d = dir('*.csv');
filenames = {};
index = 0;
k = 0;
for i = 1:length(d)
% Using regexp with lookaround on the file name to find
% a sequence of digits preceeded by '_' and followed by '.':
no = regexp(d(i).name,'(?<=_)\d*(?=\.)','match');
if ~isempty(no)
k = k+1;
filenames{k} = d(i).name;
index(k) = str2num(no{end});
end
end
% filenames contains all filenames of the correct form
% index contains the numeric value
% Now sort:
[~,ix] = sort(index);
filenames = filenames(ix);
% You can now process files in the correct order:
for i = 1:length(index)
disp(filenames{i}); % or whatever...
end
