IMPROVED ANT ALGORITHM FOR DISTRIBUTION NETWORK PLANNING

高炜欣,罗先觉

DOI: https://doi.org/10.3321/j.issn:0258-8013.2004.09.019

2004-01-01

Abstract:This paper introduces ant algorithm for distribution network planning, which optimize power loss and investment cost of feeders simultaneously. The load node is treated as food, which gives each street a smell. The smell calculation is similar to power loss calculation. The smell reflects the power loss. The possibility of ant's selecting a street is directly proportional to the smell of the street at first. After each street has pheromone, the possibility of ant's selecting a street is directly proportional to the pheromone of the street. A real example shows the effectiveness of the algorithm. And it is easier to get an optimal or nearly optimal distribution network by using the presented algorithm.