基于BSS-HHT优化数学模型的旋转机械故障特征提取与诊断  

Fault feature extraction and diagnosis of rotating machinery based on BSS-HHT optimization mathematical model

作　　者：陶明珠 Tao Mingzhu(Basic Course Teaching Department,Hefei University of Economics,Anhui Hefei,230031,China)

机构地区：[1]合肥经济学院基础课教学部,安徽合肥230031

出　　处：《机械设计与制造工程》2023年第11期67-72,共6页Machine Design and Manufacturing Engineering

基　　金：安徽省省级质量工程项目(2020szsfkc0473)。

摘　　要：针对现有旋转机械故障特征提取算法存在的故障特征分类准确率低和效率低等问题,提出一种基于BSS-HHT的故障特征提取与诊断数学模型。首先基于小波软阈值降噪,同时对原始故障信号进行预处理,消除噪声和故障信号之间的二阶性特征;然后构建BSS算法模型并以负熵作为目标函数,从源信号中分离有效的故障信号;最后通过引入HHT算法对输出信号做经验模态分解,在消除端点效应的同时实现对故障特征点的定位与诊断。实验数据显示:所提算法模型在训练集和测试集中,故障分类准确率分别达99.8%和99.7%,特征提取效率也优于现有算法。Aiming at the problems of poor accuracy and low efficiency of existing fault feature extraction algorithms for rotating machinery,a mathematical model of fault feature extraction and diagnosis based on BSS-HHT is proposed.Firstly,based on wavelet soft threshold,the original fault signal is whitened to eliminate the second-order characteristics between fault signals.The BSS algorithm model is constructed and negative entropy is used as the objective function to separate effective fault signals from source signals.Based on SA algorithm to optimize the performance of global optimization model,HHT algorithm is introduced to do empirical mode decomposition of the output signal,which eliminates the end effect and realizes the location and diagnosis of fault feature points.The experimental data show that the fault classification accuracy of the proposed algorithm is 99.8%and 99.7%respectively in the training set and the test set,and the feature extraction efficiency is better than the existing algorithm.

关 键 词：盲源分离-希尔伯特黄变换 数学模型 旋转机械 经验模态分解 

分 类 号：TH164[机械工程—机械制造及自动化]