New neighborhood and tabu search algorithm for the single machine scheduling problem

Motivated by the real-life scheduling problem in a steel-wire factory, the single machine scheduling problem with family setups is studied. With the properties of the neighborhood, it is shown that there are many non-improving neighbors in the traditional neighborhood and inefficient for searching. Then through analyzing the structure of objective function, a batch-destruction-construction method was proposed to generate new neighborhood, which generates better neighbors by reducing the setup times. Base on the new neighborhood, a new tabu search (TS) algorithm were proposed. The proposed TS algorithms were tested both on generated random problems and on the real-lift problems. Experimental results show that the proposed algorithm is better than the existing algorithm and can obtain near optimal solutions for the real-life problems within one second.