湖南电力 ›› 2025, Vol. 45 ›› Issue (4): 83-89.doi: 10.3969/j.issn.1008-0198.2025.04.012

• 电网运行与控制 • 上一篇    下一篇

基于改进遗传算法和分支定界法的行波检测装置双层优化配置方法

孙晓敏1, 温志超1, 谢浩南1, 吴宏曜1, 宋冬然2   

  1. 1.广东电网有限责任公司肇庆供电局,广东 肇庆 526000;
    2.中南大学自动化学院,湖南 长沙 410083
  • 收稿日期:2025-05-12 修回日期:2025-06-20 出版日期:2025-08-25 发布日期:2025-09-05
  • 通信作者: 温志超(1989),男,本科,研究方向为电力系统运行与优化。
  • 作者简介:孙晓敏(1983),女,博士,研究方向为电力系统规划。
  • 基金资助:
    广东电网有限公司科技项目(031200KC23110014)

Dual Layer Optimization Configuration Method for Traveling Wave Detection Device Based on Improved Genetic Algorithm and Branch & Bound Method Algorithm

SUN Xiaomin1, WEN Zhichao1, XIE Haonan1, WU Hongyao1, SONG Dongran2   

  1. 1. Guangdong Power Grid Co., Ltd., Zhaoqing Power Supply Bureau, Zhaoqing 526000, China;
    2. School of Automation, Central South University, Changsha 410083, China
  • Received:2025-05-12 Revised:2025-06-20 Online:2025-08-25 Published:2025-09-05

摘要: 针对电力系统行波故障定位中测量装置安装位置优化问题,提出一种基于改进遗传算法和分支定界法的行波检测装置双层优化配置方法,以解决电力系统中经济性与可靠性的协同优化问题。该装置上层模型以最小化装置安装成本为目标,考虑预算与防干扰约束;下层模型在给定布局下最大化故障检测覆盖率,嵌入动态调整层实现全线路检测覆盖。通过改进遗传算法与分支定界法相结合的方式求解最优布局方案,降低模型受到通信时滞的影响,加快模型在非理想场景下的收敛速度,实现层级间动态反馈修正。基于IEEE 39节点系统实验验证结果表明,该方法在保证95%故障检测率的前提下,较传统优化方法的设备成本降低了24.5%,覆盖率波动降低了2.46%,具有较强的鲁棒性。

关键词: 行波测量, 双层优化, 故障定位, 改进遗传算法, 分支定界法

Abstract: Aiming at optimizing the installation position of measurement devices in traveling wave fault location of power systems, a dual layer optimization configuration method for traveling wave detection devices based on improved genetic algorithm (IGA) and branch and bound (B&B) method is proposed to solve the collaborative optimization problem of economy and reliability in power systems. The upper level model aims to minimize the installation cost of the device, considering budget and anti-interference constraints. The lower level model maximizes fault detection coverage under a given layout and embeds a dynamic adjustment layer to achieve full line detection coverage. By combining improved genetic algorithm and branch and bound method, the optimal layout scheme is solved to reduce the influence of communication delay on the model, accelerate the convergence speed of the model in non ideal scenarios, and achieve dynamic feedback correction between levels. The validation experiment results based on IEEE 39 node system show that this method reduces equipment cost by 24.5% and coverage fluctuation by 2.46% compared with traditional optimization methods under the premise of ensuring a 95% fault detection rate, and demonstrates strong robustness.

Key words: traveling wave measurement, dual layer optimization, fault location, improved genetic algorithm, branch and bound method

中图分类号: