Finding max force values in macro with 100 cycles, in a specific interval

I have a actuator that moves from postition 0 to postion 5000 and back to postion 0 again, this cycle is repeated 100 times

i want to find the Max Force value for each cycle in a specific position interval.

Here is my code, i'm a noob at matlab:

filename='Cycle 0-100test.csv';
data = dlmread(filename);  
Force = data(109:end,2);
Position =data(109:end,4);
d = [1;diff(Position)];
Position(d<0) = -Position(d<0);
1interval=find(Position>3000 & Position<4400);
2interval=find(Position>-5000 & Position<-3000);
Firstpeak=max(Force(1interval)); 
Firstpeak2=min(Force(2interval));

I know i have to make some sort of loop. But i cant figure out how to do it :/

[dpb]

Be helpful to see plot of what the cycle looks like...I'd guess probably findpeaks and return the optional location of the peaks will solve your problem almost directly.

BTW, the variables 1interval, etc., are not valid Matlab variable names; can't begin with a number.

And, switching to interval1, etc., is NOT the correct answer, either! :)

The if going to do by a search and divide technique is, as you note, use a loop but use arrays to keep track of multiples of the same thing.

But, even more so, with Matlab one wants to try to think of how to use the matrix operations built in to the syntax and likely can avoid loops altoghether, or at least in large part.

[Morten Nielsen]

Thank you

Unfountenly I do not have access to "findpeaks" toolbox

Picture of 2 cycles as a function of time.

[Star Strider]

If you have R2017b or later, you can use the islocalmax function to identify the peaks.

[Morten Nielsen]

Yes, i know islocalmax, but im only interested in finding these peaks (marked with a circle)

I need to find them for each cycle (100 cycles) and their respective index number

sorry I have only been using matlab for a couple of days.

[dpb]

I'd do something like compute the overall mean and then find exceedance of it by a variable factor; the magnitude of the factor being determined empirically to return only peaks of some preset distance apart. If the sample waveform is representative, this should be reasonably robust with the possible exception of eliminating the overshoot recovery peak that looks like can get to be reasonably close to the desired pre-peak of interest.

The undershoot peak looks to be quite simple to isolate that way on the negative side as it is quite sizable and noise free.

You can't get access to the Signal Processing Toolbox? Is a real hamstring if going to be doing much signal processing...