👨🎓個人首頁:研學社的部落格
💥💥💞💞歡迎來到本部落格❤️❤️💥💥
🏆部落客優勢:🌞🌞🌞部落格内容盡量做到思維缜密,邏輯清晰,為了友善讀者。
⛳️座右銘:行百裡者,半于九十。
📋📋📋本文目錄如下:🎁🎁🎁
目錄
💥1 概述
📚2 運作結果
🎉3 參考文獻
🌈4 Matlab代碼實作
💥1 概述
RLMD 是一種改進的局部均值分解,由一組優化政策提供支援。優化政策可以處理LMD中的邊界條件、包絡估計和篩選停止準則。它同時從混合信号中提取一組單分量信号(稱為乘積函數)及其相關的解調信号(即AM信号和FM信号),與其他自适應信号處理方法(如EMD)相比,這是最吸引人的特征。RLMD可用于時頻分析。
📚2 運作結果
部分代碼:
[x, display, stop_thre, sifting_stopping_mode, max_iter, max_pfs, smooth_mode,...
ma_span, ma_iter_mode, extd_r, x_energy, pfs, ams, fms, iterNum, fvs]...
= initial(x,varargin{:});
% Initialize main loop
i = 0;
xs = x; % copy x to xs for sifting process, reserve original input as x.
nx = length(x);
while i < max_pfs && ~stoplmd(xs, x_energy) % outer loop for PF selection
i = i+1;
% initialize variables used in PF sifting loop
a_i = ones(1,nx);
s_j = zeros(max_iter,nx);
a_ij = zeros(max_iter, nx);
% PF sifting iteration loop
j = 0;
stop_sifting = 0;
s = xs;
while j < max_iter && ~stop_sifting % inner loop for sifting process
j = j+1;
[m_j, a_j, n_extr] = lmd_mean_amp(s, smooth_mode, ma_span, ma_iter_mode,...
extd_r);
% force to stop iter if number of extrema of s is smaller than 3.
if n_extr < 3
break;
end
h_j = s-m_j; % remove mean.
s = h_j./a_j; % demodulate amplitude.
a_i = a_i .* a_j; % mulitiply every ai
a_ij(j, :) = a_i;
s_j(j, :) = s;
[stop_sifting,fvs(i,:)] = is_sifting_stopping(a_j, j, fvs(i,:), sifting_stopping_mode, stop_thre);
end % sift iteration loop
switch sifting_stopping_mode
case {'liu'}
[~, opt0] = min(fvs(i,1:j)); % ***Critical Step***
opt_IterNum = min(j, opt0); % in case iteration stop for n_extr<3
% opt_IterNum = min(j-2, opt0);
otherwise
error('No specifications for sifting_stopping_mode.');
end
ams(i, :) = a_ij(opt_IterNum, :); % save each amplitude modulation function in ams.
fms(i, :) = s_j(opt_IterNum, :); % save each pure frequency modulation function in fms.
pfs(i, :) = ams(i, :).*fms(i, :); % gain Product Funcion.
xs = xs-pfs(i, :); % remove PF just obtained from input signal;
iterNum(i) = opt_IterNum; % record the iteration times taken by of each PF sifing.
end % main loop
pfs(i+1, :) = xs; % save residual in the last row of PFs matrix.
ams(i+1:end,:) = []; fms(i+1:end,:) = []; pfs(i+2:end,:) = []; fvs(i+1:end,:) = [];
ort = io(x, pfs);
% Output visualization
if display == 1
lmdplot(pfs, ams, fms, smooth_mode);
end
end
%--------------------------- built-in functions ---------------------------
% initialize signal and options
function [x, display, stop_thre, sifting_stopping_mode, max_iter, max_pfs, smooth_mode,...
ma_span, ma_iter_mode, extd_r, x_energy, pfs, ams, fms, iterNum, fvs]...
= initial(x,varargin)