基于自组织RBFN神经网络的预想事故分类器的设计

THE DESIGN OF CONTINGENCY CLASSIFIER BY SELF ORGANIZED RBFN NEURAL NETWORK

提出了自组织径向基函数网络（ＲＢＦＮ） 神经网络的结构和学习算法，从而提高了ＲＢＦＮ 神经网络的学习精度和效率。在此基础上，设计出了基于自组织ＲＢＦＮ 神经网络的电网预想事故分类器。用此预想事故分类器对ＩＥＥＥＲＴＳ６系统的一阶、二阶和三阶预想事故进行安全与不安全的分类。

The paper presents a structure and a learning algorithm of the self organized radial basis functions network(RBFN). The algorithm improves RBFN's learning precision and efficiency. A contingency classifier of power system is designed by using the algorithm. This classifier is used in classifying contingency of IEEE RTS 6 system into security and unsecurity, and it's effective is proved.