I'm fully aware that the code below does not quality as an appropriate teaching example - I'm simply a fan of a little bit of bit magic when it comes to string comparison.

#!/usr/bin/env perl
use warnings;
use strict;

my $q = "UGGAGUGUGACAAUGGUGUUUG";
my @r = qw(
    UAUGUACCGACCUUAUUCUCCU
    AUGUACCGACCUUAUUCUCCUG
    UGUACCGACCUUAUUCUCCUGU
);

foreach my $r (@r) {
    my $xb = $r ^ $q; # compare
    $xb =~ tr/\0\377/\0\377/c; # bitmask with mm=1, m=0
    my $xs = $r & $xb; # mismatches
    $xs =~ tr/\0/ /; # add gaps

    print "ref: ", $r,"\n";
    print "qry: ", $q,"\n";
    print "mis: ", $xs,"\n\n";
}

produces:

ref: UAUGUACCGACCUUAUUCUCCU
qry: UGGAGUGUGACAAUGGUGUUUG
mis:  AUGUACC   CU AU C CCU

ref: AUGUACCGACCUUAUUCUCCUG
qry: UGGAGUGUGACAAUGGUGUUUG
mis: AU UACCGAC UUAUUCUCC  

ref: UGUACCGACCUUAUUCUCCUGU
qry: UGGAGUGUGACAAUGGUGUUUG
mis:   U CC ACCUU  UC CC GU

[EDIT] Not a working solution - see comments above [/EDIT]

The following code snippet runs and gives you the desired result even though your approach is highly inefficient.

However, as @thackl mentioned, you made some horrible severe mistakes in your code! I do not want to be rude, but: read an introduction to Perl and learn some basics in programming before you proceed! Without that, your code will continue to be messy, fail, and you will waste time - yours and ours. Sorry.

#!/usr/bin/perl
use strict;
use warnings;

# Split the string sequence into an array
my @seq1 = split('', "UAUGUACCGACCUUAUUCUCCUAUGUACCGACCUUAUUCUCCUGUGU..."); 
my @seq2 = split('', "UGGAGUGUGACAAUGGUGUUUG");

# Determine the length of the shorter sequence
my $length;
if( scalar(@seq1) < scalar(@seq2) ){
   $length = scalar(@seq1);
}
else {
   $length = scalar(@seq2);
}

# Iterate the arrays and store results
my $result = '';
foreach (0 .. ($length - 1)){
   if($seq1[$_] ne $seq2[$_]){
      $result .= $seq2[$_];
   }
   else {
      $result .= " ";
   }
}
print $result, "\n";