How to calculate mean of this signal in the image below.

How to calculate Black waveform from brown waveform.

[jonas]

Looks like a moving average filter. Impossible to know without having access to the data. Can you upload?

[Dimitris Kalogiros]

Are you using "fixed point arithetic" ? From what you are showing us, it seems you have "wrapping around" phenomena

[Hammad]

This is an image from a research paper I was studying. I have data similar to that in Brown waveform but I am not able to figure out the black output.

[jonas]

Give some sample data so we do not have to guess. If its a scientific paper, then I bet they define the mean either in the figure caption or the text.

[Dimitris Kalogiros]

Does your paper talk about synchronization or PLLs or something like that?

[Hammad]

Title of paper: A Simulation Study of Crazy-PSO Controller For Direct Matrix Converter. I thinks its done by moving average.

Answer 1

[Dimitris Kalogiros]

I tried to generate your input data.

If you set noise_amp=0, you will observe the influence of moving average on your useful data.

% input signal generation
t=0:0.001:40;
signal_amp=1;
noise_amp=2; 
clip_amp=4;

x=signal_amp*sin(2*pi*0.25*t);
x=x+noise_amp*randn(size(x));
x(x>clip_amp)=clip_amp;
x(x<-clip_amp)=-clip_amp;

% moving average
y=zeros(size(x));
y(1)=x(1);
gain=0.001;
for n=2:length(x)
    y(n)=(y(n-1)+gain*x(n))/(1+gain);
end

% compare signals
figure; 
plot(t,x,'-r','color',[1 .4 .2]); hold on; 
plot(t,y,'-k');
zoom on; grid on;
xlabel('time')
legend('input data', 'moving average');