基于灵敏度分析的线性混合量测状态估计

Sensitivity-analysis-based linear state estimation with mixed measurements

提出一种基于灵敏度矩阵并以相量测量单元(PMU)采样为周期的混合量测线性估计方法。把PMU的电流量测转换成功率量测,在初始计算时刻与基于监控和数据采集系统的量测一起进行非线性运算,得到状态估计值和功率量测量之间的灵敏度矩阵。在后续的PMU采样时刻,将转换得到的功率量测和通过负荷预报补充的伪量测组成混合量测,然后根据求得的灵敏度矩阵进行以PMU采样为周期的线性跟踪计算。当估计误差积累到一定程度时,重新进行一次混合非线性计算以更新灵敏度矩阵。通过IEEE 118节点系统,给出了负荷变化速率、预报误差和PMU量测数对所提估计方法计算精度的影响。结果表明,负荷变化越慢、预报误差越小、PMU的个数越多,其跟踪计算精度越高。

A linear state estimation method is proposed based on sensitivity analysis,which takes the sampling cycle of PMU（Phasor Measurement Unit） as its computation period. The current phasor measured by PMU is transformed into power,which is then combined with the measurements of SCADA （Superiority Control And Data Acquisition） at the initial time to carry out nonlinear calculation,thus the sensitivity matrix between state estimation value and power measurement is obtained. In the flowing sampling cycle of PMU, the transformed power measurements are mingled with the load forecasts to compose mixed -measurements. Based on the obtained sensitivity matrix,the linear tracking state estimation is performed in accordance with the sampling period of PMU. When the estimation error is accumulated to some extent,the non-linear state estimation is implemented once again to update the sensitivity matrix. The influences of load variation,forecast error and PMU number on estimation error are investigated for IEEE 118-bus system. Results show that,slower load changes,smaller forecast error and more PMUs result in better tracking precision.