发电机神经网络模型用于电力系统暂态稳定计算

Transient stability study using artificial neural networks models of generation unit

用递归神经网络建立发电机模型，多层输出反馈网络建立励磁系统和调速系统模型，并将此三部分集结构成为一个发电机组的详细模型。将此详细模型写入电力系统的网络方程，联立求解电力系统暂态稳定过程。详细模型写入暂态稳定程序中６机２２节点网络。暂态过程计算结果表明，由人工神经网络模型仿真计算与机理模型常规稳定计算比较，使用人工神经网络模型和机理模型的计算结果非常接近，说明人工神经网络模型可以用于电力系统暂态稳态计算，为今后人工神经网络并行、快速、在线处理电力系统实时计算提供应用途径。

The importance of models of power system has long been recognized. A set of accurate models can be obtained through field tests by means of modern identification methods. In this paper, power system models established by artificial neural networks (ANN) including generator, excitation system, governor were presented. Meanwhile, three parts of generation unit were connected together as a detailed model which was written into power system network equations; and power system transient process was calculated by them. The calculation results demonstrate that artificial neural network models can give precise description of generator's transient process and so provide a novel way for ANN used in transient stability calculation.