基于预估-校正的外网量测采样误差修正方法

Correction Method for Sampling Errors of External Network Measurements Based on Predictor-Corrector

针对分布式状态估计全网一体化计算时,由于不同控制中心提供的量测断面间可能有时差而导致外网估计误差的问题,基于不同控制中心分布式并行计算系统的主从机制,提出了预估—校正的外网量测采样误差修正方法。该方法应用于分布式状态估计计算之前,利用相邻控制中心重叠建模区域的相同测点在分布式状态估计时可能具有不同量测值这一特点,首先预估出两控制中心量测断面时差,再根据此时差校正外网控制中心量测,以保证发起计算的控制中心(主子系统)状态估计的正确性。以邢台电网数据为例,验证了该方法的正确性和有效性。

For the problem of external network estimation error caused by time differences existing between measurement sections provided by different control centers when distributed state estimation of the entire network is started, this paper proposes a correcting method for sampling errors in external network measurements based on predictor-corrector. The method can be used prior to the calculation of distributed state estimation. According to the feature that the same measuring point in the modeling overlap region of adjacent control centers may own different measurement values, the time difference between two control center measurement sections is predicted. Then the control center measurements in the external network are corrected according to the time difference, so as to ensure the state estimation correctness of the control center launching calculation （main subsystem）. Finally, data on Xingtai grid is used as an example to verify the validity and performance of the method.