摘要
为提高遗传算法求解配电网故障恢复问题的效率,提出了一种染色体混合编码模型。对配电网的网络拓扑进行节点深度法编码,通过相应的交叉变异操作,产生可行的配电网生成树。非故障失电区的可中断负荷采用0-1编码,实现切负荷控制,增加故障恢复的灵活性。这种混合编码方式可以大大降低不可行解的数量,加快算法收敛速度。采用多目标优化算法NSGA-Ⅱ,减少权重系数主观性对最优解的影响,实现各个优化目标的协同进化,可以为调度人员提供多个最优恢复方案。算例计算结果验证了所提出的模型和算法的正确性和有效性。
A chromosome hybrid encoding model is established to enhance the genetic algorithm＇s efficiency of solving problems of distribution system service restoration. After node-depth encoding of distribution network topology and corresponding crossover and mutation operation, feasible distribution network spanning trees are generated. 0-1 encoding is applied to interrupttable load in the non-fault load loss regions to control load shedding and increase the flexibility of fault restoration. Using the hybrid encobing mode, the infeasible solutions can be reduced greatly, and the algorithm convergence is speeded up. Multi-objective optimization algorithm NSGA-II reduces the influence of weight coefficient subjectivity on optimum solution, and attains co-evolution of each optimization objective. It provides dispatcher with more than one service restoration plan. Simulation of the proposed model and the algorithm has verified their feasibility and correctness.
出处
《电力系统保护与控制》
EI
CSCD
北大核心
2012年第6期104-108,共5页
Power System Protection and Control
基金
城市轨道交通电力系统运行安全关键技术示范项目(10DZ1200200)
关键词
故障恢复
节点深度编码
混合编码
多目标优化
切负荷控制
配电网
service restoration
node-depth encoding
hybrid encoding
multi-objective optimization
load shedding control
distribution system
