基于蝙蝠-蛙跳神经网络的提升机故障诊断研究  

Research on Fault Diagnosis for Hoist Based on Neural Network and BA-SFLA

作　　者：乔维德 凌兴宏[2] 周晓谋[3] 

机构地区：[1]无锡开放大学科研处,江苏无锡214011 [2]苏州大学计算机科学与技术学院,江苏苏州215006 [3]中国矿业大学机电工程学院,江苏徐州221116

出　　处：《台州学院学报》2016年第6期47-52,共6页Journal of Taizhou University

基　　金：无锡市社会事业领军人才项目(WX530/2016/023)

摘　　要：针对目前提升机故障诊断方法存在的局限性,在分析提升机制动系统的故障征兆和故障类型基础上,提出一种基于蝙蝠-蛙跳神经网络的故障诊断策略,构建提升机制动系统的神经网络故障诊断模型,采用蝙蝠—蛙跳算法优化神经网络的结构参数。仿真实验分析表明,该方法用于提升机制动系统故障诊断,诊断速度快、准确性高、可靠性好。In view of the limitation of the current elevator fault diagnosis method, based on analyzing the fault symptom and fault type of the hoist braking system,this paper presents a fault diagnosis strategy based on BA- SFLA and Neural Network, a model of neural network fault diagnosis is constructed for the braking system of hoist,and the structure parameters of the neural network are optimized with BA-SFL＆ Simulation experiment analysis shows that this method is applied to the fault diagnosis for hoist braking system with advantages of fast speed, high accuracy and good reliability.

关 键 词：提升机制动系统 蝙蝠-蛙跳算法 故障诊断 

分 类 号：TD534[矿业工程—矿山机电] TP18[自动化与计算机技术—控制理论与控制工程]