摘要
提出了使用分层分布式神经网络实现大规模电力系统故障诊断的方法。该方法较用单一神经网络的方法,所需训练时间大大缩短,结构上能适应电网结构的不断变化,诊断性能更有效。仿真结果表明,该方法是可行的。
This paper presents an approach of fault diagnosis for large scale power systems based on hierarchical distributed neural networks. Compared with single neural networks approach, this approach needs much shorter training time, is easier to be changed as power system topology changes,and can give more efficient diagnosis results. Simulation results show that this approach is practical.
出处
《系统工程学报》
CSCD
1992年第1期153-160,共8页
Journal of Systems Engineering
