Single machine scheduling minimizing non-linear cost
or $1||\sum w_j sgn(\beta)C_j^\beta$
We are given n jobs. Each job has a processing time pj and a weight wj. The goal is to find an ordering of the jobs that minimizes Σwjsgn(β)Cjβ, where β is some arbitrary real constant, and Cj is the completion time of job j. Also sgn(β) is defined as β/|β|.
The complexity status of this problem is open.
[src]
$jobs = $_REQUEST[jobs] ;
$b= $_REQUEST[b] ;
if (!$jobs) {
$n=10;
for($i=0; $i<$n; $i+=1) {
$p = rand(5,55);
$w = $p + rand(0,5);
$jobs .= "$p/$w ";
}
$b = 2;
}
?>
Input
Output
The diagram below shows the optimal schedule. Vertical lines indicate for all job pairs i,j the separation point t*. (See this article for explanation).
echo "
";
?>