摘要
针对BP神经网络在柴油机故障诊断中,提取训练数据的盲目性及网络收敛速度慢、精度低的问题,提出一种基于Petri网与萤火虫神经网络的故障诊断方法.通过Petri网建模归纳出柴油机所有故障模式,提取神经网络的训练数据,利用萤火虫算法来优化BP神经网络的权值和阈值,改善BP神经网络的性能.仿真实验表明,采用Petri网建模并用萤火虫算法优化BP神经网络的方法,有效地提高了神经网络的收敛速度和诊断精度,在柴油机故障诊断中得到了较好的应用.
Aiming at the blindness of BP neural network for extracting training data in diesel engine fault diagnosis and the problem of slow convergence and low precision,a fault diagnosis method based on Petri net and firefly neural network is proposed.The training data of all the fault modes of diesel engine are extracted by Petri nets modeling,and the weights and thresholds of BP neural network are optimized by using firefly algorithm to improve the performance of BP neural network.Simulation results show that using Petri net modeling method and algorithm to optimize BP neural network with firefly enhances the convergence speed and the diagnostic accuracy of the neural network has been applied in fault diagnosis of diesel engine.
作者
卓宏明
徐鹏
毛攀峰
ZHUO Hongming;XU Peng;MAO Panfeng(Ship Engineering Institute, Zhejiang International Maritime College, Zhoushan 316021, Zhejiang,China)
出处
《中国工程机械学报》
北大核心
2018年第2期178-182,共5页
Chinese Journal of Construction Machinery
基金
舟山市科技局科技资助项目(2016C31050)
