论文部分内容阅读
经典多目标决策方法通过直接或利用层次分析法给定各目标权因子,再进行多目标综合。这种算法存在局限性,如缩放因子或模糊集参数实时调整难、获取最优解效率低等。现有舰船电网故障恢复方法仅给出负荷投切方案,且不考虑舰船线损。在严格定义“支配”、“目标支配”和“约束支配”3个概念的基础上,利用解的非支配关系,提出采用本质多目标进化算法并面向开关的故障智能恢复决策,该方法毋需权因子、缩放因子和模糊集参数,并实现了故障恢复后线损最小,在获得Pareto解集后,再作目标偏好选择,给出最优开关动作序列,提高了获取最优解的效率,解决了多约束条件下解的支配关系判别难题。算例分析证实了该方法可行、有效,且优于经典多目标方法。
The classical multi-objective decision-making method gives each objective weight factor directly or by AHP, and then carries out multi-objective synthesis. This algorithm has its limitations, such as scaling factor or fuzzy set parameters real-time adjustment difficult to obtain the optimal solution and low efficiency. The existing method of grid fault recovery of the ship only gives the load switching scheme and does not consider the ship line loss. Based on the three definitions of “dominance”, “dominance” and “dominance”, the intelligent multi-objective evolutionary algorithm and switch-oriented intelligent fault recovery This method does not need weight factor, scaling factor and fuzzy set parameter, and achieves the least line loss after fault recovery. After obtaining the Pareto solution set, it makes a choice of target preference and gives the optimal switching action sequence, The efficiency of the optimal solution, to solve the problem of dominance of solution under multi-constraint conditions. The case study shows that this method is feasible, effective and superior to the classical multi-objective method.