💥💥💥💞💞💞歡迎來到本部落格❤️❤️❤️💥💥💥
📝目前更新：🌟🌟🌟智能優化算法、神經網絡預測、信号處理、元胞自動機、圖像處理、路徑規劃、無人機等多種領域的Matlab仿真。
🎉🎉歡迎您的到來🎉🎉
⛅⛅⛅ 📃CSDN首頁：Matlab科研室🌈🌈🌈
📚📚📚📋所有代碼目錄見微信公衆号：天天Matlab👨•💻👨•💻👨•💻 1 内容介紹
為準确提取軸承故障特征資訊,提出以樣本熵為綜合目标函數的變分模态分解(variational mode decomposition,VMD)參數優化方法,并改進了診斷流程實作了無需指定參數優化範圍的自适應參數優化算法.通過麻雀算法對綜合目标函數最小值進行搜尋,以确定模态分量個數及懲罰參數的最佳組合.原始故障信号經最佳參數組合下的VMD方法分解為若幹個本征模态函數,選擇最小綜合目标函數值對應的模态分量進行包絡解調分析,進而通過模态分量的包絡譜判斷軸承故障類型.通過實測故障信号分析表明,該方法能夠從噪聲幹擾中有效提取到早期故障信号的微弱故障特征,實作了軸承故障類型的準确判定,驗證了該方法的有效性.
2 仿真代碼
function SSA_VMD(varargin)
SearchAgents_no = cell2mat(varargin(1));
Max_iteration = cell2mat(varargin(2));
dim = cell2mat(varargin(3));
lb = cell2mat(varargin(4));
ub = cell2mat(varargin(5));
try
imf_write = cell2mat(varargin(6));
catch
imf_write = 0;
end
t = evalin('base', 't');
s = evalin('base', 's');
fs = evalin('base', 'fs');
ST = 0.6;
PD = 0.7;
SD = 0.2;
PDNumber = round(SearchAgents_no * PD);
SDNumber = round(SearchAgents_no - SearchAgents_no * SD);
if (max(size(ub)) == 1)
ub = ub .* ones(1, dim);
lb = lb .* ones(1, dim);
end
X0 = initialization(SearchAgents_no, dim, ub, lb);
X = X0;
fitness = zeros(1, SearchAgents_no);
for i = 1:SearchAgents_no
fitness(i) = fun(X(i, :), s);
end
[fitness, index] = sort(fitness);
BestF = fitness(1);
WorstF = fitness(end);
GBestF = fitness(1);
for i = 1:SearchAgents_no
X(i, :) = X0(index(i), :);
end
Convergence_curve = zeros(Max_iteration, 1);
GBestX = X(1, :);
X_new = X;
gbest_array = zeros(Max_iteration, 2);
for i = 1:Max_iteration
BestF = fitness(1);
WorstF = fitness(end);
R2 = rand(1);
for j = 1:PDNumber
if (R2 < ST)
X_new(j, :) = X(j, :) .* exp(- j / (rand(1) * Max_iteration));
else
X_new(j, :) = X(j, :) + randn() * ones(1, dim);
end
end
for j = PDNumber + 1:SearchAgents_no
if (j > (SearchAgents_no - PDNumber) / 2 + PDNumber)
X_new(j, :) = randn() .* exp((X(end, :) - X(j, :)) / j^2);
else
A = ones(1, dim);
for a = 1:dim
if (rand() > 0.5)
A(a) =- 1;
end
end
AA = A' * inv(A * A');
X_new(j, :) = X(1, :) + abs(X(j, :) - X(1, :)) .* AA';
end
end
Temp = randperm(SearchAgents_no);
SDchooseIndex = Temp(1:SDNumber);
for j = 1:SDNumber
if (fitness(SDchooseIndex(j)) > BestF)
X_new(SDchooseIndex(j), :) = X(1, :) + randn() .* abs(X(SDchooseIndex(j), :) - X(1, :));
elseif (fitness(SDchooseIndex(j)) == BestF)
K = 2 * rand() - 1;
X_new(SDchooseIndex(j), :) = X(SDchooseIndex(j), :) + K .* (abs(X(SDchooseIndex(j), :) - X(end, :)) ./ (fitness(SDchooseIndex(j)) - fitness(end) + 10^ - 8));
end
end
for j = 1:SearchAgents_no
for a = 1:dim
if (X_new(j, a) > ub(a))
X_new(j, a) = ub(a);
end
if (X_new(j, a) < lb(a))
X_new(j, a) = lb(a);
end
end
end
for j = 1:SearchAgents_no
fitness_new(j) = fun(X_new(j, :), s);
end
for j = 1:SearchAgents_no
if (fitness_new(j) < GBestF)
GBestF = fitness_new(j);
GBestX = X_new(j, :);
end
end
X = X_new;
fitness = fitness_new;
[fitness, index] = sort(fitness);
BestF = fitness(1);
WorstF = fitness(end);
for j = 1:SearchAgents_no
X(j, :) = X(index(j), :);
end
disp(['current iteration is: ', num2str(i), ', best fitness is: ', num2str(GBestF)])
gbest_array(i, :) = GBestX;
Convergence_curve(i) = GBestF;
end
Best_pos = GBestX;
flag = 1;
figure(1)
plot(Convergence_curve, 'k-.o', 'linewidth', 1)
legend('最佳适應度')
grid on
xlabel('進化代數')
ylabel('包絡熵')
title('SSA進化曲線')
figure(2)
plot(gbest_array(:, 1), 'ks-.', 'linewidth', 1)
grid on
xlabel('進化代數')
ylabel('懲罰因子')
title('懲罰因子的優化過程曲線')
figure(3)
plot(round(gbest_array(:, 2)), 'kv-.', 'linewidth', 1)
grid on
xlabel('進化代數')
ylabel('分解的模态數')
title('分解模态數的優化過程曲線')
alpha = Best_pos(1);
K = round(Best_pos(2));
tau = 0;
DC = 0;
init = 1;
tol = 1e-7;
[u, u_hat, omega] = VMD(s, alpha, tau, K, DC, init, tol);
[m, ~] = size(u);
m = m + 1;
figure(4)
for i = 1:m - 1
subplot(m, 1, i)
plot(t, u(i, :), 'b-', 'linewidth', 1)
ylabel(['IMF', num2str(i)]);
end
res = s' - sum(u, 1);
subplot(m, 1, m)
plot(t, res, 'b-', 'linewidth', 1)
ylabel('Res');
xlabel('t/s')
figure(5)
for i = 1:m - 1
subplot(m, 1, i)
[cc, y_f] = plot_fft(u(i, :), fs, 1);
plot(y_f, cc, 'b', 'LineWIdth', 1);
ylabel(['FFT of IMF', num2str(i)]);
end
subplot(m, 1, m)
[cc, y_f] = plot_fft(res, fs, 1);
plot(y_f, cc, 'b', 'LineWIdth', 1);
ylabel('FFT of Res');
xlabel('f/Hz')
fig_text(flag)
% figure(6)
%
% imshow(a. personal_id);
% title( 'Color', 'red', 'FontSize', 16,...
% 'HorizontalAlignment', 'center',...
% 'Interpreter', 'none')
%
% fprintf_result(Best_pos, s);
%
% if imf_write == 1
% xlswrite('imf結果.xlsx', u', 'u', 'A1');
% xlswrite('imf結果.xlsx', u_hat, 'uhat', 'A1');
% xlswrite('imf結果.xlsx', omega, 'omega', 'A1');
% end
3 運作結果
編輯
4 參考文獻
[1]胡鴻志, 覃暢, 管芳,等. 基于麻雀搜尋算法優化支援向量機的刀具磨損識别[J]. 科學技術與工程, 2021, 21(25):7.