【信号增强】基于广义结构化收缩算法(GSSA)实现振动信号弱特征增强附matlab代码和复现论文

简介: 【信号增强】基于广义结构化收缩算法(GSSA)实现振动信号弱特征增强附matlab代码和复现论文

1 简介

Vibration signal analysis has become one of the important methods for machinery fault diagnosis. Extraction of weak fault features from vibration signals with heavy background noise remains a challenging problem. In this paper, we first introduce the idea of algorithm-aware sparsity-assisted methods for fault feature enhancement, which extends model-aware sparsity-assisted fault diagnosis and allows a more flexible and convenient algorithm design. In the framework of algorithm-aware methods, we define the generalized structured shrinkage operators and construct the generalized structured shrinkage algorithm (GSSA) to overcome the disadvantages of l1-norm regularization based fault feature enhancement methods. We then perform a series of simulation studies and two experimental cases to verify the effectiveness of the proposed method. Additionally, comparisons with model-aware methods, including basis pursuit denoising and windowed-group-lasso, and fast kurtogram further verify the advantages of GSSA for weak fault feature enhancement.

2 部分代码

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% clcclear allclose alladdpath(genpath(fileparts(mfilename('fullpath'))));rng('default')rng(19)  %% Figure initializationglobal FontSize FontName;Tstring =  'Time (s)'; Fstring = 'Frequency (Hz)';Astring =  'Amp (m/s^2)';FontSize = 11;   FontName = 'Times New Roman';MarkerSize = 4;  LineWidth = 1;%%FlagFigureAutoSave = 1;currentFolder = pwd;%% SimulationFs = 20480;N = 4096;mode = 'outer';% [Sig_Impulse , t] = MakeSignalBearing( Fs, N, 'outer');Sig_Impulse = QuasiPeiodicImpulseResponse_AM(N, Fs);t = (0 : N-1) / Fs;Sig_Cos = 0.5 * cos(2*pi*160*t') + 0.3 * cos(2*pi*320*t');Sigma = 0.6;Noise = Sigma * randn(N , 1);Sig_Combine = Sig_Cos + Sig_Impulse' + Noise;% Sig_Combine = Sig_Cos + Sig_Impulse + Noise;[ yf2, f2 ] = Dofft( Sig_Combine , Fs , 0);K = 10 : 10 : 1000;for i = 1 : length(K)    %% Setting the parameters    i    Q = 2;    r = 5;    J =10;    now = ComputeNow(N,Q,r,J,'radix2');    AH = @(Sig) tqwt_radix2(Sig, Q, r, J);       A = @(w) itqwt_radix2(w, Q, r , N);    lam = 1.0 * now;    rho = 1;    %% method1 : Generalized Structured Shrinkage    K1 = K(i);     Method1.Name = 'WGL';    Method1.Initial_Size = 5;    Method1.SubName = 'MC';    Method1.gamma = 2;    Method1.window = 'gausswin';    z1 = IterGSS(Sig_Combine, A, AH, lam, rho, K1, Method1);    %% method2 : TQWT-L1 based    Method2.Name = 'L1';    K2 = K(i);       z2 = IterGSS(Sig_Combine, A, AH, lam, rho, K2, Method2);    %% method3 : Neighbor thresholding    params.Q = 2;    params.r = 5;    params.J =10;    K3 = K(i);       z3 = TQWTDe( Sig_Combine, params , 'nc', K3);    %% method4 : Structured Shrinkage    Method3.Name = 'WGL';    Method3.Initial_Size = 5;    Method3.SubName = 'L1';    Method3.window = 'gausswin';    K4 = K(i);       z4 = IterGSS(Sig_Combine, A, AH, lam, rho, K4, Method3);     z1 = real(A(z1));    z2 = real(A(z2));    % z3 = real(A(z3));    z4 = real(A(z4));    %% Calculate RMSE    GST_RMSE(i) = RMSE(z1, Sig_Impulse);    L1_RMSE(i) = RMSE(z2, Sig_Impulse);    NC_RMSE(i) = RMSE(z3', Sig_Impulse);    ST_RMSE(i) = RMSE(z4, Sig_Impulse);endsave('Harmonic_Inference_Best_K.mat', 'K', 'GST_RMSE', 'L1_RMSE', 'NC_RMSE', 'ST_RMSE')

3 仿真结果

4 参考文献

[1] Zhao Z ,  Wang S ,  Xu W , et al. Sparsity-Assisted Fault Feature Enhancement: Algorithm-Aware Versus Model-Aware[J]. IEEE Transactions on Instrumentation and Measurement, 2020, PP(99):1-1.

博主简介:擅长智能优化算法、神经网络预测、信号处理、元胞自动机、图像处理、路径规划、无人机等多种领域的Matlab仿真,相关matlab代码问题可私信交流。

部分理论引用网络文献,若有侵权联系博主删除。


相关文章
|
30天前
|
机器学习/深度学习 人工智能 算法
【图像版权】论文阅读:CRMW 图像隐写术+压缩算法
【图像版权】论文阅读:CRMW 图像隐写术+压缩算法
19 0
|
30天前
|
数据安全/隐私保护
地震波功率谱密度函数、功率谱密度曲线,反应谱转功率谱,matlab代码
地震波格式转换、时程转换、峰值调整、规范反应谱、计算反应谱、计算持时、生成人工波、时频域转换、数据滤波、基线校正、Arias截波、傅里叶变换、耐震时程曲线、脉冲波合成与提取、三联反应谱、地震动参数、延性反应谱、地震波缩尺、功率谱密度
|
30天前
|
数据安全/隐私保护
耐震时程曲线,matlab代码,自定义反应谱与地震波,优化源代码,地震波耐震时程曲线
地震波格式转换、时程转换、峰值调整、规范反应谱、计算反应谱、计算持时、生成人工波、时频域转换、数据滤波、基线校正、Arias截波、傅里叶变换、耐震时程曲线、脉冲波合成与提取、三联反应谱、地震动参数、延性反应谱、地震波缩尺、功率谱密度
基于混合整数规划的微网储能电池容量规划(matlab代码)
基于混合整数规划的微网储能电池容量规划(matlab代码)
|
30天前
|
算法 调度
面向配电网韧性提升的移动储能预布局与动态调度策略(matlab代码)
面向配电网韧性提升的移动储能预布局与动态调度策略(matlab代码)
|
30天前
|
算法 调度
含多微网租赁共享储能的配电网博弈优化调度(含matlab代码)
含多微网租赁共享储能的配电网博弈优化调度(含matlab代码)
|
30天前
|
运维 算法
基于改进遗传算法的配电网故障定位(matlab代码)
基于改进遗传算法的配电网故障定位(matlab代码)
|
30天前
|
Serverless
基于Logistic函数的负荷需求响应(matlab代码)
基于Logistic函数的负荷需求响应(matlab代码)
|
1天前
|
算法 安全
基于龙格库塔算法的SIR病毒扩散预测matlab仿真
该程序使用龙格库塔算法实现SIR模型预测病毒扩散,输出易感、感染和康复人群曲线。在MATLAB2022a中运行显示预测结果。核心代码设置时间区间、参数,并定义微分方程组,通过Runge-Kutta方法求解。SIR模型描述三类人群动态变化,常微分方程组刻画相互转化。模型用于预测疫情趋势,支持公共卫生决策,但也存在局限性,如忽略空间结构和人口异质性。
|
1天前
|
机器学习/深度学习 监控 算法
基于yolov2深度学习网络的昆虫检测算法matlab仿真,并输出昆虫数量和大小判决
YOLOv2算法应用于昆虫检测,提供实时高效的方法识别和定位图像中的昆虫,提升检测精度。核心是统一检测网络,预测边界框和类别概率。通过预测框尺寸估算昆虫大小,适用于农业监控、生态研究等领域。在matlab2022A上运行,经过关键升级,如采用更优网络结构和损失函数,保证速度与精度。持续优化可增强对不同昆虫的检测能力。![image.png](https://ucc.alicdn.com/pic/developer-ecology/3tnl7rfrqv6tw_e760ff6682a3420cb4e24d1e48b10a2e.png)