|
|
|
| Fault diagnosis of hydraulic breaking hammer based on Fruit Fly Algorithm optimized fuzzy RBF neural network |
| LI Xiao-huo, WENG Zheng-yang, QIANG Ya-sen, SHI Shang-wei, LI Yan |
| College of Mechanical Engineering, Liaoning Technical University, Fuxin 123000, China |
|
|
|
Abstract Focusing on the variety and uncertainty of the fault reason of a hydraulic breaking hammer, in order to avoid the problems of traditional fuzzy BP neural network, such as poor convergence rate in fault diagnosis and easy to fall into a local minimum, a new method of fault diagnosis of hydraulic breaking hammer by using Fruit Fly Algorithm for optimization of fuzzy RBF neural network was proposed. By synthesizing the neural network's associative memory, processing ability, and fuzzy logic system’s qualitative knowledge, fuzzy reasoning ability, optimizing fuzzy RBF neural network expansion parameter by Fruit Fly Optimization Algorithm, a relations between the network fault information and fault reasons was established. The simulation tested by MATLAB indicated that fuzzy RBF neural network optimized by Fruit Fly Algorithm worked accurate and fast. The result of the diagnosis agrees with target outputs, which proves the feasibility of this method.
|
|
Received: 30 October 2014
Published: 28 December 2015
|
基于果蝇算法优化模糊RBF网络的液压破碎锤故障诊断
针对液压破碎锤故障原因具有多样性和不确定性，为避免传统模糊BP网络故障诊断存在收敛速度慢、易陷入局部极小值等缺陷，提出将果蝇算法优化模糊RBF网络方法用于液压破碎锤故障诊断.综合神经网络的联想记忆、并行处理能力和模糊理论的定性知识、模糊推理能力，同时利用果蝇算法对模糊RBF网络的扩展参数进行全局优化，建立了液压破碎锤系统输入故障征兆与输出故障原因间的映射.利用MATLAB软件编程进行仿真实验，结果表明：果蝇算法优化模糊RBF网络方法精度高，收敛速度快.利用该法对液压破碎锤故障诊断，结果与目标输出相符，证明该方法可行.
关键词：
液压破碎锤,
故障诊断,
果蝇算法,
模糊RBF神经网络,
优化
|
Viewed |
|
|
|
Full text
|
|
|
|
|
Abstract
|
|
|
|
|
Cited |
|
|
|
|
| |
Shared |
|
|
|
|
| |
Discussed |
|
|
|
|
