See the File::Find module, I believe the doucmentation contains an example which is very similar to your requirements.

File::Find will help you to find the files themselves based on criteria you define. That's already been suggested. I wanted to add one sanity-check piece of advice:

Whatever you do, make sure that while you're in your testing phases the only action you take is to print a list of delete candidates, or to generate a log file of delete candidates. That way you will have every opportunity to check to make sure that the correct files are being identified before you release dogs. I imagine many a costly mistake has been made by writing a script to act in some irrevocable way on a large data set only to find that it acted in an unintended way.

Now back to your regularly scheduled programming....

use warnings;
use strict;
use DateTime;
use File::Find::Rule;

my $time_zone = DateTime::TimeZone->new(name => 'local');
my $a_year_ago = DateTime->now(time_zone => $time_zone)->subtract(year
+s => 1);

File::Find::Rule->file
    ->name(qr/\A\d{7}_\d-\d{7}-\d{2}_\d{10}\.para\z/)
    ->mtime('<' . $a_year_ago->epoch)
    ->exec(
        sub {
            ### Your logic.
        }
    )->in(qw(
        /paths
        /to
        /search
    ));
[download]

One good tip, at least for Windows boxes, is that you should accumulate a list of candidates that you want to delete, and only then go back and delete them.   In my experience, directory-searches are rather easily disrupted by other file-operations in Windows.