A heuristic algorithm SDS for scheduling with timed Petri nets

The paper proposes a new heuristic algorithm SDS for the scheduling problem with timed Petri nets: Given a timed Petri net (a Petri net in which delay is associated with transition firing), an initial marking M and a firing count vector X, find a firing sequence (a sequence of transitions) which is legal on M with respect to X and which optimizes a given objective function. Such a sequence is called a scheduling. The objective function in the paper is to minimize its completion time. So far SDS has been applied to 200 test problems such that, for each of them, existence of an exact solution is guaranteed, and it finds a solution to each of 162(81) test problems. This is better than the result (144 test ones: 72) by the algorithm YWPLS that is known to have highest capability among existing ones.