改进遗传算法在含分布式电源的配电网规划中的应用

An Improved Genetic Algorithm for Distribution Network Planning With Distributed Generation

首先,为充分体现在配电网中引入分布式电源后,分布式电源规划与配电网规划的相互影响,建立了协调规划框架,并在规划目标中引入了分布式电源的环境效益奖励因子;其次,针对传统遗传算法易收敛于局部最优解、迭代次数多、译码过程复杂、迭代时间长的缺点,提出了一种全新矩阵编码模式下的混合整数、0-1遗传算法,使分布式电源和配电网能在同一矩阵中协调、并行迭代,无需译码;最后,针对迭代过程中产生大量不可行解的问题,阐述了一种简单的修复方案,通过一次循环即可完成全网孤岛、孤链、环网的修复。通过IEEE54节点的算例验证了改进算法的先进性、分布式电源与配电网协调规划方法的有效性以及考虑环境效益的必要性。

A coordinated framework is established first between distribution network and distributed generation planning. The environmental benefits-award factor, as an improvement of the function, is introduced at the same time. In order to avoid the defects of easily getting into a local optimal solution, have more iteration number, high complexity of iteration decoding and long computing time in traditional algorithms, this paper proposes an improved genetic algorithm, which can iteration parallel and needs,no process of decoding, based on a new matrix-code which combined the integer code of distributed generation with 0-1 code of the distribution network. At last, a simple restoration scheme that can revise the isolated chain, node and the closed loop all in once is put forward to make the grid into a connected and radial network. The comparison of the results from calculation of IEEE example shows that the proposed method is advanced and effective.