Job-Shop Scheduling by Simulated Annealing Combined with Deterministic Local Search

The Job-Shop Scheduling Problem (JSSP) is one of the most difficult NP-hard combinatorial optimization problems. This paper proposes a new method for solving JSSPs based on simulated annealing (SA), a stochastic local search, enhanced by shifting bottleneck (SB), a problem specific deterministic local search. In our method new schedules are generated by a variant of Giffler and Thompson’s active scheduler with operation permutations on the critical path. SA selects a new schedule and probabilistically accepts or rejects it. The modified SB is applied to repair the rejected schedule; the new schedule is accepted if an improvement is made. Experimental results showed the proposed method found near optimal schedules for the difficult benchmark problems and outperformed other existing local search algorithms.