基于蚂蚁算法的配电网网络规划

利用蚂蚁算法进行配电网网架结构规划。为了使优化过程同时考虑到网损最小和投资最小两个因素，提出将各个负荷点作为食物。食物给各条街道赋予一定的“味道”，“味道”的求取公式和网损的计算公式类似。“味道”在一定程度上反映了网损的情况。蚂蚁在计算的开始阶段将根据“味道”决定自己选择街道的概率，在计算的中间阶段将根据信息素决定自己选择街道的概率。实例计算表明该方法可行、有效，可以方便地求得配电网网络规划问题的最优或近似最优解。

IMPROVED ANT ALGORITHM FOR DISTRIBUTION NETWORK PLANNING

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.