I have a problem in performing a loop calculation

I have this code which reads data from several excel files and extracts them into an array and performs a calculation on them then presents the results in a table, the problem now is that I need this calculation to be performed for for every 2000 data points (from 0.005 to 10 sec, note that the time vector in the excel file is from 0.005 to 81.92 with a step of 0.005). However, from my basic understanding, MATLAB uses double precision and there is a floating point so maybe using machine eps is a good approach instead of logical statements, the issue is that I have multiple files and I need the code to loop through each file and perform the below calculation plus and extra interval-by-interval calculation (the same calculation but for each interval instead of the total set of data) and store the results for the total and the intervals then presnt them in a table (so loop the files, then loop the vectors in the array to establish the intervals). I have had some AI sugeestions but I was not abble to determine how correct were they. Help is appreciated. I attached 3 data files for reference but I have like 400 files.

%Prompt the user to select a folder
folder_path = uigetdir('', 'Select the folder');

%Get a list of files in the folder
file_list = dir(fullfile(folder_path, '*.xlsx'));

%Initialize the output table
output_table = {};

%Loop through each file
for i = 1:numel(file_list)
    % Get the file name
    file_name = file_list(i).name;
    
    %Total Torque RMS Calculation
    
    %Load the data
    tq = readmatrix(fullfile(folder_path, file_name));
    %TRUE if values change (down to machine eps)
    d = [true; diff(tq(:, 2)) ~= 0];
    %Torque values at change points
    TQ = tq(d, 2);
    %Location of changes in logical vector
    ix = find([d(:); true]);
    %Difference in locations (number of occurrences)
    N = diff(ix);
    %Calculating the torque rms by groups weighted by time
    %Removing roundoff in dt calculation to get the overall dt
    dt = mean(diff(tq(:, 1)));
    %Total Time
    t_tot = dt * (size(tq, 1));
    %Calculate the total tq_rms
    tq_rms = sqrt(sum(TQ.^2 .* (N - 1) * dt) / t_tot);
    newRow = {tq_rms};
    %Add the newRow to the output_table
    output_table = [output_table; newRow];
    
end

column_name = {'TQ_RMS'};
%Convert the output table to a table
output_table = cell2table(output_table, 'VariableNames', column_name);
%Display the results table
disp(output_table);

Answer 1

[Star Strider]

I cannot understand exactly what you want to do or what the logic vector ‘d’ does. If you simply want to get the RMS value of the torque signal every 10 seconds, that is actually straightforward.

Try this —

% Files = dir(pwd)
Files = dir('*.xlsx');
Files.name

ans = 'H5T06Z010RPM.xlsx'

ans = 'H5T06Z020RPM.xlsx'

ans = 'H5T06Z510RPM.xlsx'

for k = 1:numel(Files)
    T1 = readtable(Files(k).name, 'VariableNamingRule','preserve');

    VN = T1.Properties.VariableNames;
    t = T1{:,1};
    t_bfr = buffer(t, 2000);
    tq = T1{:,2};
    tq_bfr = buffer(tq, 2000);
    
    tq_rms = rms(tq_bfr);
    
    
    figure
    yyaxis left
    plot(T1{:,1}, T1{:,2})
    hold on
    plot(t_bfr, (ones(2000,1)*tq_rms))
    hold off
    ylabel(VN{2})
    yyaxis right
    plot(T1{:,1}, T1{:,end})
    ylabel(VN{3})
    grid
    xlabel(VN{1})
    xlim([min(t) max(t)])
    title(Files(k).name)
    
    timeseg = [t_bfr(1,:); t_bfr(end,:)].';
    DisplayTable = table(timeseg(:,1), timeseg(:,2), tq_rms(:), 'VariableNames',{'Start  Time','End Time','Torque RMS'})
    Results{k,:} = DisplayTable;
end

DisplayTable = 9×3 table
    Start  Time    End Time    Torque RMS
    ___________    ________    __________

       0.005          10        0.36623  
      10.005          20        0.41522  
      20.005          30        0.38779  
      30.005          40        0.38848  
      40.005          50        0.41168  
      50.005          60        0.38471  
      60.005          70        0.38019  
      70.005          80        0.41781  
      80.005           0        0.14823  

DisplayTable = 9×3 table
    Start  Time    End Time    Torque RMS
    ___________    ________    __________

       0.005          10        0.51982  
      10.005          20        0.47605  
      20.005          30         0.5089  
      30.005          40        0.48032  
      40.005          50        0.50803  
      50.005          60        0.51684  
      60.005          70         0.5473  
      70.005          80        0.48055  
      80.005           0        0.25016  

DisplayTable = 9×3 table
    Start  Time    End Time    Torque RMS
    ___________    ________    __________

       0.005          10         0.2419  
      10.005          20         0.3432  
      20.005          30        0.32467  
      30.005          40        0.30427  
      40.005          50        0.32322  
      50.005          60        0.32414  
      60.005          70        0.27453  
      70.005          80          0.265  
      80.005           0        0.11693  

.

[Star Strider]

I do not understand what you are doing, and the code does not really enlighten me in that respect. I do not see anything that could be considered a ‘change point’ in the torque values. They are quite noisy (with what I expect would be a low signal-to-noise ratio were we to calculate it) that appears to be due to the relatively low resolution of the ADC with a signal of quite small amplitude.

With respect to RMS, it is calculated as the square root of the mean of the square of the signal in a given segment of the signal or the entire signal, depending on what is desired. In my code, I use the Signal Processing Toolbox buffer function to divide the torque signal up into columns of 2000 samples (10 seconds) each, and then take the RMS value of each segment. I believe that this is what you want to do, and this way is significantly more efficient. (I did the same buffer operation with the time vector, both to plot the RMS values of the signal and to create the start and end times for tthe ‘Results’ table.)

The torque vectors are plotted in blue in these plots, with the RMS value being the interrupted blue line near the top of each plot, referenced to the left yyaxis. (The plot algorithm chooses the line style for each segment.)

[ABDALLA AL KHALEDI]

@Star Strider thank you for your input, don't worry about the data. I am trying to improve and your input was helpful.

[Star Strider]

As always, my pleasure!

Thank you!