摘要
针对汽轮发电机组的振动故障诊断的特点,分析了BP网络和遗传算法原理和缺陷。为了克服BP算法中存在的网络学习速度慢,容易陷入极小的问题,在运用改进遗传算法的基础上,探讨了一种自适应遗传神经网络算法,并将其应用于汽轮发电机组故障识别。实验数据表明,该算法可以克服BP网络的不利影响,完成故障信号分析,收敛速度快,能有效地识别故障,具有一定的参考价值。
To the problem of failure diagnosis of steam turbine generator set, principle and existing defects of genetic algorithm and BP network are analyzed. In order to overcome the problems of slow rate of convergence and falling easily into local minimum in BP algorithm, the adaptive genetic neural algorithm and combining model are discussed. The algorithm is applied to steam turbine-generators fault recognition, and the experiment results show that the algorithmcar can overcome BP network＇s influence of the capacitors and can converges quickly and can recognize faults efficiently. It is a reference for faults recognition.
出处
《控制工程》
CSCD
2007年第B05期175-177,195,共4页
Control Engineering of China
基金
山东省自然科学基金(Y2004F15)
关键词
遗传算法
神经网络
故障诊断
汽轮发电机
genetic algorithm
neural network
faults diagnosis
steam turbine generator
