基于时序数据相关性挖掘的WAMS/SCADA数据融合方法

WAMS/SCADA Data Fusion Method Based on Time-series Data Correlation Mining

在电力系统多运行方式的背景下,研究WAMS/SCADA等量测数据融合是解决大电网在线稳定分析的关键点之一。为此,基于理论分析,从2者数据相关性角度研究了WAMS/SCADA相关性系数,提出了基于时序数据相关性挖掘的WAMS/SCADA数据融合方法。通过构建Pearson相关性系数,对WAMS/SCADA的相关性进行评估;运用广义EM算法对量测数据曲线时差问题进行函数模型求解;在考虑量测权值的情况下对量测有效性进行分析。结果表明:对3种不同数据进行状态估计后,经过数据融合后的曲线结果在系统稳定时段与出现扰动时段均保持平稳;基于时序数据相关性融合法所得到的状态估计曲线变化趋势与其他算法相同,混合量测状态估计结果误差<5%。IEEE 118节点母线系统算例的仿真结果验证了上述方法的有效性与稳定性。

Under the background of multi operation mode of power system, it is one of the key points to solve the online stability analysis of large power grid by studying the integration of WAMS/SCADA measurement data. To this end, by the theoretical analysis, we researched the correlation coefficient of WAMS/SCADA from the point of view of the data corre- lation of two parts, and proposed the WAMS/SCADA data fusion method based on the correlation mining of time series data. The correlation of WAMS/SCADA was evaluated by constructing the Pearson correlation coefficient; then we used the generalized EM algorithm to solve the time difference problem of measured data curve. Finally, we analyzed the measurement validity when considering the weight measurement. The results showed that,after the state estimation of three different kinds of data, the curves of the data fusion are kept stable during the period of system stable period and the disturbance period; the variation trend of the time series data correlation is the same as that of the other algorithms, and the error of the mixed measurement state estimation is 〈5%. The simulation results Of IEEE 118 bus system have verified the effectiveness and stability of the above methods and conclusions.