基于破圈法与改进布谷鸟算法的两阶段配电网重构

Two-stage distribution network reconstruction based on breaking circle method and improved cuckoo algorithm

针对传统智能优化方法在配电网重构过程中初始种群效果较差,并且难以获得全局最优解以及效率较低的问题,提出破圈法与改进布谷鸟算法相结合的两阶段重构方法。首先,以配电网的系统网损、负荷均衡和电压偏差三大综合指标为目标函数,建立多目标数学模型。其次,使用两阶段重构方法寻优,第一阶段使用破圈法快速求解配电网重构的初步解集,并将其作为第二阶段的初始种群;第二阶段对布谷鸟算法进行改进,使用基本环网编码策略,有效地减小了解空间,避免了大量的不可行解,并引入鲸鱼优化算法的气泡网搜索和螺旋搜索来避免早熟现象,有效地提高了算法的全局搜索能力。最后,使用含分布式电源的IEEE 33节点系统和Taipower 84进行仿真,结果表明该两阶段重构方法具有初始种群优、搜索能力强、收敛速度快的优点。

Aiming at the problem that the traditional intelligent optimization method has poor initial population effect in the process of reconfiguration of the distribution network,and it is difficult to obtain the global optimal solution and the efficiency is low,a two-stage reconstruction method in combination of the breaking circle method and the improved cuckoo algorithm is proposed.Firstly,a multi-objective mathematical model is established with the three comprehensive indicators of the distribution network system network loss,load balance and voltage deviation as the objective function.Then,the two-stage reconstruction method is used for optimization.In the first stage,the breaking circle method is used to quickly solve the initial solution set of the distribution network reconstruction,and takes it as the initial population of the second stage;In the second stage,the cuckoo algorithm is improved.The basic ring network coding strategy is used to effectively reduce the understanding space,avoid a large number of infeasible solutions,and introduce the bubble network search and spiral search of the whale optimization algorithm to avoid premature phenomenon and effectively improve the global search ability of the algorithm.The IEEE 33 node system with distributed power generation and Taipower 84 are used for simulation.The results show that the twostage reconstruction method has the advantages of optimal initial population,strong search ability and fast convergence speed.