Abstract: To get a better nonpermutation flow shop(NPFS) scheduling， buffers， machine availability constraints and sequencedependent setup time were considered in building the mathematical model and disjunctive graph model. A NPFSoriented list heuristic algorithm was constructed to minimize the makespan. Allowed list and candidate list were built to record heuristic process information. Quantum ant colony and SPT heuristic rule were applied to search and select the feasible movement of related disjunction arcs to get an acyclic conjunctive graph. The algorithm parameters were verified through orthogonal test. The optimal disjunctive scheduling solution was the same with the exact CPLEX solution. The proposed algorithm has been critically compared and assessed by eight benchmark problems from Demirkol with the MHDACS and ACO ones， which confirms its good effectiveness and robustness.

Key words: buffers, nonpermutation, machine availability, disjunctive graph, quantum ant colony