基于非参数估计的在线电压预测

摘要:

随着电网规模扩大、复杂度加深,对在线潮流计算确定节点电压提出严峻挑战,通过电压对无功的响应数据来快速精确预测电压发展趋势具有重要意义。提出基于非参数估计的节点电压快速预测方法,以系统负荷水平、无功激励为输入,节点电压为输出,以均方误差作为电压预测精度的指标,衡量预测效果。最后将该方法的预测结果与传统神经网络、自适应神经网络的预测结果作比较分析。通过IEEE 24节点系统标准算例验证表明,非参数估计方法具有较强的电压拟合能力和外推能力,其预测精度与神经网络算法的预测精度相当。

Abstract:

With the expansion of grid scale and complexity，it is difficult to determine the node voltage by on － line power flow calculation，so it is very important to predict the trend of voltage quickly and accurately by the voltage response data． A fast voltage prediction method based on nonparametric estimation is proposed，which takes the system load level and reactive power as the input and the node voltage as the output． The mean square error is used as the index of the voltage prediction accuracy to measure the prediction effect． Finally，the prediction results of the proposed method are compared with those of the traditional neural networks and adaptive neural networks． The results of standard IEEE 24 － bus system shows that the nonparametric estimation method has strong ability of voltage fitting and extrapolation，and its prediction accuracy is equivalent to the prediction accuracy of neural network algorithm．