摘要
为了对刮板输送机减速器故障进行准确诊断研究,提出了一种基于改进萤火虫算法优化神经网络故障诊断方法。首先对刮板输送机减速器故障特征参数进行特征提取,其次应用特征数据样本进行基于神经网络的故障诊断模型训练,利用改进萤火虫算法对神经网络权值、阈值进行优化,加快目标的优化求解,得到最优的网络模型。初步研究表明将改进萤火虫算法与BP(back propagation)神经网络结合可以有效地解决神经网络收敛速度慢,易陷入局部最优等问题,可以对刮板输送机减速器的故障进行准确诊断。
In order to make accurate diagnosis for the scraper conveyor speed reducer failure study.This paper proposes a improved firefly algorithm to optimize neural network based fault diagnosis method.Firstly,the characteristics of the fault characteristic parameters of the blade conveyor are extracted.The second application feature data sample for fault diagnosis model based on neural network training.Using the improved firefly algorithm to optimize neural network weights and threshold,to speed up the optimum value of,get the optimal model of the network.Preliminary studies suggest that the improved firefly algorithm combined with BP(back propagation) neural network can effectively solve the neural network slow convergence speed,easily falling into the master problem,can make accurate diagnosis for the failure of scraper conveyor speed reducer.
作者
毛君
郭浩
陈洪月
MAO Jun;GUO Hao;CHEN HongYue(School of Mechanical Engineering,Liaoning Technical University,Fuxin 123000,China;China National Coal Association,Dynamic Research for high-end complete Integrated Coal Mining Equipmentand Big Data Analysis Center,Fuxin 123000,China;Combined Mining Technology and Equipment Engineering National Research Center,Fuxin 123000,China)
出处
《机械强度》
CAS
CSCD
北大核心
2019年第3期544-550,共7页
Journal of Mechanical Strength
基金
国家能源研发(实验)中心重大项目(2010_215)资助~~
