A novel distribution system reconfiguration for loss minimization using symbiotic organism search algorithm

Abstract

This paper presents a new evolutionary approach for reconfiguration of radial systems. The framework applied for optimization is Symbiotic Organism Search Algorithm (SOSA). The algorithm is impressed by the interactive behavior opted by the living organisms for surviving and to propagate in the ecosystem. This concept aims for optimal survivability in the ecosystem involving the harm and benefits received from other organisms. The aim is to find optimal reconfiguration and to reduce the real power loss in the distribution side. This approach is examined on 16-bus and 33-bus systems. The results show a significant reduction of real power loss. The time required for execution is less when compared to other approaches. Based on the results calculated with distribution load flow algorithm the SOSA gives better results in terms of real power loss reduction and it is best suitable for digital automation systems.

Keywords

Symbiotic organism search algorithm reconfiguration power loss reduction

1 Introduction

The growth of demand for electricity is becoming a challenge for the society. This needs an optimization in the production, distribution and the utilization of electricity [1]. The installation of new distribution power network is very difficult and not economical. So it becomes necessary to optimize the power economically for consumer satisfaction [2]. Several methods are there to optimize the distribution system, the main technique is reconfiguration. It is defined as adjusting the structure of the feeder with the two commonly used switches in the distribution system [3]. The sectional switches (closed) which is connected on the same feeder while the tie switches (open) which are connected between two feeders are used to change the framework of the feeder [4]. This reconfiguration reduces the real power loss in a significant way. For the reconfiguration to be carried out, many different algorithms have been used till now for optimization of distribution system [5]. The reconfiguration is done without disturbing the radial structure of the feeder [6, 7].

In the year 1975 Merlin and Back insisted on the reconfiguration concept for minimization of real power loss [8]. The optimal algorithm is an approach which involves less number of steps and less complexity [9, 11]. The real life application demands simultaneous optimization for the complicated multiple objective functions [12]. Therefore solution provided by a specific multi objective approaches may suit the applications but cannot be treated as a universal problem solution [13, 26]. Thus there is always a requirement before the researchers to come up with new algorithms which can handle the challenging task for the real life problems [14, 25]. It is important to notice that the naturally inspired multi objective approaches are generally unconstrained in nature and it also require additional techniques to solve constrained problems [24, 32].

This paper proposes a method for minimization of losses using symbiotic organism search algorithm (SOSA) for optimization of the power distribution system. This algorithm is based on the relationship between several organisms which live together and survives in the ecosystem [15, 18]. The goodness about SOSA is the algorithm is parameter free. It does not require any adjustment in any of the parameter and the only step is to initialize the population and number of generation [16 , 23]. This method is tested with two standard systems 16 and 33 buses. The power loss results, comparison tables and convergence graph are displayed in chapter 5. Load flow values are calculated using modified Newton Raphson method and these values are substituted in SOSA algorithm for 16 and 33 bust test systems and the reduced real power loss values have been calculated with radial new reconfiguration.

The paper is organized as follows. Section 2 begins with the problem identification and related works on constrained multi-objective optimization [19 –21].

The proposed multi-objective SOSA are outlined in Section 3. Implementation of SOSA is outlined in Section 4 with detailed steps the simulation environment, performance evaluation methods and standard algorithms used for comparison are described in Section 5. The applications of the proposed algorithm to structural truss design problem robustness are given in Section 6. The concluding remarks are narrated in Section 7.

2 Problem identification

The net power loss is expressed as $Minimize f = min (PT, Loss)$ (1) $P_{T}, Loss = \sum_{k = 1}^{n} P_{T}, Loss (k, k + 1)$ (2) $MinP = \sum_{i = 1}^{nbr} Ri \frac{Pi \land 2 + Qi \land 2}{Vi}$ (3)

Where,

PT, Loss = Total power loss

Nbr = Number of branches

Ri = Resistance of branch i

Pi = Real power flow through branch i

Qi = Reactive power through branch i

Vi = Voltage magnitude at bus i

n = total number of line switches

K = kth number of bus

Subjected to the following constraints

The voltages limits of the bus should not cross the permissible limits

Vmin≤Vi≤Vmax

where Vmin and Vmax are the lower and higher potentials.

The bus current must lie within the rated values.

Radiality of the network is secured always.

All feeders need to be served.

3 Symbiotic organism search algorithm (SOSA)

In the year 2014, Cheng and Prayogo developed this algorithm. This algorithm was impressed by interaction of natural organisms living together in an ecosystem [27]. Organisms rarely lives in isolation since it depends mostly on other organisms for their living. So, it develops symbiotic relationship for adaptation to the changes in their environment [28].

The three symbiotic relationships are

Mutualism - relationship among two different species which benefit mutually from that relation. The interaction between two different species provides advantage to both of them. the relationship between flower and pollinator is mutualism [27].

Commensalism - relationship among two different species in which one benefits and the other is unaffected or neutral. The relationship between Remora fish and Shark is commensalism. The small Remora fish attach itself behind the huge body of Shark and consumes its leftover food thus get benefited from the interaction. The activity of huge Shark does not get affected by the activities of the small fish [28].

Parasitism - relationship among two different species in which one benefits and the other is actively harmed. The interaction between the mosquito and human is a phenomenon of parasitism. Once the mosquito bite the human, the process create parasite in the human body. The germs present in the human body reproduce themselves and the human usually gets affected by the disease and possibly die. But if the human body has better immunity power then they can protect and the parasite gets eliminated from the body [29].

The above process is shown in Fig. 3.1.

Fig.3.1

Symbiotic process in Eco- System.

Fig.3.2

General flowchart for SOSA.

The steps followed in this algorithm are:

Step 1. Ecosystem is initialized.

Step 2. Mutualism- Random organism of different species are selected Xi and Xj. The mutual vector is determined by using Equation 3 and benefit factor which lies between 0 to 1. Calculating the fitness function for each new organism and compare them.

$\begin{matrix} X_{inew} & = & X_{i} + rand (0, 1) * (x_{best} \\ - {Mutual}_{vector} * {BF}_{1}) \end{matrix}$ (4)

$\begin{matrix} X_{inew} & = & X_{j} + rand (0, 1) * (x_{best} \\ - {Mutual}_{vector} * {BF}_{1}) \end{matrix}$ (5) $(X_{i} + X_{j}) / 2$ (6)

Step 3. Commensalism - The new candidate solution of X_i is computed. X_i is updated only if its new fitness is better. $X_{inew} = X_{i} + rand (- 1, 1) * (X_{best} - X_{j})$ (7)

Step 4. Parasitism – Xi and Xj of different species are selected in which Xi is a parasitic species and Xj is the host for parasitic species. If Parasite species has a better fitness value, it kills organism Xj and assume its position in the ecosystem. If the fitness value of Xj is better, then Xj survives in the ecosystem.

Step 5. Updating the best organism.

3.1 Implementation of ICBOA fore configuration problem

Step 1. Initializing the parameters used in optimization process (Bus data, Line data, Switching status, number of iterations, maximum number of iterations, etc,).

Step 2. Using load flow analysis post fault voltage values for each switching configuration is obtained.

Step 3. Initializing the organisms in the ecosystem which contains voltage values obtained from the above step.

Step 4. Mutualism phase is executed. Randomly selecting one organism Xj and finding the mutual vector using equation 6 and benefit factor which lies between 0 and 1. Modification is done based on their mutual relation using Equations 4 and 5.

Step 5. Calculating the fitness function for the modified organisms.

Step 6. Updating the best fitness function.

Step 7. Termination condition is checked.

Step 8. Commensalism Phase is executed. Randomly selecting one organism Xj. Modification of Xj in assistance with Xi is carried out using eqn. 6.

Step 9. Calculating the fitness function for the modified organisms.

Step 10. Updating the best fitness function.

Step 11. Termination condition is checked.

Step 12. Parasitism Phase is executed. Randomly selecting one organism Xj. Create a vector of Parasite by duplicating Xi and modification is done for randomly selected dimensions with uniform random numbers.

Step 13. Calculating the fitness function for the modified organisms.

Step 14. Updating the best fitness function.

Step 15. Termination condition is checked else repeat from step3.

Step 16. Global best optimal solution (minimum power loss) is displayed.

Step 17. Stop the process.

Radiality and voltage constraints are incorporated in this both the algorithms to avoid over loading, voltage reversal problems. This will also maintain the stability of a system.

4 Results and discussion

The Symbiotic Optimization Search (SOS) is an evolutionary approach where the organisms of ecosystem are considered as elements of population (evolutionary strategies) like genes of chromosomes in genetic algorithm.

The real power loss values with different configurations have been calculated for 16 bus and 33 bus standard systems. The 16 and 33 bus standard systems are selected for testing the SOSA method. It is shown in Figs. 4.1 and 4.4

Fig.4.1

16-bus diagram.

Table 4.2 gives the system data for a 16 bus distribution system given in Fig. 4.1.

From the above Table 4.1 it is noticed that the value of loss is reduced to 421.33kW when the switches 9, 14, 16 are in open condition. The computing time is 1.722 sec which is less. Before the reconfiguration the base loss is found to be 511.41kW and after reconfiguration it is lowered to 421.33kW. Figure 4.2 depicts the comparison chart for power loss of SOSA with other relevant approaches. It is observed that the suggested SOSA gives less power loss value for 16 bus when compared to the methods such as RGA, BFOA, Bee colony, VHDE and HDE. Figure 4.3 depicts the plot for convergence using the proposed SOSA method.

Fig.4.2

Comparison chart of power loss for 16-bus.

Fig.4.3

Plot for convergence for 16-bus using SOSA.

Fig.4.4

33-Bus diagram.

Table 4.1

Results for SOSA - 16 bus

Tie (OFF)	Loss Value	Calculated
switches	(KW)	time (sec)
9, 14, 16	421.33	1.722
5, 7, 16	424.85	1.813
7, 9, 16	426.43	1.847
7, 11, 15	429.62	1.862
5, 11, 16	432.41	1.883

Table 4.2

16 Bus Network data

From bus	To bus	R	X	P kW	Q kW
1	4	0.075	0.1	2	1.6
4	5	0.08	0.11	3	1.5
4	6	0.09	0.18	2	0.8
6	7	0.04	0.04	1.5	1.2
2	8	0.11	0.11	4	2.7
8	9	0.08	0.11	5	3
8	10	0.11	0.11	1	0.9
9	11	0.11	0.11	0.6	0.1
9	12	0.08	0.11	4.5	2
3	13	0.09	0.11	1	0.9
13	14	0.09	0.12	1	0.7
13	15	0.08	0.11	1	0.9
15	16	0.04	0.04	2.1	1
5	11	0	0	–	–
10	14	0	0	–	–
7	16	0	0	–	–

Table 4.3 shows the loss results for SOSA for 33 bus standard system. The value before reconfiguration is known as base configuration and loss value was found to be 202.71kW and it is reduced to 132.81kW after reconfiguration using SOSA when 7, 9, 13, 23, 36 switches are at open condition. The computational time was found to be 2.194 sec which is very low. Figure 4.5 depicts the comparison for power loss of SOSA with other relevant approaches. It is observed that the proposed SOSA gives less power loss value for 33 bus when compared to the approaches such as RGA, BFOA, ACSA, HSA, Bee colony, AG and Heuristics. Figure 4.6 depicts the graph of convergence for the proposed method.

Fig.4.5

Comparison chart of power loss for 33-bus.

Fig.4.6

Plot for convergence for 33-bus using SOSA.

Fig.4.7

69-Bus diagram.

Table 4.3

Results for SOSA - 33 bus

Tie (OFF)	Loss Value	Calculated
switches	(KW)	time (sec)
7, 9, 13, 23, 36	132.81	2.194
9, 15, 20, 28, 31	136.65	2.212
7, 9, 13, 24, 29	139.51	2.322
13, 28, 31, 33, 35	143.73	2.469
12, 21, 35, 36, 37	151.42	2.511
7, 14, 23, 29, 35	156.23	2.562

Table 4.4 shows the loss results for SOSA for 69 bus standard system. The value before reconfiguration is known as base configuration and loss value was found to be 225kW and it is reduced to 98.32kW after reconfiguration using SOSA when 18, 47, 56, 61, 69 switches are at open condition. The computational time was found to be 2.578 sec which is very low. Figure 4.8 depicts the comparison for power loss of SOSA with other relevant approaches. It is observed that the proposed SOSA gives less power loss value for 69 bus when compared to the approaches such as MBFOA and HSA. Figure 4.9 depicts the graph of convergence for the proposed method.

Fig.4.8

Comparison chart of power loss for 69-bus.

Fig.4.9

Plot for convergence for 69-bus using SOSA.

Table 4.4

Results for SOSA - 69 bus

Tie (OFF)	Loss Value	Calculated
switches	(KW)	time (sec)
18, 47, 56, 61, 69	98.32	2.578
18, 49, 55, 67, 69	118.56	2.654
14, 35, 65, 67, 69	133.72	2.737
11, 14, 32, 55, 62	146.49	2.824
12, 47, 39, 55, 69	153.58	2.858
19, 47, 55, 67, 63	136.31	2.767

5 Robustness of SOSA

To check the robustness of SOSA it has been tested with 16 and 33 test systems by taking 50 iterations and some random initializations. A software tool have been used in MATLAB to execute this algorithm in Intel Core i5, 4200U CPU with 1.60 GHz processor and 8 GB RAM for calculating the distribution power loss. The minimum value of loss is suggested when [9 , 16] switches are OFF for 16 bus and when [7, 9, 13, 23, 36] switches are OFF for 33 bus the loss value is found to be minimum with less calculation time. The loss is decreased by 17% for 16 bus and 34% for 33 bus test system approximately.

6 Conclusion

In this paper a recently developed meta-heuristics Symbiotic Organism Search algorithm (SOSA) has been formulated to solve the multi objective problems. The goodness of SOSA is that it is a parameter free algorithm. Thus the developed multi objective optimization problem is also parameter free. This method is tested using standard 16 and 33 bus systems. The end results led to an optimal reconfiguration which minimizes the true loss. Comparison is also made with other existing algorithms in which SOSA proves the superiority in losses minimization and faster execution time. The real power loss values for 16 bus and 33 bus distribution systems are calculated as 421.33 and 132.81 kW. With reference to the results obtained this method can be implemented for real time larger systems also.Among the nature inspired algorithms thi SOSA algorithm is computationally efficient.

References

Kendrew

T.J.

and Marks

J.A.

, Automated distribution comes of age, IEEE Comput Appl Power (1989), 7–10.

Carreno

E.M.

, Romero

and Padilha-Feltrin

, An efficient codification to solve distribution network reconfiguration for loss reduction problem, IEEE Trans Power Syst23(4) (2008), 1542–51.

Ravibabu

, Ramya

M.V.S.

, Sandeep

, Karthik

M.V.

and Harsha

, Implementation of improved genetic algorithm in distribution system with feeder reconfiguration to minimize real power losses, In: Computer Engineering and Technology (ICCET), 2010 2nd International Conference on4 (2010), V4320–V4-323.

Goswami

S.K.

and Basu

S.K.

, A new algorithm for the reconfiguration of distribution feeders for loss minimization, IEEE Trans Power Delivery7(3) (1992), 1484–1491.

Kashem

M.A.

, Ganapathy

, Jasmon

G.B.

and Buhari

M.I.

, A novel model for loss minimization in distribution networks, In: IEEE International Power Technologies 2000 Conference on Electric Utility Deregulation, Restructuring at City University London, 2000, pp. 251–256.

Jaswanti

and Thakur

, Minimum loss configuration of power distribution system. Power electronics, drives and energy systems, In: PEDES ‘06 International Conference, 2006, pp. 1–6.

Shirmohammadi

and Hong

W.H.

, Reconfiguration of electric distribution networks for resistive line loss reduction, IEEE Trans Power Delivery4(1) (1989), 1492–1498.

Merlin

and Back

, Search for a minimal loss operating spanning tree configuration in an urban power distribution system, Proceedings 5th Power System Computation Conference (PSCC), Cambridge, UK, 1975, p. 118.

Baran

M.E.

and Wu

F.F.

, Network reconfiguration in distribution systems for loss reduction and load balancing, IEEE Trans Power Delivery4(2) (1989), 1401–1407.

10.

Andrés

M.J.

and José

G.A.

, A new heuristic approach for distribution systems loss reduction, Elect Power Syst Res November78(11) (2008), 1953–1958. https://dx-doi-org.web.bisu.edu.cn/10.1016/j.epsr.2008.04.001. ISSN 0378-7796.

11.

Civanlar

, Grainger

, Yin

and Lee

, Distribution feeder reconfiguration for loss reduction, IEEE Trans Power Delivery3(3) (1988), 1217–1223.

12.

Dasgupta

, Das

, Abraham

and Biswas

, Adaptive computational chemotaxis in bacterial foraging optimization: An analysis, IEEE Trans Evol Comput13(4) (2009).

13.

Nara

, Shiose

, Kitagowoa

and Ishihare

, Implementation of genetic algorithm for distribution systems loss minimization by reconfiguration, IEEE Trans Power Syst August7(3) (1992), 1044–1051.

14.

Zhu

J.Z.

, Optimal reconfiguration of electric distribution network using refined genetic algorithm, Electr Power Syst Res62 (2002), 37–42.

15.

Sathish Kumar

and Jayabarathi

, Power system reconfiguration and loss minimization for an distribution systems using bacterial foraging optimization algorithm, Int Jour Electr Power Energy Syst36(1) (2012), 13–17. https://dx-doi-org.web.bisu.edu.cn/10.1016/j.ijepes.2011.10.016. ISSN 0142-0615.

16.

Mohammed Imran

and Kowsalya

, A new power system reconfiguration scheme for power loss minimization and voltage profile enhancement using fireworks algorithm, Int J Elect Power Energy Syst62 (2014), 312–322.

17.

Taher

, Mohsen

, Jamshid

and Azad

F.E.

, A new hybrid evolutionary optimization algorithm for distribution feeder reconfiguration;, 2013.

18.

Mezura-Montes

and Coello-Coello

C.A.

, A simple multimembered evolution strategy to solve constrained optimization problems, IEEE Trans Evol Comput9(1) (2005), 1–17.

19.

Y.-K.

, Lee

C.-Y.

and Liu Shao-Hong Tsai

L.-C.

, Study of reconfiguration for the distribution system with distributed generators, IEEE Trans Power Delivery25(3) (2010).

20.

Gomes

F.V.

, Carneiro

, Pereira

J.L.R.

, Vinagre

M.P.

and Garcia

P.A.N.

, A new heuristic reconfiguration algorithm for large distribution systems, IEEE Trans Power Syst20(3) (2005), 1373–1378.

21.

Srinivasa Rao

, Ravindra

, Satish

and Narasimham

S.V.L.

, Power loss minimization in distribution system using network reconfiguration in the presence of distributed generation, IEEE Trans Power Syst28(1) (2013), 317.

22.

Mcdermott

T.E.

, Drezga

and Broadwater

R.P.

, A heuristic nonlinear constructive method for distribution system reconfiguration, IEEE Trans Power Syst14(2) (1999), 478–483.

23.

Martín

J.A.

and Gil

A.J.

, A new heuristic approach for distribution systems loss reduction, Electr Power Syst Res78(11) (2008), 1953–1958.

24.

Swarnkar

, Gupta

and Niazi

K.R.

, Minimal loss configuration for largescale radial distribution systems using adaptive genetic algorithms, 16th National Power Systems Conference, 2010.

25.

Chiou

J.-P.

, Chang

C.-F.

and Su

C.-T.

, Variable scaling hybrid differential evolution for solving network reconfiguration of distribution systems, Senior Member IEEE Trans Power System20(2) (2005).

26.

C.-T.

, Network reconfiguration of distribution systems using improved mixed-integer hybrid differential evolution, Senior Member, IEEE, and Chu-Sheng Lee IEEE Trans Power Delivery18(3) (2003).

27.

Prasad

and Mukherjee

, A novel symbiotic organisms search algorithm for optimal power flow of power system with FACTS devices, An International Journal of Engg, Science and Tech19 (2015), 79–89.

28.

Panda

and Pani

, A Symbiotic Organisms Search algorithm with adaptive penalty function to solve multi-objective constrained optimization problems, An International Journal of Applied Soft Computing46 (2016), 344–360.

29.

C.-T.

, Chang

C.-F.

and Chiou

J.-P.

, Distribution Network reconfiguration for loss reduction by ant colony search algorithm, Int J Electric Power System Research75 (2005), 190–199. doi: 10.1016/j.epsr.2005.03.002.

30.

Murthy Sivanagaraju

G.V.K.

, Satyanarayana

and Hanumantha Rao

, Artificial Bee Colony Algorithm for distribution system reconfiguration with distributed generation, Int J Eng, Sci, Emerg, Tech3(2), 50–59.

31.

Naveen

, Sathishkumar

and Rajalakshmi

, Distribution system reconfiguration for loss minimization using modified bacterial foraging optimization algorithm, Electrical Power and Energy Systems69 (2015), 90–97.