用于供水系统直接优化调度的蚁群改进算法

在城市供水系统中建立了多目标在线直接优化调度模型，并对影响优化调度的各方面因素进行了系统的分析和挑选。使用化多为一的乘除法，将该多目标决策问题转化为单目标问题求解，提出了使用乘法形式的罚函数将模型中的约束函数转化为目标函数。采用蚁群算法求解调度模型。为了更好地得到全局最优解，对算法进行了改进，加入了更多的决策点，实现蚁群算法的二进制编码方法，并采用单只最优蚂蚁更新路径上的外激素值、外激素值限定在一定范围内等改进方法。使用改进算法实现了某小区供水系统的直接优化调度，并与遗传算法优化调度的过程进行了对比，新算法在优化时间及得到最优解的次数上都优于遗传算法。

Ant colony optimization algorithms for direct optimal scheduling of water distribution system

The direct-optimal dispatch model on multi-objective mixed discrete variables for city's water distribution system was established and the factors affecting the optimal dispatch of Water Distribution System(WDS) were systematically analyzed and selected.The Multi-objective decision problem was transformed to the mono-objective problem by the multiply-divide method,and its constraint functions were transformed to the objective functions by a multiplication penalty function. Ant Colony Optimization Algorithms(ACOAs) were appropriate for this optimization problem.To obtain the better total solution,some modifications to ACOAs were adopted,in which,more decision points were put in for the binary-coded method and only the best ant was allowed to provide a feedback mechanism by updating the pheromone trails which were limited to an interval between some maximum and minimum possible values.The new methods were applied to a simple WDS optimal scheduling problem and the better optimizations than that using the genetic algorithms were obtained.