电力系统负荷恢复问题的混合遗传算法求解

对电力恢复过程中最后一个阶段的负荷恢复问题进行了研究,把电力系统的负荷恢复问题建模为带众多约束条件的0?1背包问题,并设计了一种将贪心算法与改进遗传算法相结合的改进混合遗传算法来对问题进行求解.在遗传算法之前,先用贪心算法生成该问题的贪心解,然后让每一代中有着最差适应度的个体无条件的变为此贪心解,使最终结果至少不会比贪心法差.解决了系统在负荷恢复过程中的潮流计算问题,采用先求系统的频率变化,然后再计算潮流分布的方法,将约束条件和目标函数融合在一起,通过建立一种偏序关系,避免了罚函数选择参数的困难.利用贪心算法求解背包问题的快速性和多父体杂交的非凸组合技术,使算法具有求解的快速性和在解空间内搜索的遍历性.算例求解结果表明了该算法在负荷恢复问题中的有效性.

Hybrid Genetic Algorithm for Power System Load Restoration

The problem of the load restoration at the last stage of the power system restoration was studied in the paper. The problem of load restoration was modeled as a zero-one knapsack problem with many constraints and a hybrid genetic algorithm that combined the greedy algorithm and modified genetic algorithm together was introduced to solve the problem. In the algorithm greedy algorithm was used first to produce the greed solution and then the worst solution of each generation produced by the genetic algorithm was replaced by the greed one on any condition. This made the final solution better than the greed one. To obtain the power flow during the load restoration, frequency was calculated before power flow is done in the proposed method. In the paper the constrained conditions and the objective functions were combined together. By the means of defining a partial-order relation it need not to implement a penalty function parameter constraint-handling approach. The algorithm had the characteristic of quick and global searching because of the non-convex combination technique in the multi-parents crossover and the quick character of the greedy algorithm for knapsack problem. The simulation results show the feasibility and availability of the algorithm.