of renewable energy sources based on distributed generation: Recent trends

Abstract Numerous potential advantages to the requirements and effectiveness of the supplied electricity can be accomplished by the installation of distributed generation units. In order to take full advantage of these benefits, it is essential to position the Distributed Generation (DG) units in appropriate locations. Otherwise, their installation may have an adverse effect on the quality of energy and system operation. Several optimization techniques have been created over the years to optimize distributed generation integration. Optimization techniques are therefore constantly changing and have been the main attention of many fresh types of research lately. This article evaluates cutting-edge techniques of optimizing the issue of positioning and sizing distributed generation units from renewable energy sources based on recent papers that have already been applied to distribution system optimization. Furthermore, this article pointed out the environmental, economic, technological and regulatory drivers that lead to a rapid interest in the DG system based on renewable sources. A summary of popular meta-heuristic optimization tools discussed in table form with merits and demerits to increase fresh prospective paths to multi-approach that have not yet been studied.


PUBLIC INTEREST STATEMENT
Several optimization techniques have been created over the years to optimize the integration of distributed generation. Optimization techniques are therefore constantly changing and have been the main attention of many fresh types of research lately. This article evaluates cuttingedge techniques of optimizing the issue of positioning and sizing distributed generation units from renewable energy sources based on recent papers that have already been applied to the distribution system. Furthermore, this article pointed out the environmental, economic, technological and regulatory drivers that lead to a rapid interest in the DG system based on renewable sources. According to the investigation carried out, the area of artificial intelligence techniques is still receiving attention than conventional optimization techniques for optimal DG planning in a power distribution network system from different point of view. Computational techniques in hybrid optimization techniques convergence take place faster than the conventional optimization techniques.

Introduction
Traditional power generation is currently unable to satisfy the ever-increasing worldwide demand for electricity. About 16% of the world's population still lives without electrical energy due to poor network construction (Report, Global Status, 2017), Subsequently, a power supply system needs to accommodate these changes for a better quality of user experience ("Ekpa, T. K. 1*, Sani, S.2, Hasssan, A. S.3 and Kalyankolo, Z.4 1, Journal of the Nigerian Association of Mathematical Physics Volume 48 (Sept. & Nov., 2018 Issue), Pp347-352 © J. of NAMP ON," n.d.). Unfortunately, distributed generation (DG) has proven to be a feasible alternative solution in this view were electricity is produced close to the load centers. Although DGs have several environmental and economic advantages, in the distribution system, they impose various operational problems. These may include, but not restricted to, power relaying problems created by inverse power flow, the problem of voltage increase and power quality issues (Lam & Varbanov, 2011;Manfren et al., 2011;Quadri et al., 2018). DG is a small scale electrical power generation units connected directly to the loads or consumer meter side (Hydro, Alternate and Energy Centre, 2016). Various researchers and countries have adopted a different definition for DG (Approach, 2007) defines distribution system as a failure of a digital system that has not existed in your system and can solve other unstable decisions within the given computer system. Historical background on the concept of DG was proposed by (Public et al., 2014), which categorized DG as a small device that allows electrical power energy to be generated by using renewable energy sources. In (Friedman, 2002), DG defined as a small modular plant place close to the load for the purpose of electricity generation and storage technologies within the grids. Ref (Martin, 2009) classified DG as a point where electricity is generated at the point of its use. The electrical power research institute (EPRI) defined DG as a few kW to 50 MW (Rajkumar Viral & Khatod, 2012). For Details in various DG, definition see Martin (2009), Abou El-Ela et al. (2010) and van Gerwen (2006). However, different types of renewable energy DG technology categories were highlighted in (Østergaard et al., 2001;. Many authors characterized DG as a generation connected on the distribution side that ranges from a few kW to a few tens of MW, as shown in Table 1 Treatment, Vacuumsteam-vacuum Decontamination, 2015 (Ehsan & Yang, 2018a) as shown in Figure 1. DG has become a feasible solution to rural areas where the cost of transmission and distribution is extremely high, and this making DG more popular. In (Davis & Ieee, 2002), comparison between distributed resources and the traditional power system in terms of transmission and distribution system, to check efficiency, losses, voltage profile, reliability, emissions, and power quality was studied and the author findings concluded that investment cost in distributed resources is lower compared to the traditional system. The technologies adopted in distributed generation comprises a small gas turbine, fuel cells, micro-turbines, wind, solar energy, and hydro-power. Figure 2, shows a distinct difference between the central utility of today and distributed utility of tomorrow. In the development of traditional power plants, there are technical and environmental limitations such as fuel sourcing, global sourcing, emissions and localized pollution. Moreover, the unsafe market for fossil fuel has pushed the electricity market in searching for new sources of energy (Mahmoud et al., 2016). Adding DG to the energy scheme decreases electrical power losses to achieve greater reliability of the system (Hassan et al., 2018). However, a U-shape trajectory is presented by the DG penetration rate versus power loss. The non-optimal positioning of DG can, therefore, boost power losses and  thus enhanced the voltage profile to support the permissible limit (Quezada et al., 2006). However, utilities are already subjected to technical and non-technical issues. Therefore, an optimum placement and sizing renewable DG are required to minimize the power system losses, boost reliability and stability and enhancement of voltage profiles. Authors in this field have contested that DG capacity and locations are very important in improving the distribution system performance (Hemdan & Kurrat, 2008), and the authors finding were based on the presence of renewable DG such as (solar, wind and fuel cell), increases the load of the network system.
There are various approaches carried out in literature to find the optimal position for the integration of DG into the distribution network system. Various objective functions (Bus cumulative magnitude and voltage deviation) are developed and solved. In (Borges & Falcão, 2006), the objective function formulation is the maximization of cost-benefit of DG integration to the investment cost. The benefits were to reduce the net electrical losses and the investment cost as well as installation costs. A multi-objective technique based on a genetic algorithm was studied by (Ochoa & Harrison, 2011) to determine optimal power flow to accommodate renewable energy sources by minimizing total energy losses. In (Q. Qianyu Zhao et al., 2019), the objective function is to investigate the uncertainty of DGs and loads, power generation cost and environmental cost of different renewable DG integration. A multi-objective optimization problem applied based on a double trade-off procedureε À constrained approach for simultaneous minimization of DG integration by considering, cost energy, cost of energy not served and cost of grid energy purchased. The proposed methodology is capable of solving of maximizing network performance by optimizing some elements like voltage quality and harmonic distortion (Carpinelli et al., n.d.). In (""Optimal Placement of Multi-Distributed Generation Units Including Different Load Models Using Particle Swarm Optimisation.", 2011), a multi-objective index used to optimize the short-circuit level parameter which will affect the sizing and location of the DG. The effects of load model size investigated by D. Singh et al. (2009), based on multi-objective function using various indices. It is observed based on the model that the effects of load models can significantly disturb planning and sizing DG. Fuzzy logic-based planning by considering uncertainty modeling to incorporate the DG within the electrical power distribution is proposed in (Ganguly et al., 2013), the objective was the application of a Petro-based approach for total optimization of cost of planning, reliability and risk of, constraints fully minimized. In (Hydro, Alternate and Energy Centre, 2016) analytical approach-based techniques proposed for optimal sizing and sizing DG for a balanced radial system. The model identifies the buses in the network to be compensated by reducing the power losses with single DG placement in DN. Growing interest in the application of DG optimization techniques deploy by the use of renewable energy sources is extended all over the globe, it is examined that one quarter of that energy is witnessed in Europe, Asia and part of United States of America (Abdmouleh, Gastli, Ben-brahim et al., 2017). Despite the advantages derived from DG integration from renewable energy source research shows that utilities suffer a great system loss, from inappropriate placement and sizing (Griffin et al., 2000;Mithulananthan et al., 2004). Proper mathematical optimization techniques provide solutions for boosting the reliability of this utilities deployed. Table 5 summarized DG technologies by bringing out benefits of each sources in terms of (emission, voltage profile, cost of packing, cost of installation, reliability improvement and power quality).
This research paper presents a vast choice on the current diversity of optimization methods applied to the aspect of planning and integration of renewable-based DG. The major objective is to focus on solving the problem of optimal placement and sizing DG from renewable energy sources. In addition, other factors affecting DG planning are reviewed and discussed in detail. This study will provide general knowledge to researchers for further exploration.

Prevalence weakness in DG growth
Optimal allocation of distributed generation has received much attention recently due to its various importance. However, it becomes an exacting task in integrating the DG into an existing network system. This challenging effect arises, because the integration of DG changes the entire system behavior from active to passive. Many authors have pointed out the importance of placing DG in an optimum position. In (Murty & Kumar, 2015;Tan et al., 2013;Ugranli & Karatepe, 2013), DG properly installed in an optimum position will enhance the voltage profile of the electrical power distribution network. If installed at the best location and the proper size, they will minimize power losses and maximization of system voltage stability (Aman et al., 2012;Kalambe & Agnihotri, 2014;Kansal et al., 2013;Murty & Kumar, 2015;Pepermans et al., 2005;Vijay & Singh, 2015). Ref (Lopes et al., 2007), pointed out the benefits of integrating the DG by using renewable energy sources (wind, solar, biomass and so on), which regulates environmental effects like emission control. In (Allan et al., 2015), review various literature on different DG technologies on how to control the emission. DG serves as an alternative source in assisting the rural side, where the cost of transmitting and distribution of power system is high see details (Karki et al., 2008). If properly installed at the proper location will relieve the issues of uncertainty loading of the feeders (Zeinalzadeh et al., 2015;Muthukumar & Jayalalitha, 2016). Inappropriate optimum allocating and sizing the DG can lead to a negative impact of all the advantages mentioned (P. Prem Prakash & Khatod, 2016). It has become necessary to size and allocates the DG in an optimum position, since the advancement in the technology is rapidly growing by cutting various costs, boosting efficiency and enhancing the voltage profile within the power distribution network. Despite the availability and environmental friendly of these energy sources, the institute of electrical electronics engineering (IEEE) has set various rules to follow in integrating distributed energy sources (DER) in the distribution system. Table 3 provides the IEEE 1547 series towards achieving a sustainable environment, Table 4 provides voltage and frequency acceptable rate in operating DER in the distribution system. Most of the benefits of employing DG in existing distribution networks have both economic and technical implications and they are agnate. However, the major driving factors increased in penetration of DG integration issues are classified into three categories: environmental, regulatory and economic factors.

A. environmental aspect
In environmental aspects, the broad penetration of distributed generation in the distribution network is of major significance. DG-based fossils fuel dissipates a lot of environmental emissions and leads to unresolved issues. However, the DG implementation and only restricted measurements are characterized by a wide range of fuel energies, advanced techniques and operating patterns (Greene & Hammerschlag, 2000). The utilization of fossils-based DG energy sources leads to the largest emission of Greenhouse (GHG) gas emission such as (water vapor, carbon dioxide, nitrous oxide and hydrofluorocarbons), which resulted in environmental concerns and climate change. Environments and creatures are affected worldwide, this has led to the search of non-polluting resources and more efficient technologies that will solve environmental problems and reduction in the price of fossil energy (Chaitusaney, 2014). The integration of DG on large scale produces various types of emissions, as electricity generation is always the major contributor to these emissions. Statistics have shown that in the united states of America that electricity production from non-renewable DG sources has resulted in one-quarter of nitrogen oxide (NOx), carbon monoxide and sulphur dioxide (SO 2 ) (States, n.d.). However, despite this negative environmental impact from the non-renewable energy sources, DG could have positive impacts of integrating renewable energy sources into the electrical power network which will reduce the gaseous emission mentioned above. Various literature exists on the use of renewable energy sources to reduce gaseous emission see (M. Chen & Cheng, 2012;Liew et al., 2017;Di;Somma et al., 2016;Cao et al., 2016;Akorede et al., 2010;Abdallah & El-Shennawy, 2013). Table 3(a-b) are environmental standard value for pollutant and penalty emission control (M. Chen & Cheng, 2012) and Table 4, shows potential emission reduction of Carbon (IV) oxide and electricity emission to be achieved in 2030 ([PNNL] Pratt et al., 2010).

B. economic aspects
DG's interconnection to the traditional electrical power system needs additional underground cable installation so that DG can be integrated into the new scheme. In addition, the incorporation of DG into the new scheme can lead to certain problems such as; voltage flicker, harmonics introduction, the inverse operation of the energy system flows and protection challenges see (Akorede et al., 2010;Barker & De Mello, 2002;Hadjsaid et al., 1999), for details. Such issues must be properly addressed to achieve the maximum benefits of the reliability of the distribution system (A. K. Singh & Parida, 2017a). Summary of these challenges is better explained in Table 4(a) (Zubo et al., 2016). In addition to these economic benefits, DG still represents the world energy through reduction of cost, saving transmission and distribution.

C. Technical aspect
DG technical element covers a broad range of issues such as peak load shaving, excellent voltage profile, decreased system losses, enhanced system continuity and reliability, and some power quality issues that are filtered out. The total reduction in the loss of energy system may be of concern to some utilities in developing nations, as some of these utilities lose up to 20% of their complete energy generation (A. K. Singh & Parida, 2017b). The major technical benefit to be derived from DG, if well addressed is as follow: • The total reduction in line losses.
• Boosting system reliability and power system security.
• Increased system efficiency.
• Relieving congestion in the transmission and distribution system.

D. Regulatory policy
It appears that the creating of suitable approaches is so significant to help in the coordination of the distributed generation into an appropriate distribution system because of the nonappearance of clear legislative guidelines (Niwas et al., 2009). Nowadays particular attention is paid in the European zone on how to tackle the effects of climate change and have a strong body that will promote the use of clean energy towards achieving secured energy (Cossent et al., 2009). Table 2 summarizes some weaknesses of DG.

Review of optimization techniques applied on DG's based on optimal placement and sizing
Optimal allocation techniques solve problems linked to sizing and appropriate positioning of DG's by enforcing various mathematical optimization objective features with many technical operational limitations, implementing distinct computational methods with the account of distinct DG kind based on power factor (PF) and multiple numbers of units. This study presents a trend on research publications since the early 2000 s to date and further summarized as following in block diagram for easy understanding. These objectives are considered as single or multiple for Optimal Allocation of DG in the presence of equality and non-equality constraints. Figure 3 summarizes the review approach of the optimal allocation of DG in the form of a block diagram.
Optimal allocation and sizing of DG are not a linear method; rather they are solved as a nonlinear optimization method. The optimal allocations of DG are mathematically modeled in the form of mixed-integer non-linear programming (MILP). The majority of the literature survey carried out are basically methodologies to investigate the optimal allocation of renewable-based DGs in the distribution system for minimization power loss and cost associated.
The authors in (Wong et al., 2019), present a novel progress approach for optimal and sizing distributed generation with the placement of control storage in order to investigate the challenges of deploying Energy storage system in the distribution system. In Ref (Sedighi, 2010), authors applied Particle Swarm Optimization (PSO), for optimal sitting and sizing of DG in the distribution system as

Indirect (%)
Transmutation effect of consumer information and feedback systems.

-
Joint marketing of electrical energy efficiency and demand response application. -0 Categorization of analytics in the residential, small and medium building.

-
Quantification and ratification for energy Efficiency programs (M&V).   well as boost the voltage profile and total harmonic distortion reduction. Also in Rugthaicharoencheep and Sirisumrannukul (n.d.), an investigation was carried out by the authors to determine the importance of optimal reconfiguration of feeders within a given distribution system using a fuzzy logic approach, in order to minimize power loss, feeder balancing, and switches operation.
Ref (Soudi, 2013); studied the optimal planning of DG considering the significant and cost associate in order to improve power quality and greater reliability of the system. (Kaur et al., 2014); proposed a mixed-integer nonlinear programming technique for optimal placement of multiple DGs in a distribution network power loss minimization. Proposes a multi-objective optimization for sitting, sizing and placement of the DG in an optimal position to minimize power loss, voltage variation as well as stability. Angarita et al. (2016) also presented a novel for optimal allocation of DG in distribution system the major objective is the minimization of electrical power loss and regulate voltage profile. Subramanyam et al. (2018); presented an optimized dual-stage approach for setting and sizing fuel cell in DGs for effective system minimization. Ref (Hung et al., 2010), proposes an analytical optimization to determine size and power factor correction in the active primary distribution network to reduce losses associated with deploying DGs in DN. Hassanzadehfard and Jalilian (2018) proposes an optimization technique for proper sizing and locating renewable-based DG considering load growth in the active distribution system. The major finding is the placement of solar and wind in optimal place in DN to reduce operational cost, maintenance cost as well as Greenhouse gas emission. Badran et al. (2017) focused on optimization techniques for network reconfiguration considering different methodologies to alleviate total power loss in the distribution system. Also, Abou El-Ela et al.'s (2010) major findings were an investigation on using genetic algorithms for maximal placement of DG in the network system in order to boost voltage profile, spinning reserve control as well as a reduction in total power line losses. Karimi et al. (2016) proposes a multiobjective stochastic approach for integrating DG in the distribution network by considering economical, technical and environmental aspects by considering uncertainty. Major findings by authors were how safe is it to integrate DGs in DN to determine optimal location and size by considering some environmental parameters. Taylor (n.d.) carried out a cost-benefit analysis to determine the optimal position to place DG in DN. The investigation is in terms of placement of DG in an optimal position by considering the size of the DG and capacitor. In ref (Kumawat et al., 2017), perform a swarm intelligence-based optimization for the planning of DG in DN to minimize annually energy loss. Findings by the author were to observe the behavior of a nonlinear electrical load with time-varying characteristics to obtain the real load in the DN. Also in (Kanwar et al., 2016;Tlbo et al., 2017), carried out optimization techniques for optimal placement of DG in the distribution system considering network reconfiguration, shunt capacitor and DG in the radial distribution system. Hassanzadehfard and Jalilian (2016) proposes a novel met the heuristic mathematical objective for placement of multiple DGs in an optimal position within the distribution system to minimize the cost associated with deploying DG. (S. Singh, Figure 3. Summarized techniques for optimal allocation of multiple DG. 2016), applied a particle swarm optimization algorithm to find the best location to place their DG within the given electrical distribution network, the aim is to minimize total power losses and enhanced voltage profile. Reliability indices topology was adopted by Siddappaji and Thippeswamy (2017), to determine the best placement of DG in a distribution system for total power loss reduction using a fast decoupled method. Authors in (A. R. Gupta, 2017), investigated the effect of placement of multiple DG and STATCOM in the radial distribution system. This will help in mitigating the effect of power quality, and harmonics introduce by non-linear loads in a radial system. Authors in (Y. Aien et al., 2016Aien et al., , 2014Grechuk & Zabarankin, 2018;Jordehi, 2018;Soroudi & Amraee, 2013;Yuan Zhao et al., 2015), carried out a review on a different aspect of uncertainty modeling of DGs. The applied method can be summarized in figure 4.

A. Various DG planning models
Uncertainties and fluctuation are the primary difficulties related to Renewable energy technologies, particularly with non-consistent accessibility of wind, sun based and hydro sources. To suit the incorporation of an enormous offer of variable energy sources, it is critical to have fitting arranging devices ready to enhance the reconciliation of variable renewable energy sources. Numerous advancement systems identified with vitality issues when all is said exist in the literature, for example, intelligent search techniques. In central, searching for the ideal site and limit the search of Distributed Generation is typically modeled as a non-linear optimization hurdle. Different limitations and objective constraints are first set. The advancement strategy helps in decision making by creating an optimal solution from a decrease in initial set up of variables. Comprehensively, there are two ways to deal with the issues, by precise strategies, for example, Mixed-Integer Linear Programming (MILP) which is generally exceptionally modeling, however, requires inordinate processing time and difficult to actualize on genuine size issues, and heuristic techniques which depends on improving the issue and offering fulfilling arrangements. In this area, a straightforward definition of the most widely recognized issue is introduced, which is to locate the ideal DG size and location for the bus that minimizes the total power loss (Abdmouleh, Gastli, Ben-brahim et al., 2017). It is important to assign distributed generation units at optimal places with appropriate sizes to reduce negative impact in terms of economic, technical and environmental factors. Several advantages of placing DG in an optimum position include; Enhancing voltage profile, stability and reliability increased, minimization of total power loss and power quality issues. In this section, various approaches will be discussed in detail.
DG's are classified into four major distinct group based on their terminal performance in terms of real and reactive power delivering capability as follows: (1) Type 1: DG capable of injecting P only.
(3) Type 3: DG capable of injecting both P and Q.
(4) Type 4: DG capable of injecting P but consuming Q.
Photovoltaic, microturbines, fuel cells, which are integrated into the main grid with the help of converters/inverters are good examples of Type 1. Type 2 could be synchronous compensators such as gas turbines. DG units that are based on synchronous machines (cogeneration, gas turbine, etc.) fall in Type 3. Type 4 is mainly induction generators that are used in wind farms.

B. Analytical approach
Analytical approaches usually used for numerical approximation. The analytical algorithm approach for determining optimal DG size, location and installation has been reported in the literature by different authors (Aman et al., 2012;Gözel & Hakan Hocaoglu, 2009;Hedayati et al., 2006;Hung et al., 2010). Yarahmadi and Shakarami (2018) propose an analytical solution for optimal allocation of wind-based in radial distribution by partial derivative taking into account distinct time and voltagedependent load using a multi-objective index for efficient placement of the DG wind in an optimum location. The objective goals were based on the use of the Rayleigh pdf model to determine the nature of the wind penetration, the result obtained was tested on 33 and 69 bus system. In ref (Wang & Hashem Nehrir, 2004), the authors propose an iterative method based on analytical to determine the best location to place a DG in a radial distribution network. The objective was to minimize power loss in the network system. Placement of the DG was analyzed in a radial feeder system, the location of the bus site was obtained from the different combination of the load's sources, and finally, an analytical system was applied to generate the bus admittance matrix in the distribution system and tested on IEEE 6 and 30 bus system. Ref (Acharya et al., 2006), used exact loss formula-based analytical expression for placement of the DG in an optimum location to minimize total power loss in the primary distribution system. The basic objective was to calculate optimum size and methodology to place the DG in the best location to minimize system losses. A mathematical objective was formulated in (Khatod, 2015), to determine the optimal setting and sizing of the DG based on the analytical approach in the radial distribution network to minimize power loss on the system. The novelty of the work is the application of a simple analytical formula to minimize power system losses, with the presence of active and reactive components, and the result obtained was tested on 15 bus systems and found to fit the objective as proposed.

C. 2/3 rule
The 2/3 rule also known as the golden rule, it is a well-known analytical technique used for optimal allocation of DG. Optimally used for placement of shunt capacitors in a radial distribution system based on power flow. Here the size of the capacitor placement based on the DG is 2/3 of the incoming capacity generated is size to fit 2/3 length of the line in the system. This rule is applicable where the load is uniformly distributed in a radial network see (Rau & Wan, 1994;Willis, 2002) for details on the golden rule.

D. Sensitivity analysis-based approach
These techniques are basically used to find the exact location of the distributed generation based on sensitivity index node identification, and tend to find the feasible location for our DG in order to minimize total power loss (Murty & Kumar, 2015), also applied in (Rau & Wan, 1994). However, various literature has reported on the use of analytical analysis based on the sensitivity factor due to its simplicity (Acharya et al., 2006;Gozel et al., 2005;Hung et al., 2010;Kashem et al., 2006;Murty & Kumar, 2015). Authors in (Khatod et al., 2013), propose a sensitivity analysis technique based on active and reactive power, to give the best location for placement of the DG thereby reducing the computational time.

Linear and non-linear programming (MINLP)
The first set of iteration to solve non-linear programming is setting the direction of the search a linear programming problem is characterized by linear functions of unknowns, the objective of linear programming is that the unknown is linear and the constraints are linear inequalities. However, the popularity of linear programming depends on the primary formulation of the analysis. The concept of non-linear programming has been reported by different authors in the literature (Luenberger, 1973;Jaravel et al., 2019;Ballesteros-Pérez et al., 2019;Rueda-medina et al., 2013;A. Kumar & Gao, 2010;Felix F. Wu et al., 2005;Cerone et al., n.d.). However, the mathematical concept of solving advanced models is mixed-integer non-linear programming (MILNP). Mixedinteger non-linear programming refers to optimization problems with continuous and discrete variables, the constraints imposed are in the form of a non-linear function. MINLP has a wide range of applications including finance, engineering, and manufacturing process. MINLP has been used by the various authors to determine the optimal size and location of DG in an electric power distribution system as discussed by (Home-Ortiz et al., 2019;Nemati et al., 2018;Paaso et al., 2014;Popović et al., 2014). They can be used to solve load models having the time-varying function by converting loads in the form of discrete probabilistic to deterministic generating load model as reported by (El-saadany & Atwa, 2010). Despite the benefits of MINLP in solving the DG allocation and sizing problem, there exist some drawbacks of having a large number of decision variables and longtime computation.

E. Dynamic programming
Dynamic programming was first introduced in the 1940 s by Richard bellman aiming in solving the problem in which the optimum decision variable is sequential in nature. The dynamic word represents a time-varying function problem applied in mathematical optimization to solve problems in the form of optimal decisions. In (Khalesi et al., 2011), proposes a multi-objective function-based dynamic programming to determine a feasible location to place our DG within the distribution network to minimize power loss and improvement in the reliability of the system. Also, Chen et al. (2017) develops an optimization strategy based on dynamic programming to address energy management problems, as they suffer recently due to the expansion of a number of variables. The objective was to design a new optimization strategy for energy management in combined heat and power (CHP) with hybrid energy storage and delivered based on the Dynamic programming approach. The DG allocation problem can be formulated as follows:

F. Meta-heuristic approach
A meta-heuristic approach is an iterative approach that guides a subornative heuristic, combine artificially intelligent and mathematical optimization for solving a complex problem vastly than the classical method, generally used to find an approximate solution where the classical have failed. However, meta-heuristic approach can be summarized in Figure 4. Several meta-heuristic approaches for optimal allocation and planning of DG have been studied by various authors in the distribution network system. Such as, artificial bee colony ( (Lakum & Mahajan, 2019;Yahiaoui et al., 2017), Firefly algorithm (Fister et al., 2013;Othman et al., 2016). Equation (2) is applied for solving classical optimization and is applied to the position of vectors of PSO.
where X id is the dimension of the vector particle; v max d is the dth dimension of the maximum vector.
K is the random variable selection in [0,1] and γ is a constant value in the range of [0,1].

G. Simulated annealing
Simulated annealing process begins in 1983 by Galett, vecchi and since its inception to date; it has reported being used widely due to its simplicity. It is in the form of probability function which allows new solutions to pass or reject, in order to avoid been trapped in a non-optimum position. Various literature has captured the use of simulated annealing for details ("A Solution to the Optimal Power -Ow Using Simulated Annealing" 2003; Friesz et al., 2008;Popović et al., 2014) Kamal EL-Sayed (2017) has applied simulated annealing optimization for minimization of power losses and enhancing the voltage profile. In ref (Sutthibun & Bhasaputra, n.d.), a multi-objective optimal placement of DG based on simulated Annealing was performed in order to reduce power system losses, control emission and contingency. The model was tested on IEEE 30 bus and there was a total power reduction up to 43% compared with the system without DG. Multi-objective optimal planning of DG using simulated annealing was proposed in (Dharageshwari & Nayanatara, 2015), to reduce total power loss and improvement in voltage profile. Suitable fast convergence was attained by the placement of multiple DG and tested on the IEEE 33 bus and efficiency of the system improved, reduction in system losses and cost reduction.

H. Tabu search (TS)
Tabu search is a global Meta-heuristic optimization for controlling embedded systems. The search was proposed in 1986 by Glover and McMillian, based on human memory efficiency to solve the optimization problem. The basic principle of this search is on adaptive memory exportation that allows an anomic way of searching a solution until a better result is obtained (Hilal, 2017;McMillan & Glover, 1986). A genetic-based on tabu search was proposed by (Sutthibun & Bhasaputra, n.d.), for optimal placement and allocation of DG in the distribution network. The objective of the algorithm is to minimize power frequency losses and harmonic power losses in a distribution network system, and the proposed methodology was applied to the IEEE-14 and 34-node test systems. Thus, the proposed methodology is efficient for solving the DG allocation problem in the distribution system.

Shuffled frog leaping (SFLA)
SFLA is used based on the behavior of the frogs in rotation to search for food (Eusuff et al., 2006). In (Afzalan et al., 2012), optimal placement of the DG in radial distribution was achieved by SFLA and the result obtained was tested on 33 radial bus systems. In ref (Moazzami, 2017), a proposed methodology based on an improved shuffled frog leaping algorithm (MSFLA) to achieve optimal placement and size of DG and distributed statistic-synchronized compensator (D-STATCOM). The result was tested on the IEEE-33 bus and compared with the ones derived from the genetic algorithm and shows to be more effective than the existing one.

J. Artificial bee colony (ABC)
Artificial bee colony (ABC) algorithm is an optimization technique that simulates the foraging behavior of honey bees, and have been applied successfully too many problems. ABC is classified under swarm intelligence algorithms and develop in 2005 by karaboga. Ref (C. K. Choton K Das et al., 2018), proposed an algorithm based on ABC for optimal placement of distributed energy storage system in the distribution system to minimize power losses, line loading, mitigation of network demand and general improvement of the voltage profile. The result obtained using the ABC algorithm was compared with particle swarm optimization (PSO) and found to perform better and faster convergence. ABC algorithm was proposed in (Abu-mouti & Member, 2011), for optimal sizing of DG, power factor and location to minimize total real power loss in the distribution system and was compared with the PSO approach and found ABC offered a better quality solution with the fastest convergence.

K. Other promising heuristic optimization algorithms
However, other authors proposed a heuristic algorithm which is a powerful tool for optimal allocation, sizing, and placement of distributed generation problems in the distribution system.
• Biogeography-Based Optimization (BBO) is an evolutionary algorithm that optimizes a function by stochastically iterative function. It describes a number of behaviors linked to that of fish, birds, and insects. BBO was introduced by Dan Simon launched BBO in 2008. An optimal location based on BBO for sizing of solar photovoltaic DG in the radial distribution system was proposed by (Duong et al., 2019), by minimizing power loss and maintain a normal voltage, while controlling the effect of harmonics distortion not to exceed the limit. Results obtained were compared with genetic algorithm and particle swarm optimization and artificial bee colony it shows that the new algorithm is faster and less time to converge.
• Bacterial Foraging Optimization Algorithm (BFOA) The algorithm is stimulated by foraging properties of E. coli bacteria. Permitting to this, bacteria search the food in such a manner to maximize the obtained energy per unit time. The isolated bacterium also convey to others by transporting a signals. In this process, bacterium take decision for searching food after examine two preceding factors in this, the bacterium moves by taking small steps at the time of searching the nutrients known as chemotaxis (Prabha et al., 2015).
• Invasive Weed Optimization Algorithm (IWO) This algorithm was first presented by Mehrabian and Lucas in 2006. It is based on mathematical stochastic optimization algorithm. The technique is inspired by sensation of inhabitancy of invasive weeds in nature is based on weed biology and ecology Invading of weeds of cropping system is done by means of dispersal. Every invading weed takes the unused resources in the field and matures to the flowering weed and yields new weed autonomously (Prabha et al., 2015).
• Imperialist Competitive Algorithm (ICA) This is one of the latest meta-heuristic algorithms suggested to address socio-politically inspired mathematically optimization issues. Ref (Eisapour-moarref & Amir, 2017), ICA was proposed for multi-objective location and sizing DSTATCOM in the distribution system considering uncertainty in loads. The proposed algorithm tested on IEEE 33 and 69 bus system and there was total loss reduction, voltage profile improvement, feeder load balancing and cost reduction in the distribution system.
• Lightning Search Algorithm (LSA) is a new effective meta-heuristic optimization method for solving real numerical optimization concepts. LSA is inspired by the natural phenomenon of lightning and based on the mechanism of step ladder propagation. Ref (Thangaraj & Kuppan, 2017), proposed a multi-objective simultaneous placement of DG and DSTATCOM based on LSA for minimizing total power loss and voltage deviation, tested on IEEE 33 and 69 bus system and there was a significant improvement when the new algorithm was tested.
• Firefly Algorithm (FA) is a meta-heuristic algorithm that is inspired based on the flashing behavior of fireflies. The firefly's flash acts as a signal system to seduce other fireflies (Nadhir, 2013). In (Taylor et al., n.d.), optimal planning of distributed generation in the distribution system was achieved through the FA algorithm with the objective of minimizing power loss and voltage control.
• Dragonfly Algorithm (DA) is a new meta-heuristic optimization, which is based on simulating the swarming behavior of dragonfly and was developed by mirjalili in 2016. In (Suresh & Belwin, 2018), DA was used for optimal placement of DG unit size at different power factor in order to reduce power loss in the distribution system and enhancement of the voltage profile of the system.

Overall review of studies on optimization of the distribution system
Considering and reviewing the study work on the issue of optimizing the distribution system, the following threads are described as deficiencies and directions for future works. Additionally, Table 6 includes merit and de-merit of different research work discussed.
• The majority of the research work carried out the optimization of distribution networks based on renewable energy sources did not consider uncertainty only a few. However, taken these uncertainties into consideration gives the distribution network system a strong realistic solution that paves a way to a practical distribution network.
• Meta-heuristic optimization algorithms control parameters are not well fit and this has a significant impact on their computing efficiency. For the optimization problem on hand, there is a need to revise in literature in order to have a strong parameter for the meta-heuristic problem.
• In distribution system nowadays used plug-in vehicles as a new concept in the distribution network configuration. They are large loads that can have a drastic effect on the effectiveness of the distribution system. However, some studies were conducted to assess their impact on the problems of optimizing the distribution system. There is still a gap for a full assessment of the effects of the plug-in car scheme used in the distribution network system.
• Many literature studies focused on Lagrangian relaxation (LR), but the solution obtained in each iteration is not really feasible. Therefore, there is a need for more research on the LR parameters because the performance of the distributed algorithm depends on the parameters to achieve faster convergence.
• The impacts of communication inertia on the convergence performance of the algorithm have not been fully explored in literature studies. The problem of optimal resistance existing in communication infrastructure for designing strong communication needs to be responded to in order to have wide adoption in distributed optimization in the future.
• Generally, distribution network system is large-scale integration, most of the existing optimization techniques study focuses on small-scale distribution. For fitness and validation, purpose optimization techniques should be considered.
• Regardless of the significance offered by distribution optimization planning, most of the utilities present in advanced countries experience challenges in integration. Educating modern distribution system utilities with merits and de-merits will lead to a more practical system.
• However, newer heuristic techniques such as Bacterial Foraging Optimization Algorithm (BFOA), Shuffled Frog Leap Algorithm (SLFA), Invasive Weed Optimization Algorithm (IWO) and Simulated Annealing (SA) Algorithm may appear to be promising in the future. Simulated Annealing -statistical guarantee in finding an optimal solution -They often give a good solution -relatively easy for handling complex problems -Relatively slow if the cost function is expensive to compute -The method sometimes can tell if it has to find an optimal solution (Aly et al., 2010;Vallem et al., 2006;Nahman & Perić, 2008) Particle swarm optimization (PSO) -Higher probability efficiency in obtaining the optimal solution -Less computational time -Few parameters adjustments -Convergence is faster here -However, sometimes difficult to define initial parameters -Difficulty in convergence complex algorithm (Guadix et al., 2018;Nazari-heris et al., 2018) Tabu search (TS) -capable of handling complex iteration problems -Applications are found in continuous and discrete variables -High computational time due to many iterations. (Maciel & Padilha-Feltrin, 2009;Nara et al., 2002) Ant colony search optimization (ACSO) -capable of giving a rapid solution -convergence is guaranteed -Difficulty in theoretical analysis. -Probability distribution changes due to iteration. (Sheidaei et al., 2008) Artificial bee colony optimization (ABCO) -simplicity, flexibility and robustness. -easy implementation.
-easy to define the objective function.
-Effectively slow in sequential processing. -A higher number of objective function evaluations required for fitness.

Conclusions and future scope work
In this research work, the existing work carried out on optimization of the distribution network system has been reviewed from its point of view, in terms of the optimization algorithm, objectives, decision variables employ and models applied.
The sequence of this research has regarded over the past 10 years an extensive evaluation of mathematical optimization in the power energy system and implementation of these techniques in planning and operating issues of DG's. Additional to the various trends, it has been noted that the area of artificial intelligence is still the best technique evolving mathematical optimization of distributed energy sources (DER). However, the following conclusion is drawn from this article as follows: • The area of artificial intelligence techniques is still receiving attention than conventional optimization techniques for optimal DG planning in a power distribution network system from a different point of view.
• Computational techniques in hybrid optimization techniques are faster and convergence faster than the conventional optimization techniques employed in optimal planning Distributed generation in a distribution network system.
• An analytical approach is not computationally difficult for a simple system but not suitable for a system with a large and complex network system.
• Meta-heuristic and hybrid optimization techniques are observed to be more suitable for a large and complex system as they provide optimum solutions to both single and multi-objective problems.
• It is further observed that for optimal distributed generation and sizing based on metaheuristic optimization techniques such as: ABC, BBO, SFLA, ICA, and LSA are performing significantly well and may seem to be reassuring in the future.
The following recommendation for future scope area of research is as follows based on the literature survey.
• The stochastic study should be adopted for optimal planning of the distribution network with the sporadic nature of distributed generation should be considered.
• Assessment of different types of DG and flexible ac transmissions (FACTS) controller planning in a static and realistic model by artificial intelligent (AI) techniques should be considered.
• A comparison of various different types of DG and FACTS planning via static and real-time models by the concept of hybrid AI should be considered in further research.
• Renewable energy sources based on DG units with battery storage option and their importance were not considered in this research.
• Operations of the DG in standalone mode should be considered in the future research scope.

Merits Demerits References
Cuckoo search -Requires fewer parameters for setting up.
-slow in convergence and few solutions in literature.