How to find peak coordinates of a signal within a specific range?

Question

Anonymous45 2017-3-26

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/332022-how-to-find-peak-coordinates-of-a-signal-within-a-specific-range

评论： Anonymous45 2017-4-3

The code below is of two signals. I am trying to take the peak/max value of "signal_1", then create a range: (max_value_x_component - 0.1e+9) to (max_value_x_component + 0.1e+9).

Then for signal_5 I want to find the x and y coordinates of the peak/max value between the rang: (max_value_x_component - 0.1e+9) to (max_value_x_component + 0.1e+9).

I am looking to run this in a loop, so I need to actually find the peak values within the range through coding and not just by looking at a graph.

The code below will produce two figures with two graphs on each figure as seen in image 1 below. The graph that I am interested in analyzing is the second two graphs labeled "Graph of signal Fourier Transform". When looking at the graph you only see two large peaks/spikes one on the left near the origin and one on the far right. You need zoom into the left peak to see the waveform being analyzed as seen in image 2 below.

 h = 1.00E-12;
t = 0:h:40E-9;
A = 1;
A2 = 1.1111111;
A3 = 2.2222222;
A4 = 10;
A5 = 20;
f=3E9;
nfft = 2^(nextpow2(length(t))+5);
 %Graphed signal_1
signal = cos(2*pi*f*t).*(A.*(heaviside(t)-heaviside(t-39E-9)));
snrdB = 0;
noise = 10^(-snrdB/20)*randn(size(signal));
sn = noise+signal;
y1 = fft(sn,nfft);
m1 = abs(y1);
er1 = (0:length(y1)-1)*(1/h)/length(y1);
figure (1)
subplot(2,1,1)
plot(t,sn)
title('Graph of signal_1')
xlabel('Time') % x-axis label
ylabel('Amplitude') % y-axis label
subplot(2,1,2)
plot(er1,m1)
grid on
title('Graph of signal_1 Fourier Transform')
xlabel('Frequency') % x-axis label
ylabel('Amplitude') % y-axis label
 %Graphed signal_5
signal_5 = [cos(2*pi*f*t).*(A5.*(heaviside(t)-heaviside(t-1E-9)))]+[cos(2*pi*f*t).*(A5.*(heaviside(t-4E-9)-heaviside(t-5E-9)))];
snrdB = 0;
noise = 10^(-snrdB/20)*randn(size(signal_5));
s5n = noise+signal_5;
y5 = fft(s5n,nfft);
m5 = abs(y5);
er5 = (0:length(y5)-1)*(1/h)/length(y5);
indexmax = find(max(m5) == m5)
xmax = er5(indexmax)
ymax = m5(indexmax)
figure (5)
subplot(2,1,1)
plot(t,s5n)
title('Graph of signal_5')
xlabel('Time') % x-axis label
ylabel('Amplitude') % y-axis label
subplot(2,1,2)
plot(er5,m5)
grid on
title('Graph of signal_5 Fourier Transform')
xlabel('Frequency') % x-axis label
ylabel('Amplitude') % y-axis label

Image 1. This is the figure that will be produced.

Image 2. This is the zoomed in waveform for the Four Transform graph of signal 3.

4 个评论
显示 2更早的评论隐藏 2更早的评论

Anonymous45 2017-4-3

在 MATLAB Online 中打开

For example this is where I am with my code:

 h = 1.00E-12;
t = 0:h:40E-9;
A = 1;
A2 = 1.1111111;
A3 = 2.2222222;
A4 = 5;
A5 = 20;
f=3E9;
nfft = 2^(nextpow2(length(t))+5);
 signal = cos(2*pi*f*t).*(A.*(heaviside(t)-heaviside(t-39E-9)));
data1 = zeros(500,1);
index1 = 0;
data1_1 = zeros(16,1);
index1_1 = 0;
for k = 0:1:30
    for i = 1:1000
        snrdB = k;
        noise = 10^(-snrdB/20)*randn(size(signal));
        sn = noise+signal;
        y1 = fft(sn,nfft);
        m1 = abs(y1);
        er1 = (0:length(y1)-1)*(1/h)/length(y1); 
        indexmax1 = find(max(m1) == m1);
        xmax1 = er1(indexmax1);
        ymax1 = m1(indexmax1);
        index1 = index1 + 1;
        data1(index1, 1:1) = xmax1(1,1);
    end
    index1 = 0;
    V = std(data1);
    index1_1 = index1_1 + 1;
    data1_1(index1_1, 1:1) = V(1,1);
end
 %Graphed signal_5
signal_5 = [cos(2*pi*f*t).*(A5.*(heaviside(t)-heaviside(t-1E-9)))]+[cos(2*pi*f*t).*(A5.*(heaviside(t-4E-9)-heaviside(t-5E-9)))];
data5 = zeros(500,1);
index5 = 0;
data5_2 = zeros(30,1);
index5_2 = 0;
for k = 0:1:30
    for i = 1:1000
        snrdB = k;
        noise5 = 10.^(-snrdB./20).*randn(size(signal_5));
        s5n = noise5+signal_5;
        y5 = fft(s5n,nfft);
        m5 = abs(y5);
        er5 = (0:length(y5)-1)*(1/h)/length(y5);
        indexmax5 = find(max(m5) == m5);
        xmax5 = er5(indexmax5);
        ymax5 = m5(indexmax5);
        index5 = index5 + 1;
        data5(index5, 1:1) = xmax5(1,1);
    end
    index5 = 0;
    V = std(data5);
    index5_2 = index5_2 + 1;
    data5_2(index5_2, 1:1) = V(1,1);
end
figure(1)
plot (0:1:30, data1_1, 0:1:30, data5_2)

As you can see I am finding the max(m1) and indexing xmax1(1,1). Now I want to take that index value and make it into a boundary for the second signal. Instead of finding the max value of signal 5 I want to find the max value of signal 5 in between the range xmax1(1,1)-5 to xmax1(1,1)+5.

I hope this helps explain more. If you have any questions please don't hesitate to ask.

Anonymous45 2017-4-3