Scheduling of tasks to a single resource using Prolog
- by Reed Debaets
I searched through here as best I could and though I found some relevant questions, I don't think they covered the question at hand:
Assume a single resource and a known list of requests to schedule a task. Each request includes a start_after, start_by, expected_duration, and action.
The goal is to schedule the tasks for execution as soon as possible while keeping each task scheduled between start_after and start_by.
I coded up a simple prolog example that I "thought" should work but I've been unfortunately getting errors during run time: "=/2: Arguments are not sufficiently instantiated".
Any help or advice would be greatly appreciated
startAfter(1,0).
startAfter(2,0).
startAfter(3,0).
startBy(1,100).
startBy(2,500).
startBy(3,300).
duration(1,199).
duration(2,199).
duration(3,199).
action(1,'noop1').
action(2,'noop2').
action(3,'noop3').
can_run(R,T) :- startAfter(R,TA),startBy(R,TB),T>=TA,T=<TB.
conflicts(T,R1,T1) :- duration(R1,D1),T=<D1+T1,T>T1.
schedule(R1,T1,R2,T2,R3,T3) :-
can_run(R1,T1),\+conflicts(T1,R2,T2),\+conflicts(T1,R3,T3),
can_run(R2,T2),\+conflicts(T2,R1,T1),\+conflicts(T2,R3,T3),
can_run(R3,T3),\+conflicts(T3,R1,T1),\+conflicts(T3,R2,T2).
% when traced I *should* see T1=0, T2=400, T3=200
Edit:
conflicts goal wasn't quite right: needed extra TT1 clause.
Edit:
Apparently my schedule goal works if I supply valid Request,Time pairs ... but I'm stucking trying to force prolog to find valid values for T1..3 when given R1..3?
