基于神经网络优化法的故障诊断应用研究

在机械设备的故障诊断中，常采用BP网络算法对故障进行诊断计算，但由于BP网络易于收敛于局部极小点，且在初始参数与网络结构选取不当时。网络将出现发散现象．为此提出了将神经网络优化算法应用于汽轮发电机组的故障诊断中，实现了神经网络权值和阈值的快速计算，并以汽轮发电机组的故障诊断为背景。将两种算法的结果进行比较，证明该方法比BP算法精度高且收敛速度快、可靠性好．

Comparison between powell algorithm and BP algorithm in training neural network

BP algorithm is often used in faults diagnosis of mechanical equipment.But BP network is apt to converge at the local minimum point.And,if the selected original parameters and network structure are not suitable,the divergent phenomenon will turn up in the network.SO this article advances that the neural network optimization algorithm is used in faults diagnosis of the steam generating set (rotating mechanical equipment),and the fast calculation in weights values and threshold values of neural network are gained.The neural network optimization algorithm result is compared with BP network algorithm result.It is shown that this method is faster and has higher accuracy than BP algorithm,and the neural network is absolutedly,convergent.