00000ctm a22000004a 4500 UP-99796217610395987 Buklod 20120706085730.0 m |o d | ta 120706s2007 xx d r |||| u| (iLib)UPD-00187104160 DML eng DMLUC LG 995 2007 A64 D45 De Jesus, Ma. Rosario P. A genetic algorithm solution to a large-scale power distribution network expansion planning problem by Ma. Rosario P. De Jesus. 2007. xi, 125 leaves ill. Thesis (M.S. Applied Mathematics)--University of the Philippines Diliman. This paper presents a simple yet efficient approach on how to implement genetic algorithm in solving a large-scale power distribution network expansion planning problem. The problem is to determine the siting and sizing of the substations and feeders to be built and upgrading of the sizes of the substations and feeders that are already existing or operational. Given the forecasted power demands at different locations, the goal is to find the substation and feeder layout or design with the least overall planning cost (comprising of investment and operational costs). Spanning trees were used to represent power distribution networks so that the problem constraint that the network must have no loops can be easily handled. Two crossover techniques for the genetic algorithm were implemented and compared in terms of the optimal objective function value obtained and the solution time. Another technique which is linearizing the feeder cost function was implemented to reduce solution time significantly. Genetic Algorithms. Electric power distribution. Electric networks. FI UP Thesis UPD DARCHIVES LG 995 2007 A64 D45 Thesis