电力系统状态估计算法的缺陷探究

Explore of Defect on Power System State Estimation Algorithm

电力系统的常用状态估计方法有加权最小二乘法估计和快速分解状态估计等。由于受到函数局部单调性和凹凸性的影响,即使给出了离真实根很近的初值,随着每次的迭代,状态估计方法结果将离真实根越来越远。因此考虑到在对目标函数进行极值求解中可以舍弃牛顿迭代法,采用其他的(例如粒子群进化算法)最优化方法进行尝试。

Power system state estimation method is commonly used method of weighted least squares estimation and fast decoupled state estimation.Because of the function of local monotonicity and convexity of the influence,even given away from the actual root very close to the initial value,with each iteration,state estimation results will be more and more far away from the real root.Therefore the objective function is taken into account in the extreme solution can abandon the Newton-Raphson method,using the other optimization method to try.（e.g.evolutionary particle swarm algorithm optimization method）.