摘要
针对液压支架顶梁可靠度详细计算方法计算量大的问题,提出了基于神经网络模型的液压支架顶梁可靠度近似计算方法。采用误差反向传播算法作为权值学习算法,选用平方型函数作为误差函数建立了顶梁可靠度近似计算神经网络模型。以ZY18000/25/45D型液压支架的顶梁为例进行可靠度近似计算,结果表明,顶梁可靠度计算相对误差在6.3%以内。
In view of problem that detailed reliability calculation method of roof beam of hydraulic support has large amount of calculation,a reliabillty approximation calculation method of roof beam of hydraulic support based on neural network model was proposed.Neural network model for reliability approximation calculation of roof beam was established.The back propagation algorithm is used as weight learning algorithm,and square type function is used as error function.Taking roof beam of ZY18000/25/45 Dhydraulic support as an example to calculate the reliability,the results show that relative error of roof beam reliability calculation is within 6.3%.
作者
王雷
陆金桂
陈涌
WANG Lei;LU Jingui;CHEN Yong(School of Mechanical and Power Engineering,Nanjing University of Technology,Nanjing 211816,China;Dewei Intelligent Technology Company of Jiangsu,Jingjiang 214500,China)
出处
《工矿自动化》
北大核心
2018年第7期88-91,共4页
Journal Of Mine Automation
基金
"十二五"国家科技支撑计划项目(2013BAF02B11)
关键词
液压支架
顶梁可靠度
结构优化设计
近似计算
神经网络模型
hydraulic support
reliability of roof beam
optimized design of structure
approximationcalculation
neural network model
