DocumentCode :
424242
Title :
A parallel evolutionary programming based optimal power flow algorithm and its implementation
Author :
Lo, C.H. ; Chung, C.Y. ; Nguyen, D.M. ; Wong, K.P.
Author_Institution :
Dept. of Electr. Eng., Hong Kong Polytech. Univ., China
Volume :
4
fYear :
2004
fDate :
26-29 Aug. 2004
Firstpage :
2543
Abstract :
This paper develops a parallel evolutionary programming based optimal power flow solution algorithm. The proposed approach is less sensitive to the choice of starting points and types of generator cost curves. To improve the robustness and speed of convergence of the algorithm, population and gradient acceleration techniques are incorporated. The developed algorithm is implemented on a thirty-six-processor Beowulf cluster. The proposed approach has been tested on the IEEE 118-bus system under master-slave, dual-direction ring and 2D-mesh topologies. Computational speedup and generation costs for each parallel topology with different number of processors are then compared to those of the sequential EP approach.
Keywords :
convergence; electricity supply industry; evolutionary computation; load flow; optimisation; parallel programming; workstation clusters; 2D-mesh topology; IEEE 118-bus system; gradient acceleration technique; optimal power flow algorithm; parallel evolutionary programming; parallel topology; thirty-six-processor Beowulf cluster; Acceleration; Clustering algorithms; Costs; Genetic programming; Load flow; Master-slave; Parallel programming; Robustness; System testing; Topology;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Machine Learning and Cybernetics, 2004. Proceedings of 2004 International Conference on
Print_ISBN :
0-7803-8403-2
Type :
conf
DOI :
10.1109/ICMLC.2004.1382232
Filename :
1382232
Link To Document :
بازگشت