Dispatching Rules
These are fast algorithms for solving a scheduling problem exactly or
heuristically. Use a dispatching rule if it solves your problem
exactly or if you need a very fast algorithm.
What you have to do:
Make sure that both a
problem type and all
parameters
are defined. Choose Heuristic Algortihms|Dispatching Rules from
the menu Algorithms and after editing the options press OK.
Dependent on the current problem type, special dispatching rules may
also be listed as exact algorithms.
Options:
- construct active schedule
-
Choose an algorithm:
- TRUE
- At each time a job or machine becomes ready schedule
the ready unscheduled operation with highest priority. The
resulting schedule is active.
- FALSE
- To extend a partial schedule, among all unscheduled
operations without unscheduled predecessor the one with highest
priority is appended to the schedule. The resulting
schedule is semiactive, but in general not activ.
- Priority
-
Choose one of the following definitions of priority. The algorithm
will schedule the operation
- RAND
- which is randomly determined,
- FCFS
- which became ready first,
- EDD
- with the smallest due date,
- LQUE
- with smallest difference du date-(processing time +
tail),
- SPT
- with smallest processing time
- WSPT
- with smallest ratio (processing time / weight)
- ECT
- which can be completed first
- WI
- with biggest weight,
- LPT
- with longest processing time
But...
- If you call a dispatching rule as exact algorithm, normally you
cannot choose the priority or rule type. If you want to choose, call
Algorithms|Heuristic Algorithms|Dispatching Rules.
- Dispatching rules will not appear in any menue if the current
problem type cannot be handled by the implemented algorithm.
Table of Contents
29.10.99 TAU