摘要
提出一种应用双种群二进制微分进化算法求解输电网扩展规划的新方法。算法中通过海明距离比较个体之间的差异并定义了全局变异和局部变异两种算子。结合两种变异算子的优点,将种群划分为两个子种群,分别采用不同的进化模式同时进化,有效地平衡了种群在解空间的全局探索能力和局部开发能力。通过种群间的相互移民来进行信息交换,维持了种群的多样性,降低了算法陷入局部最优的风险。将该算法应用于18节点系统和巴西南部46节点系统,并对控制参数的选择进行了分析比较,计算结果验证了所提算法的可行性和有效性。
By applying the binary differential evolution algorithm with dual subpopulations, a novel approach of solving transmission network expansion planning is presented. The differences between individuals are obtained via using the Hamming distance, and the global mutation operator and the local mutation operator are defined. Corsidering the advantages of two mutation operators the population is divided into two subpopulations, their evolution is performed simultaneously by employing different mutation schemes, so that the global exploration and the local exploitation ability of the population in search space are balanced. The diversity of the population is maintained through immigration strategy, the risk of the algorithm being trapped in local optimal solution is reduced. The algorithm is applied to an 18-bus system and Southern Brazilian 46-bus network, and the control parameters selection is analyzed and compared. The result shows the feasibility and effectiveness of the proposed algorithm.
出处
《电力系统及其自动化学报》
CSCD
北大核心
2009年第6期56-61,共6页
Proceedings of the CSU-EPSA
基金
北京市教委科技成果转化与产业化项目
关键词
输电网扩展规划
二进制编码
微分进化算法
双种群
transmission network expansion planning
binary encoding
differential evolution algorithm
dual subpopulations
