DocumentCode :
1347370
Title :
Optimum Configurations of an n-Section System
Author :
Gupta, Sushil Das ; AI-Musawi, Muhammad Jawad
Author_Institution :
Jadavpur University, Calcutta; 40B, Southern Avenue; Calcutta-700 029 INDIA.
Issue :
2
fYear :
1981
fDate :
6/1/1981 12:00:00 AM
Firstpage :
185
Lastpage :
188
Abstract :
Reliability optimisation methods are used for computer-aided design of power-cable system such as in RTS (Rapid Transit System) on the basis of minimizing inconvenience hours per year (IHY), the users viewpoint, with respect to a set of admissible configurations. For reasons of simpler inventory the RTS often constructs its power lines using but one cable size. Each section uses as many line lengths (in parallel) as needed to handle its current. The problem then is to compute the IHY and the number of parallel cables in each section of the line. For an n section system and a alternatives per section, there will be ¿n configurations and the volume of computation will be proportional to (4¿n). An algorithm reduces this number of configurations and determines the number of reasonable alternatives in each section. In practice, apart from a mathematically tractable cost function, it is necessary to consider nonmathematical constraints. It is then convenient to list the optimal and the near optimal solutions, with respect to some cost function, e.g., total cable length used. Economic considerations dictate that reliability should increase following a cost increment. However, for the set of say 20 optimal and near optimal IHY values, the reliability does not always increase with increasing cost. An heuristic 2-stage search algorithm determines the set of optimal and near optimal configurations where reliability monotonically increases with cost.
Keywords :
Availability; Cables; Cost function; Design automation; Design optimization; Optimization methods; Power system reliability; Programming; Reliability engineering; Statistics; Computer-aided design; Optimization;
fLanguage :
English
Journal_Title :
Reliability, IEEE Transactions on
Publisher :
ieee
ISSN :
0018-9529
Type :
jour
DOI :
10.1109/TR.1981.5221028
Filename :
5221028
Link To Document :
بازگشت