click here to download - UniKL MIMET Official Website
click here to download - UniKL MIMET Official Website
click here to download - UniKL MIMET Official Website
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
Knowledge based system is a computer system<br />
that is programmed <strong>to</strong> imitate human problem‐<br />
solving by means of artificial intelligence and<br />
reference <strong>to</strong> a database of knowledge on a par‐<br />
ticular subject .Jung et al [8] proposed an artifi‐<br />
cial intelligent based reconfiguration methodol‐<br />
ogy for load balancing in a distribution system.<br />
An expert system was applied <strong>to</strong> the heuristic<br />
search in order <strong>to</strong> reduce the search space and<br />
reduce the computational time for the recon‐<br />
figuration.<br />
Wu et al [9] proposed a Petri net based recon‐<br />
figuration methodology for res<strong>to</strong>ration of the<br />
power system. A <strong>to</strong>ken passing and a backward<br />
search processes are used <strong>to</strong> identify the se‐<br />
quence of res<strong>to</strong>ration actions and their time.<br />
This method can help <strong>to</strong> estimate the time re‐<br />
quired <strong>to</strong> res<strong>to</strong>re a subsystem and obtain a sys‐<br />
tematical method for identification of the se‐<br />
quence of actions. Y.L.Ke [10] proposed a Petri<br />
net base approach for reconfiguring a distribu‐<br />
tion system <strong>to</strong> enhance the performance of the<br />
power system by considering the daily load char‐<br />
acteristics and the variations among cus<strong>to</strong>mers<br />
due <strong>to</strong> the temperature increase in the power<br />
system.<br />
Jiang and Baldick [11] proposed a comprehen‐<br />
sive reconfiguration algorithm for distribution<br />
system reconfiguration. They employed simu‐<br />
lated annealing <strong>to</strong> optimize the switch configura‐<br />
tion of a distribution system. The objective of<br />
the reconfiguration is <strong>to</strong> decrease the power loss<br />
in the distribution system. Ma<strong>to</strong>s and Melo [12]<br />
put forward a simulated annealing based multi<br />
objective reconfiguration for power system for<br />
loss reduction and service res<strong>to</strong>ration. A recon‐<br />
figuration for enhancing the reliability of the<br />
power system was proposed by Brown [13]. A<br />
predictive reliability model is used <strong>to</strong> compute<br />
reliability indices for the distribution system and<br />
a simulated annealing algorithm is used <strong>to</strong> find a<br />
<strong>MIMET</strong> Technical Bulletin Volume 1 (2) 2010<br />
reconfiguration solution.<br />
Shu and Sun [14] proposed a reconfiguration<br />
methodology <strong>to</strong> maintain the load and genera‐<br />
tion balance during the res<strong>to</strong>ration of a power<br />
system. An ant colony optimization algorithm<br />
was used <strong>to</strong> search the proper reconfiguration<br />
sequence based on the Petri net model. Daniel<br />
et al [16] proposed an ant colony based recon‐<br />
figuration for a distribution system. The objec‐<br />
tive of the reconfiguration was <strong>to</strong> reduce the<br />
power loss in the power system.<br />
Salazar et al [16] proposed a feeder reconfigura‐<br />
tion methodology for distribution system <strong>to</strong><br />
minimize the power loss. A reconfiguration algo‐<br />
rithm was proposed based on the artificial neu‐<br />
ral network theory. Clustering techniques <strong>to</strong> de‐<br />
termine the best training set for a single neural<br />
network with generalization ability are also pre‐<br />
sented in that work. Hsu and Huang [17] put for‐<br />
ward another artificial neural network based<br />
reconfiguration for a distribution system. The<br />
reconfiguration can achieve service res<strong>to</strong>ration<br />
by using artificial neural network and pattern<br />
recognition method.<br />
Wang and Zhang [18] proposed a particle swarm<br />
optimization algorithm based reconfiguration<br />
methodology for distribution system. A modified<br />
particle swarm algorithm has been presented <strong>to</strong><br />
solve the complex optimization problem. The<br />
objective of the methodology was <strong>to</strong> minimize<br />
the power loss in the power system. Jin et al [19]<br />
introduced a binary particle swam optimization<br />
based reconfiguration methodology for distribu‐<br />
tion system. The objective of the reconfiguration<br />
was load balancing. The reconfiguration method‐<br />
ology proposed in that work can only be applied<br />
in the power system with radial configuration.<br />
Heo and Lee [20] proposed MAS based intelli‐<br />
gent identification system for power plant con‐<br />
trol and fault diagnosis. The proposed methodol‐<br />
| MARINE FRONTIER @ <strong>UniKL</strong><br />
89