| 引用本文: | 陈艳波,马 进,陈 茜.混合整数线性规划形式的抗差状态估计方法[J].电力自动化设备,2015,35(7): |
| CHEN Yanbo,MA Jin,CHEN Qian.Robust state estimation in form of mixed integer linear programming[J].Electric Power Automation Equipment,2015,35(7): |
|
| 摘要: |
| 现有的抗差状态估计方法一般需要求解非线性非凸优化问题，并用基于梯度的方法予以求解，难以保证获得全局最优解，且可能存在收敛性问题。基于精确线性化量测方程提出了一种混合整数线性规划形式的抗差状态估计方法。首先通过引进辅助状态向量和辅助量测向量，得到了线性量测方程；通过引入代表量测量是否为正常量测量的二值变量，将线性量测方程变为线性量测不等式；最后求取一个可使尽可能多的正常量测量来支持的状态向量。所提方法抗差能力强，从数学上可保证获得全局解；无需非线性迭代。仿真算例验证了所提方法的有效性和高效性。 |
| 关键词: 电力系统 抗差估计 状态估计 不良数据辨识 混合整数线性规划 混合整数非线性规划 收敛性 |
| DOI： |
| 分类号: |
| 基金项目:国家自然科学基金资助项目（51407069） |
|
| Robust state estimation in form of mixed integer linear programming |
|
CHEN Yanbo1, MA Jin2, CHEN Qian1
|
|
1.State Key Laboratory of Alternate Electrical Power System with Renewable Energy Sources，North China Electric Power University，Beijing 102206，China;2.School of Electrical and Information Engineering，University of Sydney，Sydney NSW2006，Australia
|
| Abstract: |
| The existing robust state estimation approaches normally apply the gradient-based methods to solve the problems of nonlinear and non-convex optimization，which cannot guarantee the globally optimized solution and may not be convergent. A method of robust state estimation in the form of mixed integer linear programming is proposed based on the exact linear measurement equations. An auxiliary state vector and an auxiliary measurement vector are introduced to obtain the linear measurement equations；a binary variable used to identify a measurement as normal or abnormal is then introduced to transform the linear measurement equations into the linear measurement inequalities；a state vector supported by the normal measurements as more as possible is finally estimated. Without nonlinear iteration，the proposed approach has excellent robustness to mathematically guarantee the globally optimized solution. Simulative results demonstrate its effectiveness and efficiency. |
| Key words: electric power systems robust estimation state estimation bad-data identification mixed integer linear programming mixed integer nonlinear programming convergence |
