基于聚类的差分进化算法的两阶段最优潮流方法  

Group-based Differential Evolution Method and Its Application to Two-stage Optimal Power Flow

作　　者：田玮 江晓东 TIAN Wei;CHIANG Hsiao-dong(School of Electrical and Information Engineering,Tianjin University,Tianjin 300072,China)

机构地区：[1]天津大学电气自动化与信息工程学院,天津300072

出　　处：《电力系统及其自动化学报》2021年第11期50-55,共6页Proceedings of the CSU-EPSA

摘　　要：差分进化算法是一种广泛应用于求解非线性优化问题的全局最优解的元启发式方法,但存在容易找到次优解或近似局部最优解的问题。为此,提出了一种求解高质量局部最优解甚至全局最优解的基于聚类的差分进化算法的两阶段方法,并将该方法应用于电力系统最优潮流问题。所提方法由基于聚类的差分进化算法和局部优化算法组成。第Ⅰ阶段是基于聚类的差分进化算法利用强大的全局搜索能力快速确定包含局部最优解的区域;第Ⅱ阶段是局部优化算法利用局部寻优能力为非线性优化问题高效寻找高质量的局部最优解甚至全局最优解。在一组基准函数上测试了该两阶段优化方法的求解性能,并通过对IEEE 118节点电力系统最优潮流的计算,验证了所提两阶段优化方法的有效性和实用性。As a popular meta-heuristic method,the differential evolution(DE)method can compute the global optimal solutions of nonlinear optimization problems.However,it may find sub-optimal or near local optimal solutions.To solve this problem,a two-stage group-based DE method is proposed in this paper,which can compute the high-quality local optimal or even the global optimal solution.In addition,it is applied to solve the optimal power flow problem in the power system.The proposed method consists of a group-based DE method and a local optimization method.At the first stage,the group-based DE method exploits its powerful global search capability to quickly identify the regions containing local optimal solutions.At the second stage,the local optimization method exploits its local search capability to efficiently find the high-quality local optimal or even the global optimal solution.The performance of the proposed two-stage optimization method is tested on a set of benchmark functions.Moreover,the effectiveness and practicability of the proposed method are verified by calculating the optimal power flow in an IEEE 118-bus power system.

关 键 词：非线性优化问题 差分进化算法 局部优化方法 最优潮流计算 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]