Page 19 - IJEEE-2023-Vol19-ISSUE-1

15 | Hadaeghi & Abdollahi
R = F$% (11)
?'&() F&
Where fBH is the fitness value of the black hole and fi is the
fitness value of the star i.
6) In this paper, the criterion of maximum number of
iterations is considered as the criterion of stopping.
C. Hybrid TLBO-BH Algorithm
In this paper, the two mentioned algorithms are used in two
separate steps to find the appropriate answer. Since the BH
algorithm is more accurate than the TLBO algorithm and the
TLBO algorithm is faster than the BH algorithm, the search
space is first reduced using the TLBO and then the BH algorithm
is used to obtain a more accurate answer. In fact, first the TLBO
algorithm is executed and completed, then the answers obtained
by the TLBO algorithm are used as the inputs of the BH
algorithm.
The implementation process of TLBO algorithm is as
follows:
First, the input data (e.g., load data, line data, DGs data, etc.) determined using hybrid TLBO-BA algorithm. It can be seen
is imported. Then the initial population is generated. In the next
step, based on the network configuration, power flow is
performed and the corresponding objective function is
calculated. The population is then updated and again based on
the network configuration, power flow is carried out and the
corresponding objective function is evaluated. In this step, if the
convergence criterion is satisfied, the configuration with the
minimum objective function is selected as the final solution,
otherwise the population is updated and the process continues
until the convergence criterion is met.
IV. SIMULATION AND RESULTS
To illustrate the performance of the proposed method, it is
tested on a standard 33-bus radial distribution system according
to “Fig. 1”. The data of this 33-bus network is available in
reference [19]. This network has a nominal voltage of 12.66 kV
and the active and reactive loads installed in this network is
equal to 3715 kW and 2300 kVar, respectively. The total active
power loss is equal to 202.6 kW. The system has 37 branches, 32
sectionalizing switches and 5 tie switches. The switches 37, 36,
35, 34 and are open before the reconfiguration of the system.
In this paper numerical results are calculated in four different
cases. The first case is related to normal network conditions, the
second case is when network reconfiguration is applied
separately, in the third case, DG placement and sizing is
considered separately, and in the fourth case, DG placement and
sizing and network reconfiguration are considered
simultaneously.
A. Case I (Normal Network Conditions)
Table I, shows the results of applying the proposed method
in all cases. As mentioned before, in the first case, network
reconfiguration and DG placement are not considered. In this
case, the network is in the normal condition and the switches 37,
36, 35, 34 and 33 are open. The network loss in this case is 202.6
kW.
B. Case II (Network Reconfiguration Using TLBO-BA
Algorithm)
According to Table I, it can be seen that after network
reconfiguration, the opened switches are 37, 32, 14, 9 and 7 and
the active power losses reduced from 202.6 kW to 124.8 kW
(38.4% reduction) which indicates the effective and useful role
of reconfiguration. The convergence curve of the hybrid TLBO-
BH algorithm for this case can be seen in “Fig. 2”. It can be seen
that the hybrid TLBO-BH algorithm converges to the optimal
global solution after 10 iterations. “Fig. 3” shows the results of
applying the proposed method in this case on network voltage
profile. As shown in “Fig. 3”, the network voltage profile has
significantly improved after reconfiguration. For instance, in bus
18, the voltage was 0.913 P.U. before the reconfiguration, while
after reconfiguration it reaches 0.947 P.U. That is nearly 3.7%
increase in bus voltage 18.
C. Case III (Optimal DG Placement and Sizing Using hybrid
TLBO-BA Algorithm)
In this case, optimal placement and sizing of DG is
from the simulation that bus No. 6 is the optimal location for DG
placement and its optimal capacity is 2575 kW. According to
Table I, after optimal DG placement and sizing, the opened
switches are 11, 14, 17, 33 and 37 and the active power losses
reduced from 202.6 kW to 103.97 kW. The convergence curve
of the TLBO-BH algorithm and voltage profile for case III can
be seen in “Fig. 4” and “Fig.5” respectively. According to “Fig.
4”, after 8 iterations, the hybrid TLBO-BH algorithm finds the
best global solution. Comparing the results obtained in “Fig. 5”,
it is clear that the system voltage profile in the presence of DG
has been effectively improved.
D. Case IV (Network Reconfiguration and DG Placement and
Sizing Using hybrid TLBO-BA Algorithm)
In case IV, network reconfiguration and optimal DG
placement and sizing are considered simultaneously. According
to Table I, in this case the opened switches are 9, 14, 16, 25 and
32 and the active power losses reduced to 79.67 kW which
shows a 61% decrease compared to the first case. In this case,
the bus No. 29 is the optimal location for DG placement and its
optimal capacity is 1925 kW.
Figure 6 shows the convergence curve of the hybrid TLBO-
BH algorithm for case IV. It can be seen that the hybrid TLBO-
BH algorithm converges to the optimal global solution after 11
iterations. The result of applying the proposed method in this
case on network voltage profile is shown in “Fig. 7”. According
to “Fig. 7”, significant improvement of the network voltage
profile is clearly visible after network reconfiguration and DG
placement simultaneously. For instance, the voltage of bus 18
reaches from 0.913 P.U. to 0.967 P.U., which means an increase
of nearly 5.91% in bus 18 voltage.