基于数据挖掘技术的电力网络参数估计方法研究

错误的网络参数对状态估计的结果产生较大影响,现在一般认为参数值是常量,但实际上网络参数随周围环境、天气状况及负荷水平等变化而变化。提出基于数据挖掘技术的电力网络参数估计方法,利用大量样本数据估计网络的真实参数值。首先,利用聚类分析技术对SCADA历史数据进行分类,分成不同类型的样本数据;其次,利用数据处理技术对各个样本数据中的孤立点、空缺值、噪声等进行处理;最后,利用线性回归技术估计满足一定条件的网络参数。经过两个算例验证,证明基于数据挖掘技术的状态估计方法是非常有效的。

The State Estimation of Power System Parameters based on Data Mining Technology

The false network parameters impair the correctness of state estimation. Parameters, regarded as constants usually, vary along with the status of environment, weather and loads actually. This paper puts forward the grid parameters estimation method based on data mining and estimates the real value of parameters through enormous sample data. Firstly, the SCADA historical data is sorted into kinds of sample ones by cluster analysis technology; Secondly, the outliers, missing value and noise are treated by data processing methods; Finally, the needed grid parameters are estimated through liner regression and the state estimation based on data mining technology is proved to work well in the two calculation examples.