清空筛选条件
清空筛选条件

Estimate Power Spectral Density (PSD)

29 次查看（过去 30 天）
Gisela Clemente
Gisela Clemente 2024-4-6
评论： Gisela Clemente 2024-4-6
Hello! I want to estimate the PSD of a signal from the CWT coefficients. How can I do it? Can you provide me with some algorithm to do this? I have Matlab R2016a installed. Thank you so much!
  2 个评论
Manikanta Aditya
Manikanta Aditya 2024-4-6
Hey,
Check this example code, I feel this can help.
% Assuming 'cwtCoeffs' is your matrix of CWT coefficients and 'Fs' is your sampling frequency
% Compute the absolute square of the CWT coefficients
cwtCoeffsSquared = abs(cwtCoeffs).^2;
% Compute the time-averaged power, which gives the scalogram
scalogram = mean(cwtCoeffsSquared, 2);
% Normalize the scalogram to get the relative energy distribution as a function of frequency, which is the PSD
psd = scalogram / sum(scalogram);
% Plot the PSD
frequencies = Fs * (0:(length(psd)/2))/length(psd);
plot(frequencies, 10*log10(psd(1:length(frequencies))));
xlabel('Frequency (Hz)');
ylabel('Power Spectral Density (dB/Hz)');
Gisela Clemente
Gisela Clemente 2024-4-6
Thanks for your answer. But when I compare your CWT-PSD with the clasical FFT-PSD, they aren`t similar. It is the code:
% Constants and Signal Definition
Fs = 1000; % Sampling frequency in Hz
duracion = 1; % Duration in seconds
N = Fs * duracion; % Total number of samples
delta = zeros(1, N); % Initialize delta function
delta(round(N/2)) = sqrt(N); % Impulse in the center
scales=1:16;
cwtCoeffs = cwt(delta, scales, 'db6'); % CWT
% Compute the absolute square of the CWT coefficients
cwtCoeffsSquared = abs(cwtCoeffs).^2;
% Compute the time-averaged power, which gives the scalogram
scalogram = mean(cwtCoeffsSquared, 2);
% Normalize the scalogram to get the relative energy distribution as a function of frequency, which is the PSD
psd = scalogram / sum(scalogram);
% Plot the PSD
frequencies = Fs * (0:(length(psd)/2))/length(psd);
plot(frequencies, 10*log10((psd(1:length(frequencies)))));
xlabel('Frequency (Hz)');
ylabel('Power Spectral Density (dB/Hz)');
% FFT-based PSD
X = fft(delta);
Pxx1_delta = 2*abs(X(1:N/2)).^2/(N*Fs); % PSD computation
frecuencias = linspace(0, Fs/2, N/2); % Frequency vector for FFT
hold on
plot(frecuencias, 10*log10(Pxx1_delta))

请先登录，再进行评论。

请先登录，再回答此问题。

回答（0 个）

类别

在 Help Center 和 File Exchange 中查找有关 Parametric Spectral Estimation 的更多信息

标签

产品


版本

R2016a

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by