基于记忆的改进克隆算法及其PMU配置多目标优化应用

Modified Clonal Algorithm Based on Memory and Its Application in Multiobjective Optimization of PMU Placement

以PMU安装数、量测系统可观测性和基于混合量测的状态估计精度三者为优化目标的PMU优化配置(OPP)是二层规划问题。该文证明了用单次状态估计精度评价量测系统性能的可行性,提出精度加权估算公式。将二层规划目标简化为分段函数,提出基于记忆的改进克隆算法。除模仿生物免疫系统的克隆选择和受体编辑机制外,该算法引入记忆加速算子以强化邻域搜索,并分段调整循环补充规模、高频变异与重组操作概率,从而显著加快和稳定进化进程,避免搜索陷入局部最优解。基于IEEE14/57节点系统的算例表明,该算法能快速稳定地求出全局最优解及近似解,比原克隆算法等更适用。

The optimal PMU placement （OPP） with three optimal objects of the PMU number, the observability and the precision of state estimation based on PMUs and conventional measurements, is a bilevel programming problem. To evaluate the precision which is proved to be determined by the measurement configuration mainly, a weighted formula is proposed. A simplified objective function and a modified clonal algorithm （CLONALG） based on memory are proposed. In addition to imitating the clonal selection and the receptor editing mechanism of immune systems, the algorithm introduces an accelerating operator based on memory and stepwisely adjusts the number of cycle supplement population and the probabilities of hypermutation and recombination operators. The modifications can quicken and stabilize the optimizing process and prevent the search from locally optimal traps. The examples based on IEEE 14-bus/57-bus indicate that the proposed algorithm is more applicable than the original CLONALG.