摘要
综合遗传算法的全局优化和神经网络的并行计算等特点 ,提出了一种基于遗传 -神经网络的凝汽器故障诊断的方法。用遗传算法来优化神经网络权值 ,克服了神经网络易陷入局部解的缺陷 ,使神经网络具有较好的全局性和收敛速度。具体故障诊断实例表明 ,该方法诊断准确 。
The global search capabilities of the genetic algorithm and the parallel computation of the neural networks is combined and Genetic-based neural networks(GNNs) for the fault diagnosis of steam condenser is presented. The network connection weights are optimized by the genetic algorithm,so the network is in whole and rapid convergence. A real condenser fault is reliably diagnosed by this method.
出处
《电站辅机》
2004年第3期8-11,共4页
Power Station Auxiliary Equipment
关键词
神经网络
遗传算法
凝汽器
故障诊断
neural network
genetic algorithm
steam condenser
fault diagnosis