We consider the problem of scheduling n jobs on a single machine that
is continuously available from time zero onward and that can handle no
more than one job at a time. Each job requires processing during a gi
ven positive uninterrupted time. The cost of each job is measured by K
(K = 2,3) nondecreasing penalty functions; the quality of a schedule
is computed on the basis of K performance criteria, the kth one being
given by the maximum value of the kth penalty function over all jobs.
We wish to minimize an objective function that is a nondecreasing func
tion of these K performance criteria. We present a polynomial algorith
m for both problems and we show that these can also be used if precede
nce constraints exist between the jobs or if all penalty functions are
nonincreasing in the job completion times. (C) 1996 Academic Press, I
nc.