求解负荷分配问题的多目标加权优化方法

Multi-objective weighting optimization for solving load distribution

针对节能减排计划下多目标优化的负荷分配问题,考虑到实际调度中对阶段性及目标值的追求,建立了多目标期望控制的负荷分配模型,并采用动态加权算法求解寻优。动态加权算法可减少主观性赋权,并能在每次迭代过程中根据目标偏差率大小自动调整权重系数,使得三类目标函数实现同时趋近于期望值。但求解时发现,动态加权求解存在多目标优化问题,为此提出了权值预估方法,经仿真算例分析验证,该方法能较好地解决复杂的电力系统负荷分配问题。

Aiming at the multi-objective optimization of load distribution under the plan of energy saving and emission reduction,this paper,taking the pursuit of stage and target value in the practical dispatching,establishes the multi-objective expected control model of load distribution,and adopts dynamic weighting to find the optimal solution. Although dynamic weighting is able to reduce subjective weight determination and adjust automatically weight coefficient according to the ratio of target deviation during the iterative process so as to make the three kinds of objective functions approach to the expected value at the same time,this algorithm is found that there is the problem of multi-objective optimization during the solution. Based on the analysis,this paper proposes the weight value prediction method which is proved by simulation example to be able to solve complicated power system load distribution.