引用本文: | 李 智,杨洪耕.基于邻近中心算法的无功优化分解协调计算[J].电力自动化设备,2012,32(12): |
| LI Zhi,YANG Honggeng.Decomposition and coordination algorithm based on proximal center algorithm for reactive power optimization[J].Electric Power Automation Equipment,2012,32(12): |
|
摘要: |
针对大规模电网无功优化存在的计算速度慢和数据传输瓶颈问题,提出了一种基于邻近中心算法的无功优化分解协调算法。通过邻近函数构造平滑的拉格朗日函数,避免了增广拉格朗日函数的不可分问题;通过最优梯度更新拉格朗日乘子,大幅减少了迭代次数,并且可以直接确定平滑参数等计算所用参数,仅需要通信交换边界节点和拉格朗日乘子信息即可实现全网无功优化的分解协调计算。算例结果表明,所提算法可以有效提高全网无功优化的计算效率,并且与基于辅助问题原理的分解协调算法相比,其收敛速度更快、计算效率更高。 |
关键词: 无功功率 优化 分解 协调 拉格朗日乘子 邻近中心算法 平滑 最优梯度 |
DOI: |
分类号: |
基金项目: |
|
Decomposition and coordination algorithm based on proximal center algorithm for reactive power optimization |
LI Zhi, YANG Honggeng
|
College of Electrical Engineering and Information Technology,Sichuan University,Chengdu 610065,China
|
Abstract: |
A decomposition and coordination algorithm based on proximal center algorithm is proposed to enhance the computing speed and data transmission of centralized reactive power optimization of large-scale grid. The prox-function is adopted in the construction of smooth Lagrange function to avoid the inseparable augmented Lagrange function and the Lagrange multipliers are updated with the optimal gradient to greatly reduce the iteration times. The parameters used,such as the smoothing parameter,can be directly determined and only the boundary node information and the Lagrange multipliers are needed via data communication in the decomposition and coordination of reactive power optimization for the whole network. Case study shows that,the proposed algorithm improves the computing efficiency significantly,and compared with the decomposition and coordination algorithm based on auxiliary problem principle,it has faster convergence rate and higher computing efficiency. |
Key words: reactive power optimization decomposition coordination Lagrange multipliers proximal center algorithm smoothness optimal gradient |