how to hold positioning vector to obtain maximum power?

Hi,

Currently my plot looks like this (the code is attached below):

where the red line represents an position vector as the power is increasing. What I would like to do is following (used MS paint for visualization) :

In other words, as the Gaussian profile reaches its maximum, the only way to keep maximum power all time by holding the position at the same place. And I am stuck with the coding here.

So far, my code looks like this:

x = -3:0.1:3;

norm = normpdf(x,0,1);

m = 0;
mPos = zeros(length(norm),1);

for i = 2:length(norm)
    if(norm(i)>norm(i-1))
        m = m + 0.1;
    else
        m = m - 0.1;
    end
    mPos(i) =  m;

end

plot(x, mPos, 'r')
hold on
plot(x, norm, 'b')
grid on
hold off

Answer 1

[Image Analyst]

Try this:

% Initialization steps.
clc;    % Clear the command window.
close all;  % Close all figures (except those of imtool.)
clear;  % Erase all existing variables. Or clearvars if you want.
workspace;  % Make sure the workspace panel is showing.
format long g;
format compact;
fontSize = 20;

x = -3:0.1:3;
numPoints = length(x);
norm = normpdf(x,0,1);

m = 0;
mPos = zeros(1, numPoints);
normMax = norm(1) * ones(1, numPoints);
mPosMax = zeros(1, numPoints);
for k = 2:length(norm)
    if(norm(k)>norm(k-1))
        m = m + 0.1;
    else
        m = m - 0.1;
    end
    mPos(k) =  m;
  if norm(k) >= normMax(k-1)
    normMax(k) = norm(k);
  else
    normMax(k) = normMax(k-1);
  end
  if mPos(k) >= mPosMax(k-1)
    mPosMax(k) = mPos(k);
  else
    mPosMax(k) = mPosMax(k-1);
  end
end
subplot(2, 1, 1);
plot(x, mPos, 'r-', 'LineWidth', 2)
hold on
plot(x, norm, 'b-', 'LineWidth', 2)
grid on
hold off
title('Original Signals', 'FontSize', fontSize);
xlabel('x', 'FontSize', fontSize);
ylabel('Original Signal Value', 'FontSize', fontSize);
legend('mPos', 'norm', 'Location', 'east');

subplot(2, 1, 2);
plot(x, mPosMax, 'r-', 'LineWidth', 2);
hold on;
plot(x, normMax, 'b-', 'LineWidth', 2);
grid on
hold off
title('Max (Peak Detection) Signals', 'FontSize', fontSize);
xlabel('x', 'FontSize', fontSize);
ylabel('Max Signal Value', 'FontSize', fontSize);
legend('mPos', 'norm', 'Location', 'east');

% Enlarge figure to full screen.
set(gcf, 'Units', 'Normalized', 'OuterPosition', [0, 0.04, 1, 0.96]);

[kat001]

Thank you very much for the help. So, I have tweaked the code a little so that it suits my task. This is what's been done:

%%Simulation of Gaussian profile
x = -3:0.0001:3;
numPoints = length(x);
norm = normpdf(x,0,0.3925);

%%Vector assignment
t = 0:1:100; %time
xi = 1;  % initial position of x
xPos = zeros(1, length(t)); % creates 100 columns containing zeros
normMax = norm(1) * ones(1, length(t)); % creates 100 columns containing norm(1)
xPosMax = zeros(1, length(t)); 
normT = zeros(1, length(t));

%init movement
xPos(1)=xi;
normT(1) = norm(find(abs(x-xi)<min(diff(x))/2)); % provide 'norm' value at 'xi'. There has to be a better to write it. Any suggestion?
xi=xi+0.1;
xPos(2)=xi;
normT(2) = norm(find(abs(x-xi)<min(diff(x))/2));

%%Power profile comparison & tracking 
for k = 3:length(t)
    if(normT(k-1)>normT(k-2))
        xi = xi + (xPos(k-1) - xPos(k-2));  % position increment x axis
    else
        xi =xi - (xPos(k-1) - xPos(k-2));   % position decrement x axis
    end

      xPos(k) =  xi;      % saving increment/decrement of position in xPos for each step
      normT(k) = norm(find(abs(x-xi)<min(diff(x))/2));   

end

%%Plot

plot(t, xPos, 'r-', 'LineWidth', 2);
ylabel('X Position', 'FontSize', fontSize);
hold on;
plot(t, normT, 'b-', 'LineWidth', 2);
grid on
hold off
ylabel('Gaussian', 'FontSize', fontSize);
title('Peak Power Detection', 'FontSize', fontSize);
xlabel('t', 'FontSize', fontSize);
legend('xPos', 'norm', 'Location', 'east');

btw, in 'init movement' section, I have written the following code, which was just a work around:

normT(1) = norm(find(abs(x-xi)<min(diff(x))/2));

Do you have a better suggestion?

Now, the major is question how exactly I can run this in Simulink.

I have generated a Gaussian profile in the code. But in Simulink, it is already given as an input. So I have to skip the power creation part in SImulink. However, the input power is a continuous signal. The output is still the same, i.e. the xPos is affecting the input power. How to proceed in Simulink? Is there a way to convert the entire code to simulink? Any idea is appreciated.

[Image Analyst]

I don't know what you're doing with the line of code you put in. I don't know what normT is supposed to represent - you forgot to tell me. All you said is you've written a line of code and then asked for suggestions about it. What can I say? If it does what you want, then fine. Otherwise say what you want to do.

I don't know Simulink - sorry.

[kat001]

So, Now this code seems to work for horizontal or x axis. However, I have used the same code in a new function for vertical or y axis.

I have set t = 0:1:5 for the x axis fucntion. And for y axis function, it is t2 = t(end):1:10. 't' is an input for the 'y axis' function. I also want to feed the same power from 'x axis'-function to 'y axis'-function, i.e. the same power value will be picked at t2(1) and will continue until t2 = 10.

I am facing the problem with writing the code for 'normT' which is the input to y axis the vector assignment section of the code. The rest of the code won't be changing so much other than variable name changes.

Any idea?

Answer 2

[Image Analyst]

Try movmax(). Write back if you can't figure it out.

[kat001]

Well, this is not really what I had in my mind.

What I was thinking of is that the position vector checks the power level all the time. From the first plot, you may see the Gaussian profile increases to maximum then decreases and so does the position vector.

What I would like to do is to hold the power level at maximum by controlling the position vector. Hence the second plot drawn in 'paint' to show what I had in my mind.