求解状态估计问题的内点半定规划法

An Semi-Definite Programming Method for State Estimation Problem

为更好地进行状态估计,提出了基于内点半定规划求解状态估计问题全局最优解的方法。该方法基于求解状态估计常用的加权最小二乘模型,通过引入辅助变量,将非凸估计问题等价转化为半定规划问题,然后应用原始–对偶内点法求解。根据状态估计半定规划模型的特点,采用基于半定规划的稀疏技术,提高存储效率和计算性能。算例结果验证了该方法的有效性。

To attain better state estimation results,based on interior-point semi-difinite programming（SDP）,a method to obtain global optimization solution of state estimation problem is proposed.The proposed method is based on weighted least square model,which is commonly used in state estimation,and by means of leading in auxiliary variables the non-convex estimation problem is equivalently turned into a semi-definite programming problem and then solved by primal-dual interior point method.According to the features of SDP model for state estimation the sparsity technique based on SDP is adopted to improve both storage efficiency and calculation performance.The testing of the proposed method is performed by four IEEE sytems,namely IEEE 14-bus system,IEEE 30-bus sytem,IEEE 57-bus system and IEEE 118-bus system,and simulation results of these testi systems show that the proposed method is available.