Use a hash of arrays to save the SQL statements and their running times, then use standard routines from List::Util to compute all the data you need (in other words, don't reinvent the wheel):

#!/usr/bin/perl

use feature qw/say/;
use strict;
use warnings;
use List::Util qw/min max reduce/;

my %results = ();

while(<>) {
    chomp;
    my ($sql, $time) = m/^SQL :: (.*), Time Taken :: (\d+)$/;

    push @{ $results{$sql} }, $time;
}

foreach(sort keys %results) {
    my $occ = scalar                @{ $results{$_} };
    my $min = min                   @{ $results{$_} };
    my $max = max                   @{ $results{$_} };
    my $tot = reduce { $a + $b } 0, @{ $results{$_} };

    say "$_ $occ $min $max $tot";
}
[download]

This outputs:

$ perl test.pl
select * from emp ...etc  2 5 30 35
select * from home ...etc  2 20 20 40
select * from person ...etc  3 10 50 90
$
[download]

P.S. I see that you crossposted your question to Stack Overflow. It is generally considered polite to inform people of this to avoid duplicated effort.